Merge branch 'drm-nouveau-next' of git://git.freedesktop.org/git/nouveau/linux-2.6 into drm-next

* 'drm-nouveau-next' of git://git.freedesktop.org/git/nouveau/linux-2.6: (353 commits) drm/nouveau: remove allocations from gart populate() hook drm/nvc0/fb: slightly improve PMFB intr handling, move out of nvc0_graph.c drm/nvc0/fifo: avoid touching missing subfifos drm/nvd9/disp: bail out of mode_set_base if no fb bound to crtc drm/nvd9/disp: stub some more api hooks so we don't oops on resume drm/nouveau: fix printk typo in ioremap failure path drm/nvc0/pm: minor clock readback fixes drm/nv40/pm: execute memory reset script from vbios drm/nv50/gr: refactor initialisation drm/nouveau: if requested, try harder at disabling sysmem pushbufs drm/nv50/gr: enable ctxprog xfer only when we need it to save power drm/nouveau/dp: add support for displayport table 0x30 drm/nouveau/dp: return master dp table pointer too when looking up encoder drm/nouveau/bios: simplify U/d table hash matching func to just match drm/nouveau/dp: preserve non-pattern bits in DP_TRAINING_PATTERN_SET drm/nvc0/gr: remove MODULE_FIRMWARE() lines drm/nouveau/dp: use alternate lane mask for nvaf drm/nouveau/dp: link rate scripts are selected with a comparison table drm/nv40/pm: write nv40-specific reclocking routines drm/nv40/pm: parse geometric delta clock from vbios ...
author: Dave Airlie <airlied@redhat.com> 2011-09-20 09:35:22 +0100
committer: Dave Airlie <airlied@redhat.com> 2011-09-20 09:35:22 +0100
commit: b2d108ba333cdff80d9e7645d7697cbb6bb0fc29 (patch)
tree: f62b2464a35e4676b5e3e9ef340758a17d0b80f8
parent: fcf4de5acf09889e3f0c131ebe385c983006d71b (diff)
parent: a0d9a8feb928465f3cb525a19e5fafd06ef66ced (diff)
281 files changed, 6233 insertions, 2521 deletions
diff --git a/Documentation/DocBook/media/v4l/controls.xml b/Documentation/DocBook/media/v4l/controls.xml
index 85164016ed2..23fdf79f8cf 100644
--- a/Documentation/DocBook/media/v4l/controls.xml
+++ b/Documentation/DocBook/media/v4l/controls.xml
@@ -1455,7 +1455,7 @@ Applicable to the H264 encoder.</entry>
 	      </row>
 
 	      <row><entry></entry></row>
-	      <row>
+	      <row id="v4l2-mpeg-video-h264-vui-sar-idc">
 		<entry spanname="id"><constant>V4L2_CID_MPEG_VIDEO_H264_VUI_SAR_IDC</constant>&nbsp;</entry>
 		<entry>enum&nbsp;v4l2_mpeg_video_h264_vui_sar_idc</entry>
 	      </row>
@@ -1561,7 +1561,7 @@ Applicable to the H264 encoder.</entry>
 	      </row>
 
 	      <row><entry></entry></row>
-	      <row>
+	      <row id="v4l2-mpeg-video-h264-level">
 		<entry spanname="id"><constant>V4L2_CID_MPEG_VIDEO_H264_LEVEL</constant>&nbsp;</entry>
 		<entry>enum&nbsp;v4l2_mpeg_video_h264_level</entry>
 	      </row>
@@ -1641,7 +1641,7 @@ Possible values are:</entry>
 	      </row>
 
 	      <row><entry></entry></row>
-	      <row>
+	      <row id="v4l2-mpeg-video-mpeg4-level">
 		<entry spanname="id"><constant>V4L2_CID_MPEG_VIDEO_MPEG4_LEVEL</constant>&nbsp;</entry>
 		<entry>enum&nbsp;v4l2_mpeg_video_mpeg4_level</entry>
 	      </row>
@@ -1689,9 +1689,9 @@ Possible values are:</entry>
 	      </row>
 
 	      <row><entry></entry></row>
-	      <row>
+	      <row id="v4l2-mpeg-video-h264-profile">
 		<entry spanname="id"><constant>V4L2_CID_MPEG_VIDEO_H264_PROFILE</constant>&nbsp;</entry>
-		<entry>enum&nbsp;v4l2_mpeg_h264_profile</entry>
+		<entry>enum&nbsp;v4l2_mpeg_video_h264_profile</entry>
 	      </row>
 	      <row><entry spanname="descr">The profile information for H264.
 Applicable to the H264 encoder.
@@ -1774,9 +1774,9 @@ Possible values are:</entry>
 	      </row>
 
 	      <row><entry></entry></row>
-	      <row>
+	      <row id="v4l2-mpeg-video-mpeg4-profile">
 		<entry spanname="id"><constant>V4L2_CID_MPEG_VIDEO_MPEG4_PROFILE</constant>&nbsp;</entry>
-		<entry>enum&nbsp;v4l2_mpeg_mpeg4_profile</entry>
+		<entry>enum&nbsp;v4l2_mpeg_video_mpeg4_profile</entry>
 	      </row>
 	      <row><entry spanname="descr">The profile information for MPEG4.
 Applicable to the MPEG4 encoder.
@@ -1820,9 +1820,9 @@ Applicable to the encoder.
 	      </row>
 
 	      <row><entry></entry></row>
-	      <row>
+	      <row id="v4l2-mpeg-video-multi-slice-mode">
 		<entry spanname="id"><constant>V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MODE</constant>&nbsp;</entry>
-		<entry>enum&nbsp;v4l2_mpeg_multi_slice_mode</entry>
+		<entry>enum&nbsp;v4l2_mpeg_video_multi_slice_mode</entry>
 	      </row>
 	      <row><entry spanname="descr">Determines how the encoder should handle division of frame into slices.
 Applicable to the encoder.
@@ -1868,9 +1868,9 @@ Applicable to the encoder.</entry>
 	      </row>
 
 	      <row><entry></entry></row>
-	      <row>
+	      <row id="v4l2-mpeg-video-h264-loop-filter-mode">
 		<entry spanname="id"><constant>V4L2_CID_MPEG_VIDEO_H264_LOOP_FILTER_MODE</constant>&nbsp;</entry>
-		<entry>enum&nbsp;v4l2_mpeg_h264_loop_filter_mode</entry>
+		<entry>enum&nbsp;v4l2_mpeg_video_h264_loop_filter_mode</entry>
 	      </row>
 	      <row><entry spanname="descr">Loop filter mode for H264 encoder.
 Possible values are:</entry>
@@ -1913,9 +1913,9 @@ Applicable to the H264 encoder.</entry>
 	      </row>
 
 	      <row><entry></entry></row>
-	      <row>
+	      <row id="v4l2-mpeg-video-h264-entropy-mode">
 		<entry spanname="id"><constant>V4L2_CID_MPEG_VIDEO_H264_ENTROPY_MODE</constant>&nbsp;</entry>
-		<entry>enum&nbsp;v4l2_mpeg_h264_symbol_mode</entry>
+		<entry>enum&nbsp;v4l2_mpeg_video_h264_entropy_mode</entry>
 	      </row>
 	      <row><entry spanname="descr">Entropy coding mode for H264 - CABAC/CAVALC.
 Applicable to the H264 encoder.
@@ -2140,9 +2140,9 @@ previous frames. Applicable to the H264 encoder.</entry>
 	      </row>
 
 	      <row><entry></entry></row>
-	      <row>
+	      <row id="v4l2-mpeg-video-header-mode">
 		<entry spanname="id"><constant>V4L2_CID_MPEG_VIDEO_HEADER_MODE</constant>&nbsp;</entry>
-		<entry>enum&nbsp;v4l2_mpeg_header_mode</entry>
+		<entry>enum&nbsp;v4l2_mpeg_video_header_mode</entry>
 	      </row>
 	      <row><entry spanname="descr">Determines whether the header is returned as the first buffer or is
 it returned together with the first frame. Applicable to encoders.
@@ -2320,9 +2320,9 @@ Valid only when H.264 and macroblock level RC is enabled (<constant>V4L2_CID_MPE
 Applicable to the H264 encoder.</entry>
 	      </row>
 	      <row><entry></entry></row>
-	      <row>
+	      <row id="v4l2-mpeg-mfc51-video-frame-skip-mode">
 		<entry spanname="id"><constant>V4L2_CID_MPEG_MFC51_VIDEO_FRAME_SKIP_MODE</constant>&nbsp;</entry>
-		<entry>enum&nbsp;v4l2_mpeg_mfc51_frame_skip_mode</entry>
+		<entry>enum&nbsp;v4l2_mpeg_mfc51_video_frame_skip_mode</entry>
 	      </row>
 	      <row><entry spanname="descr">
 Indicates in what conditions the encoder should skip frames. If encoding a frame would cause the encoded stream to be larger then
@@ -2361,9 +2361,9 @@ the stream will meet tight bandwidth contraints. Applicable to encoders.
 </entry>
 	      </row>
 	      <row><entry></entry></row>
-	      <row>
+	      <row id="v4l2-mpeg-mfc51-video-force-frame-type">
 		<entry spanname="id"><constant>V4L2_CID_MPEG_MFC51_VIDEO_FORCE_FRAME_TYPE</constant>&nbsp;</entry>
-		<entry>enum&nbsp;v4l2_mpeg_mfc51_force_frame_type</entry>
+		<entry>enum&nbsp;v4l2_mpeg_mfc51_video_force_frame_type</entry>
 	      </row>
 	      <row><entry spanname="descr">Force a frame type for the next queued buffer. Applicable to encoders.
 Possible values are:</entry>
diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt
index c4a6e148732..4dc46547766 100644
--- a/Documentation/feature-removal-schedule.txt
+++ b/Documentation/feature-removal-schedule.txt
@@ -592,3 +592,11 @@ Why:    In 3.0, we can now autodetect internal 3G device and already have
 	interface that was used by acer-wmi driver. It will replaced by
 	information log when acer-wmi initial.
 Who:    Lee, Chun-Yi <jlee@novell.com>
+
+----------------------------
+What:	The XFS nodelaylog mount option
+When:	3.3
+Why:	The delaylog mode that has been the default since 2.6.39 has proven
+	stable, and the old code is in the way of additional improvements in
+	the log code.
+Who:	Christoph Hellwig <hch@lst.de>
diff --git a/Documentation/hwmon/max16065 b/Documentation/hwmon/max16065
index 44b4f61e04f..c11f64a1f2a 100644
--- a/Documentation/hwmon/max16065
+++ b/Documentation/hwmon/max16065
@@ -62,6 +62,13 @@ can be safely used to identify the chip. You will have to instantiate
 the devices explicitly. Please see Documentation/i2c/instantiating-devices for
 details.
 
+WARNING: Do not access chip registers using the i2cdump command, and do not use
+any of the i2ctools commands on a command register (0xa5 to 0xac). The chips
+supported by this driver interpret any access to a command register (including
+read commands) as request to execute the command in question. This may result in
+power loss, board resets, and/or Flash corruption. Worst case, your board may
+turn into a brick.
+
 
 Sysfs entries
 -------------
diff --git a/Documentation/ioctl/ioctl-number.txt b/Documentation/ioctl/ioctl-number.txt
index 845a191004b..54078ed96b3 100644
--- a/Documentation/ioctl/ioctl-number.txt
+++ b/Documentation/ioctl/ioctl-number.txt
@@ -319,4 +319,6 @@ Code  Seq#(hex)	Include File		Comments
 					<mailto:thomas@winischhofer.net>
 0xF4	00-1F	video/mbxfb.h		mbxfb
 					<mailto:raph@8d.com>
+0xF6	all	LTTng			Linux Trace Toolkit Next Generation
+					<mailto:mathieu.desnoyers@efficios.com>
 0xFD	all	linux/dm-ioctl.h
diff --git a/MAINTAINERS b/MAINTAINERS
index 1a8cc600067..28f65c249b9 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -2649,11 +2649,11 @@ F:	drivers/net/wan/dlci.c
 F:	drivers/net/wan/sdla.c
 
 FRAMEBUFFER LAYER
-M:	Paul Mundt <lethal@linux-sh.org>
+M:	Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
 L:	linux-fbdev@vger.kernel.org
 W:	http://linux-fbdev.sourceforge.net/
 Q:	http://patchwork.kernel.org/project/linux-fbdev/list/
-T:	git git://git.kernel.org/pub/scm/linux/kernel/git/lethal/fbdev-2.6.git
+T:	git git://github.com/schandinat/linux-2.6.git fbdev-next
 S:	Maintained
 F:	Documentation/fb/
 F:	Documentation/devicetree/bindings/fb/
@@ -4450,8 +4450,8 @@ M:	"David S. Miller" <davem@davemloft.net>
 L:	netdev@vger.kernel.org
 W:	http://www.linuxfoundation.org/en/Net
 W:	http://patchwork.ozlabs.org/project/netdev/list/
-T:	git git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6.git
-T:	git git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6.git
+T:	git git://git.kernel.org/pub/scm/linux/kernel/git/davem/net.git
+T:	git git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git
 S:	Maintained
 F:	net/
 F:	include/net/
diff --git a/Makefile b/Makefile
index c3e90c530a6..522fa4784e6 100644
--- a/Makefile
+++ b/Makefile
@@ -1,7 +1,7 @@
 VERSION = 3
 PATCHLEVEL = 1
 SUBLEVEL = 0
-EXTRAVERSION = -rc4
+EXTRAVERSION = -rc6
 NAME = "Divemaster Edition"
 
 # *DOCUMENTATION*
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 5ebc5d922ea..3269576dbfa 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -1271,6 +1271,18 @@ config ARM_ERRATA_754327
 	  This workaround defines cpu_relax() as smp_mb(), preventing correctly
 	  written polling loops from denying visibility of updates to memory.
 
+config ARM_ERRATA_364296
+	bool "ARM errata: Possible cache data corruption with hit-under-miss enabled"
+	depends on CPU_V6 && !SMP
+	help
+	  This options enables the workaround for the 364296 ARM1136
+	  r0p2 erratum (possible cache data corruption with
+	  hit-under-miss enabled). It sets the undocumented bit 31 in
+	  the auxiliary control register and the FI bit in the control
+	  register, thus disabling hit-under-miss without putting the
+	  processor into full low interrupt latency mode. ARM11MPCore
+	  is not affected.
+
 endmenu
 
 source "arch/arm/common/Kconfig"
diff --git a/arch/arm/boot/compressed/mmcif-sh7372.c b/arch/arm/boot/compressed/mmcif-sh7372.c
index b6f61d9a5a1..672ae95db5c 100644
--- a/arch/arm/boot/compressed/mmcif-sh7372.c
+++ b/arch/arm/boot/compressed/mmcif-sh7372.c
@@ -82,7 +82,7 @@ asmlinkage void mmc_loader(unsigned char *buf, unsigned long len)
 
 
 	/* Disable clock to MMC hardware block */
-	__raw_writel(__raw_readl(SMSTPCR3) & (1 << 12), SMSTPCR3);
+	__raw_writel(__raw_readl(SMSTPCR3) | (1 << 12), SMSTPCR3);
 
 	mmc_update_progress(MMC_PROGRESS_DONE);
 }
diff --git a/arch/arm/boot/compressed/sdhi-sh7372.c b/arch/arm/boot/compressed/sdhi-sh7372.c
index d403a8b24d7..d279294f238 100644
--- a/arch/arm/boot/compressed/sdhi-sh7372.c
+++ b/arch/arm/boot/compressed/sdhi-sh7372.c
@@ -85,7 +85,7 @@ asmlinkage void mmc_loader(unsigned short *buf, unsigned long len)
 		goto err;
 
         /* Disable clock to SDHI1 hardware block */
-        __raw_writel(__raw_readl(SMSTPCR3) & (1 << 13), SMSTPCR3);
+        __raw_writel(__raw_readl(SMSTPCR3) | (1 << 13), SMSTPCR3);
 
 	mmc_update_progress(MMC_PROGRESS_DONE);
 
diff --git a/arch/arm/include/asm/hardware/cache-l2x0.h b/arch/arm/include/asm/hardware/cache-l2x0.h
index 16bd4803158..99a6ed7e1bf 100644
--- a/arch/arm/include/asm/hardware/cache-l2x0.h
+++ b/arch/arm/include/asm/hardware/cache-l2x0.h
@@ -45,8 +45,13 @@
 #define L2X0_CLEAN_INV_LINE_PA		0x7F0
 #define L2X0_CLEAN_INV_LINE_IDX		0x7F8
 #define L2X0_CLEAN_INV_WAY		0x7FC
-#define L2X0_LOCKDOWN_WAY_D		0x900
-#define L2X0_LOCKDOWN_WAY_I		0x904
+/*
+ * The lockdown registers repeat 8 times for L310, the L210 has only one
+ * D and one I lockdown register at 0x0900 and 0x0904.
+ */
+#define L2X0_LOCKDOWN_WAY_D_BASE	0x900
+#define L2X0_LOCKDOWN_WAY_I_BASE	0x904
+#define L2X0_LOCKDOWN_STRIDE		0x08
 #define L2X0_TEST_OPERATION		0xF00
 #define L2X0_LINE_DATA			0xF10
 #define L2X0_LINE_TAG			0xF30
@@ -64,7 +69,7 @@
 #define L2X0_AUX_CTRL_MASK			0xc0000fff
 #define L2X0_AUX_CTRL_ASSOCIATIVITY_SHIFT	16
 #define L2X0_AUX_CTRL_WAY_SIZE_SHIFT		17
-#define L2X0_AUX_CTRL_WAY_SIZE_MASK		(0x3 << 17)
+#define L2X0_AUX_CTRL_WAY_SIZE_MASK		(0x7 << 17)
 #define L2X0_AUX_CTRL_SHARE_OVERRIDE_SHIFT	22
 #define L2X0_AUX_CTRL_NS_LOCKDOWN_SHIFT		26
 #define L2X0_AUX_CTRL_NS_INT_CTRL_SHIFT		27
diff --git a/arch/arm/include/asm/pmu.h b/arch/arm/include/asm/pmu.h
index 67c70a31a1b..b7e82c4aced 100644
--- a/arch/arm/include/asm/pmu.h
+++ b/arch/arm/include/asm/pmu.h
@@ -41,7 +41,7 @@ struct arm_pmu_platdata {
  * encoded error on failure.
  */
 extern struct platform_device *
-reserve_pmu(enum arm_pmu_type device);
+reserve_pmu(enum arm_pmu_type type);
 
 /**
  * release_pmu() - Relinquish control of the performance counters
@@ -62,26 +62,26 @@ release_pmu(enum arm_pmu_type type);
  * the actual hardware initialisation.
  */
 extern int
-init_pmu(enum arm_pmu_type device);
+init_pmu(enum arm_pmu_type type);
 
 #else /* CONFIG_CPU_HAS_PMU */
 
 #include <linux/err.h>
 
 static inline struct platform_device *
-reserve_pmu(enum arm_pmu_type device)
+reserve_pmu(enum arm_pmu_type type)
 {
 	return ERR_PTR(-ENODEV);
 }
 
 static inline int
-release_pmu(struct platform_device *pdev)
+release_pmu(enum arm_pmu_type type)
 {
 	return -ENODEV;
 }
 
 static inline int
-init_pmu(enum arm_pmu_type device)
+init_pmu(enum arm_pmu_type type)
 {
 	return -ENODEV;
 }
diff --git a/arch/arm/kernel/pmu.c b/arch/arm/kernel/pmu.c
index 2b70709376c..c53474fe84d 100644
--- a/arch/arm/kernel/pmu.c
+++ b/arch/arm/kernel/pmu.c
@@ -31,7 +31,7 @@ static int __devinit pmu_register(struct platform_device *pdev,
 {
 	if (type < 0 || type >= ARM_NUM_PMU_DEVICES) {
 		pr_warning("received registration request for unknown "
-				"device %d\n", type);
+				"PMU device type %d\n", type);
 		return -EINVAL;
 	}
 
@@ -112,17 +112,17 @@ static int __init register_pmu_driver(void)
 device_initcall(register_pmu_driver);
 
 struct platform_device *
-reserve_pmu(enum arm_pmu_type device)
+reserve_pmu(enum arm_pmu_type type)
 {
 	struct platform_device *pdev;
 
-	if (test_and_set_bit_lock(device, &pmu_lock)) {
+	if (test_and_set_bit_lock(type, &pmu_lock)) {
 		pdev = ERR_PTR(-EBUSY);
-	} else if (pmu_devices[device] == NULL) {
-		clear_bit_unlock(device, &pmu_lock);
+	} else if (pmu_devices[type] == NULL) {
+		clear_bit_unlock(type, &pmu_lock);
 		pdev = ERR_PTR(-ENODEV);
 	} else {
-		pdev = pmu_devices[device];
+		pdev = pmu_devices[type];
 	}
 
 	return pdev;
@@ -130,11 +130,11 @@ reserve_pmu(enum arm_pmu_type device)
 EXPORT_SYMBOL_GPL(reserve_pmu);
 
 int
-release_pmu(enum arm_pmu_type device)
+release_pmu(enum arm_pmu_type type)
 {
-	if (WARN_ON(!pmu_devices[device]))
+	if (WARN_ON(!pmu_devices[type]))
 		return -EINVAL;
-	clear_bit_unlock(device, &pmu_lock);
+	clear_bit_unlock(type, &pmu_lock);
 	return 0;
 }
 EXPORT_SYMBOL_GPL(release_pmu);
@@ -182,17 +182,17 @@ init_cpu_pmu(void)
 }
 
 int
-init_pmu(enum arm_pmu_type device)
+init_pmu(enum arm_pmu_type type)
 {
 	int err = 0;
 
-	switch (device) {
+	switch (type) {
 	case ARM_PMU_DEVICE_CPU:
 		err = init_cpu_pmu();
 		break;
 	default:
-		pr_warning("attempt to initialise unknown device %d\n",
-				device);
+		pr_warning("attempt to initialise PMU of unknown "
+			   "type %d\n", type);
 		err = -EINVAL;
 	}
 
diff --git a/arch/arm/kernel/relocate_kernel.S b/arch/arm/kernel/relocate_kernel.S
index 9cf4cbf8f95..d0cdedf4864 100644
--- a/arch/arm/kernel/relocate_kernel.S
+++ b/arch/arm/kernel/relocate_kernel.S
@@ -57,7 +57,8 @@ relocate_new_kernel:
 	mov r0,#0
 	ldr r1,kexec_mach_type
 	ldr r2,kexec_boot_atags
-	mov pc,lr
+ ARM(	mov pc, lr	)
+ THUMB(	bx lr		)
 
 	.align
 
diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c
index 70bca649e92..e514c76043b 100644
--- a/arch/arm/kernel/setup.c
+++ b/arch/arm/kernel/setup.c
@@ -280,18 +280,19 @@ static void __init cacheid_init(void)
 	if (arch >= CPU_ARCH_ARMv6) {
 		if ((cachetype & (7 << 29)) == 4 << 29) {
 			/* ARMv7 register format */
+			arch = CPU_ARCH_ARMv7;
 			cacheid = CACHEID_VIPT_NONALIASING;
 			if ((cachetype & (3 << 14)) == 1 << 14)
 				cacheid |= CACHEID_ASID_TAGGED;
-			else if (cpu_has_aliasing_icache(CPU_ARCH_ARMv7))
-				cacheid |= CACHEID_VIPT_I_ALIASING;
-		} else if (cachetype & (1 << 23)) {
-			cacheid = CACHEID_VIPT_ALIASING;
 		} else {
-			cacheid = CACHEID_VIPT_NONALIASING;
-			if (cpu_has_aliasing_icache(CPU_ARCH_ARMv6))
-				cacheid |= CACHEID_VIPT_I_ALIASING;
+			arch = CPU_ARCH_ARMv6;
+			if (cachetype & (1 << 23))
+				cacheid = CACHEID_VIPT_ALIASING;
+			else
+				cacheid = CACHEID_VIPT_NONALIASING;
 		}
+		if (cpu_has_aliasing_icache(arch))
+			cacheid |= CACHEID_VIPT_I_ALIASING;
 	} else {
 		cacheid = CACHEID_VIVT;
 	}
diff --git a/arch/arm/kernel/smp_twd.c b/arch/arm/kernel/smp_twd.c
index 2c277d40cee..01c186222f3 100644
--- a/arch/arm/kernel/smp_twd.c
+++ b/arch/arm/kernel/smp_twd.c
@@ -137,8 +137,8 @@ void __cpuinit twd_timer_setup(struct clock_event_device *clk)
 	clk->max_delta_ns = clockevent_delta2ns(0xffffffff, clk);
 	clk->min_delta_ns = clockevent_delta2ns(0xf, clk);
 
+	clockevents_register_device(clk);
+
 	/* Make sure our local interrupt controller has this enabled */
 	gic_enable_ppi(clk->irq);
-
-	clockevents_register_device(clk);
 }
diff --git a/arch/arm/mach-at91/at91sam9261.c b/arch/arm/mach-at91/at91sam9261.c
index d522b47e30b..6c8e3b5f669 100644
--- a/arch/arm/mach-at91/at91sam9261.c
+++ b/arch/arm/mach-at91/at91sam9261.c
@@ -157,7 +157,7 @@ static struct clk_lookup periph_clocks_lookups[] = {
 	CLKDEV_CON_DEV_ID("spi_clk", "atmel_spi.1", &spi1_clk),
 	CLKDEV_CON_DEV_ID("t0_clk", "atmel_tcb.0", &tc0_clk),
 	CLKDEV_CON_DEV_ID("t1_clk", "atmel_tcb.0", &tc1_clk),
-	CLKDEV_CON_DEV_ID("t2_clk", "atmel_tcb.0", &tc1_clk),
+	CLKDEV_CON_DEV_ID("t2_clk", "atmel_tcb.0", &tc2_clk),
 	CLKDEV_CON_DEV_ID("pclk", "ssc.0", &ssc0_clk),
 	CLKDEV_CON_DEV_ID("pclk", "ssc.1", &ssc1_clk),
 	CLKDEV_CON_DEV_ID("pclk", "ssc.2", &ssc2_clk),
diff --git a/arch/arm/mach-cns3xxx/include/mach/entry-macro.S b/arch/arm/mach-cns3xxx/include/mach/entry-macro.S
index 6bd83ed90af..d87bfc397d3 100644
--- a/arch/arm/mach-cns3xxx/include/mach/entry-macro.S
+++ b/arch/arm/mach-cns3xxx/include/mach/entry-macro.S
@@ -8,7 +8,6 @@
  * published by the Free Software Foundation.
  */
 
-#include <mach/hardware.h>
 #include <asm/hardware/entry-macro-gic.S>
 
 		.macro	disable_fiq
diff --git a/arch/arm/mach-cns3xxx/include/mach/system.h b/arch/arm/mach-cns3xxx/include/mach/system.h
index 58bb03ae3cf..4f16c9b79f7 100644
--- a/arch/arm/mach-cns3xxx/include/mach/system.h
+++ b/arch/arm/mach-cns3xxx/include/mach/system.h
@@ -13,7 +13,6 @@
 
 #include <linux/io.h>
 #include <asm/proc-fns.h>
-#include <mach/hardware.h>
 
 static inline void arch_idle(void)
 {
diff --git a/arch/arm/mach-cns3xxx/include/mach/uncompress.h b/arch/arm/mach-cns3xxx/include/mach/uncompress.h
index de8ead9b91f..a91b6058ab4 100644
--- a/arch/arm/mach-cns3xxx/include/mach/uncompress.h
+++ b/arch/arm/mach-cns3xxx/include/mach/uncompress.h
@@ -8,7 +8,6 @@
  */
 
 #include <asm/mach-types.h>
-#include <mach/hardware.h>
 #include <mach/cns3xxx.h>
 
 #define AMBA_UART_DR(base)	(*(volatile unsigned char *)((base) + 0x00))
diff --git a/arch/arm/mach-cns3xxx/pcie.c b/arch/arm/mach-cns3xxx/pcie.c
index 06fd25d70ae..0f8fca48a5e 100644
--- a/arch/arm/mach-cns3xxx/pcie.c
+++ b/arch/arm/mach-cns3xxx/pcie.c
@@ -49,7 +49,7 @@ static struct cns3xxx_pcie *sysdata_to_cnspci(void *sysdata)
 	return &cns3xxx_pcie[root->domain];
 }
 
-static struct cns3xxx_pcie *pdev_to_cnspci(struct pci_dev *dev)
+static struct cns3xxx_pcie *pdev_to_cnspci(const struct pci_dev *dev)
 {
 	return sysdata_to_cnspci(dev->sysdata);
 }
diff --git a/arch/arm/mach-davinci/board-da850-evm.c b/arch/arm/mach-davinci/board-da850-evm.c
index bd5394537c8..008d51407cd 100644
--- a/arch/arm/mach-davinci/board-da850-evm.c
+++ b/arch/arm/mach-davinci/board-da850-evm.c
@@ -115,6 +115,32 @@ static struct spi_board_info da850evm_spi_info[] = {
 	},
 };
 
+#ifdef CONFIG_MTD
+static void da850_evm_m25p80_notify_add(struct mtd_info *mtd)
+{
+	char *mac_addr = davinci_soc_info.emac_pdata->mac_addr;
+	size_t retlen;
+
+	if (!strcmp(mtd->name, "MAC-Address")) {
+		mtd->read(mtd, 0, ETH_ALEN, &retlen, mac_addr);
+		if (retlen == ETH_ALEN)
+			pr_info("Read MAC addr from SPI Flash: %pM\n",
+				mac_addr);
+	}
+}
+
+static struct mtd_notifier da850evm_spi_notifier = {
+	.add	= da850_evm_m25p80_notify_add,
+};
+
+static void da850_evm_setup_mac_addr(void)
+{
+	register_mtd_user(&da850evm_spi_notifier);
+}
+#else
+static void da850_evm_setup_mac_addr(void) { }
+#endif
+
 static struct mtd_partition da850_evm_norflash_partition[] = {
 	{
 		.name           = "bootloaders + env",
@@ -1244,6 +1270,8 @@ static __init void da850_evm_init(void)
 	if (ret)
 		pr_warning("da850_evm_init: sata registration failed: %d\n",
 				ret);
+
+	da850_evm_setup_mac_addr();
 }
 
 #ifdef CONFIG_SERIAL_8250_CONSOLE
diff --git a/arch/arm/mach-davinci/include/mach/psc.h b/arch/arm/mach-davinci/include/mach/psc.h
index 47fd0bc3d3e..fa59c097223 100644
--- a/arch/arm/mach-davinci/include/mach/psc.h
+++ b/arch/arm/mach-davinci/include/mach/psc.h
@@ -243,7 +243,7 @@
 #define PSC_STATE_DISABLE	2
 #define PSC_STATE_ENABLE	3
 
-#define MDSTAT_STATE_MASK	0x1f
+#define MDSTAT_STATE_MASK	0x3f
 #define MDCTL_FORCE		BIT(31)
 
 #ifndef __ASSEMBLER__
diff --git a/arch/arm/mach-davinci/sleep.S b/arch/arm/mach-davinci/sleep.S
index fb5e72b532b..5f1e045a3ad 100644
--- a/arch/arm/mach-davinci/sleep.S
+++ b/arch/arm/mach-davinci/sleep.S
@@ -217,7 +217,11 @@ ddr2clk_stop_done:
 ENDPROC(davinci_ddr_psc_config)
 
 CACHE_FLUSH:
-	.word	arm926_flush_kern_cache_all
+#ifdef CONFIG_CPU_V6
+	.word	v6_flush_kern_cache_all
+#else
+	.word   arm926_flush_kern_cache_all
+#endif
 
 ENTRY(davinci_cpu_suspend_sz)
 	.word	. - davinci_cpu_suspend
diff --git a/arch/arm/mach-ep93xx/include/mach/ts72xx.h b/arch/arm/mach-ep93xx/include/mach/ts72xx.h
index 0eabec62cd9..f1397a13e76 100644
--- a/arch/arm/mach-ep93xx/include/mach/ts72xx.h
+++ b/arch/arm/mach-ep93xx/include/mach/ts72xx.h
@@ -6,7 +6,7 @@
  * TS72xx memory map:
  *
  * virt		phys		size
- * febff000	22000000	4K	model number register
+ * febff000	22000000	4K	model number register (bits 0-2)
  * febfe000	22400000	4K	options register
  * febfd000	22800000	4K	options register #2
  * febf9000	10800000	4K	TS-5620 RTC index register
@@ -20,6 +20,9 @@
 #define TS72XX_MODEL_TS7200		0x00
 #define TS72XX_MODEL_TS7250		0x01
 #define TS72XX_MODEL_TS7260		0x02
+#define TS72XX_MODEL_TS7300		0x03
+#define TS72XX_MODEL_TS7400		0x04
+#define TS72XX_MODEL_MASK		0x07
 
 
 #define TS72XX_OPTIONS_PHYS_BASE	0x22400000
@@ -51,19 +54,34 @@
 
 #ifndef __ASSEMBLY__
 
+static inline int ts72xx_model(void)
+{
+	return __raw_readb(TS72XX_MODEL_VIRT_BASE) & TS72XX_MODEL_MASK;
+}
+
 static inline int board_is_ts7200(void)
 {
-	return __raw_readb(TS72XX_MODEL_VIRT_BASE) == TS72XX_MODEL_TS7200;
+	return ts72xx_model() == TS72XX_MODEL_TS7200;
 }
 
 static inline int board_is_ts7250(void)
 {
-	return __raw_readb(TS72XX_MODEL_VIRT_BASE) == TS72XX_MODEL_TS7250;
+	return ts72xx_model() == TS72XX_MODEL_TS7250;
 }
 
 static inline int board_is_ts7260(void)
 {
-	return __raw_readb(TS72XX_MODEL_VIRT_BASE) == TS72XX_MODEL_TS7260;
+	return ts72xx_model() == TS72XX_MODEL_TS7260;
+}
+
+static inline int board_is_ts7300(void)
+{
+	return ts72xx_model()  == TS72XX_MODEL_TS7300;
+}
+
+static inline int board_is_ts7400(void)
+{
+	return ts72xx_model() == TS72XX_MODEL_TS7400;
 }
 
 static inline int is_max197_installed(void)
diff --git a/arch/arm/mach-exynos4/clock.c b/arch/arm/mach-exynos4/clock.c
index 851dea01857..1561b036a9b 100644
--- a/arch/arm/mach-exynos4/clock.c
+++ b/arch/arm/mach-exynos4/clock.c
@@ -520,7 +520,7 @@ static struct clk init_clocks_off[] = {
 		.ctrlbit	= (1 << 21),
 	}, {
 		.name		= "ac97",
-		.id		= -1,
+		.devname	= "samsung-ac97",
 		.enable		= exynos4_clk_ip_peril_ctrl,
 		.ctrlbit	= (1 << 27),
 	}, {
diff --git a/arch/arm/mach-exynos4/cpu.c b/arch/arm/mach-exynos4/cpu.c
index 2d8a40c9e6e..746d6fc6d39 100644
--- a/arch/arm/mach-exynos4/cpu.c
+++ b/arch/arm/mach-exynos4/cpu.c
@@ -24,12 +24,13 @@
 #include <plat/exynos4.h>
 #include <plat/adc-core.h>
 #include <plat/sdhci.h>
-#include <plat/devs.h>
 #include <plat/fb-core.h>
 #include <plat/fimc-core.h>
 #include <plat/iic-core.h>
+#include <plat/reset.h>
 
 #include <mach/regs-irq.h>
+#include <mach/regs-pmu.h>
 
 extern int combiner_init(unsigned int combiner_nr, void __iomem *base,
 			 unsigned int irq_start);
@@ -128,6 +129,11 @@ static void exynos4_idle(void)
 	local_irq_enable();
 }
 
+static void exynos4_sw_reset(void)
+{
+	__raw_writel(0x1, S5P_SWRESET);
+}
+
 /*
  * exynos4_map_io
  *
@@ -241,5 +247,8 @@ int __init exynos4_init(void)
 	/* set idle function */
 	pm_idle = exynos4_idle;
 
+	/* set sw_reset function */
+	s5p_reset_hook = exynos4_sw_reset;
+
 	return sysdev_register(&exynos4_sysdev);
 }
diff --git a/arch/arm/mach-exynos4/include/mach/irqs.h b/arch/arm/mach-exynos4/include/mach/irqs.h
index 934d2a49398..f8952f8f375 100644
--- a/arch/arm/mach-exynos4/include/mach/irqs.h
+++ b/arch/arm/mach-exynos4/include/mach/irqs.h
@@ -80,9 +80,8 @@
 #define IRQ_HSMMC3		IRQ_SPI(76)
 #define IRQ_DWMCI		IRQ_SPI(77)
 
-#define IRQ_MIPICSI0		IRQ_SPI(78)
-
-#define IRQ_MIPICSI1		IRQ_SPI(80)
+#define IRQ_MIPI_CSIS0		IRQ_SPI(78)
+#define IRQ_MIPI_CSIS1		IRQ_SPI(80)
 
 #define IRQ_ONENAND_AUDI	IRQ_SPI(82)
 #define IRQ_ROTATOR		IRQ_SPI(83)
diff --git a/arch/arm/mach-exynos4/include/mach/regs-pmu.h b/arch/arm/mach-exynos4/include/mach/regs-pmu.h
index fa49bbb8e7b..cdf9b47c303 100644
--- a/arch/arm/mach-exynos4/include/mach/regs-pmu.h
+++ b/arch/arm/mach-exynos4/include/mach/regs-pmu.h
@@ -29,6 +29,8 @@
 #define S5P_USE_STANDBY_WFE1			(1 << 25)
 #define S5P_USE_MASK				((0x3 << 16) | (0x3 << 24))
 
+#define S5P_SWRESET				S5P_PMUREG(0x0400)
+
 #define S5P_WAKEUP_STAT				S5P_PMUREG(0x0600)
 #define S5P_EINT_WAKEUP_MASK			S5P_PMUREG(0x0604)
 #define S5P_WAKEUP_MASK				S5P_PMUREG(0x0608)
diff --git a/arch/arm/mach-exynos4/irq-eint.c b/arch/arm/mach-exynos4/irq-eint.c
index 9d87d2ac7f6..badb8c66fc9 100644
--- a/arch/arm/mach-exynos4/irq-eint.c
+++ b/arch/arm/mach-exynos4/irq-eint.c
@@ -23,6 +23,8 @@
 
 #include <mach/regs-gpio.h>
 
+#include <asm/mach/irq.h>
+
 static DEFINE_SPINLOCK(eint_lock);
 
 static unsigned int eint0_15_data[16];
@@ -184,8 +186,11 @@ static inline void exynos4_irq_demux_eint(unsigned int start)
 
 static void exynos4_irq_demux_eint16_31(unsigned int irq, struct irq_desc *desc)
 {
+	struct irq_chip *chip = irq_get_chip(irq);
+	chained_irq_enter(chip, desc);
 	exynos4_irq_demux_eint(IRQ_EINT(16));
 	exynos4_irq_demux_eint(IRQ_EINT(24));
+	chained_irq_exit(chip, desc);
 }
 
 static void exynos4_irq_eint0_15(unsigned int irq, struct irq_desc *desc)
@@ -193,6 +198,7 @@ static void exynos4_irq_eint0_15(unsigned int irq, struct irq_desc *desc)
 	u32 *irq_data = irq_get_handler_data(irq);
 	struct irq_chip *chip = irq_get_chip(irq);
 
+	chained_irq_enter(chip, desc);
 	chip->irq_mask(&desc->irq_data);
 
 	if (chip->irq_ack)
@@ -201,6 +207,7 @@ static void exynos4_irq_eint0_15(unsigned int irq, struct irq_desc *desc)
 	generic_handle_irq(*irq_data);
 
 	chip->irq_unmask(&desc->irq_data);
+	chained_irq_exit(chip, desc);
 }
 
 int __init exynos4_init_irq_eint(void)
diff --git a/arch/arm/mach-exynos4/mach-universal_c210.c b/arch/arm/mach-exynos4/mach-universal_c210.c
index 0e280d12301..b3b5d891100 100644
--- a/arch/arm/mach-exynos4/mach-universal_c210.c
+++ b/arch/arm/mach-exynos4/mach-universal_c210.c
@@ -79,7 +79,7 @@ static struct s3c2410_uartcfg universal_uartcfgs[] __initdata = {
 };
 
 static struct regulator_consumer_supply max8952_consumer =
-	REGULATOR_SUPPLY("vddarm", NULL);
+	REGULATOR_SUPPLY("vdd_arm", NULL);
 
 static struct max8952_platform_data universal_max8952_pdata __initdata = {
 	.gpio_vid0	= EXYNOS4_GPX0(3),
@@ -105,7 +105,7 @@ static struct max8952_platform_data universal_max8952_pdata __initdata = {
 };
 
 static struct regulator_consumer_supply lp3974_buck1_consumer =
-	REGULATOR_SUPPLY("vddint", NULL);
+	REGULATOR_SUPPLY("vdd_int", NULL);
 
 static struct regulator_consumer_supply lp3974_buck2_consumer =
 	REGULATOR_SUPPLY("vddg3d", NULL);
diff --git a/arch/arm/mach-exynos4/setup-usb-phy.c b/arch/arm/mach-exynos4/setup-usb-phy.c
index 0883c1b824b..39aca045f66 100644
--- a/arch/arm/mach-exynos4/setup-usb-phy.c
+++ b/arch/arm/mach-exynos4/setup-usb-phy.c
@@ -82,7 +82,7 @@ static int exynos4_usb_phy1_init(struct platform_device *pdev)
 
 	rstcon &= ~(HOST_LINK_PORT_SWRST_MASK | PHY1_SWRST_MASK);
 	writel(rstcon, EXYNOS4_RSTCON);
-	udelay(50);
+	udelay(80);
 
 	clk_disable(otg_clk);
 	clk_put(otg_clk);
diff --git a/arch/arm/mach-footbridge/Kconfig b/arch/arm/mach-footbridge/Kconfig
index dc26fff22cf..c8e7afcf14e 100644
--- a/arch/arm/mach-footbridge/Kconfig
+++ b/arch/arm/mach-footbridge/Kconfig
@@ -62,6 +62,7 @@ config ARCH_EBSA285_HOST
 config ARCH_NETWINDER
 	bool "NetWinder"
 	select CLKSRC_I8253
+	select CLKEVT_I8253
 	select FOOTBRIDGE_HOST
 	select ISA
 	select ISA_DMA
diff --git a/arch/arm/mach-footbridge/dc21285.c b/arch/arm/mach-footbridge/dc21285.c
index 1331fff51ae..18c32a5541d 100644
--- a/arch/arm/mach-footbridge/dc21285.c
+++ b/arch/arm/mach-footbridge/dc21285.c
@@ -18,6 +18,7 @@
 #include <linux/irq.h>
 #include <linux/io.h>
 #include <linux/spinlock.h>
+#include <video/vga.h>
 
 #include <asm/irq.h>
 #include <asm/system.h>
diff --git a/arch/arm/mach-imx/mach-cpuimx27.c b/arch/arm/mach-imx/mach-cpuimx27.c
index 87887ac5806..f851fe90368 100644
--- a/arch/arm/mach-imx/mach-cpuimx27.c
+++ b/arch/arm/mach-imx/mach-cpuimx27.c
@@ -310,7 +310,7 @@ static struct sys_timer eukrea_cpuimx27_timer = {
 	.init = eukrea_cpuimx27_timer_init,
 };
 
-MACHINE_START(CPUIMX27, "EUKREA CPUIMX27")
+MACHINE_START(EUKREA_CPUIMX27, "EUKREA CPUIMX27")
 	.boot_params = MX27_PHYS_OFFSET + 0x100,
 	.map_io = mx27_map_io,
 	.init_early = imx27_init_early,
diff --git a/arch/arm/mach-imx/mach-cpuimx35.c b/arch/arm/mach-imx/mach-cpuimx35.c
index f39a478ba1a..4bd083ba9af 100644
--- a/arch/arm/mach-imx/mach-cpuimx35.c
+++ b/arch/arm/mach-imx/mach-cpuimx35.c
@@ -192,7 +192,7 @@ struct sys_timer eukrea_cpuimx35_timer = {
 	.init	= eukrea_cpuimx35_timer_init,
 };
 
-MACHINE_START(EUKREA_CPUIMX35, "Eukrea CPUIMX35")
+MACHINE_START(EUKREA_CPUIMX35SD, "Eukrea CPUIMX35")
 	/* Maintainer: Eukrea Electromatique */
 	.boot_params = MX3x_PHYS_OFFSET + 0x100,
 	.map_io = mx35_map_io,
diff --git a/arch/arm/mach-imx/mach-eukrea_cpuimx25.c b/arch/arm/mach-imx/mach-eukrea_cpuimx25.c
index da36da52969..2442d5da883 100644
--- a/arch/arm/mach-imx/mach-eukrea_cpuimx25.c
+++ b/arch/arm/mach-imx/mach-eukrea_cpuimx25.c
@@ -161,7 +161,7 @@ static struct sys_timer eukrea_cpuimx25_timer = {
 	.init   = eukrea_cpuimx25_timer_init,
 };
 
-MACHINE_START(EUKREA_CPUIMX25, "Eukrea CPUIMX25")
+MACHINE_START(EUKREA_CPUIMX25SD, "Eukrea CPUIMX25")
 	/* Maintainer: Eukrea Electromatique */
 	.boot_params = MX25_PHYS_OFFSET + 0x100,
 	.map_io = mx25_map_io,
diff --git a/arch/arm/mach-integrator/integrator_ap.c b/arch/arm/mach-integrator/integrator_ap.c
index 2fbbdd5eac3..fcf0ae95651 100644
--- a/arch/arm/mach-integrator/integrator_ap.c
+++ b/arch/arm/mach-integrator/integrator_ap.c
@@ -337,15 +337,15 @@ static unsigned long timer_reload;
 static void integrator_clocksource_init(u32 khz)
 {
 	void __iomem *base = (void __iomem *)TIMER2_VA_BASE;
-	u32 ctrl = TIMER_CTRL_ENABLE;
+	u32 ctrl = TIMER_CTRL_ENABLE | TIMER_CTRL_PERIODIC;
 
 	if (khz >= 1500) {
 		khz /= 16;
-		ctrl = TIMER_CTRL_DIV16;
+		ctrl |= TIMER_CTRL_DIV16;
 	}
 
-	writel(ctrl, base + TIMER_CTRL);
 	writel(0xffff, base + TIMER_LOAD);
+	writel(ctrl, base + TIMER_CTRL);
 
 	clocksource_mmio_init(base + TIMER_VALUE, "timer2",
 		khz * 1000, 200, 16, clocksource_mmio_readl_down);
diff --git a/arch/arm/mach-omap2/clock3xxx_data.c b/arch/arm/mach-omap2/clock3xxx_data.c
index ffd55b1c439..b9b84468314 100644
--- a/arch/arm/mach-omap2/clock3xxx_data.c
+++ b/arch/arm/mach-omap2/clock3xxx_data.c
@@ -3078,6 +3078,7 @@ static struct clk gpt12_fck = {
 	.name		= "gpt12_fck",
 	.ops		= &clkops_null,
 	.parent		= &secure_32k_fck,
+	.clkdm_name	= "wkup_clkdm",
 	.recalc		= &followparent_recalc,
 };
 
@@ -3085,6 +3086,7 @@ static struct clk wdt1_fck = {
 	.name		= "wdt1_fck",
 	.ops		= &clkops_null,
 	.parent		= &secure_32k_fck,
+	.clkdm_name	= "wkup_clkdm",
 	.recalc		= &followparent_recalc,
 };
 
diff --git a/arch/arm/mach-omap2/clock44xx_data.c b/arch/arm/mach-omap2/clock44xx_data.c
index 2af0e3f00ce..c0b6fbda340 100644
--- a/arch/arm/mach-omap2/clock44xx_data.c
+++ b/arch/arm/mach-omap2/clock44xx_data.c
@@ -3376,10 +3376,18 @@ int __init omap4xxx_clk_init(void)
 	} else if (cpu_is_omap446x()) {
 		cpu_mask = RATE_IN_4460;
 		cpu_clkflg = CK_446X;
+	} else {
+		return 0;
 	}
 
 	clk_init(&omap2_clk_functions);
-	omap2_clk_disable_clkdm_control();
+
+	/*
+	 * Must stay commented until all OMAP SoC drivers are
+	 * converted to runtime PM, or drivers may start crashing
+	 *
+	 * omap2_clk_disable_clkdm_control();
+	 */
 
 	for (c = omap44xx_clks; c < omap44xx_clks + ARRAY_SIZE(omap44xx_clks);
 									  c++)
diff --git a/arch/arm/mach-omap2/clockdomain.c b/arch/arm/mach-omap2/clockdomain.c
index ab7db083f97..8f0890685d7 100644
--- a/arch/arm/mach-omap2/clockdomain.c
+++ b/arch/arm/mach-omap2/clockdomain.c
@@ -747,6 +747,7 @@ int clkdm_wakeup(struct clockdomain *clkdm)
 	spin_lock_irqsave(&clkdm->lock, flags);
 	clkdm->_flags &= ~_CLKDM_FLAG_HWSUP_ENABLED;
 	ret = arch_clkdm->clkdm_wakeup(clkdm);
+	ret |= pwrdm_state_switch(clkdm->pwrdm.ptr);
 	spin_unlock_irqrestore(&clkdm->lock, flags);
 	return ret;
 }
@@ -818,6 +819,7 @@ void clkdm_deny_idle(struct clockdomain *clkdm)
 	spin_lock_irqsave(&clkdm->lock, flags);
 	clkdm->_flags &= ~_CLKDM_FLAG_HWSUP_ENABLED;
 	arch_clkdm->clkdm_deny_idle(clkdm);
+	pwrdm_state_switch(clkdm->pwrdm.ptr);
 	spin_unlock_irqrestore(&clkdm->lock, flags);
 }
 
diff --git a/arch/arm/mach-omap2/omap_hwmod_2430_data.c b/arch/arm/mach-omap2/omap_hwmod_2430_data.c
index 16743c7d6e8..408193d8e04 100644
--- a/arch/arm/mach-omap2/omap_hwmod_2430_data.c
+++ b/arch/arm/mach-omap2/omap_hwmod_2430_data.c
@@ -192,6 +192,7 @@ static struct omap_hwmod_addr_space omap2430_usbhsotg_addrs[] = {
 		.pa_end		= OMAP243X_HS_BASE + SZ_4K - 1,
 		.flags		= ADDR_TYPE_RT
 	},
+	{ }
 };
 
 /*  l4_core ->usbhsotg  interface */
diff --git a/arch/arm/mach-omap2/pm.c b/arch/arm/mach-omap2/pm.c
index 3feb35911a3..472bf22d5e8 100644
--- a/arch/arm/mach-omap2/pm.c
+++ b/arch/arm/mach-omap2/pm.c
@@ -130,7 +130,6 @@ int omap_set_pwrdm_state(struct powerdomain *pwrdm, u32 state)
 		} else {
 			hwsup = clkdm_in_hwsup(pwrdm->pwrdm_clkdms[0]);
 			clkdm_wakeup(pwrdm->pwrdm_clkdms[0]);
-			pwrdm_wait_transition(pwrdm);
 			sleep_switch = FORCEWAKEUP_SWITCH;
 		}
 	}
@@ -156,7 +155,6 @@ int omap_set_pwrdm_state(struct powerdomain *pwrdm, u32 state)
 		return ret;
 	}
 
-	pwrdm_wait_transition(pwrdm);
 	pwrdm_state_switch(pwrdm);
 err:
 	return ret;
diff --git a/arch/arm/mach-omap2/powerdomain.c b/arch/arm/mach-omap2/powerdomain.c
index 9af08473bf1..ef71fdd40fc 100644
--- a/arch/arm/mach-omap2/powerdomain.c
+++ b/arch/arm/mach-omap2/powerdomain.c
@@ -195,28 +195,35 @@ static int _pwrdm_post_transition_cb(struct powerdomain *pwrdm, void *unused)
 
 /**
  * pwrdm_init - set up the powerdomain layer
- * @pwrdm_list: array of struct powerdomain pointers to register
+ * @pwrdms: array of struct powerdomain pointers to register
  * @custom_funcs: func pointers for arch specific implementations
  *
- * Loop through the array of powerdomains @pwrdm_list, registering all
- * that are available on the current CPU. If pwrdm_list is supplied
- * and not null, all of the referenced powerdomains will be
- * registered.  No return value.  XXX pwrdm_list is not really a
- * "list"; it is an array.  Rename appropriately.
+ * Loop through the array of powerdomains @pwrdms, registering all
+ * that are available on the current CPU.  Also, program all
+ * powerdomain target state as ON; this is to prevent domains from
+ * hitting low power states (if bootloader has target states set to
+ * something other than ON) and potentially even losing context while
+ * PM is not fully initialized.  The PM late init code can then program
+ * the desired target state for all the power domains.  No return
+ * value.
  */
-void pwrdm_init(struct powerdomain **pwrdm_list, struct pwrdm_ops *custom_funcs)
+void pwrdm_init(struct powerdomain **pwrdms, struct pwrdm_ops *custom_funcs)
 {
 	struct powerdomain **p = NULL;
+	struct powerdomain *temp_p;
 
 	if (!custom_funcs)
 		WARN(1, "powerdomain: No custom pwrdm functions registered\n");
 	else
 		arch_pwrdm = custom_funcs;
 
-	if (pwrdm_list) {
-		for (p = pwrdm_list; *p; p++)
+	if (pwrdms) {
+		for (p = pwrdms; *p; p++)
 			_pwrdm_register(*p);
 	}
+
+	list_for_each_entry(temp_p, &pwrdm_list, node)
+		pwrdm_set_next_pwrst(temp_p, PWRDM_POWER_ON);
 }
 
 /**
diff --git a/arch/arm/mach-orion5x/dns323-setup.c b/arch/arm/mach-orion5x/dns323-setup.c
index a6eddae82a0..c105556a0ee 100644
--- a/arch/arm/mach-orion5x/dns323-setup.c
+++ b/arch/arm/mach-orion5x/dns323-setup.c
@@ -77,7 +77,7 @@ static int __init dns323_pci_map_irq(const struct pci_dev *dev, u8 slot, u8 pin)
 	/*
 	 * Check for devices with hard-wired IRQs.
 	 */
-	irq = orion5x_pci_map_irq(const dev, slot, pin);
+	irq = orion5x_pci_map_irq(dev, slot, pin);
 	if (irq != -1)
 		return irq;
 
diff --git a/arch/arm/mach-orion5x/pci.c b/arch/arm/mach-orion5x/pci.c
index 28b8760ab9f..bc4a920e26e 100644
--- a/arch/arm/mach-orion5x/pci.c
+++ b/arch/arm/mach-orion5x/pci.c
@@ -14,6 +14,7 @@
 #include <linux/pci.h>
 #include <linux/slab.h>
 #include <linux/mbus.h>
+#include <video/vga.h>
 #include <asm/irq.h>
 #include <asm/mach/pci.h>
 #include <plat/pcie.h>
diff --git a/arch/arm/mach-prima2/clock.c b/arch/arm/mach-prima2/clock.c
index f9a2aaf63f7..615a4e75cea 100644
--- a/arch/arm/mach-prima2/clock.c
+++ b/arch/arm/mach-prima2/clock.c
@@ -481,6 +481,7 @@ static void __init sirfsoc_clk_init(void)
 
 static struct of_device_id clkc_ids[] = {
 	{ .compatible = "sirf,prima2-clkc" },
+	{},
 };
 
 void __init sirfsoc_of_clk_init(void)
diff --git a/arch/arm/mach-prima2/irq.c b/arch/arm/mach-prima2/irq.c
index c3404cbb6ff..7af254d046b 100644
--- a/arch/arm/mach-prima2/irq.c
+++ b/arch/arm/mach-prima2/irq.c
@@ -51,6 +51,7 @@ static __init void sirfsoc_irq_init(void)
 
 static struct of_device_id intc_ids[]  = {
 	{ .compatible = "sirf,prima2-intc" },
+	{},
 };
 
 void __init sirfsoc_of_irq_init(void)
diff --git a/arch/arm/mach-prima2/rstc.c b/arch/arm/mach-prima2/rstc.c
index d074786e83d..492cfa8d261 100644
--- a/arch/arm/mach-prima2/rstc.c
+++ b/arch/arm/mach-prima2/rstc.c
@@ -19,6 +19,7 @@ static DEFINE_MUTEX(rstc_lock);
 
 static struct of_device_id rstc_ids[]  = {
 	{ .compatible = "sirf,prima2-rstc" },
+	{},
 };
 
 static int __init sirfsoc_of_rstc_init(void)
diff --git a/arch/arm/mach-prima2/timer.c b/arch/arm/mach-prima2/timer.c
index 44027f34a88..ed7ec48d11d 100644
--- a/arch/arm/mach-prima2/timer.c
+++ b/arch/arm/mach-prima2/timer.c
@@ -190,6 +190,7 @@ static void __init sirfsoc_timer_init(void)
 
 static struct of_device_id timer_ids[] = {
 	{ .compatible = "sirf,prima2-tick" },
+	{},
 };
 
 static void __init sirfsoc_of_timer_map(void)
diff --git a/arch/arm/mach-realview/include/mach/system.h b/arch/arm/mach-realview/include/mach/system.h
index a30f2e3ec17..6657ff23116 100644
--- a/arch/arm/mach-realview/include/mach/system.h
+++ b/arch/arm/mach-realview/include/mach/system.h
@@ -44,6 +44,7 @@ static inline void arch_reset(char mode, const char *cmd)
 	 */
 	if (realview_reset)
 		realview_reset(mode);
+	dsb();
 }
 
 #endif
diff --git a/arch/arm/mach-s3c64xx/pm.c b/arch/arm/mach-s3c64xx/pm.c
index 8bad6437068..055e2858b0d 100644
--- a/arch/arm/mach-s3c64xx/pm.c
+++ b/arch/arm/mach-s3c64xx/pm.c
@@ -16,6 +16,7 @@
 #include <linux/suspend.h>
 #include <linux/serial_core.h>
 #include <linux/io.h>
+#include <linux/gpio.h>
 
 #include <mach/map.h>
 #include <mach/irqs.h>
diff --git a/arch/arm/mach-s5p64x0/irq-eint.c b/arch/arm/mach-s5p64x0/irq-eint.c
index 69ed4545112..fe7380f5c3c 100644
--- a/arch/arm/mach-s5p64x0/irq-eint.c
+++ b/arch/arm/mach-s5p64x0/irq-eint.c
@@ -129,7 +129,7 @@ static int s5p64x0_alloc_gc(void)
 	}
 
 	ct = gc->chip_types;
-	ct->chip.irq_ack = irq_gc_ack;
+	ct->chip.irq_ack = irq_gc_ack_set_bit;
 	ct->chip.irq_mask = irq_gc_mask_set_bit;
 	ct->chip.irq_unmask = irq_gc_mask_clr_bit;
 	ct->chip.irq_set_type = s5p64x0_irq_eint_set_type;
diff --git a/arch/arm/mach-s5pv210/pm.c b/arch/arm/mach-s5pv210/pm.c
index 309e388a8a8..f149d278377 100644
--- a/arch/arm/mach-s5pv210/pm.c
+++ b/arch/arm/mach-s5pv210/pm.c
@@ -88,7 +88,7 @@ static struct sleep_save s5pv210_core_save[] = {
 	SAVE_ITEM(S3C2410_TCNTO(0)),
 };
 
-void s5pv210_cpu_suspend(unsigned long arg)
+static int s5pv210_cpu_suspend(unsigned long arg)
 {
 	unsigned long tmp;
 
diff --git a/arch/arm/mach-shmobile/board-ag5evm.c b/arch/arm/mach-shmobile/board-ag5evm.c
index ce5c2513c6c..cdfdd624d21 100644
--- a/arch/arm/mach-shmobile/board-ag5evm.c
+++ b/arch/arm/mach-shmobile/board-ag5evm.c
@@ -341,6 +341,7 @@ static struct platform_device mipidsi0_device = {
 static struct sh_mobile_sdhi_info sdhi0_info = {
 	.dma_slave_tx	= SHDMA_SLAVE_SDHI0_TX,
 	.dma_slave_rx	= SHDMA_SLAVE_SDHI0_RX,
+	.tmio_flags	= TMIO_MMC_HAS_IDLE_WAIT,
 	.tmio_caps	= MMC_CAP_SD_HIGHSPEED,
 	.tmio_ocr_mask	= MMC_VDD_27_28 | MMC_VDD_28_29,
 };
@@ -382,7 +383,7 @@ void ag5evm_sdhi1_set_pwr(struct platform_device *pdev, int state)
 }
 
 static struct sh_mobile_sdhi_info sh_sdhi1_info = {
-	.tmio_flags	= TMIO_MMC_WRPROTECT_DISABLE,
+	.tmio_flags	= TMIO_MMC_WRPROTECT_DISABLE | TMIO_MMC_HAS_IDLE_WAIT,
 	.tmio_caps	= MMC_CAP_NONREMOVABLE | MMC_CAP_SDIO_IRQ,
 	.tmio_ocr_mask	= MMC_VDD_32_33 | MMC_VDD_33_34,
 	.set_pwr	= ag5evm_sdhi1_set_pwr,
diff --git a/arch/arm/mach-shmobile/board-mackerel.c b/arch/arm/mach-shmobile/board-mackerel.c
index 0ea71f8d4b8..17c19dc2560 100644
--- a/arch/arm/mach-shmobile/board-mackerel.c
+++ b/arch/arm/mach-shmobile/board-mackerel.c
@@ -641,6 +641,8 @@ static struct usbhs_private usbhs0_private = {
 		},
 		.driver_param = {
 			.buswait_bwait	= 4,
+			.d0_tx_id	= SHDMA_SLAVE_USB0_TX,
+			.d1_rx_id	= SHDMA_SLAVE_USB0_RX,
 		},
 	},
 };
@@ -810,6 +812,8 @@ static struct usbhs_private usbhs1_private = {
 			.buswait_bwait	= 4,
 			.pipe_type	= usbhs1_pipe_cfg,
 			.pipe_size	= ARRAY_SIZE(usbhs1_pipe_cfg),
+			.d0_tx_id	= SHDMA_SLAVE_USB1_TX,
+			.d1_rx_id	= SHDMA_SLAVE_USB1_RX,
 		},
 	},
 };
diff --git a/arch/arm/mach-shmobile/clock-sh7372.c b/arch/arm/mach-shmobile/clock-sh7372.c
index e6e11e4e2d4..66975921e64 100644
--- a/arch/arm/mach-shmobile/clock-sh7372.c
+++ b/arch/arm/mach-shmobile/clock-sh7372.c
@@ -503,16 +503,17 @@ static struct clk *late_main_clks[] = {
 	&sh7372_fsidivb_clk,
 };
 
-enum { MSTP001,
+enum { MSTP001, MSTP000,
        MSTP131, MSTP130,
        MSTP129, MSTP128, MSTP127, MSTP126, MSTP125,
        MSTP118, MSTP117, MSTP116, MSTP113,
        MSTP106, MSTP101, MSTP100,
        MSTP223,
-       MSTP218, MSTP217, MSTP216,
-       MSTP207, MSTP206, MSTP204, MSTP203, MSTP202, MSTP201, MSTP200,
-       MSTP329, MSTP328, MSTP323, MSTP322, MSTP314, MSTP313, MSTP312,
-       MSTP423, MSTP415, MSTP413, MSTP411, MSTP410, MSTP406, MSTP403,
+       MSTP218, MSTP217, MSTP216, MSTP214, MSTP208, MSTP207,
+       MSTP206, MSTP205, MSTP204, MSTP203, MSTP202, MSTP201, MSTP200,
+       MSTP328, MSTP323, MSTP322, MSTP314, MSTP313, MSTP312,
+       MSTP423, MSTP415, MSTP413, MSTP411, MSTP410, MSTP407, MSTP406,
+       MSTP405, MSTP404, MSTP403, MSTP400,
        MSTP_NR };
 
 #define MSTP(_parent, _reg, _bit, _flags) \
@@ -520,6 +521,7 @@ enum { MSTP001,
 
 static struct clk mstp_clks[MSTP_NR] = {
 	[MSTP001] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR0, 1, 0), /* IIC2 */
+	[MSTP000] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR0, 0, 0), /* MSIOF0 */
 	[MSTP131] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 31, 0), /* VEU3 */
 	[MSTP130] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 30, 0), /* VEU2 */
 	[MSTP129] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 29, 0), /* VEU1 */
@@ -538,14 +540,16 @@ static struct clk mstp_clks[MSTP_NR] = {
 	[MSTP218] = MSTP(&div4_clks[DIV4_HP], SMSTPCR2, 18, 0), /* DMAC1 */
 	[MSTP217] = MSTP(&div4_clks[DIV4_HP], SMSTPCR2, 17, 0), /* DMAC2 */
 	[MSTP216] = MSTP(&div4_clks[DIV4_HP], SMSTPCR2, 16, 0), /* DMAC3 */
+	[MSTP214] = MSTP(&div4_clks[DIV4_HP], SMSTPCR2, 14, 0), /* USBDMAC */
+	[MSTP208] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 8, 0), /* MSIOF1 */
 	[MSTP207] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 7, 0), /* SCIFA5 */
 	[MSTP206] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 6, 0), /* SCIFB */
+	[MSTP205] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 5, 0), /* MSIOF2 */
 	[MSTP204] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 4, 0), /* SCIFA0 */
 	[MSTP203] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 3, 0), /* SCIFA1 */
 	[MSTP202] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 2, 0), /* SCIFA2 */
 	[MSTP201] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 1, 0), /* SCIFA3 */
 	[MSTP200] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 0, 0), /* SCIFA4 */
-	[MSTP329] = MSTP(&r_clk, SMSTPCR3, 29, 0), /* CMT10 */
 	[MSTP328] = MSTP(&div6_clks[DIV6_SPU], SMSTPCR3, 28, 0), /* FSI2 */
 	[MSTP323] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR3, 23, 0), /* IIC1 */
 	[MSTP322] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR3, 22, 0), /* USB0 */
@@ -557,8 +561,12 @@ static struct clk mstp_clks[MSTP_NR] = {
 	[MSTP413] = MSTP(&pllc1_div2_clk, SMSTPCR4, 13, 0), /* HDMI */
 	[MSTP411] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR4, 11, 0), /* IIC3 */
 	[MSTP410] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR4, 10, 0), /* IIC4 */
+	[MSTP407] = MSTP(&div4_clks[DIV4_HP], SMSTPCR4, 7, 0), /* USB-DMAC1 */
 	[MSTP406] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR4, 6, 0), /* USB1 */
+	[MSTP405] = MSTP(&r_clk, SMSTPCR4, 5, 0), /* CMT4 */
+	[MSTP404] = MSTP(&r_clk, SMSTPCR4, 4, 0), /* CMT3 */
 	[MSTP403] = MSTP(&r_clk, SMSTPCR4, 3, 0), /* KEYSC */
+	[MSTP400] = MSTP(&r_clk, SMSTPCR4, 0, 0), /* CMT2 */
 };
 
 static struct clk_lookup lookups[] = {
@@ -609,6 +617,7 @@ static struct clk_lookup lookups[] = {
 
 	/* MSTP32 clocks */
 	CLKDEV_DEV_ID("i2c-sh_mobile.2", &mstp_clks[MSTP001]), /* IIC2 */
+	CLKDEV_DEV_ID("spi_sh_msiof.0", &mstp_clks[MSTP000]), /* MSIOF0 */
 	CLKDEV_DEV_ID("uio_pdrv_genirq.4", &mstp_clks[MSTP131]), /* VEU3 */
 	CLKDEV_DEV_ID("uio_pdrv_genirq.3", &mstp_clks[MSTP130]), /* VEU2 */
 	CLKDEV_DEV_ID("uio_pdrv_genirq.2", &mstp_clks[MSTP129]), /* VEU1 */
@@ -629,14 +638,16 @@ static struct clk_lookup lookups[] = {
 	CLKDEV_DEV_ID("sh-dma-engine.0", &mstp_clks[MSTP218]), /* DMAC1 */
 	CLKDEV_DEV_ID("sh-dma-engine.1", &mstp_clks[MSTP217]), /* DMAC2 */
 	CLKDEV_DEV_ID("sh-dma-engine.2", &mstp_clks[MSTP216]), /* DMAC3 */
+	CLKDEV_DEV_ID("sh-dma-engine.3", &mstp_clks[MSTP214]), /* USB-DMAC0 */
+	CLKDEV_DEV_ID("spi_sh_msiof.1", &mstp_clks[MSTP208]), /* MSIOF1 */
 	CLKDEV_DEV_ID("sh-sci.5", &mstp_clks[MSTP207]), /* SCIFA5 */
 	CLKDEV_DEV_ID("sh-sci.6", &mstp_clks[MSTP206]), /* SCIFB */
+	CLKDEV_DEV_ID("spi_sh_msiof.2", &mstp_clks[MSTP205]), /* MSIOF2 */
 	CLKDEV_DEV_ID("sh-sci.0", &mstp_clks[MSTP204]), /* SCIFA0 */
 	CLKDEV_DEV_ID("sh-sci.1", &mstp_clks[MSTP203]), /* SCIFA1 */
 	CLKDEV_DEV_ID("sh-sci.2", &mstp_clks[MSTP202]), /* SCIFA2 */
 	CLKDEV_DEV_ID("sh-sci.3", &mstp_clks[MSTP201]), /* SCIFA3 */
 	CLKDEV_DEV_ID("sh-sci.4", &mstp_clks[MSTP200]), /* SCIFA4 */
-	CLKDEV_DEV_ID("sh_cmt.10", &mstp_clks[MSTP329]), /* CMT10 */
 	CLKDEV_DEV_ID("sh_fsi2", &mstp_clks[MSTP328]), /* FSI2 */
 	CLKDEV_DEV_ID("i2c-sh_mobile.1", &mstp_clks[MSTP323]), /* IIC1 */
 	CLKDEV_DEV_ID("r8a66597_hcd.0", &mstp_clks[MSTP322]), /* USB0 */
@@ -650,10 +661,14 @@ static struct clk_lookup lookups[] = {
 	CLKDEV_DEV_ID("sh-mobile-hdmi", &mstp_clks[MSTP413]), /* HDMI */
 	CLKDEV_DEV_ID("i2c-sh_mobile.3", &mstp_clks[MSTP411]), /* IIC3 */
 	CLKDEV_DEV_ID("i2c-sh_mobile.4", &mstp_clks[MSTP410]), /* IIC4 */
+	CLKDEV_DEV_ID("sh-dma-engine.4", &mstp_clks[MSTP407]), /* USB-DMAC1 */
 	CLKDEV_DEV_ID("r8a66597_hcd.1", &mstp_clks[MSTP406]), /* USB1 */
 	CLKDEV_DEV_ID("r8a66597_udc.1", &mstp_clks[MSTP406]), /* USB1 */
 	CLKDEV_DEV_ID("renesas_usbhs.1", &mstp_clks[MSTP406]), /* USB1 */
+	CLKDEV_DEV_ID("sh_cmt.4", &mstp_clks[MSTP405]), /* CMT4 */
+	CLKDEV_DEV_ID("sh_cmt.3", &mstp_clks[MSTP404]), /* CMT3 */
 	CLKDEV_DEV_ID("sh_keysc.0", &mstp_clks[MSTP403]), /* KEYSC */
+	CLKDEV_DEV_ID("sh_cmt.2", &mstp_clks[MSTP400]), /* CMT2 */
 
 	CLKDEV_ICK_ID("hdmi", "sh_mobile_lcdc_fb.1",
 		      &div6_reparent_clks[DIV6_HDMI]),
diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
index 6db2ccabc2b..61a846bb30f 100644
--- a/arch/arm/mach-shmobile/clock-sh73a0.c
+++ b/arch/arm/mach-shmobile/clock-sh73a0.c
@@ -365,7 +365,7 @@ void __init sh73a0_clock_init(void)
 	__raw_writel(0x108, SD2CKCR);
 
 	/* detect main clock parent */
-	switch ((__raw_readl(CKSCR) >> 24) & 0x03) {
+	switch ((__raw_readl(CKSCR) >> 28) & 0x03) {
 	case 0:
 		main_clk.parent = &sh73a0_extal1_clk;
 		break;
diff --git a/arch/arm/mach-shmobile/include/mach/sh7372.h b/arch/arm/mach-shmobile/include/mach/sh7372.h
index ce595cee86c..24e63a85e66 100644
--- a/arch/arm/mach-shmobile/include/mach/sh7372.h
+++ b/arch/arm/mach-shmobile/include/mach/sh7372.h
@@ -459,6 +459,10 @@ enum {
 	SHDMA_SLAVE_SDHI2_TX,
 	SHDMA_SLAVE_MMCIF_RX,
 	SHDMA_SLAVE_MMCIF_TX,
+	SHDMA_SLAVE_USB0_TX,
+	SHDMA_SLAVE_USB0_RX,
+	SHDMA_SLAVE_USB1_TX,
+	SHDMA_SLAVE_USB1_RX,
 };
 
 extern struct clk sh7372_extal1_clk;
diff --git a/arch/arm/mach-shmobile/intc-sh7372.c b/arch/arm/mach-shmobile/intc-sh7372.c
index 3b28743c77e..739315e30eb 100644
--- a/arch/arm/mach-shmobile/intc-sh7372.c
+++ b/arch/arm/mach-shmobile/intc-sh7372.c
@@ -379,7 +379,7 @@ enum {
 	/* BBIF2 */
 	VPU,
 	TSIF1,
-	_3DG_SGX530,
+	/* 3DG */
 	_2DDMAC,
 	IIC2_ALI2, IIC2_TACKI2, IIC2_WAITI2, IIC2_DTEI2,
 	IPMMU_IPMMUR, IPMMU_IPMMUR2,
@@ -436,7 +436,7 @@ static struct intc_vect intcs_vectors[] = {
 	/* BBIF2 */
 	INTCS_VECT(VPU, 0x980),
 	INTCS_VECT(TSIF1, 0x9a0),
-	INTCS_VECT(_3DG_SGX530, 0x9e0),
+	/* 3DG */
 	INTCS_VECT(_2DDMAC, 0xa00),
 	INTCS_VECT(IIC2_ALI2, 0xa80), INTCS_VECT(IIC2_TACKI2, 0xaa0),
 	INTCS_VECT(IIC2_WAITI2, 0xac0), INTCS_VECT(IIC2_DTEI2, 0xae0),
@@ -521,7 +521,7 @@ static struct intc_mask_reg intcs_mask_registers[] = {
 	    RTDMAC_1_DEI3, RTDMAC_1_DEI2, RTDMAC_1_DEI1, RTDMAC_1_DEI0 } },
 	{ 0xffd20198, 0xffd201d8, 8, /* IMR6SA / IMCR6SA */
 	  { 0, 0, MSIOF, 0,
-	    _3DG_SGX530, 0, 0, 0 } },
+	    0, 0, 0, 0 } },
 	{ 0xffd2019c, 0xffd201dc, 8, /* IMR7SA / IMCR7SA */
 	  { 0, TMU_TUNI2, TMU_TUNI1, TMU_TUNI0,
 	    0, 0, 0, 0 } },
@@ -561,7 +561,6 @@ static struct intc_prio_reg intcs_prio_registers[] = {
 					      TMU_TUNI2, TSIF1 } },
 	{ 0xffd2001c, 0, 16, 4, /* IPRHS */ { 0, 0, VEU, BEU } },
 	{ 0xffd20020, 0, 16, 4, /* IPRIS */ { 0, MSIOF, TSIF0, IIC0 } },
-	{ 0xffd20024, 0, 16, 4, /* IPRJS */ { 0, _3DG_SGX530, 0, 0 } },
 	{ 0xffd20028, 0, 16, 4, /* IPRKS */ { 0, 0, LMB, 0 } },
 	{ 0xffd2002c, 0, 16, 4, /* IPRLS */ { IPMMU, 0, 0, 0 } },
 	{ 0xffd20030, 0, 16, 4, /* IPRMS */ { IIC2, 0, 0, 0 } },
diff --git a/arch/arm/mach-shmobile/setup-sh7372.c b/arch/arm/mach-shmobile/setup-sh7372.c
index 79f0413d872..2d9b1b1a253 100644
--- a/arch/arm/mach-shmobile/setup-sh7372.c
+++ b/arch/arm/mach-shmobile/setup-sh7372.c
@@ -169,35 +169,35 @@ static struct platform_device scif6_device = {
 };
 
 /* CMT */
-static struct sh_timer_config cmt10_platform_data = {
-	.name = "CMT10",
-	.channel_offset = 0x10,
-	.timer_bit = 0,
+static struct sh_timer_config cmt2_platform_data = {
+	.name = "CMT2",
+	.channel_offset = 0x40,
+	.timer_bit = 5,
 	.clockevent_rating = 125,
 	.clocksource_rating = 125,
 };
 
-static struct resource cmt10_resources[] = {
+static struct resource cmt2_resources[] = {
 	[0] = {
-		.name	= "CMT10",
-		.start	= 0xe6138010,
-		.end	= 0xe613801b,
+		.name	= "CMT2",
+		.start	= 0xe6130040,
+		.end	= 0xe613004b,
 		.flags	= IORESOURCE_MEM,
 	},
 	[1] = {
-		.start	= evt2irq(0x0b00), /* CMT1_CMT10 */
+		.start	= evt2irq(0x0b80), /* CMT2 */
 		.flags	= IORESOURCE_IRQ,
 	},
 };
 
-static struct platform_device cmt10_device = {
+static struct platform_device cmt2_device = {
 	.name		= "sh_cmt",
-	.id		= 10,
+	.id		= 2,
 	.dev = {
-		.platform_data	= &cmt10_platform_data,
+		.platform_data	= &cmt2_platform_data,
 	},
-	.resource	= cmt10_resources,
-	.num_resources	= ARRAY_SIZE(cmt10_resources),
+	.resource	= cmt2_resources,
+	.num_resources	= ARRAY_SIZE(cmt2_resources),
 };
 
 /* TMU */
@@ -602,6 +602,150 @@ static struct platform_device dma2_device = {
 	},
 };
 
+/*
+ * USB-DMAC
+ */
+
+unsigned int usbts_shift[] = {3, 4, 5};
+
+enum {
+	XMIT_SZ_8BYTE		= 0,
+	XMIT_SZ_16BYTE		= 1,
+	XMIT_SZ_32BYTE		= 2,
+};
+
+#define USBTS_INDEX2VAL(i) (((i) & 3) << 6)
+
+static const struct sh_dmae_channel sh7372_usb_dmae_channels[] = {
+	{
+		.offset = 0,
+	}, {
+		.offset = 0x20,
+	},
+};
+
+/* USB DMAC0 */
+static const struct sh_dmae_slave_config sh7372_usb_dmae0_slaves[] = {
+	{
+		.slave_id	= SHDMA_SLAVE_USB0_TX,
+		.chcr		= USBTS_INDEX2VAL(XMIT_SZ_8BYTE),
+	}, {
+		.slave_id	= SHDMA_SLAVE_USB0_RX,
+		.chcr		= USBTS_INDEX2VAL(XMIT_SZ_8BYTE),
+	},
+};
+
+static struct sh_dmae_pdata usb_dma0_platform_data = {
+	.slave		= sh7372_usb_dmae0_slaves,
+	.slave_num	= ARRAY_SIZE(sh7372_usb_dmae0_slaves),
+	.channel	= sh7372_usb_dmae_channels,
+	.channel_num	= ARRAY_SIZE(sh7372_usb_dmae_channels),
+	.ts_low_shift	= 6,
+	.ts_low_mask	= 0xc0,
+	.ts_high_shift	= 0,
+	.ts_high_mask	= 0,
+	.ts_shift	= usbts_shift,
+	.ts_shift_num	= ARRAY_SIZE(usbts_shift),
+	.dmaor_init	= DMAOR_DME,
+	.chcr_offset	= 0x14,
+	.chcr_ie_bit	= 1 << 5,
+	.dmaor_is_32bit	= 1,
+	.needs_tend_set	= 1,
+	.no_dmars	= 1,
+};
+
+static struct resource sh7372_usb_dmae0_resources[] = {
+	{
+		/* Channel registers and DMAOR */
+		.start	= 0xe68a0020,
+		.end	= 0xe68a0064 - 1,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		/* VCR/SWR/DMICR */
+		.start	= 0xe68a0000,
+		.end	= 0xe68a0014 - 1,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		/* IRQ for channels */
+		.start	= evt2irq(0x0a00),
+		.end	= evt2irq(0x0a00),
+		.flags	= IORESOURCE_IRQ,
+	},
+};
+
+static struct platform_device usb_dma0_device = {
+	.name		= "sh-dma-engine",
+	.id		= 3,
+	.resource	= sh7372_usb_dmae0_resources,
+	.num_resources	= ARRAY_SIZE(sh7372_usb_dmae0_resources),
+	.dev		= {
+		.platform_data	= &usb_dma0_platform_data,
+	},
+};
+
+/* USB DMAC1 */
+static const struct sh_dmae_slave_config sh7372_usb_dmae1_slaves[] = {
+	{
+		.slave_id	= SHDMA_SLAVE_USB1_TX,
+		.chcr		= USBTS_INDEX2VAL(XMIT_SZ_8BYTE),
+	}, {
+		.slave_id	= SHDMA_SLAVE_USB1_RX,
+		.chcr		= USBTS_INDEX2VAL(XMIT_SZ_8BYTE),
+	},
+};
+
+static struct sh_dmae_pdata usb_dma1_platform_data = {
+	.slave		= sh7372_usb_dmae1_slaves,
+	.slave_num	= ARRAY_SIZE(sh7372_usb_dmae1_slaves),
+	.channel	= sh7372_usb_dmae_channels,
+	.channel_num	= ARRAY_SIZE(sh7372_usb_dmae_channels),
+	.ts_low_shift	= 6,
+	.ts_low_mask	= 0xc0,
+	.ts_high_shift	= 0,
+	.ts_high_mask	= 0,
+	.ts_shift	= usbts_shift,
+	.ts_shift_num	= ARRAY_SIZE(usbts_shift),
+	.dmaor_init	= DMAOR_DME,
+	.chcr_offset	= 0x14,
+	.chcr_ie_bit	= 1 << 5,
+	.dmaor_is_32bit	= 1,
+	.needs_tend_set	= 1,
+	.no_dmars	= 1,
+};
+
+static struct resource sh7372_usb_dmae1_resources[] = {
+	{
+		/* Channel registers and DMAOR */
+		.start	= 0xe68c0020,
+		.end	= 0xe68c0064 - 1,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		/* VCR/SWR/DMICR */
+		.start	= 0xe68c0000,
+		.end	= 0xe68c0014 - 1,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		/* IRQ for channels */
+		.start	= evt2irq(0x1d00),
+		.end	= evt2irq(0x1d00),
+		.flags	= IORESOURCE_IRQ,
+	},
+};
+
+static struct platform_device usb_dma1_device = {
+	.name		= "sh-dma-engine",
+	.id		= 4,
+	.resource	= sh7372_usb_dmae1_resources,
+	.num_resources	= ARRAY_SIZE(sh7372_usb_dmae1_resources),
+	.dev		= {
+		.platform_data	= &usb_dma1_platform_data,
+	},
+};
+
 /* VPU */
 static struct uio_info vpu_platform_data = {
 	.name = "VPU5HG",
@@ -818,7 +962,7 @@ static struct platform_device *sh7372_early_devices[] __initdata = {
 	&scif4_device,
 	&scif5_device,
 	&scif6_device,
-	&cmt10_device,
+	&cmt2_device,
 	&tmu00_device,
 	&tmu01_device,
 };
@@ -829,6 +973,8 @@ static struct platform_device *sh7372_late_devices[] __initdata = {
 	&dma0_device,
 	&dma1_device,
 	&dma2_device,
+	&usb_dma0_device,
+	&usb_dma1_device,
 	&vpu_device,
 	&veu0_device,
 	&veu1_device,
diff --git a/arch/arm/mach-vexpress/v2m.c b/arch/arm/mach-vexpress/v2m.c
index 9e6b93b1a04..d0d267a8d3f 100644
--- a/arch/arm/mach-vexpress/v2m.c
+++ b/arch/arm/mach-vexpress/v2m.c
@@ -318,6 +318,10 @@ static struct clk v2m_sp804_clk = {
 	.rate	= 1000000,
 };
 
+static struct clk v2m_ref_clk = {
+	.rate   = 32768,
+};
+
 static struct clk dummy_apb_pclk;
 
 static struct clk_lookup v2m_lookups[] = {
@@ -348,6 +352,9 @@ static struct clk_lookup v2m_lookups[] = {
 	}, {	/* CLCD */
 		.dev_id		= "mb:clcd",
 		.clk		= &osc1_clk,
+	}, {	/* SP805 WDT */
+		.dev_id		= "mb:wdt",
+		.clk		= &v2m_ref_clk,
 	}, {	/* SP804 timers */
 		.dev_id		= "sp804",
 		.con_id		= "v2m-timer0",
diff --git a/arch/arm/mm/abort-macro.S b/arch/arm/mm/abort-macro.S
index 52162d59407..2cbf68ef0e8 100644
--- a/arch/arm/mm/abort-macro.S
+++ b/arch/arm/mm/abort-macro.S
@@ -17,7 +17,7 @@
 	cmp	\tmp, # 0x5600			@ Is it ldrsb?
 	orreq	\tmp, \tmp, #1 << 11		@ Set L-bit if yes
 	tst	\tmp, #1 << 11			@ L = 0 -> write
-	orreq	\psr, \psr, #1 << 11		@ yes.
+	orreq	\fsr, \fsr, #1 << 11		@ yes.
 	b	do_DataAbort
 not_thumb:
 	.endm
diff --git a/arch/arm/mm/cache-l2x0.c b/arch/arm/mm/cache-l2x0.c
index 44c086710d2..9ecfdb51195 100644
--- a/arch/arm/mm/cache-l2x0.c
+++ b/arch/arm/mm/cache-l2x0.c
@@ -277,6 +277,25 @@ static void l2x0_disable(void)
 	spin_unlock_irqrestore(&l2x0_lock, flags);
 }
 
+static void __init l2x0_unlock(__u32 cache_id)
+{
+	int lockregs;
+	int i;
+
+	if (cache_id == L2X0_CACHE_ID_PART_L310)
+		lockregs = 8;
+	else
+		/* L210 and unknown types */
+		lockregs = 1;
+
+	for (i = 0; i < lockregs; i++) {
+		writel_relaxed(0x0, l2x0_base + L2X0_LOCKDOWN_WAY_D_BASE +
+			       i * L2X0_LOCKDOWN_STRIDE);
+		writel_relaxed(0x0, l2x0_base + L2X0_LOCKDOWN_WAY_I_BASE +
+			       i * L2X0_LOCKDOWN_STRIDE);
+	}
+}
+
 void __init l2x0_init(void __iomem *base, __u32 aux_val, __u32 aux_mask)
 {
 	__u32 aux;
@@ -328,6 +347,8 @@ void __init l2x0_init(void __iomem *base, __u32 aux_val, __u32 aux_mask)
 	 * accessing the below registers will fault.
 	 */
 	if (!(readl_relaxed(l2x0_base + L2X0_CTRL) & 1)) {
+		/* Make sure that I&D is not locked down when starting */
+		l2x0_unlock(cache_id);
 
 		/* l2x0 controller is disabled */
 		writel_relaxed(aux, l2x0_base + L2X0_AUX_CTRL);
diff --git a/arch/arm/mm/init.c b/arch/arm/mm/init.c
index 91bca355cd3..cc7e2d8be9a 100644
--- a/arch/arm/mm/init.c
+++ b/arch/arm/mm/init.c
@@ -298,7 +298,7 @@ static void __init arm_bootmem_free(unsigned long min, unsigned long max_low,
 #ifdef CONFIG_HAVE_ARCH_PFN_VALID
 int pfn_valid(unsigned long pfn)
 {
-	return memblock_is_memory(pfn << PAGE_SHIFT);
+	return memblock_is_memory(__pfn_to_phys(pfn));
 }
 EXPORT_SYMBOL(pfn_valid);
 #endif
diff --git a/arch/arm/mm/proc-arm920.S b/arch/arm/mm/proc-arm920.S
index 92bd102e398..2e6849b41f6 100644
--- a/arch/arm/mm/proc-arm920.S
+++ b/arch/arm/mm/proc-arm920.S
@@ -379,7 +379,7 @@ ENTRY(cpu_arm920_set_pte_ext)
 
 /* Suspend/resume support: taken from arch/arm/plat-s3c24xx/sleep.S */
 .globl	cpu_arm920_suspend_size
-.equ	cpu_arm920_suspend_size, 4 * 3
+.equ	cpu_arm920_suspend_size, 4 * 4
 #ifdef CONFIG_PM_SLEEP
 ENTRY(cpu_arm920_do_suspend)
 	stmfd	sp!, {r4 - r7, lr}
diff --git a/arch/arm/mm/proc-arm926.S b/arch/arm/mm/proc-arm926.S
index 2bbcf053dff..cd8f79c3a28 100644
--- a/arch/arm/mm/proc-arm926.S
+++ b/arch/arm/mm/proc-arm926.S
@@ -394,7 +394,7 @@ ENTRY(cpu_arm926_set_pte_ext)
 
 /* Suspend/resume support: taken from arch/arm/plat-s3c24xx/sleep.S */
 .globl	cpu_arm926_suspend_size
-.equ	cpu_arm926_suspend_size, 4 * 3
+.equ	cpu_arm926_suspend_size, 4 * 4
 #ifdef CONFIG_PM_SLEEP
 ENTRY(cpu_arm926_do_suspend)
 	stmfd	sp!, {r4 - r7, lr}
diff --git a/arch/arm/mm/proc-sa1100.S b/arch/arm/mm/proc-sa1100.S
index 07219c2ae11..69e7f2ef738 100644
--- a/arch/arm/mm/proc-sa1100.S
+++ b/arch/arm/mm/proc-sa1100.S
@@ -182,11 +182,11 @@ ENDPROC(cpu_sa1100_do_suspend)
 
 ENTRY(cpu_sa1100_do_resume)
 	ldmia	r0, {r4 - r7}			@ load cp regs
-	mov	r1, #0
-	mcr	p15, 0, r1, c8, c7, 0		@ flush I+D TLBs
-	mcr	p15, 0, r1, c7, c7, 0		@ flush I&D cache
-	mcr	p15, 0, r1, c9, c0, 0		@ invalidate RB
-	mcr	p15, 0, r1, c9, c0, 5		@ allow user space to use RB
+	mov	ip, #0
+	mcr	p15, 0, ip, c8, c7, 0		@ flush I+D TLBs
+	mcr	p15, 0, ip, c7, c7, 0		@ flush I&D cache
+	mcr	p15, 0, ip, c9, c0, 0		@ invalidate RB
+	mcr	p15, 0, ip, c9, c0, 5		@ allow user space to use RB
 
 	mcr	p15, 0, r4, c3, c0, 0		@ domain ID
 	mcr	p15, 0, r5, c2, c0, 0		@ translation table base addr
diff --git a/arch/arm/mm/proc-v6.S b/arch/arm/mm/proc-v6.S
index 219138d2f15..a923aa0fd00 100644
--- a/arch/arm/mm/proc-v6.S
+++ b/arch/arm/mm/proc-v6.S
@@ -223,6 +223,22 @@ __v6_setup:
 	mrc	p15, 0, r0, c1, c0, 0		@ read control register
 	bic	r0, r0, r5			@ clear bits them
 	orr	r0, r0, r6			@ set them
+#ifdef CONFIG_ARM_ERRATA_364296
+	/*
+	 * Workaround for the 364296 ARM1136 r0p2 erratum (possible cache data
+	 * corruption with hit-under-miss enabled). The conditional code below
+	 * (setting the undocumented bit 31 in the auxiliary control register
+	 * and the FI bit in the control register) disables hit-under-miss
+	 * without putting the processor into full low interrupt latency mode.
+	 */
+	ldr	r6, =0x4107b362			@ id for ARM1136 r0p2
+	mrc	p15, 0, r5, c0, c0, 0		@ get processor id
+	teq	r5, r6				@ check for the faulty core
+	mrceq	p15, 0, r5, c1, c0, 1		@ load aux control reg
+	orreq	r5, r5, #(1 << 31)		@ set the undocumented bit 31
+	mcreq	p15, 0, r5, c1, c0, 1		@ write aux control reg
+	orreq	r0, r0, #(1 << 21)		@ low interrupt latency configuration
+#endif
 	mov	pc, lr				@ return to head.S:__ret
 
 	/*
diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S
index a30e78542cc..9049c0764db 100644
--- a/arch/arm/mm/proc-v7.S
+++ b/arch/arm/mm/proc-v7.S
@@ -66,6 +66,7 @@ ENDPROC(cpu_v7_proc_fin)
 ENTRY(cpu_v7_reset)
 	mrc	p15, 0, r1, c1, c0, 0		@ ctrl register
 	bic	r1, r1, #0x1			@ ...............m
+ THUMB(	bic	r1, r1, #1 << 30 )		@ SCTLR.TE (Thumb exceptions)
 	mcr	p15, 0, r1, c1, c0, 0		@ disable MMU
 	isb
 	mov	pc, r0
@@ -247,13 +248,16 @@ ENTRY(cpu_v7_do_resume)
 	mcr	p15, 0, r7, c2, c0, 0	@ TTB 0
 	mcr	p15, 0, r8, c2, c0, 1	@ TTB 1
 	mcr	p15, 0, ip, c2, c0, 2	@ TTB control register
-	mcr	p15, 0, r10, c1, c0, 1	@ Auxiliary control register
+	mrc	p15, 0, r4, c1, c0, 1	@ Read Auxiliary control register
+	teq	r4, r10			@ Is it already set?
+	mcrne	p15, 0, r10, c1, c0, 1	@ No, so write it
 	mcr	p15, 0, r11, c1, c0, 2	@ Co-processor access control
 	ldr	r4, =PRRR		@ PRRR
 	ldr	r5, =NMRR		@ NMRR
 	mcr	p15, 0, r4, c10, c2, 0	@ write PRRR
 	mcr	p15, 0, r5, c10, c2, 1	@ write NMRR
 	isb
+	dsb
 	mov	r0, r9			@ control register
 	mov	r2, r7, lsr #14		@ get TTB0 base
 	mov	r2, r2, lsl #14
diff --git a/arch/arm/mm/proc-xsc3.S b/arch/arm/mm/proc-xsc3.S
index 28c72a2006a..755e1bf2268 100644
--- a/arch/arm/mm/proc-xsc3.S
+++ b/arch/arm/mm/proc-xsc3.S
@@ -406,7 +406,7 @@ ENTRY(cpu_xsc3_set_pte_ext)
 	.align
 
 .globl	cpu_xsc3_suspend_size
-.equ	cpu_xsc3_suspend_size, 4 * 8
+.equ	cpu_xsc3_suspend_size, 4 * 7
 #ifdef CONFIG_PM_SLEEP
 ENTRY(cpu_xsc3_do_suspend)
 	stmfd	sp!, {r4 - r10, lr}
@@ -418,12 +418,12 @@ ENTRY(cpu_xsc3_do_suspend)
 	mrc	p15, 0, r9, c1, c0, 1	@ auxiliary control reg
 	mrc 	p15, 0, r10, c1, c0, 0	@ control reg
 	bic	r4, r4, #2		@ clear frequency change bit
-	stmia	r0, {r1, r4 - r10}	@ store v:p offset + cp regs
+	stmia	r0, {r4 - r10}		@ store cp regs
 	ldmia	sp!, {r4 - r10, pc}
 ENDPROC(cpu_xsc3_do_suspend)
 
 ENTRY(cpu_xsc3_do_resume)
-	ldmia	r0, {r1, r4 - r10}	@ load v:p offset + cp regs
+	ldmia	r0, {r4 - r10}		@ load cp regs
 	mov	ip, #0
 	mcr	p15, 0, ip, c7, c7, 0	@ invalidate I & D caches, BTB
 	mcr	p15, 0, ip, c7, c10, 4	@ drain write (&fill) buffer
diff --git a/arch/arm/plat-omap/omap_device.c b/arch/arm/plat-omap/omap_device.c
index 9a6a5385491..02609eee056 100644
--- a/arch/arm/plat-omap/omap_device.c
+++ b/arch/arm/plat-omap/omap_device.c
@@ -615,6 +615,9 @@ static int _od_resume_noirq(struct device *dev)
 
 	return pm_generic_resume_noirq(dev);
 }
+#else
+#define _od_suspend_noirq NULL
+#define _od_resume_noirq NULL
 #endif
 
 static struct dev_pm_domain omap_device_pm_domain = {
diff --git a/arch/arm/plat-s5p/clock.c b/arch/arm/plat-s5p/clock.c
index 02af235298e..5f84a3f13ef 100644
--- a/arch/arm/plat-s5p/clock.c
+++ b/arch/arm/plat-s5p/clock.c
@@ -192,7 +192,7 @@ unsigned long s5p_spdif_get_rate(struct clk *clk)
 	if (IS_ERR(pclk))
 		return -EINVAL;
 
-	rate = pclk->ops->get_rate(clk);
+	rate = pclk->ops->get_rate(pclk);
 	clk_put(pclk);
 
 	return rate;
diff --git a/arch/arm/plat-s5p/irq-gpioint.c b/arch/arm/plat-s5p/irq-gpioint.c
index 327ab9f662e..f71078ef6bb 100644
--- a/arch/arm/plat-s5p/irq-gpioint.c
+++ b/arch/arm/plat-s5p/irq-gpioint.c
@@ -23,6 +23,8 @@
 #include <plat/gpio-core.h>
 #include <plat/gpio-cfg.h>
 
+#include <asm/mach/irq.h>
+
 #define GPIO_BASE(chip)		(((unsigned long)(chip)->base) & 0xFFFFF000u)
 
 #define CON_OFFSET		0x700
@@ -81,6 +83,9 @@ static void s5p_gpioint_handler(unsigned int irq, struct irq_desc *desc)
 	int group, pend_offset, mask_offset;
 	unsigned int pend, mask;
 
+	struct irq_chip *chip = irq_get_chip(irq);
+	chained_irq_enter(chip, desc);
+
 	for (group = 0; group < bank->nr_groups; group++) {
 		struct s3c_gpio_chip *chip = bank->chips[group];
 		if (!chip)
@@ -102,6 +107,7 @@ static void s5p_gpioint_handler(unsigned int irq, struct irq_desc *desc)
 			pend &= ~BIT(offset);
 		}
 	}
+	chained_irq_exit(chip, desc);
 }
 
 static __init int s5p_gpioint_add(struct s3c_gpio_chip *chip)
diff --git a/arch/arm/plat-samsung/include/plat/backlight.h b/arch/arm/plat-samsung/include/plat/backlight.h
index 51d8da846a6..ad530c78fe8 100644
--- a/arch/arm/plat-samsung/include/plat/backlight.h
+++ b/arch/arm/plat-samsung/include/plat/backlight.h
@@ -20,7 +20,7 @@ struct samsung_bl_gpio_info {
 	int func;
 };
 
-extern void samsung_bl_set(struct samsung_bl_gpio_info *gpio_info,
+extern void __init samsung_bl_set(struct samsung_bl_gpio_info *gpio_info,
 	struct platform_pwm_backlight_data *bl_data);
 
 #endif /* __ASM_PLAT_BACKLIGHT_H */
diff --git a/arch/arm/plat-samsung/irq-vic-timer.c b/arch/arm/plat-samsung/irq-vic-timer.c
index f714d060370..51583cd3016 100644
--- a/arch/arm/plat-samsung/irq-vic-timer.c
+++ b/arch/arm/plat-samsung/irq-vic-timer.c
@@ -22,9 +22,14 @@
 #include <plat/irq-vic-timer.h>
 #include <plat/regs-timer.h>
 
+#include <asm/mach/irq.h>
+
 static void s3c_irq_demux_vic_timer(unsigned int irq, struct irq_desc *desc)
 {
+	struct irq_chip *chip = irq_get_chip(irq);
+	chained_irq_enter(chip, desc);
 	generic_handle_irq((int)desc->irq_data.handler_data);
+	chained_irq_exit(chip, desc);
 }
 
 /* We assume the IRQ_TIMER0..IRQ_TIMER4 range is continuous. */
diff --git a/arch/arm/tools/mach-types b/arch/arm/tools/mach-types
index fff68d0d521..62cc8f98117 100644
--- a/arch/arm/tools/mach-types
+++ b/arch/arm/tools/mach-types
@@ -351,7 +351,7 @@ centro			MACH_CENTRO		CENTRO			1944
 nokia_rx51		MACH_NOKIA_RX51		NOKIA_RX51		1955
 omap_zoom2		MACH_OMAP_ZOOM2		OMAP_ZOOM2		1967
 cpuat9260		MACH_CPUAT9260		CPUAT9260		1973
-eukrea_cpuimx27		MACH_CPUIMX27		CPUIMX27		1975
+eukrea_cpuimx27		MACH_EUKREA_CPUIMX27	EUKREA_CPUIMX27		1975
 acs5k			MACH_ACS5K		ACS5K			1982
 snapper_9260		MACH_SNAPPER_9260	SNAPPER_9260		1987
 dsm320			MACH_DSM320		DSM320			1988
@@ -476,8 +476,8 @@ cns3420vb		MACH_CNS3420VB		CNS3420VB		2776
 omap4_panda		MACH_OMAP4_PANDA	OMAP4_PANDA		2791
 ti8168evm		MACH_TI8168EVM		TI8168EVM		2800
 teton_bga		MACH_TETON_BGA		TETON_BGA		2816
-eukrea_cpuimx25sd	MACH_EUKREA_CPUIMX25	EUKREA_CPUIMX25		2820
-eukrea_cpuimx35sd	MACH_EUKREA_CPUIMX35	EUKREA_CPUIMX35		2821
+eukrea_cpuimx25sd	MACH_EUKREA_CPUIMX25SD	EUKREA_CPUIMX25SD	2820
+eukrea_cpuimx35sd	MACH_EUKREA_CPUIMX35SD	EUKREA_CPUIMX35SD	2821
 eukrea_cpuimx51sd	MACH_EUKREA_CPUIMX51SD	EUKREA_CPUIMX51SD	2822
 eukrea_cpuimx51		MACH_EUKREA_CPUIMX51	EUKREA_CPUIMX51		2823
 smdkc210		MACH_SMDKC210		SMDKC210		2838
diff --git a/arch/openrisc/include/asm/dma-mapping.h b/arch/openrisc/include/asm/dma-mapping.h
index 052f877b52a..60b47223390 100644
--- a/arch/openrisc/include/asm/dma-mapping.h
+++ b/arch/openrisc/include/asm/dma-mapping.h
@@ -31,7 +31,6 @@
 
 #define DMA_ERROR_CODE		(~(dma_addr_t)0x0)
 
-int dma_mapping_error(struct device *dev, dma_addr_t dma_addr);
 
 #define dma_alloc_noncoherent(d, s, h, f) dma_alloc_coherent(d, s, h, f)
 #define dma_free_noncoherent(d, s, v, h) dma_free_coherent(d, s, v, h)
@@ -47,6 +46,12 @@ dma_addr_t or1k_map_page(struct device *dev, struct page *page,
 void or1k_unmap_page(struct device *dev, dma_addr_t dma_handle,
 		     size_t size, enum dma_data_direction dir,
 		     struct dma_attrs *attrs);
+int or1k_map_sg(struct device *dev, struct scatterlist *sg,
+		int nents, enum dma_data_direction dir,
+		struct dma_attrs *attrs);
+void or1k_unmap_sg(struct device *dev, struct scatterlist *sg,
+		   int nents, enum dma_data_direction dir,
+		   struct dma_attrs *attrs);
 void or1k_sync_single_for_cpu(struct device *dev,
 			      dma_addr_t dma_handle, size_t size,
 			      enum dma_data_direction dir);
@@ -98,6 +103,51 @@ static inline void dma_unmap_single(struct device *dev, dma_addr_t addr,
 	debug_dma_unmap_page(dev, addr, size, dir, true);
 }
 
+static inline int dma_map_sg(struct device *dev, struct scatterlist *sg,
+				   int nents, enum dma_data_direction dir)
+{
+	int i, ents;
+	struct scatterlist *s;
+
+	for_each_sg(sg, s, nents, i)
+		kmemcheck_mark_initialized(sg_virt(s), s->length);
+	BUG_ON(!valid_dma_direction(dir));
+	ents = or1k_map_sg(dev, sg, nents, dir, NULL);
+	debug_dma_map_sg(dev, sg, nents, ents, dir);
+
+	return ents;
+}
+
+static inline void dma_unmap_sg(struct device *dev, struct scatterlist *sg,
+				      int nents, enum dma_data_direction dir)
+{
+	BUG_ON(!valid_dma_direction(dir));
+	debug_dma_unmap_sg(dev, sg, nents, dir);
+	or1k_unmap_sg(dev, sg, nents, dir, NULL);
+}
+
+static inline dma_addr_t dma_map_page(struct device *dev, struct page *page,
+				      size_t offset, size_t size,
+				      enum dma_data_direction dir)
+{
+	dma_addr_t addr;
+
+	kmemcheck_mark_initialized(page_address(page) + offset, size);
+	BUG_ON(!valid_dma_direction(dir));
+	addr = or1k_map_page(dev, page, offset, size, dir, NULL);
+	debug_dma_map_page(dev, page, offset, size, dir, addr, false);
+
+	return addr;
+}
+
+static inline void dma_unmap_page(struct device *dev, dma_addr_t addr,
+				  size_t size, enum dma_data_direction dir)
+{
+	BUG_ON(!valid_dma_direction(dir));
+	or1k_unmap_page(dev, addr, size, dir, NULL);
+	debug_dma_unmap_page(dev, addr, size, dir, true);
+}
+
 static inline void dma_sync_single_for_cpu(struct device *dev, dma_addr_t addr,
 					   size_t size,
 					   enum dma_data_direction dir)
@@ -119,7 +169,12 @@ static inline void dma_sync_single_for_device(struct device *dev,
 static inline int dma_supported(struct device *dev, u64 dma_mask)
 {
 	/* Support 32 bit DMA mask exclusively */
-	return dma_mask == 0xffffffffULL;
+	return dma_mask == DMA_BIT_MASK(32);
+}
+
+static inline int dma_mapping_error(struct device *dev, dma_addr_t dma_addr)
+{
+	return 0;
 }
 
 static inline int dma_set_mask(struct device *dev, u64 dma_mask)
diff --git a/arch/openrisc/include/asm/sigcontext.h b/arch/openrisc/include/asm/sigcontext.h
index 54a5c50132e..b79c2b19afb 100644
--- a/arch/openrisc/include/asm/sigcontext.h
+++ b/arch/openrisc/include/asm/sigcontext.h
@@ -23,16 +23,11 @@
 
 /* This struct is saved by setup_frame in signal.c, to keep the current
    context while a signal handler is executed. It's restored by sys_sigreturn.
-
-   To keep things simple, we use pt_regs here even though normally you just
-   specify the list of regs to save. Then we can use copy_from_user on the
-   entire regs instead of a bunch of get_user's as well...
 */
 
 struct sigcontext {
-	struct pt_regs regs;  /* needs to be first */
+	struct user_regs_struct regs;  /* needs to be first */
 	unsigned long oldmask;
-	unsigned long usp;    /* usp before stacking this gunk on it */
 };
 
 #endif /* __ASM_OPENRISC_SIGCONTEXT_H */
diff --git a/arch/openrisc/kernel/dma.c b/arch/openrisc/kernel/dma.c
index 968d3ee477e..f1c8ee2895d 100644
--- a/arch/openrisc/kernel/dma.c
+++ b/arch/openrisc/kernel/dma.c
@@ -154,6 +154,33 @@ void or1k_unmap_page(struct device *dev, dma_addr_t dma_handle,
 	/* Nothing special to do here... */
 }
 
+int or1k_map_sg(struct device *dev, struct scatterlist *sg,
+		int nents, enum dma_data_direction dir,
+		struct dma_attrs *attrs)
+{
+	struct scatterlist *s;
+	int i;
+
+	for_each_sg(sg, s, nents, i) {
+		s->dma_address = or1k_map_page(dev, sg_page(s), s->offset,
+					       s->length, dir, NULL);
+	}
+
+	return nents;
+}
+
+void or1k_unmap_sg(struct device *dev, struct scatterlist *sg,
+		   int nents, enum dma_data_direction dir,
+		   struct dma_attrs *attrs)
+{
+	struct scatterlist *s;
+	int i;
+
+	for_each_sg(sg, s, nents, i) {
+		or1k_unmap_page(dev, sg_dma_address(s), sg_dma_len(s), dir, NULL);
+	}
+}
+
 void or1k_sync_single_for_cpu(struct device *dev,
 			      dma_addr_t dma_handle, size_t size,
 			      enum dma_data_direction dir)
@@ -187,5 +214,4 @@ static int __init dma_init(void)
 
 	return 0;
 }
-
 fs_initcall(dma_init);
diff --git a/arch/openrisc/kernel/signal.c b/arch/openrisc/kernel/signal.c
index 5f759c76834..95207ab0c99 100644
--- a/arch/openrisc/kernel/signal.c
+++ b/arch/openrisc/kernel/signal.c
@@ -52,31 +52,25 @@ struct rt_sigframe {
 static int restore_sigcontext(struct pt_regs *regs, struct sigcontext *sc)
 {
 	unsigned int err = 0;
-	unsigned long old_usp;
 
 	/* Alwys make any pending restarted system call return -EINTR */
 	current_thread_info()->restart_block.fn = do_no_restart_syscall;
 
-	/* restore the regs from &sc->regs (same as sc, since regs is first)
+	/*
+	 * Restore the regs from &sc->regs.
 	 * (sc is already checked for VERIFY_READ since the sigframe was
 	 *  checked in sys_sigreturn previously)
 	 */
-
-	if (__copy_from_user(regs, sc, sizeof(struct pt_regs)))
+	if (__copy_from_user(regs, sc->regs.gpr, 32 * sizeof(unsigned long)))
+		goto badframe;
+	if (__copy_from_user(&regs->pc, &sc->regs.pc, sizeof(unsigned long)))
+		goto badframe;
+	if (__copy_from_user(&regs->sr, &sc->regs.sr, sizeof(unsigned long)))
 		goto badframe;
 
 	/* make sure the SM-bit is cleared so user-mode cannot fool us */
 	regs->sr &= ~SPR_SR_SM;
 
-	/* restore the old USP as it was before we stacked the sc etc.
-	 * (we cannot just pop the sigcontext since we aligned the sp and
-	 *  stuff after pushing it)
-	 */
-
-	err |= __get_user(old_usp, &sc->usp);
-
-	regs->sp = old_usp;
-
 	/* TODO: the other ports use regs->orig_XX to disable syscall checks
 	 * after this completes, but we don't use that mechanism. maybe we can
 	 * use it now ?
@@ -137,18 +131,17 @@ static int setup_sigcontext(struct sigcontext *sc, struct pt_regs *regs,
 			    unsigned long mask)
 {
 	int err = 0;
-	unsigned long usp = regs->sp;
 
-	/* copy the regs. they are first in sc so we can use sc directly */
+	/* copy the regs */
 
-	err |= __copy_to_user(sc, regs, sizeof(struct pt_regs));
+	err |= __copy_to_user(sc->regs.gpr, regs, 32 * sizeof(unsigned long));
+	err |= __copy_to_user(&sc->regs.pc, &regs->pc, sizeof(unsigned long));
+	err |= __copy_to_user(&sc->regs.sr, &regs->sr, sizeof(unsigned long));
 
 	/* then some other stuff */
 
 	err |= __put_user(mask, &sc->oldmask);
 
-	err |= __put_user(usp, &sc->usp);
-
 	return err;
 }
 
diff --git a/arch/parisc/kernel/syscall_table.S b/arch/parisc/kernel/syscall_table.S
index e66366fd2ab..3735abd7f8f 100644
--- a/arch/parisc/kernel/syscall_table.S
+++ b/arch/parisc/kernel/syscall_table.S
@@ -259,7 +259,7 @@
 	ENTRY_SAME(ni_syscall)		/* query_module */
 	ENTRY_SAME(poll)
 	/* structs contain pointers and an in_addr... */
-	ENTRY_COMP(nfsservctl)
+	ENTRY_SAME(ni_syscall)		/* was nfsservctl */
 	ENTRY_SAME(setresgid)		/* 170 */
 	ENTRY_SAME(getresgid)
 	ENTRY_SAME(prctl)
diff --git a/arch/powerpc/boot/dts/p1023rds.dts b/arch/powerpc/boot/dts/p1023rds.dts
index bfa96aa8f2c..d9b776740a6 100644
--- a/arch/powerpc/boot/dts/p1023rds.dts
+++ b/arch/powerpc/boot/dts/p1023rds.dts
@@ -387,7 +387,7 @@
 			#size-cells = <1>;
 			compatible = "cfi-flash";
 			reg = <0x0 0x0 0x02000000>;
-			bank-width = <1>;
+			bank-width = <2>;
 			device-width = <1>;
 			partition@0 {
 				label = "ramdisk";
diff --git a/arch/powerpc/configs/85xx/p1023rds_defconfig b/arch/powerpc/configs/85xx/p1023rds_defconfig
index 980ff8f61fd..3ff5a81c709 100644
--- a/arch/powerpc/configs/85xx/p1023rds_defconfig
+++ b/arch/powerpc/configs/85xx/p1023rds_defconfig
@@ -171,3 +171,4 @@ CONFIG_CRYPTO_SHA256=y
 CONFIG_CRYPTO_SHA512=y
 CONFIG_CRYPTO_AES=y
 # CONFIG_CRYPTO_ANSI_CPRNG is not set
+CONFIG_CRYPTO_DEV_FSL_CAAM=y
diff --git a/arch/powerpc/configs/corenet32_smp_defconfig b/arch/powerpc/configs/corenet32_smp_defconfig
index 10562a5c65b..4311d02a3bf 100644
--- a/arch/powerpc/configs/corenet32_smp_defconfig
+++ b/arch/powerpc/configs/corenet32_smp_defconfig
@@ -185,3 +185,4 @@ CONFIG_CRYPTO_SHA256=y
 CONFIG_CRYPTO_SHA512=y
 CONFIG_CRYPTO_AES=y
 # CONFIG_CRYPTO_ANSI_CPRNG is not set
+CONFIG_CRYPTO_DEV_FSL_CAAM=y
diff --git a/arch/powerpc/configs/corenet64_smp_defconfig b/arch/powerpc/configs/corenet64_smp_defconfig
index d32283555b5..c92c204a204 100644
--- a/arch/powerpc/configs/corenet64_smp_defconfig
+++ b/arch/powerpc/configs/corenet64_smp_defconfig
@@ -100,5 +100,8 @@ CONFIG_DEBUG_INFO=y
 CONFIG_SYSCTL_SYSCALL_CHECK=y
 CONFIG_VIRQ_DEBUG=y
 CONFIG_CRYPTO_PCBC=m
+CONFIG_CRYPTO_SHA256=y
+CONFIG_CRYPTO_SHA512=y
+CONFIG_CRYPTO_AES=y
 # CONFIG_CRYPTO_ANSI_CPRNG is not set
-CONFIG_CRYPTO_DEV_TALITOS=y
+CONFIG_CRYPTO_DEV_FSL_CAAM=y
diff --git a/arch/powerpc/configs/mpc85xx_defconfig b/arch/powerpc/configs/mpc85xx_defconfig
index fcd85d2c72d..a3467bfb767 100644
--- a/arch/powerpc/configs/mpc85xx_defconfig
+++ b/arch/powerpc/configs/mpc85xx_defconfig
@@ -139,6 +139,7 @@ CONFIG_SND=y
 CONFIG_SND_INTEL8X0=y
 # CONFIG_SND_PPC is not set
 # CONFIG_SND_USB is not set
+CONFIG_SND_SOC=y
 CONFIG_HID_A4TECH=y
 CONFIG_HID_APPLE=y
 CONFIG_HID_BELKIN=y
diff --git a/arch/powerpc/configs/mpc85xx_smp_defconfig b/arch/powerpc/configs/mpc85xx_smp_defconfig
index 908c941fc24..9693f6ed3da 100644
--- a/arch/powerpc/configs/mpc85xx_smp_defconfig
+++ b/arch/powerpc/configs/mpc85xx_smp_defconfig
@@ -140,6 +140,7 @@ CONFIG_SND=y
 CONFIG_SND_INTEL8X0=y
 # CONFIG_SND_PPC is not set
 # CONFIG_SND_USB is not set
+CONFIG_SND_SOC=y
 CONFIG_HID_A4TECH=y
 CONFIG_HID_APPLE=y
 CONFIG_HID_BELKIN=y
diff --git a/arch/powerpc/include/asm/systbl.h b/arch/powerpc/include/asm/systbl.h
index f6736b7da46..fa0d27a400d 100644
--- a/arch/powerpc/include/asm/systbl.h
+++ b/arch/powerpc/include/asm/systbl.h
@@ -171,7 +171,7 @@ SYSCALL_SPU(setresuid)
 SYSCALL_SPU(getresuid)
 SYSCALL(ni_syscall)
 SYSCALL_SPU(poll)
-COMPAT_SYS(nfsservctl)
+SYSCALL(ni_syscall)
 SYSCALL_SPU(setresgid)
 SYSCALL_SPU(getresgid)
 COMPAT_SYS_SPU(prctl)
diff --git a/arch/sh/include/asm/ptrace.h b/arch/sh/include/asm/ptrace.h
index b97baf81a87..2d3679b2447 100644
--- a/arch/sh/include/asm/ptrace.h
+++ b/arch/sh/include/asm/ptrace.h
@@ -123,7 +123,7 @@ static inline unsigned long regs_get_kernel_stack_nth(struct pt_regs *regs,
 struct perf_event;
 struct perf_sample_data;
 
-extern void ptrace_triggered(struct perf_event *bp, int nmi,
+extern void ptrace_triggered(struct perf_event *bp,
 		      struct perf_sample_data *data, struct pt_regs *regs);
 
 #define task_pt_regs(task) \
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7757.c b/arch/sh/kernel/cpu/sh4a/setup-sh7757.c
index e915deafac8..05559295d2c 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7757.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7757.c
@@ -15,6 +15,7 @@
 #include <linux/serial_sci.h>
 #include <linux/io.h>
 #include <linux/mm.h>
+#include <linux/dma-mapping.h>
 #include <linux/sh_timer.h>
 #include <linux/sh_dma.h>
 
diff --git a/arch/sh/kernel/idle.c b/arch/sh/kernel/idle.c
index 32114e0941a..db4ecd731a0 100644
--- a/arch/sh/kernel/idle.c
+++ b/arch/sh/kernel/idle.c
@@ -22,7 +22,7 @@
 #include <linux/atomic.h>
 #include <asm/smp.h>
 
-static void (*pm_idle)(void);
+void (*pm_idle)(void);
 
 static int hlt_counter;
 
diff --git a/arch/sh/kernel/traps_32.c b/arch/sh/kernel/traps_32.c
index d9006f8ffc1..7bbef95c9d1 100644
--- a/arch/sh/kernel/traps_32.c
+++ b/arch/sh/kernel/traps_32.c
@@ -316,6 +316,35 @@ static int handle_unaligned_ins(insn_size_t instruction, struct pt_regs *regs,
 			break;
 		}
 		break;
+
+	case 9: /* mov.w @(disp,PC),Rn */
+		srcu = (unsigned char __user *)regs->pc;
+		srcu += 4;
+		srcu += (instruction & 0x00FF) << 1;
+		dst = (unsigned char *)rn;
+		*(unsigned long *)dst = 0;
+
+#if !defined(__LITTLE_ENDIAN__)
+		dst += 2;
+#endif
+
+		if (ma->from(dst, srcu, 2))
+			goto fetch_fault;
+		sign_extend(2, dst);
+		ret = 0;
+		break;
+
+	case 0xd: /* mov.l @(disp,PC),Rn */
+		srcu = (unsigned char __user *)(regs->pc & ~0x3);
+		srcu += 4;
+		srcu += (instruction & 0x00FF) << 2;
+		dst = (unsigned char *)rn;
+		*(unsigned long *)dst = 0;
+
+		if (ma->from(dst, srcu, 4))
+			goto fetch_fault;
+		ret = 0;
+		break;
 	}
 	return ret;
 
@@ -466,6 +495,7 @@ int handle_unaligned_access(insn_size_t instruction, struct pt_regs *regs,
 		case 0x0500: /* mov.w @(disp,Rm),R0 */
 			goto simple;
 		case 0x0B00: /* bf   lab - no delayslot*/
+			ret = 0;
 			break;
 		case 0x0F00: /* bf/s lab */
 			ret = handle_delayslot(regs, instruction, ma);
@@ -479,6 +509,7 @@ int handle_unaligned_access(insn_size_t instruction, struct pt_regs *regs,
 			}
 			break;
 		case 0x0900: /* bt   lab - no delayslot */
+			ret = 0;
 			break;
 		case 0x0D00: /* bt/s lab */
 			ret = handle_delayslot(regs, instruction, ma);
@@ -494,6 +525,9 @@ int handle_unaligned_access(insn_size_t instruction, struct pt_regs *regs,
 		}
 		break;
 
+	case 0x9000: /* mov.w @(disp,Rm),Rn */
+		goto simple;
+
 	case 0xA000: /* bra label */
 		ret = handle_delayslot(regs, instruction, ma);
 		if (ret==0)
@@ -507,6 +541,9 @@ int handle_unaligned_access(insn_size_t instruction, struct pt_regs *regs,
 			regs->pc += SH_PC_12BIT_OFFSET(instruction);
 		}
 		break;
+
+	case 0xD000: /* mov.l @(disp,Rm),Rn */
+		goto simple;
 	}
 	return ret;
 
diff --git a/arch/sparc/kernel/irq.h b/arch/sparc/kernel/irq.h
index 100b9c204e7..42851122bbd 100644
--- a/arch/sparc/kernel/irq.h
+++ b/arch/sparc/kernel/irq.h
@@ -88,7 +88,7 @@ BTFIXUPDEF_CALL(void, set_irq_udt, int)
 #define set_irq_udt(cpu) BTFIXUP_CALL(set_irq_udt)(cpu)
 
 /* All SUN4D IPIs are sent on this IRQ, may be shared with hard IRQs */
-#define SUN4D_IPI_IRQ 14
+#define SUN4D_IPI_IRQ 13
 
 extern void sun4d_ipi_interrupt(void);
 
diff --git a/arch/sparc/kernel/setup_64.c b/arch/sparc/kernel/setup_64.c
index 3e9daea1653..3c5bb784214 100644
--- a/arch/sparc/kernel/setup_64.c
+++ b/arch/sparc/kernel/setup_64.c
@@ -440,8 +440,14 @@ static void __init init_sparc64_elf_hwcap(void)
 			cap |= AV_SPARC_VIS;
 		if (tlb_type == cheetah || tlb_type == cheetah_plus)
 			cap |= AV_SPARC_VIS | AV_SPARC_VIS2;
-		if (tlb_type == cheetah_plus)
-			cap |= AV_SPARC_POPC;
+		if (tlb_type == cheetah_plus) {
+			unsigned long impl, ver;
+
+			__asm__ __volatile__("rdpr %%ver, %0" : "=r" (ver));
+			impl = ((ver >> 32) & 0xffff);
+			if (impl == PANTHER_IMPL)
+				cap |= AV_SPARC_POPC;
+		}
 		if (tlb_type == hypervisor) {
 			if (sun4v_chip_type == SUN4V_CHIP_NIAGARA1)
 				cap |= AV_SPARC_ASI_BLK_INIT;
diff --git a/arch/sparc/kernel/systbls_64.S b/arch/sparc/kernel/systbls_64.S
index c9296ab0b1f..edbec45d468 100644
--- a/arch/sparc/kernel/systbls_64.S
+++ b/arch/sparc/kernel/systbls_64.S
@@ -68,7 +68,7 @@ sys_call_table32:
 	.word compat_sys_fstatfs64, sys_llseek, sys_mlock, sys_munlock, sys32_mlockall
 /*240*/	.word sys_munlockall, sys32_sched_setparam, sys32_sched_getparam, sys32_sched_setscheduler, sys32_sched_getscheduler
 	.word sys_sched_yield, sys32_sched_get_priority_max, sys32_sched_get_priority_min, sys32_sched_rr_get_interval, compat_sys_nanosleep
-/*250*/	.word sys_mremap, compat_sys_sysctl, sys32_getsid, sys_fdatasync, sys32_nfsservctl
+/*250*/	.word sys_mremap, compat_sys_sysctl, sys32_getsid, sys_fdatasync, sys_nis_syscall
 	.word sys32_sync_file_range, compat_sys_clock_settime, compat_sys_clock_gettime, compat_sys_clock_getres, sys32_clock_nanosleep
 /*260*/	.word compat_sys_sched_getaffinity, compat_sys_sched_setaffinity, sys32_timer_settime, compat_sys_timer_gettime, sys_timer_getoverrun
 	.word sys_timer_delete, compat_sys_timer_create, sys_ni_syscall, compat_sys_io_setup, sys_io_destroy
diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
index a518c0a4504..c59cc97fe6c 100644
--- a/arch/x86/include/asm/pvclock.h
+++ b/arch/x86/include/asm/pvclock.h
@@ -44,7 +44,7 @@ static inline u64 pvclock_scale_delta(u64 delta, u32 mul_frac, int shift)
 		: "a" ((u32)delta), "1" ((u32)(delta >> 32)), "2" (mul_frac) );
 #elif defined(__x86_64__)
 	__asm__ (
-		"mul %[mul_frac] ; shrd $32, %[hi], %[lo]"
+		"mulq %[mul_frac] ; shrd $32, %[hi], %[lo]"
 		: [lo]"=a"(product),
 		  [hi]"=d"(tmp)
 		: "0"(delta),
diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c
index 4ee3abf20ed..cfa62ec090e 100644
--- a/arch/x86/kernel/cpu/perf_event.c
+++ b/arch/x86/kernel/cpu/perf_event.c
@@ -1900,6 +1900,9 @@ perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs)
 
 	perf_callchain_store(entry, regs->ip);
 
+	if (!current->mm)
+		return;
+
 	if (perf_callchain_user32(regs, entry))
 		return;
 
diff --git a/arch/x86/pci/acpi.c b/arch/x86/pci/acpi.c
index c95330267f0..039d91315bc 100644
--- a/arch/x86/pci/acpi.c
+++ b/arch/x86/pci/acpi.c
@@ -365,8 +365,13 @@ struct pci_bus * __devinit pci_acpi_scan_root(struct acpi_pci_root *root)
 	 */
 	if (bus) {
 		struct pci_bus *child;
-		list_for_each_entry(child, &bus->children, node)
-			pcie_bus_configure_settings(child, child->self->pcie_mpss);
+		list_for_each_entry(child, &bus->children, node) {
+			struct pci_dev *self = child->self;
+			if (!self)
+				continue;
+
+			pcie_bus_configure_settings(child, self->pcie_mpss);
+		}
 	}
 
 	if (!bus)
diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c
index df118a825f3..c3b8d440873 100644
--- a/arch/x86/xen/setup.c
+++ b/arch/x86/xen/setup.c
@@ -184,6 +184,19 @@ static unsigned long __init xen_set_identity(const struct e820entry *list,
 					PFN_UP(start_pci), PFN_DOWN(last));
 	return identity;
 }
+
+static unsigned long __init xen_get_max_pages(void)
+{
+	unsigned long max_pages = MAX_DOMAIN_PAGES;
+	domid_t domid = DOMID_SELF;
+	int ret;
+
+	ret = HYPERVISOR_memory_op(XENMEM_maximum_reservation, &domid);
+	if (ret > 0)
+		max_pages = ret;
+	return min(max_pages, MAX_DOMAIN_PAGES);
+}
+
 /**
  * machine_specific_memory_setup - Hook for machine specific memory setup.
  **/
@@ -292,6 +305,12 @@ char * __init xen_memory_setup(void)
 
 	sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map);
 
+	extra_limit = xen_get_max_pages();
+	if (extra_limit >= max_pfn)
+		extra_pages = extra_limit - max_pfn;
+	else
+		extra_pages = 0;
+
 	extra_pages += xen_return_unused_memory(xen_start_info->nr_pages, &e820);
 
 	/*
diff --git a/arch/x86/xen/smp.c b/arch/x86/xen/smp.c
index e79dbb95482..d4fc6d454f8 100644
--- a/arch/x86/xen/smp.c
+++ b/arch/x86/xen/smp.c
@@ -32,6 +32,7 @@
 #include <xen/page.h>
 #include <xen/events.h>
 
+#include <xen/hvc-console.h>
 #include "xen-ops.h"
 #include "mmu.h"
 
@@ -207,6 +208,15 @@ static void __init xen_smp_prepare_cpus(unsigned int max_cpus)
 	unsigned cpu;
 	unsigned int i;
 
+	if (skip_ioapic_setup) {
+		char *m = (max_cpus == 0) ?
+			"The nosmp parameter is incompatible with Xen; " \
+			"use Xen dom0_max_vcpus=1 parameter" :
+			"The noapic parameter is incompatible with Xen";
+
+		xen_raw_printk(m);
+		panic(m);
+	}
 	xen_init_lock_cpu(0);
 
 	smp_store_cpu_info(0);
diff --git a/arch/x86/xen/xen-asm_32.S b/arch/x86/xen/xen-asm_32.S
index 22a2093b586..b040b0e518c 100644
--- a/arch/x86/xen/xen-asm_32.S
+++ b/arch/x86/xen/xen-asm_32.S
@@ -113,11 +113,13 @@ xen_iret_start_crit:
 
 	/*
 	 * If there's something pending, mask events again so we can
-	 * jump back into xen_hypervisor_callback
+	 * jump back into xen_hypervisor_callback. Otherwise do not
+	 * touch XEN_vcpu_info_mask.
 	 */
-	sete XEN_vcpu_info_mask(%eax)
+	jne 1f
+	movb $1, XEN_vcpu_info_mask(%eax)
 
-	popl %eax
+1:	popl %eax
 
 	/*
 	 * From this point on the registers are restored and the stack
diff --git a/drivers/base/regmap/regmap.c b/drivers/base/regmap/regmap.c
index 0eef4da1ac6..20663f8dae4 100644
--- a/drivers/base/regmap/regmap.c
+++ b/drivers/base/regmap/regmap.c
@@ -168,13 +168,11 @@ struct regmap *regmap_init(struct device *dev,
 	map->work_buf = kmalloc(map->format.buf_size, GFP_KERNEL);
 	if (map->work_buf == NULL) {
 		ret = -ENOMEM;
-		goto err_bus;
+		goto err_map;
 	}
 
 	return map;
 
-err_bus:
-	module_put(map->bus->owner);
 err_map:
 	kfree(map);
 err:
@@ -188,7 +186,6 @@ EXPORT_SYMBOL_GPL(regmap_init);
 void regmap_exit(struct regmap *map)
 {
 	kfree(map->work_buf);
-	module_put(map->bus->owner);
 	kfree(map);
 }
 EXPORT_SYMBOL_GPL(regmap_exit);
diff --git a/drivers/bcma/main.c b/drivers/bcma/main.c
index 873e2e4ac55..73b7b1a18fa 100644
--- a/drivers/bcma/main.c
+++ b/drivers/bcma/main.c
@@ -15,6 +15,7 @@ MODULE_LICENSE("GPL");
 static int bcma_bus_match(struct device *dev, struct device_driver *drv);
 static int bcma_device_probe(struct device *dev);
 static int bcma_device_remove(struct device *dev);
+static int bcma_device_uevent(struct device *dev, struct kobj_uevent_env *env);
 
 static ssize_t manuf_show(struct device *dev, struct device_attribute *attr, char *buf)
 {
@@ -49,6 +50,7 @@ static struct bus_type bcma_bus_type = {
 	.match		= bcma_bus_match,
 	.probe		= bcma_device_probe,
 	.remove		= bcma_device_remove,
+	.uevent		= bcma_device_uevent,
 	.dev_attrs	= bcma_device_attrs,
 };
 
@@ -227,6 +229,16 @@ static int bcma_device_remove(struct device *dev)
 	return 0;
 }
 
+static int bcma_device_uevent(struct device *dev, struct kobj_uevent_env *env)
+{
+	struct bcma_device *core = container_of(dev, struct bcma_device, dev);
+
+	return add_uevent_var(env,
+			      "MODALIAS=bcma:m%04Xid%04Xrev%02Xcl%02X",
+			      core->id.manuf, core->id.id,
+			      core->id.rev, core->id.class);
+}
+
 static int __init bcma_modinit(void)
 {
 	int err;
diff --git a/drivers/bluetooth/ath3k.c b/drivers/bluetooth/ath3k.c
index a5854735bb2..db7cb8111fb 100644
--- a/drivers/bluetooth/ath3k.c
+++ b/drivers/bluetooth/ath3k.c
@@ -63,6 +63,7 @@ static struct usb_device_id ath3k_table[] = {
 	/* Atheros AR3011 with sflash firmware*/
 	{ USB_DEVICE(0x0CF3, 0x3002) },
 	{ USB_DEVICE(0x13d3, 0x3304) },
+	{ USB_DEVICE(0x0930, 0x0215) },
 
 	/* Atheros AR9285 Malbec with sflash firmware */
 	{ USB_DEVICE(0x03F0, 0x311D) },
diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
index 91d13a9e8c6..3ef476070ba 100644
--- a/drivers/bluetooth/btusb.c
+++ b/drivers/bluetooth/btusb.c
@@ -106,6 +106,7 @@ static struct usb_device_id blacklist_table[] = {
 	/* Atheros 3011 with sflash firmware */
 	{ USB_DEVICE(0x0cf3, 0x3002), .driver_info = BTUSB_IGNORE },
 	{ USB_DEVICE(0x13d3, 0x3304), .driver_info = BTUSB_IGNORE },
+	{ USB_DEVICE(0x0930, 0x0215), .driver_info = BTUSB_IGNORE },
 
 	/* Atheros AR9285 Malbec with sflash firmware */
 	{ USB_DEVICE(0x03f0, 0x311d), .driver_info = BTUSB_IGNORE },
@@ -256,7 +257,9 @@ static void btusb_intr_complete(struct urb *urb)
 
 	err = usb_submit_urb(urb, GFP_ATOMIC);
 	if (err < 0) {
-		if (err != -EPERM)
+		/* -EPERM: urb is being killed;
+		 * -ENODEV: device got disconnected */
+		if (err != -EPERM && err != -ENODEV)
 			BT_ERR("%s urb %p failed to resubmit (%d)",
 						hdev->name, urb, -err);
 		usb_unanchor_urb(urb);
@@ -341,7 +344,9 @@ static void btusb_bulk_complete(struct urb *urb)
 
 	err = usb_submit_urb(urb, GFP_ATOMIC);
 	if (err < 0) {
-		if (err != -EPERM)
+		/* -EPERM: urb is being killed;
+		 * -ENODEV: device got disconnected */
+		if (err != -EPERM && err != -ENODEV)
 			BT_ERR("%s urb %p failed to resubmit (%d)",
 						hdev->name, urb, -err);
 		usb_unanchor_urb(urb);
@@ -431,7 +436,9 @@ static void btusb_isoc_complete(struct urb *urb)
 
 	err = usb_submit_urb(urb, GFP_ATOMIC);
 	if (err < 0) {
-		if (err != -EPERM)
+		/* -EPERM: urb is being killed;
+		 * -ENODEV: device got disconnected */
+		if (err != -EPERM && err != -ENODEV)
 			BT_ERR("%s urb %p failed to resubmit (%d)",
 						hdev->name, urb, -err);
 		usb_unanchor_urb(urb);
diff --git a/drivers/clocksource/sh_cmt.c b/drivers/clocksource/sh_cmt.c
index dc7c033ef58..32a77becc09 100644
--- a/drivers/clocksource/sh_cmt.c
+++ b/drivers/clocksource/sh_cmt.c
@@ -26,6 +26,7 @@
 #include <linux/clk.h>
 #include <linux/irq.h>
 #include <linux/err.h>
+#include <linux/delay.h>
 #include <linux/clocksource.h>
 #include <linux/clockchips.h>
 #include <linux/sh_timer.h>
@@ -150,13 +151,13 @@ static void sh_cmt_start_stop_ch(struct sh_cmt_priv *p, int start)
 
 static int sh_cmt_enable(struct sh_cmt_priv *p, unsigned long *rate)
 {
-	int ret;
+	int k, ret;
 
 	/* enable clock */
 	ret = clk_enable(p->clk);
 	if (ret) {
 		dev_err(&p->pdev->dev, "cannot enable clock\n");
-		return ret;
+		goto err0;
 	}
 
 	/* make sure channel is disabled */
@@ -174,9 +175,38 @@ static int sh_cmt_enable(struct sh_cmt_priv *p, unsigned long *rate)
 	sh_cmt_write(p, CMCOR, 0xffffffff);
 	sh_cmt_write(p, CMCNT, 0);
 
+	/*
+	 * According to the sh73a0 user's manual, as CMCNT can be operated
+	 * only by the RCLK (Pseudo 32 KHz), there's one restriction on
+	 * modifying CMCNT register; two RCLK cycles are necessary before
+	 * this register is either read or any modification of the value
+	 * it holds is reflected in the LSI's actual operation.
+	 *
+	 * While at it, we're supposed to clear out the CMCNT as of this
+	 * moment, so make sure it's processed properly here.  This will
+	 * take RCLKx2 at maximum.
+	 */
+	for (k = 0; k < 100; k++) {
+		if (!sh_cmt_read(p, CMCNT))
+			break;
+		udelay(1);
+	}
+
+	if (sh_cmt_read(p, CMCNT)) {
+		dev_err(&p->pdev->dev, "cannot clear CMCNT\n");
+		ret = -ETIMEDOUT;
+		goto err1;
+	}
+
 	/* enable channel */
 	sh_cmt_start_stop_ch(p, 1);
 	return 0;
+ err1:
+	/* stop clock */
+	clk_disable(p->clk);
+
+ err0:
+	return ret;
 }
 
 static void sh_cmt_disable(struct sh_cmt_priv *p)
diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c
index cd3a7c726bf..467e4dcb20a 100644
--- a/drivers/dma/ste_dma40.c
+++ b/drivers/dma/ste_dma40.c
@@ -174,8 +174,10 @@ struct d40_base;
  * @tasklet: Tasklet that gets scheduled from interrupt context to complete a
  * transfer and call client callback.
  * @client: Cliented owned descriptor list.
+ * @pending_queue: Submitted jobs, to be issued by issue_pending()
  * @active: Active descriptor.
  * @queue: Queued jobs.
+ * @prepare_queue: Prepared jobs.
  * @dma_cfg: The client configuration of this dma channel.
  * @configured: whether the dma_cfg configuration is valid
  * @base: Pointer to the device instance struct.
@@ -203,6 +205,7 @@ struct d40_chan {
 	struct list_head		 pending_queue;
 	struct list_head		 active;
 	struct list_head		 queue;
+	struct list_head		 prepare_queue;
 	struct stedma40_chan_cfg	 dma_cfg;
 	bool				 configured;
 	struct d40_base			*base;
@@ -477,7 +480,6 @@ static struct d40_desc *d40_desc_get(struct d40_chan *d40c)
 
 		list_for_each_entry_safe(d, _d, &d40c->client, node)
 			if (async_tx_test_ack(&d->txd)) {
-				d40_pool_lli_free(d40c, d);
 				d40_desc_remove(d);
 				desc = d;
 				memset(desc, 0, sizeof(*desc));
@@ -644,8 +646,11 @@ static struct d40_desc *d40_first_active_get(struct d40_chan *d40c)
 	return d;
 }
 
+/* remove desc from current queue and add it to the pending_queue */
 static void d40_desc_queue(struct d40_chan *d40c, struct d40_desc *desc)
 {
+	d40_desc_remove(desc);
+	desc->is_in_client_list = false;
 	list_add_tail(&desc->node, &d40c->pending_queue);
 }
 
@@ -803,6 +808,7 @@ done:
 static void d40_term_all(struct d40_chan *d40c)
 {
 	struct d40_desc *d40d;
+	struct d40_desc *_d;
 
 	/* Release active descriptors */
 	while ((d40d = d40_first_active_get(d40c))) {
@@ -822,6 +828,21 @@ static void d40_term_all(struct d40_chan *d40c)
 		d40_desc_free(d40c, d40d);
 	}
 
+	/* Release client owned descriptors */
+	if (!list_empty(&d40c->client))
+		list_for_each_entry_safe(d40d, _d, &d40c->client, node) {
+			d40_desc_remove(d40d);
+			d40_desc_free(d40c, d40d);
+		}
+
+	/* Release descriptors in prepare queue */
+	if (!list_empty(&d40c->prepare_queue))
+		list_for_each_entry_safe(d40d, _d,
+					 &d40c->prepare_queue, node) {
+			d40_desc_remove(d40d);
+			d40_desc_free(d40c, d40d);
+		}
+
 	d40c->pending_tx = 0;
 	d40c->busy = false;
 }
@@ -1208,7 +1229,6 @@ static void dma_tasklet(unsigned long data)
 
 	if (!d40d->cyclic) {
 		if (async_tx_test_ack(&d40d->txd)) {
-			d40_pool_lli_free(d40c, d40d);
 			d40_desc_remove(d40d);
 			d40_desc_free(d40c, d40d);
 		} else {
@@ -1595,21 +1615,10 @@ static int d40_free_dma(struct d40_chan *d40c)
 	u32 event;
 	struct d40_phy_res *phy = d40c->phy_chan;
 	bool is_src;
-	struct d40_desc *d;
-	struct d40_desc *_d;
-
 
 	/* Terminate all queued and active transfers */
 	d40_term_all(d40c);
 
-	/* Release client owned descriptors */
-	if (!list_empty(&d40c->client))
-		list_for_each_entry_safe(d, _d, &d40c->client, node) {
-			d40_pool_lli_free(d40c, d);
-			d40_desc_remove(d);
-			d40_desc_free(d40c, d);
-		}
-
 	if (phy == NULL) {
 		chan_err(d40c, "phy == null\n");
 		return -EINVAL;
@@ -1911,6 +1920,12 @@ d40_prep_sg(struct dma_chan *dchan, struct scatterlist *sg_src,
 		goto err;
 	}
 
+	/*
+	 * add descriptor to the prepare queue in order to be able
+	 * to free them later in terminate_all
+	 */
+	list_add_tail(&desc->node, &chan->prepare_queue);
+
 	spin_unlock_irqrestore(&chan->lock, flags);
 
 	return &desc->txd;
@@ -2400,6 +2415,7 @@ static void __init d40_chan_init(struct d40_base *base, struct dma_device *dma,
 		INIT_LIST_HEAD(&d40c->queue);
 		INIT_LIST_HEAD(&d40c->pending_queue);
 		INIT_LIST_HEAD(&d40c->client);
+		INIT_LIST_HEAD(&d40c->prepare_queue);
 
 		tasklet_init(&d40c->tasklet, dma_tasklet,
 			     (unsigned long) d40c);
diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
index 802b61ac313..f7c6854eb4d 100644
--- a/drivers/gpu/drm/drm_fb_helper.c
+++ b/drivers/gpu/drm/drm_fb_helper.c
@@ -256,7 +256,6 @@ int drm_fb_helper_panic(struct notifier_block *n, unsigned long ununsed,
 {
 	printk(KERN_ERR "panic occurred, switching back to text console\n");
 	return drm_fb_helper_force_kernel_mode();
-	return 0;
 }
 EXPORT_SYMBOL(drm_fb_helper_panic);
 
diff --git a/drivers/gpu/drm/nouveau/Makefile b/drivers/gpu/drm/nouveau/Makefile
index 0583677e458..35ef5b1e356 100644
--- a/drivers/gpu/drm/nouveau/Makefile
+++ b/drivers/gpu/drm/nouveau/Makefile
@@ -21,16 +21,17 @@ nouveau-y := nouveau_drv.o nouveau_state.o nouveau_channel.o nouveau_mem.o \
              nv40_grctx.o nv50_grctx.o nvc0_grctx.o \
              nv84_crypt.o \
              nva3_copy.o nvc0_copy.o \
-             nv40_mpeg.o nv50_mpeg.o \
+             nv31_mpeg.o nv50_mpeg.o \
              nv04_instmem.o nv50_instmem.o nvc0_instmem.o \
-             nv50_evo.o nv50_crtc.o nv50_dac.o nv50_sor.o \
-             nv50_cursor.o nv50_display.o \
              nv04_dac.o nv04_dfp.o nv04_tv.o nv17_tv.o nv17_tv_modes.o \
              nv04_crtc.o nv04_display.o nv04_cursor.o \
+             nv50_evo.o nv50_crtc.o nv50_dac.o nv50_sor.o \
+             nv50_cursor.o nv50_display.o \
+             nvd0_display.o \
              nv04_fbcon.o nv50_fbcon.o nvc0_fbcon.o \
              nv10_gpio.o nv50_gpio.o \
 	     nv50_calc.o \
-	     nv04_pm.o nv50_pm.o nva3_pm.o \
+	     nv04_pm.o nv40_pm.o nv50_pm.o nva3_pm.o nvc0_pm.o \
 	     nv50_vram.o nvc0_vram.o \
 	     nv50_vm.o nvc0_vm.o
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_backlight.c b/drivers/gpu/drm/nouveau/nouveau_backlight.c
index 00a55dfdba8..fa22b28e877 100644
--- a/drivers/gpu/drm/nouveau/nouveau_backlight.c
+++ b/drivers/gpu/drm/nouveau/nouveau_backlight.c
@@ -37,8 +37,10 @@
 #include "nouveau_drv.h"
 #include "nouveau_drm.h"
 #include "nouveau_reg.h"
+#include "nouveau_encoder.h"
 
-static int nv40_get_intensity(struct backlight_device *bd)
+static int
+nv40_get_intensity(struct backlight_device *bd)
 {
 	struct drm_device *dev = bl_get_data(bd);
 	int val = (nv_rd32(dev, NV40_PMC_BACKLIGHT) & NV40_PMC_BACKLIGHT_MASK)
@@ -47,7 +49,8 @@ static int nv40_get_intensity(struct backlight_device *bd)
 	return val;
 }
 
-static int nv40_set_intensity(struct backlight_device *bd)
+static int
+nv40_set_intensity(struct backlight_device *bd)
 {
 	struct drm_device *dev = bl_get_data(bd);
 	int val = bd->props.brightness;
@@ -65,30 +68,8 @@ static const struct backlight_ops nv40_bl_ops = {
 	.update_status = nv40_set_intensity,
 };
 
-static int nv50_get_intensity(struct backlight_device *bd)
-{
-	struct drm_device *dev = bl_get_data(bd);
-
-	return nv_rd32(dev, NV50_PDISPLAY_SOR_BACKLIGHT);
-}
-
-static int nv50_set_intensity(struct backlight_device *bd)
-{
-	struct drm_device *dev = bl_get_data(bd);
-	int val = bd->props.brightness;
-
-	nv_wr32(dev, NV50_PDISPLAY_SOR_BACKLIGHT,
-		val | NV50_PDISPLAY_SOR_BACKLIGHT_ENABLE);
-	return 0;
-}
-
-static const struct backlight_ops nv50_bl_ops = {
-	.options = BL_CORE_SUSPENDRESUME,
-	.get_brightness = nv50_get_intensity,
-	.update_status = nv50_set_intensity,
-};
-
-static int nouveau_nv40_backlight_init(struct drm_connector *connector)
+static int
+nv40_backlight_init(struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
@@ -113,34 +94,129 @@ static int nouveau_nv40_backlight_init(struct drm_connector *connector)
 	return 0;
 }
 
-static int nouveau_nv50_backlight_init(struct drm_connector *connector)
+static int
+nv50_get_intensity(struct backlight_device *bd)
+{
+	struct nouveau_encoder *nv_encoder = bl_get_data(bd);
+	struct drm_device *dev = nv_encoder->base.base.dev;
+	int or = nv_encoder->or;
+	u32 div = 1025;
+	u32 val;
+
+	val  = nv_rd32(dev, NV50_PDISP_SOR_PWM_CTL(or));
+	val &= NV50_PDISP_SOR_PWM_CTL_VAL;
+	return ((val * 100) + (div / 2)) / div;
+}
+
+static int
+nv50_set_intensity(struct backlight_device *bd)
+{
+	struct nouveau_encoder *nv_encoder = bl_get_data(bd);
+	struct drm_device *dev = nv_encoder->base.base.dev;
+	int or = nv_encoder->or;
+	u32 div = 1025;
+	u32 val = (bd->props.brightness * div) / 100;
+
+	nv_wr32(dev, NV50_PDISP_SOR_PWM_CTL(or),
+		     NV50_PDISP_SOR_PWM_CTL_NEW | val);
+	return 0;
+}
+
+static const struct backlight_ops nv50_bl_ops = {
+	.options = BL_CORE_SUSPENDRESUME,
+	.get_brightness = nv50_get_intensity,
+	.update_status = nv50_set_intensity,
+};
+
+static int
+nva3_get_intensity(struct backlight_device *bd)
+{
+	struct nouveau_encoder *nv_encoder = bl_get_data(bd);
+	struct drm_device *dev = nv_encoder->base.base.dev;
+	int or = nv_encoder->or;
+	u32 div, val;
+
+	div  = nv_rd32(dev, NV50_PDISP_SOR_PWM_DIV(or));
+	val  = nv_rd32(dev, NV50_PDISP_SOR_PWM_CTL(or));
+	val &= NVA3_PDISP_SOR_PWM_CTL_VAL;
+	if (div && div >= val)
+		return ((val * 100) + (div / 2)) / div;
+
+	return 100;
+}
+
+static int
+nva3_set_intensity(struct backlight_device *bd)
+{
+	struct nouveau_encoder *nv_encoder = bl_get_data(bd);
+	struct drm_device *dev = nv_encoder->base.base.dev;
+	int or = nv_encoder->or;
+	u32 div, val;
+
+	div = nv_rd32(dev, NV50_PDISP_SOR_PWM_DIV(or));
+	val = (bd->props.brightness * div) / 100;
+	if (div) {
+		nv_wr32(dev, NV50_PDISP_SOR_PWM_CTL(or), val |
+			     NV50_PDISP_SOR_PWM_CTL_NEW |
+			     NVA3_PDISP_SOR_PWM_CTL_UNK);
+		return 0;
+	}
+
+	return -EINVAL;
+}
+
+static const struct backlight_ops nva3_bl_ops = {
+	.options = BL_CORE_SUSPENDRESUME,
+	.get_brightness = nva3_get_intensity,
+	.update_status = nva3_set_intensity,
+};
+
+static int
+nv50_backlight_init(struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct nouveau_encoder *nv_encoder;
 	struct backlight_properties props;
 	struct backlight_device *bd;
+	const struct backlight_ops *ops;
+
+	nv_encoder = find_encoder(connector, OUTPUT_LVDS);
+	if (!nv_encoder) {
+		nv_encoder = find_encoder(connector, OUTPUT_DP);
+		if (!nv_encoder)
+			return -ENODEV;
+	}
 
-	if (!nv_rd32(dev, NV50_PDISPLAY_SOR_BACKLIGHT))
+	if (!nv_rd32(dev, NV50_PDISP_SOR_PWM_CTL(nv_encoder->or)))
 		return 0;
 
+	if (dev_priv->chipset <= 0xa0 ||
+	    dev_priv->chipset == 0xaa ||
+	    dev_priv->chipset == 0xac)
+		ops = &nv50_bl_ops;
+	else
+		ops = &nva3_bl_ops;
+
 	memset(&props, 0, sizeof(struct backlight_properties));
 	props.type = BACKLIGHT_RAW;
-	props.max_brightness = 1025;
-	bd = backlight_device_register("nv_backlight", &connector->kdev, dev,
-				       &nv50_bl_ops, &props);
+	props.max_brightness = 100;
+	bd = backlight_device_register("nv_backlight", &connector->kdev,
+				       nv_encoder, ops, &props);
 	if (IS_ERR(bd))
 		return PTR_ERR(bd);
 
 	dev_priv->backlight = bd;
-	bd->props.brightness = nv50_get_intensity(bd);
+	bd->props.brightness = bd->ops->get_brightness(bd);
 	backlight_update_status(bd);
 	return 0;
 }
 
-int nouveau_backlight_init(struct drm_connector *connector)
+int
+nouveau_backlight_init(struct drm_device *dev)
 {
-	struct drm_device *dev = connector->dev;
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct drm_connector *connector;
 
 #ifdef CONFIG_ACPI
 	if (acpi_video_backlight_support()) {
@@ -150,21 +226,28 @@ int nouveau_backlight_init(struct drm_connector *connector)
 	}
 #endif
 
-	switch (dev_priv->card_type) {
-	case NV_40:
-		return nouveau_nv40_backlight_init(connector);
-	case NV_50:
-		return nouveau_nv50_backlight_init(connector);
-	default:
-		break;
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+		if (connector->connector_type != DRM_MODE_CONNECTOR_LVDS &&
+		    connector->connector_type != DRM_MODE_CONNECTOR_eDP)
+			continue;
+
+		switch (dev_priv->card_type) {
+		case NV_40:
+			return nv40_backlight_init(connector);
+		case NV_50:
+			return nv50_backlight_init(connector);
+		default:
+			break;
+		}
 	}
 
+
 	return 0;
 }
 
-void nouveau_backlight_exit(struct drm_connector *connector)
+void
+nouveau_backlight_exit(struct drm_device *dev)
 {
-	struct drm_device *dev = connector->dev;
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
 
 	if (dev_priv->backlight) {
diff --git a/drivers/gpu/drm/nouveau/nouveau_bios.c b/drivers/gpu/drm/nouveau/nouveau_bios.c
index b311faba34f..032a8209813 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bios.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bios.c
@@ -296,6 +296,11 @@ munge_reg(struct nvbios *bios, uint32_t reg)
 	if (dev_priv->card_type < NV_50)
 		return reg;
 
+	if (reg & 0x80000000) {
+		BUG_ON(bios->display.crtc < 0);
+		reg += bios->display.crtc * 0x800;
+	}
+
 	if (reg & 0x40000000) {
 		BUG_ON(!dcbent);
 
@@ -304,7 +309,7 @@ munge_reg(struct nvbios *bios, uint32_t reg)
 			reg += 0x00000080;
 	}
 
-	reg &= ~0x60000000;
+	reg &= ~0xe0000000;
 	return reg;
 }
 
@@ -1174,22 +1179,19 @@ init_dp_condition(struct nvbios *bios, uint16_t offset, struct init_exec *iexec)
 	 *
 	 */
 
-	struct bit_displayport_encoder_table *dpe = NULL;
 	struct dcb_entry *dcb = bios->display.output;
 	struct drm_device *dev = bios->dev;
 	uint8_t cond = bios->data[offset + 1];
-	int dummy;
+	uint8_t *table, *entry;
 
 	BIOSLOG(bios, "0x%04X: subop 0x%02X\n", offset, cond);
 
 	if (!iexec->execute)
 		return 3;
 
-	dpe = nouveau_bios_dp_table(dev, dcb, &dummy);
-	if (!dpe) {
-		NV_ERROR(dev, "0x%04X: INIT_3A: no encoder table!!\n", offset);
+	table = nouveau_dp_bios_data(dev, dcb, &entry);
+	if (!table)
 		return 3;
-	}
 
 	switch (cond) {
 	case 0:
@@ -1203,7 +1205,7 @@ init_dp_condition(struct nvbios *bios, uint16_t offset, struct init_exec *iexec)
 		break;
 	case 1:
 	case 2:
-		if (!(dpe->unknown & cond))
+		if (!(entry[5] & cond))
 			iexec->execute = false;
 		break;
 	case 5:
@@ -3221,6 +3223,49 @@ init_8d(struct nvbios *bios, uint16_t offset, struct init_exec *iexec)
 	return 1;
 }
 
+static void
+init_gpio_unknv50(struct nvbios *bios, struct dcb_gpio_entry *gpio)
+{
+	const uint32_t nv50_gpio_ctl[2] = { 0xe100, 0xe28c };
+	u32 r, s, v;
+
+	/* Not a clue, needs de-magicing */
+	r = nv50_gpio_ctl[gpio->line >> 4];
+	s = (gpio->line & 0x0f);
+	v = bios_rd32(bios, r) & ~(0x00010001 << s);
+	switch ((gpio->entry & 0x06000000) >> 25) {
+	case 1:
+		v |= (0x00000001 << s);
+		break;
+	case 2:
+		v |= (0x00010000 << s);
+		break;
+	default:
+		break;
+	}
+
+	bios_wr32(bios, r, v);
+}
+
+static void
+init_gpio_unknvd0(struct nvbios *bios, struct dcb_gpio_entry *gpio)
+{
+	u32 v, i;
+
+	v  = bios_rd32(bios, 0x00d610 + (gpio->line * 4));
+	v &= 0xffffff00;
+	v |= (gpio->entry & 0x00ff0000) >> 16;
+	bios_wr32(bios, 0x00d610 + (gpio->line * 4), v);
+
+	i = (gpio->entry & 0x1f000000) >> 24;
+	if (i) {
+		v  = bios_rd32(bios, 0x00d640 + ((i - 1) * 4));
+		v &= 0xffffff00;
+		v |= gpio->line;
+		bios_wr32(bios, 0x00d640 + ((i - 1) * 4), v);
+	}
+}
+
 static int
 init_gpio(struct nvbios *bios, uint16_t offset, struct init_exec *iexec)
 {
@@ -3235,7 +3280,6 @@ init_gpio(struct nvbios *bios, uint16_t offset, struct init_exec *iexec)
 
 	struct drm_nouveau_private *dev_priv = bios->dev->dev_private;
 	struct nouveau_gpio_engine *pgpio = &dev_priv->engine.gpio;
-	const uint32_t nv50_gpio_ctl[2] = { 0xe100, 0xe28c };
 	int i;
 
 	if (dev_priv->card_type < NV_50) {
@@ -3248,33 +3292,20 @@ init_gpio(struct nvbios *bios, uint16_t offset, struct init_exec *iexec)
 
 	for (i = 0; i < bios->dcb.gpio.entries; i++) {
 		struct dcb_gpio_entry *gpio = &bios->dcb.gpio.entry[i];
-		uint32_t r, s, v;
 
 		BIOSLOG(bios, "0x%04X: Entry: 0x%08X\n", offset, gpio->entry);
 
 		BIOSLOG(bios, "0x%04X: set gpio 0x%02x, state %d\n",
 			offset, gpio->tag, gpio->state_default);
-		if (bios->execute)
-			pgpio->set(bios->dev, gpio->tag, gpio->state_default);
 
-		/* The NVIDIA binary driver doesn't appear to actually do
-		 * any of this, my VBIOS does however.
-		 */
-		/* Not a clue, needs de-magicing */
-		r = nv50_gpio_ctl[gpio->line >> 4];
-		s = (gpio->line & 0x0f);
-		v = bios_rd32(bios, r) & ~(0x00010001 << s);
-		switch ((gpio->entry & 0x06000000) >> 25) {
-		case 1:
-			v |= (0x00000001 << s);
-			break;
-		case 2:
-			v |= (0x00010000 << s);
-			break;
-		default:
-			break;
-		}
-		bios_wr32(bios, r, v);
+		if (!bios->execute)
+			continue;
+
+		pgpio->set(bios->dev, gpio->tag, gpio->state_default);
+		if (dev_priv->card_type < NV_D0)
+			init_gpio_unknv50(bios, gpio);
+		else
+			init_gpio_unknvd0(bios, gpio);
 	}
 
 	return 1;
@@ -3737,6 +3768,10 @@ parse_init_table(struct nvbios *bios, uint16_t offset, struct init_exec *iexec)
 	int count = 0, i, ret;
 	uint8_t id;
 
+	/* catch NULL script pointers */
+	if (offset == 0)
+		return 0;
+
 	/*
 	 * Loop until INIT_DONE causes us to break out of the loop
 	 * (or until offset > bios length just in case... )
@@ -4389,86 +4424,37 @@ int nouveau_bios_parse_lvds_table(struct drm_device *dev, int pxclk, bool *dl, b
 	return 0;
 }
 
-static uint8_t *
-bios_output_config_match(struct drm_device *dev, struct dcb_entry *dcbent,
-			 uint16_t record, int record_len, int record_nr,
-			 bool match_link)
+/* BIT 'U'/'d' table encoder subtables have hashes matching them to
+ * a particular set of encoders.
+ *
+ * This function returns true if a particular DCB entry matches.
+ */
+bool
+bios_encoder_match(struct dcb_entry *dcb, u32 hash)
 {
-	struct drm_nouveau_private *dev_priv = dev->dev_private;
-	struct nvbios *bios = &dev_priv->vbios;
-	uint32_t entry;
-	uint16_t table;
-	int i, v;
+	if ((hash & 0x000000f0) != (dcb->location << 4))
+		return false;
+	if ((hash & 0x0000000f) != dcb->type)
+		return false;
+	if (!(hash & (dcb->or << 16)))
+		return false;
 
-	switch (dcbent->type) {
+	switch (dcb->type) {
 	case OUTPUT_TMDS:
 	case OUTPUT_LVDS:
 	case OUTPUT_DP:
-		break;
-	default:
-		match_link = false;
-		break;
-	}
-
-	for (i = 0; i < record_nr; i++, record += record_len) {
-		table = ROM16(bios->data[record]);
-		if (!table)
-			continue;
-		entry = ROM32(bios->data[table]);
-
-		if (match_link) {
-			v = (entry & 0x00c00000) >> 22;
-			if (!(v & dcbent->sorconf.link))
-				continue;
+		if (hash & 0x00c00000) {
+			if (!(hash & (dcb->sorconf.link << 22)))
+				return false;
 		}
-
-		v = (entry & 0x000f0000) >> 16;
-		if (!(v & dcbent->or))
-			continue;
-
-		v = (entry & 0x000000f0) >> 4;
-		if (v != dcbent->location)
-			continue;
-
-		v = (entry & 0x0000000f);
-		if (v != dcbent->type)
-			continue;
-
-		return &bios->data[table];
-	}
-
-	return NULL;
-}
-
-void *
-nouveau_bios_dp_table(struct drm_device *dev, struct dcb_entry *dcbent,
-		      int *length)
-{
-	struct drm_nouveau_private *dev_priv = dev->dev_private;
-	struct nvbios *bios = &dev_priv->vbios;
-	uint8_t *table;
-
-	if (!bios->display.dp_table_ptr) {
-		NV_ERROR(dev, "No pointer to DisplayPort table\n");
-		return NULL;
-	}
-	table = &bios->data[bios->display.dp_table_ptr];
-
-	if (table[0] != 0x20 && table[0] != 0x21) {
-		NV_ERROR(dev, "DisplayPort table version 0x%02x unknown\n",
-			 table[0]);
-		return NULL;
+	default:
+		return true;
 	}
-
-	*length = table[4];
-	return bios_output_config_match(dev, dcbent,
-					bios->display.dp_table_ptr + table[1],
-					table[2], table[3], table[0] >= 0x21);
 }
 
 int
-nouveau_bios_run_display_table(struct drm_device *dev, struct dcb_entry *dcbent,
-			       uint32_t sub, int pxclk)
+nouveau_bios_run_display_table(struct drm_device *dev, u16 type, int pclk,
+			       struct dcb_entry *dcbent, int crtc)
 {
 	/*
 	 * The display script table is located by the BIT 'U' table.
@@ -4498,7 +4484,7 @@ nouveau_bios_run_display_table(struct drm_device *dev, struct dcb_entry *dcbent,
 	uint8_t *table = &bios->data[bios->display.script_table_ptr];
 	uint8_t *otable = NULL;
 	uint16_t script;
-	int i = 0;
+	int i;
 
 	if (!bios->display.script_table_ptr) {
 		NV_ERROR(dev, "No pointer to output script table\n");
@@ -4550,30 +4536,33 @@ nouveau_bios_run_display_table(struct drm_device *dev, struct dcb_entry *dcbent,
 
 	NV_DEBUG_KMS(dev, "Searching for output entry for %d %d %d\n",
 			dcbent->type, dcbent->location, dcbent->or);
-	otable = bios_output_config_match(dev, dcbent, table[1] +
-					  bios->display.script_table_ptr,
-					  table[2], table[3], table[0] >= 0x21);
+	for (i = 0; i < table[3]; i++) {
+		otable = ROMPTR(bios, table[table[1] + (i * table[2])]);
+		if (otable && bios_encoder_match(dcbent, ROM32(otable[0])))
+			break;
+	}
+
 	if (!otable) {
 		NV_DEBUG_KMS(dev, "failed to match any output table\n");
 		return 1;
 	}
 
-	if (pxclk < -2 || pxclk > 0) {
+	if (pclk < -2 || pclk > 0) {
 		/* Try to find matching script table entry */
 		for (i = 0; i < otable[5]; i++) {
-			if (ROM16(otable[table[4] + i*6]) == sub)
+			if (ROM16(otable[table[4] + i*6]) == type)
 				break;
 		}
 
 		if (i == otable[5]) {
 			NV_ERROR(dev, "Table 0x%04x not found for %d/%d, "
 				      "using first\n",
-				 sub, dcbent->type, dcbent->or);
+				 type, dcbent->type, dcbent->or);
 			i = 0;
 		}
 	}
 
-	if (pxclk == 0) {
+	if (pclk == 0) {
 		script = ROM16(otable[6]);
 		if (!script) {
 			NV_DEBUG_KMS(dev, "output script 0 not found\n");
@@ -4581,9 +4570,9 @@ nouveau_bios_run_display_table(struct drm_device *dev, struct dcb_entry *dcbent,
 		}
 
 		NV_DEBUG_KMS(dev, "0x%04X: parsing output script 0\n", script);
-		nouveau_bios_run_init_table(dev, script, dcbent);
+		nouveau_bios_run_init_table(dev, script, dcbent, crtc);
 	} else
-	if (pxclk == -1) {
+	if (pclk == -1) {
 		script = ROM16(otable[8]);
 		if (!script) {
 			NV_DEBUG_KMS(dev, "output script 1 not found\n");
@@ -4591,9 +4580,9 @@ nouveau_bios_run_display_table(struct drm_device *dev, struct dcb_entry *dcbent,
 		}
 
 		NV_DEBUG_KMS(dev, "0x%04X: parsing output script 1\n", script);
-		nouveau_bios_run_init_table(dev, script, dcbent);
+		nouveau_bios_run_init_table(dev, script, dcbent, crtc);
 	} else
-	if (pxclk == -2) {
+	if (pclk == -2) {
 		if (table[4] >= 12)
 			script = ROM16(otable[10]);
 		else
@@ -4604,31 +4593,31 @@ nouveau_bios_run_display_table(struct drm_device *dev, struct dcb_entry *dcbent,
 		}
 
 		NV_DEBUG_KMS(dev, "0x%04X: parsing output script 2\n", script);
-		nouveau_bios_run_init_table(dev, script, dcbent);
+		nouveau_bios_run_init_table(dev, script, dcbent, crtc);
 	} else
-	if (pxclk > 0) {
+	if (pclk > 0) {
 		script = ROM16(otable[table[4] + i*6 + 2]);
 		if (script)
-			script = clkcmptable(bios, script, pxclk);
+			script = clkcmptable(bios, script, pclk);
 		if (!script) {
 			NV_DEBUG_KMS(dev, "clock script 0 not found\n");
 			return 1;
 		}
 
 		NV_DEBUG_KMS(dev, "0x%04X: parsing clock script 0\n", script);
-		nouveau_bios_run_init_table(dev, script, dcbent);
+		nouveau_bios_run_init_table(dev, script, dcbent, crtc);
 	} else
-	if (pxclk < 0) {
+	if (pclk < 0) {
 		script = ROM16(otable[table[4] + i*6 + 4]);
 		if (script)
-			script = clkcmptable(bios, script, -pxclk);
+			script = clkcmptable(bios, script, -pclk);
 		if (!script) {
 			NV_DEBUG_KMS(dev, "clock script 1 not found\n");
 			return 1;
 		}
 
 		NV_DEBUG_KMS(dev, "0x%04X: parsing clock script 1\n", script);
-		nouveau_bios_run_init_table(dev, script, dcbent);
+		nouveau_bios_run_init_table(dev, script, dcbent, crtc);
 	}
 
 	return 0;
@@ -5478,14 +5467,6 @@ parse_bit_U_tbl_entry(struct drm_device *dev, struct nvbios *bios,
 	return 0;
 }
 
-static int
-parse_bit_displayport_tbl_entry(struct drm_device *dev, struct nvbios *bios,
-				struct bit_entry *bitentry)
-{
-	bios->display.dp_table_ptr = ROM16(bios->data[bitentry->offset]);
-	return 0;
-}
-
 struct bit_table {
 	const char id;
 	int (* const parse_fn)(struct drm_device *, struct nvbios *, struct bit_entry *);
@@ -5559,7 +5540,6 @@ parse_bit_structure(struct nvbios *bios, const uint16_t bitoffset)
 	parse_bit_table(bios, bitoffset, &BIT_TABLE('L', lvds));
 	parse_bit_table(bios, bitoffset, &BIT_TABLE('T', tmds));
 	parse_bit_table(bios, bitoffset, &BIT_TABLE('U', U));
-	parse_bit_table(bios, bitoffset, &BIT_TABLE('d', displayport));
 
 	return 0;
 }
@@ -5884,9 +5864,15 @@ parse_dcb_gpio_table(struct nvbios *bios)
 			}
 
 			e->line = (e->entry & 0x0000001f) >> 0;
-			e->state_default = (e->entry & 0x01000000) >> 24;
-			e->state[0] = (e->entry & 0x18000000) >> 27;
-			e->state[1] = (e->entry & 0x60000000) >> 29;
+			if (gpio[0] == 0x40) {
+				e->state_default = (e->entry & 0x01000000) >> 24;
+				e->state[0] = (e->entry & 0x18000000) >> 27;
+				e->state[1] = (e->entry & 0x60000000) >> 29;
+			} else {
+				e->state_default = (e->entry & 0x00000080) >> 7;
+				e->state[0] = (entry[4] >> 4) & 3;
+				e->state[1] = (entry[4] >> 6) & 3;
+			}
 		}
 	}
 
@@ -6156,7 +6142,14 @@ parse_dcb20_entry(struct drm_device *dev, struct dcb_table *dcb,
 	}
 	case OUTPUT_DP:
 		entry->dpconf.sor.link = (conf & 0x00000030) >> 4;
-		entry->dpconf.link_bw = (conf & 0x00e00000) >> 21;
+		switch ((conf & 0x00e00000) >> 21) {
+		case 0:
+			entry->dpconf.link_bw = 162000;
+			break;
+		default:
+			entry->dpconf.link_bw = 270000;
+			break;
+		}
 		switch ((conf & 0x0f000000) >> 24) {
 		case 0xf:
 			entry->dpconf.link_nr = 4;
@@ -6769,7 +6762,7 @@ uint8_t *nouveau_bios_embedded_edid(struct drm_device *dev)
 
 void
 nouveau_bios_run_init_table(struct drm_device *dev, uint16_t table,
-			    struct dcb_entry *dcbent)
+			    struct dcb_entry *dcbent, int crtc)
 {
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
 	struct nvbios *bios = &dev_priv->vbios;
@@ -6777,11 +6770,22 @@ nouveau_bios_run_init_table(struct drm_device *dev, uint16_t table,
 
 	spin_lock_bh(&bios->lock);
 	bios->display.output = dcbent;
+	bios->display.crtc = crtc;
 	parse_init_table(bios, table, &iexec);
 	bios->display.output = NULL;
 	spin_unlock_bh(&bios->lock);
 }
 
+void
+nouveau_bios_init_exec(struct drm_device *dev, uint16_t table)
+{
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct nvbios *bios = &dev_priv->vbios;
+	struct init_exec iexec = { true, false };
+
+	parse_init_table(bios, table, &iexec);
+}
+
 static bool NVInitVBIOS(struct drm_device *dev)
 {
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
@@ -6863,9 +6867,8 @@ nouveau_run_vbios_init(struct drm_device *dev)
 
 	if (dev_priv->card_type >= NV_50) {
 		for (i = 0; i < bios->dcb.entries; i++) {
-			nouveau_bios_run_display_table(dev,
-						       &bios->dcb.entry[i],
-						       0, 0);
+			nouveau_bios_run_display_table(dev, 0, 0,
+						       &bios->dcb.entry[i], -1);
 		}
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_bios.h b/drivers/gpu/drm/nouveau/nouveau_bios.h
index 050c314119d..8adb69e4a6b 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bios.h
+++ b/drivers/gpu/drm/nouveau/nouveau_bios.h
@@ -289,8 +289,8 @@ struct nvbios {
 
 	struct {
 		struct dcb_entry *output;
+		int crtc;
 		uint16_t script_table_ptr;
-		uint16_t dp_table_ptr;
 	} display;
 
 	struct {
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c
index e87e24b9c0a..424dff5d0ab 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.c
@@ -956,7 +956,7 @@ nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem)
 			break;
 		}
 
-		if (dev_priv->card_type == NV_C0)
+		if (dev_priv->card_type >= NV_C0)
 			page_shift = node->page_shift;
 		else
 			page_shift = 12;
diff --git a/drivers/gpu/drm/nouveau/nouveau_channel.c b/drivers/gpu/drm/nouveau/nouveau_channel.c
index b0d753f45bb..a319d5646ea 100644
--- a/drivers/gpu/drm/nouveau/nouveau_channel.c
+++ b/drivers/gpu/drm/nouveau/nouveau_channel.c
@@ -411,13 +411,17 @@ nouveau_ioctl_fifo_alloc(struct drm_device *dev, void *data,
 		return ret;
 	init->channel  = chan->id;
 
-	if (chan->dma.ib_max)
-		init->pushbuf_domains = NOUVEAU_GEM_DOMAIN_VRAM |
-					NOUVEAU_GEM_DOMAIN_GART;
-	else if (chan->pushbuf_bo->bo.mem.mem_type == TTM_PL_VRAM)
+	if (nouveau_vram_pushbuf == 0) {
+		if (chan->dma.ib_max)
+			init->pushbuf_domains = NOUVEAU_GEM_DOMAIN_VRAM |
+						NOUVEAU_GEM_DOMAIN_GART;
+		else if (chan->pushbuf_bo->bo.mem.mem_type == TTM_PL_VRAM)
+			init->pushbuf_domains = NOUVEAU_GEM_DOMAIN_VRAM;
+		else
+			init->pushbuf_domains = NOUVEAU_GEM_DOMAIN_GART;
+	} else {
 		init->pushbuf_domains = NOUVEAU_GEM_DOMAIN_VRAM;
-	else
-		init->pushbuf_domains = NOUVEAU_GEM_DOMAIN_GART;
+	}
 
 	if (dev_priv->card_type < NV_C0) {
 		init->subchan[0].handle = NvM2MF;
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c
index 939d4df0777..e0d275e1c96 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.c
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.c
@@ -39,7 +39,7 @@
 
 static void nouveau_connector_hotplug(void *, int);
 
-static struct nouveau_encoder *
+struct nouveau_encoder *
 find_encoder(struct drm_connector *connector, int type)
 {
 	struct drm_device *dev = connector->dev;
@@ -116,10 +116,6 @@ nouveau_connector_destroy(struct drm_connector *connector)
 				      nouveau_connector_hotplug, connector);
 	}
 
-	if (connector->connector_type == DRM_MODE_CONNECTOR_LVDS ||
-	    connector->connector_type == DRM_MODE_CONNECTOR_eDP)
-		nouveau_backlight_exit(connector);
-
 	kfree(nv_connector->edid);
 	drm_sysfs_connector_remove(connector);
 	drm_connector_cleanup(connector);
@@ -712,11 +708,8 @@ nouveau_connector_mode_valid(struct drm_connector *connector,
 	case OUTPUT_TV:
 		return get_slave_funcs(encoder)->mode_valid(encoder, mode);
 	case OUTPUT_DP:
-		if (nv_encoder->dp.link_bw == DP_LINK_BW_2_7)
-			max_clock = nv_encoder->dp.link_nr * 270000;
-		else
-			max_clock = nv_encoder->dp.link_nr * 162000;
-
+		max_clock  = nv_encoder->dp.link_nr;
+		max_clock *= nv_encoder->dp.link_bw;
 		clock = clock * nouveau_connector_bpp(connector) / 8;
 		break;
 	default:
@@ -871,7 +864,6 @@ nouveau_connector_create(struct drm_device *dev, int index)
 					dev->mode_config.scaling_mode_property,
 					nv_connector->scaling_mode);
 		}
-		connector->polled = DRM_CONNECTOR_POLL_CONNECT;
 		/* fall-through */
 	case DCB_CONNECTOR_TV_0:
 	case DCB_CONNECTOR_TV_1:
@@ -888,27 +880,20 @@ nouveau_connector_create(struct drm_device *dev, int index)
 				dev->mode_config.dithering_mode_property,
 				nv_connector->use_dithering ?
 				DRM_MODE_DITHERING_ON : DRM_MODE_DITHERING_OFF);
-
-		if (connector->connector_type != DRM_MODE_CONNECTOR_LVDS) {
-			if (dev_priv->card_type >= NV_50)
-				connector->polled = DRM_CONNECTOR_POLL_HPD;
-			else
-				connector->polled = DRM_CONNECTOR_POLL_CONNECT;
-		}
 		break;
 	}
 
-	if (pgpio->irq_register) {
+	if (nv_connector->dcb->gpio_tag != 0xff && pgpio->irq_register) {
 		pgpio->irq_register(dev, nv_connector->dcb->gpio_tag,
 				    nouveau_connector_hotplug, connector);
+
+		connector->polled = DRM_CONNECTOR_POLL_HPD;
+	} else {
+		connector->polled = DRM_CONNECTOR_POLL_CONNECT;
 	}
 
 	drm_sysfs_connector_add(connector);
 
-	if (connector->connector_type == DRM_MODE_CONNECTOR_LVDS ||
-	    connector->connector_type == DRM_MODE_CONNECTOR_eDP)
-		nouveau_backlight_init(connector);
-
 	dcb->drm = connector;
 	return dcb->drm;
 
@@ -925,22 +910,13 @@ nouveau_connector_hotplug(void *data, int plugged)
 	struct drm_connector *connector = data;
 	struct drm_device *dev = connector->dev;
 
-	NV_INFO(dev, "%splugged %s\n", plugged ? "" : "un",
-		drm_get_connector_name(connector));
-
-	if (connector->encoder && connector->encoder->crtc &&
-	    connector->encoder->crtc->enabled) {
-		struct nouveau_encoder *nv_encoder = nouveau_encoder(connector->encoder);
-		struct drm_encoder_helper_funcs *helper =
-			connector->encoder->helper_private;
+	NV_DEBUG(dev, "%splugged %s\n", plugged ? "" : "un",
+		 drm_get_connector_name(connector));
 
-		if (nv_encoder->dcb->type == OUTPUT_DP) {
-			if (plugged)
-				helper->dpms(connector->encoder, DRM_MODE_DPMS_ON);
-			else
-				helper->dpms(connector->encoder, DRM_MODE_DPMS_OFF);
-		}
-	}
+	if (plugged)
+		drm_helper_connector_dpms(connector, DRM_MODE_DPMS_ON);
+	else
+		drm_helper_connector_dpms(connector, DRM_MODE_DPMS_OFF);
 
 	drm_helper_hpd_irq_event(dev);
 }
diff --git a/drivers/gpu/drm/nouveau/nouveau_crtc.h b/drivers/gpu/drm/nouveau/nouveau_crtc.h
index cb1ce2a0916..bf8e1289953 100644
--- a/drivers/gpu/drm/nouveau/nouveau_crtc.h
+++ b/drivers/gpu/drm/nouveau/nouveau_crtc.h
@@ -82,14 +82,13 @@ static inline struct drm_crtc *to_drm_crtc(struct nouveau_crtc *crtc)
 }
 
 int nv50_crtc_create(struct drm_device *dev, int index);
-int nv50_cursor_init(struct nouveau_crtc *);
-void nv50_cursor_fini(struct nouveau_crtc *);
 int nv50_crtc_cursor_set(struct drm_crtc *drm_crtc, struct drm_file *file_priv,
 			 uint32_t buffer_handle, uint32_t width,
 			 uint32_t height);
 int nv50_crtc_cursor_move(struct drm_crtc *drm_crtc, int x, int y);
 
 int nv04_cursor_init(struct nouveau_crtc *);
+int nv50_cursor_init(struct nouveau_crtc *);
 
 struct nouveau_connector *
 nouveau_crtc_connector_get(struct nouveau_crtc *crtc);
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index eb514ea2937..ddbabefb427 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -105,9 +105,12 @@ nouveau_framebuffer_init(struct drm_device *dev,
 		if (dev_priv->chipset == 0x50)
 			nv_fb->r_format |= (tile_flags << 8);
 
-		if (!tile_flags)
-			nv_fb->r_pitch = 0x00100000 | fb->pitch;
-		else {
+		if (!tile_flags) {
+			if (dev_priv->card_type < NV_D0)
+				nv_fb->r_pitch = 0x00100000 | fb->pitch;
+			else
+				nv_fb->r_pitch = 0x01000000 | fb->pitch;
+		} else {
 			u32 mode = nvbo->tile_mode;
 			if (dev_priv->card_type >= NV_C0)
 				mode >>= 4;
diff --git a/drivers/gpu/drm/nouveau/nouveau_dp.c b/drivers/gpu/drm/nouveau/nouveau_dp.c
index 7beb82a0315..de5efe71fef 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dp.c
+++ b/drivers/gpu/drm/nouveau/nouveau_dp.c
@@ -28,418 +28,619 @@
 #include "nouveau_i2c.h"
 #include "nouveau_connector.h"
 #include "nouveau_encoder.h"
+#include "nouveau_crtc.h"
+
+/******************************************************************************
+ * aux channel util functions
+ *****************************************************************************/
+#define AUX_DBG(fmt, args...) do {                                             \
+	if (nouveau_reg_debug & NOUVEAU_REG_DEBUG_AUXCH) {                     \
+		NV_PRINTK(KERN_DEBUG, dev, "AUXCH(%d): " fmt, ch, ##args);     \
+	}                                                                      \
+} while (0)
+#define AUX_ERR(fmt, args...) NV_ERROR(dev, "AUXCH(%d): " fmt, ch, ##args)
+
+static void
+auxch_fini(struct drm_device *dev, int ch)
+{
+	nv_mask(dev, 0x00e4e4 + (ch * 0x50), 0x00310000, 0x00000000);
+}
 
 static int
-auxch_rd(struct drm_encoder *encoder, int address, uint8_t *buf, int size)
+auxch_init(struct drm_device *dev, int ch)
 {
-	struct drm_device *dev = encoder->dev;
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_i2c_chan *auxch;
-	int ret;
-
-	auxch = nouveau_i2c_find(dev, nv_encoder->dcb->i2c_index);
-	if (!auxch)
-		return -ENODEV;
-
-	ret = nouveau_dp_auxch(auxch, 9, address, buf, size);
-	if (ret)
-		return ret;
+	const u32 unksel = 1; /* nfi which to use, or if it matters.. */
+	const u32 ureq = unksel ? 0x00100000 : 0x00200000;
+	const u32 urep = unksel ? 0x01000000 : 0x02000000;
+	u32 ctrl, timeout;
+
+	/* wait up to 1ms for any previous transaction to be done... */
+	timeout = 1000;
+	do {
+		ctrl = nv_rd32(dev, 0x00e4e4 + (ch * 0x50));
+		udelay(1);
+		if (!timeout--) {
+			AUX_ERR("begin idle timeout 0x%08x", ctrl);
+			return -EBUSY;
+		}
+	} while (ctrl & 0x03010000);
+
+	/* set some magic, and wait up to 1ms for it to appear */
+	nv_mask(dev, 0x00e4e4 + (ch * 0x50), 0x00300000, ureq);
+	timeout = 1000;
+	do {
+		ctrl = nv_rd32(dev, 0x00e4e4 + (ch * 0x50));
+		udelay(1);
+		if (!timeout--) {
+			AUX_ERR("magic wait 0x%08x\n", ctrl);
+			auxch_fini(dev, ch);
+			return -EBUSY;
+		}
+	} while ((ctrl & 0x03000000) != urep);
 
 	return 0;
 }
 
 static int
-auxch_wr(struct drm_encoder *encoder, int address, uint8_t *buf, int size)
+auxch_tx(struct drm_device *dev, int ch, u8 type, u32 addr, u8 *data, u8 size)
 {
-	struct drm_device *dev = encoder->dev;
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_i2c_chan *auxch;
-	int ret;
+	u32 ctrl, stat, timeout, retries;
+	u32 xbuf[4] = {};
+	int ret, i;
 
-	auxch = nouveau_i2c_find(dev, nv_encoder->dcb->i2c_index);
-	if (!auxch)
-		return -ENODEV;
+	AUX_DBG("%d: 0x%08x %d\n", type, addr, size);
 
-	ret = nouveau_dp_auxch(auxch, 8, address, buf, size);
-	return ret;
-}
+	ret = auxch_init(dev, ch);
+	if (ret)
+		goto out;
 
-static int
-nouveau_dp_lane_count_set(struct drm_encoder *encoder, uint8_t cmd)
-{
-	struct drm_device *dev = encoder->dev;
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	uint32_t tmp;
-	int or = nv_encoder->or, link = !(nv_encoder->dcb->sorconf.link & 1);
-
-	tmp  = nv_rd32(dev, NV50_SOR_DP_CTRL(or, link));
-	tmp &= ~(NV50_SOR_DP_CTRL_ENHANCED_FRAME_ENABLED |
-		 NV50_SOR_DP_CTRL_LANE_MASK);
-	tmp |= ((1 << (cmd & DP_LANE_COUNT_MASK)) - 1) << 16;
-	if (cmd & DP_LANE_COUNT_ENHANCED_FRAME_EN)
-		tmp |= NV50_SOR_DP_CTRL_ENHANCED_FRAME_ENABLED;
-	nv_wr32(dev, NV50_SOR_DP_CTRL(or, link), tmp);
-
-	return auxch_wr(encoder, DP_LANE_COUNT_SET, &cmd, 1);
-}
+	stat = nv_rd32(dev, 0x00e4e8 + (ch * 0x50));
+	if (!(stat & 0x10000000)) {
+		AUX_DBG("sink not detected\n");
+		ret = -ENXIO;
+		goto out;
+	}
 
-static int
-nouveau_dp_link_bw_set(struct drm_encoder *encoder, uint8_t cmd)
-{
-	struct drm_device *dev = encoder->dev;
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	uint32_t tmp;
-	int reg = 0x614300 + (nv_encoder->or * 0x800);
+	if (!(type & 1)) {
+		memcpy(xbuf, data, size);
+		for (i = 0; i < 16; i += 4) {
+			AUX_DBG("wr 0x%08x\n", xbuf[i / 4]);
+			nv_wr32(dev, 0x00e4c0 + (ch * 0x50) + i, xbuf[i / 4]);
+		}
+	}
 
-	tmp  = nv_rd32(dev, reg);
-	tmp &= 0xfff3ffff;
-	if (cmd == DP_LINK_BW_2_7)
-		tmp |= 0x00040000;
-	nv_wr32(dev, reg, tmp);
+	ctrl  = nv_rd32(dev, 0x00e4e4 + (ch * 0x50));
+	ctrl &= ~0x0001f0ff;
+	ctrl |= type << 12;
+	ctrl |= size - 1;
+	nv_wr32(dev, 0x00e4e0 + (ch * 0x50), addr);
+
+	/* retry transaction a number of times on failure... */
+	ret = -EREMOTEIO;
+	for (retries = 0; retries < 32; retries++) {
+		/* reset, and delay a while if this is a retry */
+		nv_wr32(dev, 0x00e4e4 + (ch * 0x50), 0x80000000 | ctrl);
+		nv_wr32(dev, 0x00e4e4 + (ch * 0x50), 0x00000000 | ctrl);
+		if (retries)
+			udelay(400);
+
+		/* transaction request, wait up to 1ms for it to complete */
+		nv_wr32(dev, 0x00e4e4 + (ch * 0x50), 0x00010000 | ctrl);
+
+		timeout = 1000;
+		do {
+			ctrl = nv_rd32(dev, 0x00e4e4 + (ch * 0x50));
+			udelay(1);
+			if (!timeout--) {
+				AUX_ERR("tx req timeout 0x%08x\n", ctrl);
+				goto out;
+			}
+		} while (ctrl & 0x00010000);
 
-	return auxch_wr(encoder, DP_LINK_BW_SET, &cmd, 1);
-}
+		/* read status, and check if transaction completed ok */
+		stat = nv_mask(dev, 0x00e4e8 + (ch * 0x50), 0, 0);
+		if (!(stat & 0x000f0f00)) {
+			ret = 0;
+			break;
+		}
 
-static int
-nouveau_dp_link_train_set(struct drm_encoder *encoder, int pattern)
-{
-	struct drm_device *dev = encoder->dev;
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	uint32_t tmp;
-	uint8_t cmd;
-	int or = nv_encoder->or, link = !(nv_encoder->dcb->sorconf.link & 1);
-	int ret;
+		AUX_DBG("%02d 0x%08x 0x%08x\n", retries, ctrl, stat);
+	}
 
-	tmp  = nv_rd32(dev, NV50_SOR_DP_CTRL(or, link));
-	tmp &= ~NV50_SOR_DP_CTRL_TRAINING_PATTERN;
-	tmp |= (pattern << 24);
-	nv_wr32(dev, NV50_SOR_DP_CTRL(or, link), tmp);
+	if (type & 1) {
+		for (i = 0; i < 16; i += 4) {
+			xbuf[i / 4] = nv_rd32(dev, 0x00e4d0 + (ch * 0x50) + i);
+			AUX_DBG("rd 0x%08x\n", xbuf[i / 4]);
+		}
+		memcpy(data, xbuf, size);
+	}
 
-	ret = auxch_rd(encoder, DP_TRAINING_PATTERN_SET, &cmd, 1);
-	if (ret)
-		return ret;
-	cmd &= ~DP_TRAINING_PATTERN_MASK;
-	cmd |= (pattern & DP_TRAINING_PATTERN_MASK);
-	return auxch_wr(encoder, DP_TRAINING_PATTERN_SET, &cmd, 1);
+out:
+	auxch_fini(dev, ch);
+	return ret;
 }
 
-static int
-nouveau_dp_max_voltage_swing(struct drm_encoder *encoder)
+static u32
+dp_link_bw_get(struct drm_device *dev, int or, int link)
 {
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct drm_device *dev = encoder->dev;
-	struct bit_displayport_encoder_table_entry *dpse;
-	struct bit_displayport_encoder_table *dpe;
-	int i, dpe_headerlen, max_vs = 0;
-
-	dpe = nouveau_bios_dp_table(dev, nv_encoder->dcb, &dpe_headerlen);
-	if (!dpe)
-		return false;
-	dpse = (void *)((char *)dpe + dpe_headerlen);
+	u32 ctrl = nv_rd32(dev, 0x614300 + (or * 0x800));
+	if (!(ctrl & 0x000c0000))
+		return 162000;
+	return 270000;
+}
 
-	for (i = 0; i < dpe_headerlen; i++, dpse++) {
-		if (dpse->vs_level > max_vs)
-			max_vs = dpse->vs_level;
+static int
+dp_lane_count_get(struct drm_device *dev, int or, int link)
+{
+	u32 ctrl = nv_rd32(dev, NV50_SOR_DP_CTRL(or, link));
+	switch (ctrl & 0x000f0000) {
+	case 0x00010000: return 1;
+	case 0x00030000: return 2;
+	default:
+		return 4;
 	}
-
-	return max_vs;
 }
 
-static int
-nouveau_dp_max_pre_emphasis(struct drm_encoder *encoder, int vs)
+void
+nouveau_dp_tu_update(struct drm_device *dev, int or, int link, u32 clk, u32 bpp)
 {
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct drm_device *dev = encoder->dev;
-	struct bit_displayport_encoder_table_entry *dpse;
-	struct bit_displayport_encoder_table *dpe;
-	int i, dpe_headerlen, max_pre = 0;
+	const u32 symbol = 100000;
+	int bestTU = 0, bestVTUi = 0, bestVTUf = 0, bestVTUa = 0;
+	int TU, VTUi, VTUf, VTUa;
+	u64 link_data_rate, link_ratio, unk;
+	u32 best_diff = 64 * symbol;
+	u32 link_nr, link_bw, r;
+
+	/* calculate packed data rate for each lane */
+	link_nr = dp_lane_count_get(dev, or, link);
+	link_data_rate = (clk * bpp / 8) / link_nr;
+
+	/* calculate ratio of packed data rate to link symbol rate */
+	link_bw = dp_link_bw_get(dev, or, link);
+	link_ratio = link_data_rate * symbol;
+	r = do_div(link_ratio, link_bw);
+
+	for (TU = 64; TU >= 32; TU--) {
+		/* calculate average number of valid symbols in each TU */
+		u32 tu_valid = link_ratio * TU;
+		u32 calc, diff;
+
+		/* find a hw representation for the fraction.. */
+		VTUi = tu_valid / symbol;
+		calc = VTUi * symbol;
+		diff = tu_valid - calc;
+		if (diff) {
+			if (diff >= (symbol / 2)) {
+				VTUf = symbol / (symbol - diff);
+				if (symbol - (VTUf * diff))
+					VTUf++;
+
+				if (VTUf <= 15) {
+					VTUa  = 1;
+					calc += symbol - (symbol / VTUf);
+				} else {
+					VTUa  = 0;
+					VTUf  = 1;
+					calc += symbol;
+				}
+			} else {
+				VTUa  = 0;
+				VTUf  = min((int)(symbol / diff), 15);
+				calc += symbol / VTUf;
+			}
 
-	dpe = nouveau_bios_dp_table(dev, nv_encoder->dcb, &dpe_headerlen);
-	if (!dpe)
-		return false;
-	dpse = (void *)((char *)dpe + dpe_headerlen);
+			diff = calc - tu_valid;
+		} else {
+			/* no remainder, but the hw doesn't like the fractional
+			 * part to be zero.  decrement the integer part and
+			 * have the fraction add a whole symbol back
+			 */
+			VTUa = 0;
+			VTUf = 1;
+			VTUi--;
+		}
 
-	for (i = 0; i < dpe_headerlen; i++, dpse++) {
-		if (dpse->vs_level != vs)
-			continue;
+		if (diff < best_diff) {
+			best_diff = diff;
+			bestTU = TU;
+			bestVTUa = VTUa;
+			bestVTUf = VTUf;
+			bestVTUi = VTUi;
+			if (diff == 0)
+				break;
+		}
+	}
 
-		if (dpse->pre_level > max_pre)
-			max_pre = dpse->pre_level;
+	if (!bestTU) {
+		NV_ERROR(dev, "DP: unable to find suitable config\n");
+		return;
 	}
 
-	return max_pre;
+	/* XXX close to vbios numbers, but not right */
+	unk  = (symbol - link_ratio) * bestTU;
+	unk *= link_ratio;
+	r = do_div(unk, symbol);
+	r = do_div(unk, symbol);
+	unk += 6;
+
+	nv_mask(dev, NV50_SOR_DP_CTRL(or, link), 0x000001fc, bestTU << 2);
+	nv_mask(dev, NV50_SOR_DP_SCFG(or, link), 0x010f7f3f, bestVTUa << 24 |
+							     bestVTUf << 16 |
+							     bestVTUi << 8 |
+							     unk);
 }
 
-static bool
-nouveau_dp_link_train_adjust(struct drm_encoder *encoder, uint8_t *config)
+u8 *
+nouveau_dp_bios_data(struct drm_device *dev, struct dcb_entry *dcb, u8 **entry)
 {
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct drm_device *dev = encoder->dev;
-	struct bit_displayport_encoder_table *dpe;
-	int ret, i, dpe_headerlen, vs = 0, pre = 0;
-	uint8_t request[2];
-
-	dpe = nouveau_bios_dp_table(dev, nv_encoder->dcb, &dpe_headerlen);
-	if (!dpe)
-		return false;
-
-	ret = auxch_rd(encoder, DP_ADJUST_REQUEST_LANE0_1, request, 2);
-	if (ret)
-		return false;
-
-	NV_DEBUG_KMS(dev, "\t\tadjust 0x%02x 0x%02x\n", request[0], request[1]);
-
-	/* Keep all lanes at the same level.. */
-	for (i = 0; i < nv_encoder->dp.link_nr; i++) {
-		int lane_req = (request[i >> 1] >> ((i & 1) << 2)) & 0xf;
-		int lane_vs = lane_req & 3;
-		int lane_pre = (lane_req >> 2) & 3;
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct nvbios *bios = &dev_priv->vbios;
+	struct bit_entry d;
+	u8 *table;
+	int i;
+
+	if (bit_table(dev, 'd', &d)) {
+		NV_ERROR(dev, "BIT 'd' table not found\n");
+		return NULL;
+	}
 
-		if (lane_vs > vs)
-			vs = lane_vs;
-		if (lane_pre > pre)
-			pre = lane_pre;
+	if (d.version != 1) {
+		NV_ERROR(dev, "BIT 'd' table version %d unknown\n", d.version);
+		return NULL;
 	}
 
-	if (vs >= nouveau_dp_max_voltage_swing(encoder)) {
-		vs  = nouveau_dp_max_voltage_swing(encoder);
-		vs |= 4;
+	table = ROMPTR(bios, d.data[0]);
+	if (!table) {
+		NV_ERROR(dev, "displayport table pointer invalid\n");
+		return NULL;
 	}
 
-	if (pre >= nouveau_dp_max_pre_emphasis(encoder, vs & 3)) {
-		pre  = nouveau_dp_max_pre_emphasis(encoder, vs & 3);
-		pre |= 4;
+	switch (table[0]) {
+	case 0x20:
+	case 0x21:
+	case 0x30:
+		break;
+	default:
+		NV_ERROR(dev, "displayport table 0x%02x unknown\n", table[0]);
+		return NULL;
 	}
 
-	/* Update the configuration for all lanes.. */
-	for (i = 0; i < nv_encoder->dp.link_nr; i++)
-		config[i] = (pre << 3) | vs;
+	for (i = 0; i < table[3]; i++) {
+		*entry = ROMPTR(bios, table[table[1] + (i * table[2])]);
+		if (*entry && bios_encoder_match(dcb, ROM32((*entry)[0])))
+			return table;
+	}
 
-	return true;
+	NV_ERROR(dev, "displayport encoder table not found\n");
+	return NULL;
 }
 
-static bool
-nouveau_dp_link_train_commit(struct drm_encoder *encoder, uint8_t *config)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct drm_device *dev = encoder->dev;
-	struct bit_displayport_encoder_table_entry *dpse;
-	struct bit_displayport_encoder_table *dpe;
-	int or = nv_encoder->or, link = !(nv_encoder->dcb->sorconf.link & 1);
-	int dpe_headerlen, ret, i;
+/******************************************************************************
+ * link training
+ *****************************************************************************/
+struct dp_state {
+	struct dcb_entry *dcb;
+	u8 *table;
+	u8 *entry;
+	int auxch;
+	int crtc;
+	int or;
+	int link;
+	u8 *dpcd;
+	int link_nr;
+	u32 link_bw;
+	u8  stat[6];
+	u8  conf[4];
+};
 
-	NV_DEBUG_KMS(dev, "\t\tconfig 0x%02x 0x%02x 0x%02x 0x%02x\n",
-		 config[0], config[1], config[2], config[3]);
+static void
+dp_set_link_config(struct drm_device *dev, struct dp_state *dp)
+{
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	int or = dp->or, link = dp->link;
+	u8 *entry, sink[2];
+	u32 dp_ctrl;
+	u16 script;
+
+	NV_DEBUG_KMS(dev, "%d lanes at %d KB/s\n", dp->link_nr, dp->link_bw);
+
+	/* set selected link rate on source */
+	switch (dp->link_bw) {
+	case 270000:
+		nv_mask(dev, 0x614300 + (or * 0x800), 0x000c0000, 0x00040000);
+		sink[0] = DP_LINK_BW_2_7;
+		break;
+	default:
+		nv_mask(dev, 0x614300 + (or * 0x800), 0x000c0000, 0x00000000);
+		sink[0] = DP_LINK_BW_1_62;
+		break;
+	}
 
-	dpe = nouveau_bios_dp_table(dev, nv_encoder->dcb, &dpe_headerlen);
-	if (!dpe)
-		return false;
-	dpse = (void *)((char *)dpe + dpe_headerlen);
+	/* offset +0x0a of each dp encoder table entry is a pointer to another
+	 * table, that has (among other things) pointers to more scripts that
+	 * need to be executed, this time depending on link speed.
+	 */
+	entry = ROMPTR(&dev_priv->vbios, dp->entry[10]);
+	if (entry) {
+		if (dp->table[0] < 0x30) {
+			while (dp->link_bw < (ROM16(entry[0]) * 10))
+				entry += 4;
+			script = ROM16(entry[2]);
+		} else {
+			while (dp->link_bw < (entry[0] * 27000))
+				entry += 3;
+			script = ROM16(entry[1]);
+		}
 
-	for (i = 0; i < dpe->record_nr; i++, dpse++) {
-		if (dpse->vs_level == (config[0] & 3) &&
-		    dpse->pre_level == ((config[0] >> 3) & 3))
-			break;
+		nouveau_bios_run_init_table(dev, script, dp->dcb, dp->crtc);
 	}
-	BUG_ON(i == dpe->record_nr);
-
-	for (i = 0; i < nv_encoder->dp.link_nr; i++) {
-		const int shift[4] = { 16, 8, 0, 24 };
-		uint32_t mask = 0xff << shift[i];
-		uint32_t reg0, reg1, reg2;
-
-		reg0  = nv_rd32(dev, NV50_SOR_DP_UNK118(or, link)) & ~mask;
-		reg0 |= (dpse->reg0 << shift[i]);
-		reg1  = nv_rd32(dev, NV50_SOR_DP_UNK120(or, link)) & ~mask;
-		reg1 |= (dpse->reg1 << shift[i]);
-		reg2  = nv_rd32(dev, NV50_SOR_DP_UNK130(or, link)) & 0xffff00ff;
-		reg2 |= (dpse->reg2 << 8);
-		nv_wr32(dev, NV50_SOR_DP_UNK118(or, link), reg0);
-		nv_wr32(dev, NV50_SOR_DP_UNK120(or, link), reg1);
-		nv_wr32(dev, NV50_SOR_DP_UNK130(or, link), reg2);
+
+	/* configure lane count on the source */
+	dp_ctrl = ((1 << dp->link_nr) - 1) << 16;
+	sink[1] = dp->link_nr;
+	if (dp->dpcd[2] & DP_ENHANCED_FRAME_CAP) {
+		dp_ctrl |= 0x00004000;
+		sink[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
 	}
 
-	ret = auxch_wr(encoder, DP_TRAINING_LANE0_SET, config, 4);
-	if (ret)
-		return false;
+	nv_mask(dev, NV50_SOR_DP_CTRL(or, link), 0x001f4000, dp_ctrl);
 
-	return true;
+	/* inform the sink of the new configuration */
+	auxch_tx(dev, dp->auxch, 8, DP_LINK_BW_SET, sink, 2);
 }
 
-bool
-nouveau_dp_link_train(struct drm_encoder *encoder)
+static void
+dp_set_training_pattern(struct drm_device *dev, struct dp_state *dp, u8 tp)
 {
-	struct drm_device *dev = encoder->dev;
-	struct drm_nouveau_private *dev_priv = dev->dev_private;
-	struct nouveau_gpio_engine *pgpio = &dev_priv->engine.gpio;
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_connector *nv_connector;
-	struct bit_displayport_encoder_table *dpe;
-	int dpe_headerlen;
-	uint8_t config[4], status[3];
-	bool cr_done, cr_max_vs, eq_done, hpd_state;
-	int ret = 0, i, tries, voltage;
+	u8 sink_tp;
 
-	NV_DEBUG_KMS(dev, "link training!!\n");
+	NV_DEBUG_KMS(dev, "training pattern %d\n", tp);
 
-	nv_connector = nouveau_encoder_connector_get(nv_encoder);
-	if (!nv_connector)
-		return false;
+	nv_mask(dev, NV50_SOR_DP_CTRL(dp->or, dp->link), 0x0f000000, tp << 24);
 
-	dpe = nouveau_bios_dp_table(dev, nv_encoder->dcb, &dpe_headerlen);
-	if (!dpe) {
-		NV_ERROR(dev, "SOR-%d: no DP encoder table!\n", nv_encoder->or);
-		return false;
-	}
+	auxch_tx(dev, dp->auxch, 9, DP_TRAINING_PATTERN_SET, &sink_tp, 1);
+	sink_tp &= ~DP_TRAINING_PATTERN_MASK;
+	sink_tp |= tp;
+	auxch_tx(dev, dp->auxch, 8, DP_TRAINING_PATTERN_SET, &sink_tp, 1);
+}
 
-	/* disable hotplug detect, this flips around on some panels during
-	 * link training.
-	 */
-	hpd_state = pgpio->irq_enable(dev, nv_connector->dcb->gpio_tag, false);
+static const u8 nv50_lane_map[] = { 16, 8, 0, 24 };
+static const u8 nvaf_lane_map[] = { 24, 16, 8, 0 };
+
+static int
+dp_link_train_commit(struct drm_device *dev, struct dp_state *dp)
+{
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	u32 mask = 0, drv = 0, pre = 0, unk = 0;
+	const u8 *shifts;
+	int link = dp->link;
+	int or = dp->or;
+	int i;
+
+	if (dev_priv->chipset != 0xaf)
+		shifts = nv50_lane_map;
+	else
+		shifts = nvaf_lane_map;
+
+	for (i = 0; i < dp->link_nr; i++) {
+		u8 *conf = dp->entry + dp->table[4];
+		u8 lane = (dp->stat[4 + (i >> 1)] >> ((i & 1) * 4)) & 0xf;
+		u8 lpre = (lane & 0x0c) >> 2;
+		u8 lvsw = (lane & 0x03) >> 0;
+
+		mask |= 0xff << shifts[i];
+		unk |= 1 << (shifts[i] >> 3);
+
+		dp->conf[i] = (lpre << 3) | lvsw;
+		if (lvsw == DP_TRAIN_VOLTAGE_SWING_1200)
+			dp->conf[i] |= DP_TRAIN_MAX_SWING_REACHED;
+		if (lpre == DP_TRAIN_PRE_EMPHASIS_9_5)
+			dp->conf[i] |= DP_TRAIN_MAX_PRE_EMPHASIS_REACHED;
+
+		NV_DEBUG_KMS(dev, "config lane %d %02x\n", i, dp->conf[i]);
+
+		if (dp->table[0] < 0x30) {
+			u8 *last = conf + (dp->entry[4] * dp->table[5]);
+			while (lvsw != conf[0] || lpre != conf[1]) {
+				conf += dp->table[5];
+				if (conf >= last)
+					return -EINVAL;
+			}
+
+			conf += 2;
+		} else {
+			/* no lookup table anymore, set entries for each
+			 * combination of voltage swing and pre-emphasis
+			 * level allowed by the DP spec.
+			 */
+			switch (lvsw) {
+			case 0: lpre += 0; break;
+			case 1: lpre += 4; break;
+			case 2: lpre += 7; break;
+			case 3: lpre += 9; break;
+			}
+
+			conf = conf + (lpre * dp->table[5]);
+			conf++;
+		}
 
-	if (dpe->script0) {
-		NV_DEBUG_KMS(dev, "SOR-%d: running DP script 0\n", nv_encoder->or);
-		nouveau_bios_run_init_table(dev, le16_to_cpu(dpe->script0),
-					    nv_encoder->dcb);
+		drv |= conf[0] << shifts[i];
+		pre |= conf[1] << shifts[i];
+		unk  = (unk & ~0x0000ff00) | (conf[2] << 8);
 	}
 
-train:
-	cr_done = eq_done = false;
+	nv_mask(dev, NV50_SOR_DP_UNK118(or, link), mask, drv);
+	nv_mask(dev, NV50_SOR_DP_UNK120(or, link), mask, pre);
+	nv_mask(dev, NV50_SOR_DP_UNK130(or, link), 0x0000ff0f, unk);
 
-	/* set link configuration */
-	NV_DEBUG_KMS(dev, "\tbegin train: bw %d, lanes %d\n",
-		 nv_encoder->dp.link_bw, nv_encoder->dp.link_nr);
+	return auxch_tx(dev, dp->auxch, 8, DP_TRAINING_LANE0_SET, dp->conf, 4);
+}
 
-	ret = nouveau_dp_link_bw_set(encoder, nv_encoder->dp.link_bw);
-	if (ret)
-		return false;
+static int
+dp_link_train_update(struct drm_device *dev, struct dp_state *dp, u32 delay)
+{
+	int ret;
 
-	config[0] = nv_encoder->dp.link_nr;
-	if (nv_encoder->dp.dpcd_version >= 0x11 &&
-	    nv_encoder->dp.enhanced_frame)
-		config[0] |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
+	udelay(delay);
 
-	ret = nouveau_dp_lane_count_set(encoder, config[0]);
+	ret = auxch_tx(dev, dp->auxch, 9, DP_LANE0_1_STATUS, dp->stat, 6);
 	if (ret)
-		return false;
+		return ret;
 
-	/* clock recovery */
-	NV_DEBUG_KMS(dev, "\tbegin cr\n");
-	ret = nouveau_dp_link_train_set(encoder, DP_TRAINING_PATTERN_1);
-	if (ret)
-		goto stop;
+	NV_DEBUG_KMS(dev, "status %02x %02x %02x %02x %02x %02x\n",
+		     dp->stat[0], dp->stat[1], dp->stat[2], dp->stat[3],
+		     dp->stat[4], dp->stat[5]);
+	return 0;
+}
 
-	tries = 0;
-	voltage = -1;
-	memset(config, 0x00, sizeof(config));
-	for (;;) {
-		if (!nouveau_dp_link_train_commit(encoder, config))
-			break;
+static int
+dp_link_train_cr(struct drm_device *dev, struct dp_state *dp)
+{
+	bool cr_done = false, abort = false;
+	int voltage = dp->conf[0] & DP_TRAIN_VOLTAGE_SWING_MASK;
+	int tries = 0, i;
 
-		udelay(100);
+	dp_set_training_pattern(dev, dp, DP_TRAINING_PATTERN_1);
 
-		ret = auxch_rd(encoder, DP_LANE0_1_STATUS, status, 2);
-		if (ret)
+	do {
+		if (dp_link_train_commit(dev, dp) ||
+		    dp_link_train_update(dev, dp, 100))
 			break;
-		NV_DEBUG_KMS(dev, "\t\tstatus: 0x%02x 0x%02x\n",
-			 status[0], status[1]);
 
 		cr_done = true;
-		cr_max_vs = false;
-		for (i = 0; i < nv_encoder->dp.link_nr; i++) {
-			int lane = (status[i >> 1] >> ((i & 1) * 4)) & 0xf;
-
+		for (i = 0; i < dp->link_nr; i++) {
+			u8 lane = (dp->stat[i >> 1] >> ((i & 1) * 4)) & 0xf;
 			if (!(lane & DP_LANE_CR_DONE)) {
 				cr_done = false;
-				if (config[i] & DP_TRAIN_MAX_PRE_EMPHASIS_REACHED)
-					cr_max_vs = true;
+				if (dp->conf[i] & DP_TRAIN_MAX_SWING_REACHED)
+					abort = true;
 				break;
 			}
 		}
 
-		if ((config[0] & DP_TRAIN_VOLTAGE_SWING_MASK) != voltage) {
-			voltage = config[0] & DP_TRAIN_VOLTAGE_SWING_MASK;
+		if ((dp->conf[0] & DP_TRAIN_VOLTAGE_SWING_MASK) != voltage) {
+			voltage = dp->conf[0] & DP_TRAIN_VOLTAGE_SWING_MASK;
 			tries = 0;
 		}
+	} while (!cr_done && !abort && ++tries < 5);
 
-		if (cr_done || cr_max_vs || (++tries == 5))
-			break;
-
-		if (!nouveau_dp_link_train_adjust(encoder, config))
-			break;
-	}
-
-	if (!cr_done)
-		goto stop;
+	return cr_done ? 0 : -1;
+}
 
-	/* channel equalisation */
-	NV_DEBUG_KMS(dev, "\tbegin eq\n");
-	ret = nouveau_dp_link_train_set(encoder, DP_TRAINING_PATTERN_2);
-	if (ret)
-		goto stop;
+static int
+dp_link_train_eq(struct drm_device *dev, struct dp_state *dp)
+{
+	bool eq_done, cr_done = true;
+	int tries = 0, i;
 
-	for (tries = 0; tries <= 5; tries++) {
-		udelay(400);
+	dp_set_training_pattern(dev, dp, DP_TRAINING_PATTERN_2);
 
-		ret = auxch_rd(encoder, DP_LANE0_1_STATUS, status, 3);
-		if (ret)
+	do {
+		if (dp_link_train_update(dev, dp, 400))
 			break;
-		NV_DEBUG_KMS(dev, "\t\tstatus: 0x%02x 0x%02x\n",
-			 status[0], status[1]);
 
-		eq_done = true;
-		if (!(status[2] & DP_INTERLANE_ALIGN_DONE))
-			eq_done = false;
-
-		for (i = 0; eq_done && i < nv_encoder->dp.link_nr; i++) {
-			int lane = (status[i >> 1] >> ((i & 1) * 4)) & 0xf;
-
-			if (!(lane & DP_LANE_CR_DONE)) {
+		eq_done = !!(dp->stat[2] & DP_INTERLANE_ALIGN_DONE);
+		for (i = 0; i < dp->link_nr && eq_done; i++) {
+			u8 lane = (dp->stat[i >> 1] >> ((i & 1) * 4)) & 0xf;
+			if (!(lane & DP_LANE_CR_DONE))
 				cr_done = false;
-				break;
-			}
-
 			if (!(lane & DP_LANE_CHANNEL_EQ_DONE) ||
-			    !(lane & DP_LANE_SYMBOL_LOCKED)) {
+			    !(lane & DP_LANE_SYMBOL_LOCKED))
 				eq_done = false;
-				break;
-			}
 		}
 
-		if (eq_done || !cr_done)
+		if (dp_link_train_commit(dev, dp))
 			break;
+	} while (!eq_done && cr_done && ++tries <= 5);
 
-		if (!nouveau_dp_link_train_adjust(encoder, config) ||
-		    !nouveau_dp_link_train_commit(encoder, config))
-			break;
-	}
+	return eq_done ? 0 : -1;
+}
 
-stop:
-	/* end link training */
-	ret = nouveau_dp_link_train_set(encoder, DP_TRAINING_PATTERN_DISABLE);
-	if (ret)
+bool
+nouveau_dp_link_train(struct drm_encoder *encoder, u32 datarate)
+{
+	struct drm_nouveau_private *dev_priv = encoder->dev->dev_private;
+	struct nouveau_gpio_engine *pgpio = &dev_priv->engine.gpio;
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct nouveau_connector *nv_connector =
+		nouveau_encoder_connector_get(nv_encoder);
+	struct drm_device *dev = encoder->dev;
+	struct nouveau_i2c_chan *auxch;
+	const u32 bw_list[] = { 270000, 162000, 0 };
+	const u32 *link_bw = bw_list;
+	struct dp_state dp;
+
+	auxch = nouveau_i2c_find(dev, nv_encoder->dcb->i2c_index);
+	if (!auxch)
 		return false;
 
-	/* retry at a lower setting, if possible */
-	if (!ret && !(eq_done && cr_done)) {
-		NV_DEBUG_KMS(dev, "\twe failed\n");
-		if (nv_encoder->dp.link_bw != DP_LINK_BW_1_62) {
-			NV_DEBUG_KMS(dev, "retry link training at low rate\n");
-			nv_encoder->dp.link_bw = DP_LINK_BW_1_62;
-			goto train;
-		}
+	dp.table = nouveau_dp_bios_data(dev, nv_encoder->dcb, &dp.entry);
+	if (!dp.table)
+		return -EINVAL;
+
+	dp.dcb = nv_encoder->dcb;
+	dp.crtc = nv_crtc->index;
+	dp.auxch = auxch->rd;
+	dp.or = nv_encoder->or;
+	dp.link = !(nv_encoder->dcb->sorconf.link & 1);
+	dp.dpcd = nv_encoder->dp.dpcd;
+
+	/* some sinks toggle hotplug in response to some of the actions
+	 * we take during link training (DP_SET_POWER is one), we need
+	 * to ignore them for the moment to avoid races.
+	 */
+	pgpio->irq_enable(dev, nv_connector->dcb->gpio_tag, false);
+
+	/* enable down-spreading, if possible */
+	if (dp.table[1] >= 16) {
+		u16 script = ROM16(dp.entry[14]);
+		if (nv_encoder->dp.dpcd[3] & 1)
+			script = ROM16(dp.entry[12]);
+
+		nouveau_bios_run_init_table(dev, script, dp.dcb, dp.crtc);
 	}
 
-	if (dpe->script1) {
-		NV_DEBUG_KMS(dev, "SOR-%d: running DP script 1\n", nv_encoder->or);
-		nouveau_bios_run_init_table(dev, le16_to_cpu(dpe->script1),
-					    nv_encoder->dcb);
+	/* execute pre-train script from vbios */
+	nouveau_bios_run_init_table(dev, ROM16(dp.entry[6]), dp.dcb, dp.crtc);
+
+	/* start off at highest link rate supported by encoder and display */
+	while (*link_bw > nv_encoder->dp.link_bw)
+		link_bw++;
+
+	while (link_bw[0]) {
+		/* find minimum required lane count at this link rate */
+		dp.link_nr = nv_encoder->dp.link_nr;
+		while ((dp.link_nr >> 1) * link_bw[0] > datarate)
+			dp.link_nr >>= 1;
+
+		/* drop link rate to minimum with this lane count */
+		while ((link_bw[1] * dp.link_nr) > datarate)
+			link_bw++;
+		dp.link_bw = link_bw[0];
+
+		/* program selected link configuration */
+		dp_set_link_config(dev, &dp);
+
+		/* attempt to train the link at this configuration */
+		memset(dp.stat, 0x00, sizeof(dp.stat));
+		if (!dp_link_train_cr(dev, &dp) &&
+		    !dp_link_train_eq(dev, &dp))
+			break;
+
+		/* retry at lower rate */
+		link_bw++;
 	}
 
-	/* re-enable hotplug detect */
-	pgpio->irq_enable(dev, nv_connector->dcb->gpio_tag, hpd_state);
+	/* finish link training */
+	dp_set_training_pattern(dev, &dp, DP_TRAINING_PATTERN_DISABLE);
 
-	return eq_done;
+	/* execute post-train script from vbios */
+	nouveau_bios_run_init_table(dev, ROM16(dp.entry[8]), dp.dcb, dp.crtc);
+
+	/* re-enable hotplug detect */
+	pgpio->irq_enable(dev, nv_connector->dcb->gpio_tag, true);
+	return true;
 }
 
 bool
@@ -447,31 +648,34 @@ nouveau_dp_detect(struct drm_encoder *encoder)
 {
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
 	struct drm_device *dev = encoder->dev;
-	uint8_t dpcd[4];
+	struct nouveau_i2c_chan *auxch;
+	u8 *dpcd = nv_encoder->dp.dpcd;
 	int ret;
 
-	ret = auxch_rd(encoder, 0x0000, dpcd, 4);
-	if (ret)
+	auxch = nouveau_i2c_find(dev, nv_encoder->dcb->i2c_index);
+	if (!auxch)
 		return false;
 
-	NV_DEBUG_KMS(dev, "encoder: link_bw %d, link_nr %d\n"
-		      "display: link_bw %d, link_nr %d version 0x%02x\n",
-		 nv_encoder->dcb->dpconf.link_bw,
-		 nv_encoder->dcb->dpconf.link_nr,
-		 dpcd[1], dpcd[2] & 0x0f, dpcd[0]);
+	ret = auxch_tx(dev, auxch->rd, 9, DP_DPCD_REV, dpcd, 8);
+	if (ret)
+		return false;
 
-	nv_encoder->dp.dpcd_version = dpcd[0];
+	nv_encoder->dp.link_bw = 27000 * dpcd[1];
+	nv_encoder->dp.link_nr = dpcd[2] & DP_MAX_LANE_COUNT_MASK;
 
-	nv_encoder->dp.link_bw = dpcd[1];
-	if (nv_encoder->dp.link_bw != DP_LINK_BW_1_62 &&
-	    !nv_encoder->dcb->dpconf.link_bw)
-		nv_encoder->dp.link_bw = DP_LINK_BW_1_62;
+	NV_DEBUG_KMS(dev, "display: %dx%d dpcd 0x%02x\n",
+		     nv_encoder->dp.link_nr, nv_encoder->dp.link_bw, dpcd[0]);
+	NV_DEBUG_KMS(dev, "encoder: %dx%d\n",
+		     nv_encoder->dcb->dpconf.link_nr,
+		     nv_encoder->dcb->dpconf.link_bw);
 
-	nv_encoder->dp.link_nr = dpcd[2] & DP_MAX_LANE_COUNT_MASK;
-	if (nv_encoder->dp.link_nr > nv_encoder->dcb->dpconf.link_nr)
+	if (nv_encoder->dcb->dpconf.link_nr < nv_encoder->dp.link_nr)
 		nv_encoder->dp.link_nr = nv_encoder->dcb->dpconf.link_nr;
+	if (nv_encoder->dcb->dpconf.link_bw < nv_encoder->dp.link_bw)
+		nv_encoder->dp.link_bw = nv_encoder->dcb->dpconf.link_bw;
 
-	nv_encoder->dp.enhanced_frame = (dpcd[2] & DP_ENHANCED_FRAME_CAP);
+	NV_DEBUG_KMS(dev, "maximum: %dx%d\n",
+		     nv_encoder->dp.link_nr, nv_encoder->dp.link_bw);
 
 	return true;
 }
@@ -480,105 +684,13 @@ int
 nouveau_dp_auxch(struct nouveau_i2c_chan *auxch, int cmd, int addr,
 		 uint8_t *data, int data_nr)
 {
-	struct drm_device *dev = auxch->dev;
-	uint32_t tmp, ctrl, stat = 0, data32[4] = {};
-	int ret = 0, i, index = auxch->rd;
-
-	NV_DEBUG_KMS(dev, "ch %d cmd %d addr 0x%x len %d\n", index, cmd, addr, data_nr);
-
-	tmp = nv_rd32(dev, NV50_AUXCH_CTRL(auxch->rd));
-	nv_wr32(dev, NV50_AUXCH_CTRL(auxch->rd), tmp | 0x00100000);
-	tmp = nv_rd32(dev, NV50_AUXCH_CTRL(auxch->rd));
-	if (!(tmp & 0x01000000)) {
-		NV_ERROR(dev, "expected bit 24 == 1, got 0x%08x\n", tmp);
-		ret = -EIO;
-		goto out;
-	}
-
-	for (i = 0; i < 3; i++) {
-		tmp = nv_rd32(dev, NV50_AUXCH_STAT(auxch->rd));
-		if (tmp & NV50_AUXCH_STAT_STATE_READY)
-			break;
-		udelay(100);
-	}
-
-	if (i == 3) {
-		ret = -EBUSY;
-		goto out;
-	}
-
-	if (!(cmd & 1)) {
-		memcpy(data32, data, data_nr);
-		for (i = 0; i < 4; i++) {
-			NV_DEBUG_KMS(dev, "wr %d: 0x%08x\n", i, data32[i]);
-			nv_wr32(dev, NV50_AUXCH_DATA_OUT(index, i), data32[i]);
-		}
-	}
-
-	nv_wr32(dev, NV50_AUXCH_ADDR(index), addr);
-	ctrl  = nv_rd32(dev, NV50_AUXCH_CTRL(index));
-	ctrl &= ~(NV50_AUXCH_CTRL_CMD | NV50_AUXCH_CTRL_LEN);
-	ctrl |= (cmd << NV50_AUXCH_CTRL_CMD_SHIFT);
-	ctrl |= ((data_nr - 1) << NV50_AUXCH_CTRL_LEN_SHIFT);
-
-	for (i = 0; i < 16; i++) {
-		nv_wr32(dev, NV50_AUXCH_CTRL(index), ctrl | 0x80000000);
-		nv_wr32(dev, NV50_AUXCH_CTRL(index), ctrl);
-		nv_wr32(dev, NV50_AUXCH_CTRL(index), ctrl | 0x00010000);
-		if (!nv_wait(dev, NV50_AUXCH_CTRL(index),
-			     0x00010000, 0x00000000)) {
-			NV_ERROR(dev, "expected bit 16 == 0, got 0x%08x\n",
-				 nv_rd32(dev, NV50_AUXCH_CTRL(index)));
-			ret = -EBUSY;
-			goto out;
-		}
-
-		udelay(400);
-
-		stat = nv_rd32(dev, NV50_AUXCH_STAT(index));
-		if ((stat & NV50_AUXCH_STAT_REPLY_AUX) !=
-			    NV50_AUXCH_STAT_REPLY_AUX_DEFER)
-			break;
-	}
-
-	if (i == 16) {
-		NV_ERROR(dev, "auxch DEFER too many times, bailing\n");
-		ret = -EREMOTEIO;
-		goto out;
-	}
-
-	if (cmd & 1) {
-		if ((stat & NV50_AUXCH_STAT_COUNT) != data_nr) {
-			ret = -EREMOTEIO;
-			goto out;
-		}
-
-		for (i = 0; i < 4; i++) {
-			data32[i] = nv_rd32(dev, NV50_AUXCH_DATA_IN(index, i));
-			NV_DEBUG_KMS(dev, "rd %d: 0x%08x\n", i, data32[i]);
-		}
-		memcpy(data, data32, data_nr);
-	}
-
-out:
-	tmp = nv_rd32(dev, NV50_AUXCH_CTRL(auxch->rd));
-	nv_wr32(dev, NV50_AUXCH_CTRL(auxch->rd), tmp & ~0x00100000);
-	tmp = nv_rd32(dev, NV50_AUXCH_CTRL(auxch->rd));
-	if (tmp & 0x01000000) {
-		NV_ERROR(dev, "expected bit 24 == 0, got 0x%08x\n", tmp);
-		ret = -EIO;
-	}
-
-	udelay(400);
-
-	return ret ? ret : (stat & NV50_AUXCH_STAT_REPLY);
+	return auxch_tx(auxch->dev, auxch->rd, cmd, addr, data, data_nr);
 }
 
 static int
 nouveau_dp_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs, int num)
 {
 	struct nouveau_i2c_chan *auxch = (struct nouveau_i2c_chan *)adap;
-	struct drm_device *dev = auxch->dev;
 	struct i2c_msg *msg = msgs;
 	int ret, mcnt = num;
 
@@ -602,19 +714,6 @@ nouveau_dp_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs, int num)
 			if (ret < 0)
 				return ret;
 
-			switch (ret & NV50_AUXCH_STAT_REPLY_I2C) {
-			case NV50_AUXCH_STAT_REPLY_I2C_ACK:
-				break;
-			case NV50_AUXCH_STAT_REPLY_I2C_NACK:
-				return -EREMOTEIO;
-			case NV50_AUXCH_STAT_REPLY_I2C_DEFER:
-				udelay(100);
-				continue;
-			default:
-				NV_ERROR(dev, "bad auxch reply: 0x%08x\n", ret);
-				return -EREMOTEIO;
-			}
-
 			ptr += cnt;
 			remaining -= cnt;
 		}
diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.c b/drivers/gpu/drm/nouveau/nouveau_drv.c
index b30ddd8d2e2..c1e01f37b9d 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drv.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drv.c
@@ -41,7 +41,7 @@ int nouveau_agpmode = -1;
 module_param_named(agpmode, nouveau_agpmode, int, 0400);
 
 MODULE_PARM_DESC(modeset, "Enable kernel modesetting");
-static int nouveau_modeset = -1; /* kms */
+int nouveau_modeset = -1;
 module_param_named(modeset, nouveau_modeset, int, 0400);
 
 MODULE_PARM_DESC(vbios, "Override default VBIOS location");
diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h
index d7d51deb34b..29837da1098 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drv.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drv.h
@@ -414,12 +414,13 @@ struct nouveau_gpio_engine {
 };
 
 struct nouveau_pm_voltage_level {
-	u8 voltage;
-	u8 vid;
+	u32 voltage; /* microvolts */
+	u8  vid;
 };
 
 struct nouveau_pm_voltage {
 	bool supported;
+	u8 version;
 	u8 vid_mask;
 
 	struct nouveau_pm_voltage_level *level;
@@ -428,17 +429,48 @@ struct nouveau_pm_voltage {
 
 struct nouveau_pm_memtiming {
 	int id;
-	u32 reg_100220;
-	u32 reg_100224;
-	u32 reg_100228;
-	u32 reg_10022c;
-	u32 reg_100230;
-	u32 reg_100234;
-	u32 reg_100238;
-	u32 reg_10023c;
-	u32 reg_100240;
+	u32 reg_0; /* 0x10f290 on Fermi, 0x100220 for older */
+	u32 reg_1;
+	u32 reg_2;
+	u32 reg_3;
+	u32 reg_4;
+	u32 reg_5;
+	u32 reg_6;
+	u32 reg_7;
+	u32 reg_8;
+	/* To be written to 0x1002c0 */
+	u8 CL;
+	u8 WR;
 };
 
+struct nouveau_pm_tbl_header{
+	u8 version;
+	u8 header_len;
+	u8 entry_cnt;
+	u8 entry_len;
+};
+
+struct nouveau_pm_tbl_entry{
+	u8 tWR;
+	u8 tUNK_1;
+	u8 tCL;
+	u8 tRP;		/* Byte 3 */
+	u8 empty_4;
+	u8 tRAS;	/* Byte 5 */
+	u8 empty_6;
+	u8 tRFC;	/* Byte 7 */
+	u8 empty_8;
+	u8 tRC;		/* Byte 9 */
+	u8 tUNK_10, tUNK_11, tUNK_12, tUNK_13, tUNK_14;
+	u8 empty_15,empty_16,empty_17;
+	u8 tUNK_18, tUNK_19, tUNK_20, tUNK_21;
+};
+
+/* nouveau_mem.c */
+void nv30_mem_timing_entry(struct drm_device *dev, struct nouveau_pm_tbl_header *hdr,
+							struct nouveau_pm_tbl_entry *e, uint8_t magic_number,
+							struct nouveau_pm_memtiming *timing);
+
 #define NOUVEAU_PM_MAX_LEVEL 8
 struct nouveau_pm_level {
 	struct device_attribute dev_attr;
@@ -448,11 +480,19 @@ struct nouveau_pm_level {
 	u32 core;
 	u32 memory;
 	u32 shader;
-	u32 unk05;
-	u32 unk0a;
-
-	u8 voltage;
-	u8 fanspeed;
+	u32 rop;
+	u32 copy;
+	u32 daemon;
+	u32 vdec;
+	u32 unk05;	/* nv50:nva3, roughly.. */
+	u32 unka0;	/* nva3:nvc0 */
+	u32 hub01;	/* nvc0- */
+	u32 hub06;	/* nvc0- */
+	u32 hub07;	/* nvc0- */
+
+	u32 volt_min; /* microvolts */
+	u32 volt_max;
+	u8  fanspeed;
 
 	u16 memscript;
 	struct nouveau_pm_memtiming *timing;
@@ -496,6 +536,11 @@ struct nouveau_pm_engine {
 	void *(*clock_pre)(struct drm_device *, struct nouveau_pm_level *,
 			   u32 id, int khz);
 	void (*clock_set)(struct drm_device *, void *);
+
+	int  (*clocks_get)(struct drm_device *, struct nouveau_pm_level *);
+	void *(*clocks_pre)(struct drm_device *, struct nouveau_pm_level *);
+	void (*clocks_set)(struct drm_device *, void *);
+
 	int (*voltage_get)(struct drm_device *);
 	int (*voltage_set)(struct drm_device *, int voltage);
 	int (*fanspeed_get)(struct drm_device *);
@@ -504,7 +549,7 @@ struct nouveau_pm_engine {
 };
 
 struct nouveau_vram_engine {
-	struct nouveau_mm *mm;
+	struct nouveau_mm mm;
 
 	int  (*init)(struct drm_device *);
 	void (*takedown)(struct drm_device *dev);
@@ -623,6 +668,7 @@ enum nouveau_card_type {
 	NV_40      = 0x40,
 	NV_50      = 0x50,
 	NV_C0      = 0xc0,
+	NV_D0      = 0xd0
 };
 
 struct drm_nouveau_private {
@@ -633,8 +679,8 @@ struct drm_nouveau_private {
 	enum nouveau_card_type card_type;
 	/* exact chipset, derived from NV_PMC_BOOT_0 */
 	int chipset;
-	int stepping;
 	int flags;
+	u32 crystal;
 
 	void __iomem *mmio;
 
@@ -721,7 +767,6 @@ struct drm_nouveau_private {
 	uint64_t vram_size;
 	uint64_t vram_sys_base;
 
-	uint64_t fb_phys;
 	uint64_t fb_available_size;
 	uint64_t fb_mappable_pages;
 	uint64_t fb_aper_free;
@@ -784,6 +829,7 @@ nouveau_bo_ref(struct nouveau_bo *ref, struct nouveau_bo **pnvbo)
 }
 
 /* nouveau_drv.c */
+extern int nouveau_modeset;
 extern int nouveau_agpmode;
 extern int nouveau_duallink;
 extern int nouveau_uscript_lvds;
@@ -824,6 +870,8 @@ extern bool nouveau_wait_eq(struct drm_device *, uint64_t timeout,
 			    uint32_t reg, uint32_t mask, uint32_t val);
 extern bool nouveau_wait_ne(struct drm_device *, uint64_t timeout,
 			    uint32_t reg, uint32_t mask, uint32_t val);
+extern bool nouveau_wait_cb(struct drm_device *, u64 timeout,
+			    bool (*cond)(void *), void *);
 extern bool nouveau_wait_for_idle(struct drm_device *);
 extern int  nouveau_card_init(struct drm_device *);
 
@@ -1006,15 +1054,15 @@ static inline int nouveau_acpi_edid(struct drm_device *dev, struct drm_connector
 
 /* nouveau_backlight.c */
 #ifdef CONFIG_DRM_NOUVEAU_BACKLIGHT
-extern int nouveau_backlight_init(struct drm_connector *);
-extern void nouveau_backlight_exit(struct drm_connector *);
+extern int nouveau_backlight_init(struct drm_device *);
+extern void nouveau_backlight_exit(struct drm_device *);
 #else
-static inline int nouveau_backlight_init(struct drm_connector *dev)
+static inline int nouveau_backlight_init(struct drm_device *dev)
 {
 	return 0;
 }
 
-static inline void nouveau_backlight_exit(struct drm_connector *dev) { }
+static inline void nouveau_backlight_exit(struct drm_device *dev) { }
 #endif
 
 /* nouveau_bios.c */
@@ -1022,7 +1070,8 @@ extern int nouveau_bios_init(struct drm_device *);
 extern void nouveau_bios_takedown(struct drm_device *dev);
 extern int nouveau_run_vbios_init(struct drm_device *);
 extern void nouveau_bios_run_init_table(struct drm_device *, uint16_t table,
-					struct dcb_entry *);
+					struct dcb_entry *, int crtc);
+extern void nouveau_bios_init_exec(struct drm_device *, uint16_t table);
 extern struct dcb_gpio_entry *nouveau_bios_gpio_entry(struct drm_device *,
 						      enum dcb_gpio_tag);
 extern struct dcb_connector_table_entry *
@@ -1030,11 +1079,8 @@ nouveau_bios_connector_entry(struct drm_device *, int index);
 extern u32 get_pll_register(struct drm_device *, enum pll_types);
 extern int get_pll_limits(struct drm_device *, uint32_t limit_match,
 			  struct pll_lims *);
-extern int nouveau_bios_run_display_table(struct drm_device *,
-					  struct dcb_entry *,
-					  uint32_t script, int pxclk);
-extern void *nouveau_bios_dp_table(struct drm_device *, struct dcb_entry *,
-				   int *length);
+extern int nouveau_bios_run_display_table(struct drm_device *, u16 id, int clk,
+					  struct dcb_entry *, int crtc);
 extern bool nouveau_bios_fp_mode(struct drm_device *, struct drm_display_mode *);
 extern uint8_t *nouveau_bios_embedded_edid(struct drm_device *);
 extern int nouveau_bios_parse_lvds_table(struct drm_device *, int pxclk,
@@ -1043,6 +1089,7 @@ extern int run_tmds_table(struct drm_device *, struct dcb_entry *,
 			  int head, int pxclk);
 extern int call_lvds_script(struct drm_device *, struct dcb_entry *, int head,
 			    enum LVDS_script, int pxclk);
+bool bios_encoder_match(struct dcb_entry *, u32 hash);
 
 /* nouveau_ttm.c */
 int nouveau_ttm_global_init(struct drm_nouveau_private *);
@@ -1053,7 +1100,9 @@ int nouveau_ttm_mmap(struct file *, struct vm_area_struct *);
 int nouveau_dp_auxch(struct nouveau_i2c_chan *auxch, int cmd, int addr,
 		     uint8_t *data, int data_nr);
 bool nouveau_dp_detect(struct drm_encoder *);
-bool nouveau_dp_link_train(struct drm_encoder *);
+bool nouveau_dp_link_train(struct drm_encoder *, u32 datarate);
+void nouveau_dp_tu_update(struct drm_device *, int, int, u32, u32);
+u8 *nouveau_dp_bios_data(struct drm_device *, struct dcb_entry *, u8 **);
 
 /* nv04_fb.c */
 extern int  nv04_fb_init(struct drm_device *);
@@ -1179,8 +1228,8 @@ extern int  nva3_copy_create(struct drm_device *dev);
 /* nvc0_copy.c */
 extern int  nvc0_copy_create(struct drm_device *dev, int engine);
 
-/* nv40_mpeg.c */
-extern int  nv40_mpeg_create(struct drm_device *dev);
+/* nv31_mpeg.c */
+extern int  nv31_mpeg_create(struct drm_device *dev);
 
 /* nv50_mpeg.c */
 extern int  nv50_mpeg_create(struct drm_device *dev);
@@ -1265,6 +1314,11 @@ extern int nv04_display_create(struct drm_device *);
 extern int nv04_display_init(struct drm_device *);
 extern void nv04_display_destroy(struct drm_device *);
 
+/* nvd0_display.c */
+extern int nvd0_display_create(struct drm_device *);
+extern int nvd0_display_init(struct drm_device *);
+extern void nvd0_display_destroy(struct drm_device *);
+
 /* nv04_crtc.c */
 extern int nv04_crtc_create(struct drm_device *, int index);
 
@@ -1374,6 +1428,8 @@ int nv50_gpio_init(struct drm_device *dev);
 void nv50_gpio_fini(struct drm_device *dev);
 int nv50_gpio_get(struct drm_device *dev, enum dcb_gpio_tag tag);
 int nv50_gpio_set(struct drm_device *dev, enum dcb_gpio_tag tag, int state);
+int nvd0_gpio_get(struct drm_device *dev, enum dcb_gpio_tag tag);
+int nvd0_gpio_set(struct drm_device *dev, enum dcb_gpio_tag tag, int state);
 int  nv50_gpio_irq_register(struct drm_device *, enum dcb_gpio_tag,
 			    void (*)(void *, int), void *);
 void nv50_gpio_irq_unregister(struct drm_device *, enum dcb_gpio_tag,
@@ -1448,6 +1504,8 @@ static inline void nv_wr08(struct drm_device *dev, unsigned reg, u8 val)
 	nouveau_wait_eq(dev, 2000000000ULL, (reg), (mask), (val))
 #define nv_wait_ne(dev, reg, mask, val) \
 	nouveau_wait_ne(dev, 2000000000ULL, (reg), (mask), (val))
+#define nv_wait_cb(dev, func, data) \
+	nouveau_wait_cb(dev, 2000000000ULL, (func), (data))
 
 /* PRAMIN access */
 static inline u32 nv_ri32(struct drm_device *dev, unsigned offset)
@@ -1514,6 +1572,7 @@ enum {
 	NOUVEAU_REG_DEBUG_RMVIO          = 0x80,
 	NOUVEAU_REG_DEBUG_VGAATTR        = 0x100,
 	NOUVEAU_REG_DEBUG_EVO            = 0x200,
+	NOUVEAU_REG_DEBUG_AUXCH          = 0x400
 };
 
 #define NV_REG_DEBUG(type, dev, fmt, arg...) do { \
diff --git a/drivers/gpu/drm/nouveau/nouveau_encoder.h b/drivers/gpu/drm/nouveau/nouveau_encoder.h
index ae69b61d93d..e5d6e3faff3 100644
--- a/drivers/gpu/drm/nouveau/nouveau_encoder.h
+++ b/drivers/gpu/drm/nouveau/nouveau_encoder.h
@@ -49,17 +49,17 @@ struct nouveau_encoder {
 
 	union {
 		struct {
-			int mc_unknown;
-			uint32_t unk0;
-			uint32_t unk1;
-			int dpcd_version;
+			u8  dpcd[8];
 			int link_nr;
 			int link_bw;
-			bool enhanced_frame;
+			u32 datarate;
 		} dp;
 	};
 };
 
+struct nouveau_encoder *
+find_encoder(struct drm_connector *connector, int type);
+
 static inline struct nouveau_encoder *nouveau_encoder(struct drm_encoder *enc)
 {
 	struct drm_encoder_slave *slave = to_encoder_slave(enc);
@@ -83,21 +83,4 @@ nouveau_encoder_connector_get(struct nouveau_encoder *encoder);
 int nv50_sor_create(struct drm_connector *, struct dcb_entry *);
 int nv50_dac_create(struct drm_connector *, struct dcb_entry *);
 
-struct bit_displayport_encoder_table {
-	uint32_t match;
-	uint8_t  record_nr;
-	uint8_t  unknown;
-	uint16_t script0;
-	uint16_t script1;
-	uint16_t unknown_table;
-} __attribute__ ((packed));
-
-struct bit_displayport_encoder_table_entry {
-	uint8_t vs_level;
-	uint8_t pre_level;
-	uint8_t reg0;
-	uint8_t reg1;
-	uint8_t reg2;
-} __attribute__ ((packed));
-
 #endif /* __NOUVEAU_ENCODER_H__ */
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c
index 8d02d875376..81116cfea27 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.c
@@ -519,7 +519,7 @@ nouveau_fence_channel_init(struct nouveau_channel *chan)
 	if (USE_SEMA(dev) && dev_priv->chipset < 0x84) {
 		struct ttm_mem_reg *mem = &dev_priv->fence.bo->bo.mem;
 
-		ret = nouveau_gpuobj_dma_new(chan, NV_CLASS_DMA_IN_MEMORY,
+		ret = nouveau_gpuobj_dma_new(chan, NV_CLASS_DMA_FROM_MEMORY,
 					     mem->start << PAGE_SHIFT,
 					     mem->size, NV_MEM_ACCESS_RW,
 					     NV_MEM_TARGET_VRAM, &obj);
@@ -530,7 +530,8 @@ nouveau_fence_channel_init(struct nouveau_channel *chan)
 		nouveau_gpuobj_ref(NULL, &obj);
 		if (ret)
 			return ret;
-	} else {
+	} else
+	if (USE_SEMA(dev)) {
 		/* map fence bo into channel's vm */
 		ret = nouveau_bo_vma_add(dev_priv->fence.bo, chan->vm,
 					 &chan->fence.vma);
diff --git a/drivers/gpu/drm/nouveau/nouveau_i2c.c b/drivers/gpu/drm/nouveau/nouveau_i2c.c
index cb389d01432..f6a27fabcfe 100644
--- a/drivers/gpu/drm/nouveau/nouveau_i2c.c
+++ b/drivers/gpu/drm/nouveau/nouveau_i2c.c
@@ -107,6 +107,13 @@ nv4e_i2c_getsda(void *data)
 	return !!((nv_rd32(dev, i2c->rd) >> 16) & 8);
 }
 
+static const uint32_t nv50_i2c_port[] = {
+	0x00e138, 0x00e150, 0x00e168, 0x00e180,
+	0x00e254, 0x00e274, 0x00e764, 0x00e780,
+	0x00e79c, 0x00e7b8
+};
+#define NV50_I2C_PORTS ARRAY_SIZE(nv50_i2c_port)
+
 static int
 nv50_i2c_getscl(void *data)
 {
@@ -130,28 +137,32 @@ static void
 nv50_i2c_setscl(void *data, int state)
 {
 	struct nouveau_i2c_chan *i2c = data;
-	struct drm_device *dev = i2c->dev;
 
-	nv_wr32(dev, i2c->wr, 4 | (i2c->data ? 2 : 0) | (state ? 1 : 0));
+	nv_wr32(i2c->dev, i2c->wr, 4 | (i2c->data ? 2 : 0) | (state ? 1 : 0));
 }
 
 static void
 nv50_i2c_setsda(void *data, int state)
 {
 	struct nouveau_i2c_chan *i2c = data;
-	struct drm_device *dev = i2c->dev;
 
-	nv_wr32(dev, i2c->wr,
-			(nv_rd32(dev, i2c->rd) & 1) | 4 | (state ? 2 : 0));
+	nv_mask(i2c->dev, i2c->wr, 0x00000006, 4 | (state ? 2 : 0));
 	i2c->data = state;
 }
 
-static const uint32_t nv50_i2c_port[] = {
-	0x00e138, 0x00e150, 0x00e168, 0x00e180,
-	0x00e254, 0x00e274, 0x00e764, 0x00e780,
-	0x00e79c, 0x00e7b8
-};
-#define NV50_I2C_PORTS ARRAY_SIZE(nv50_i2c_port)
+static int
+nvd0_i2c_getscl(void *data)
+{
+	struct nouveau_i2c_chan *i2c = data;
+	return !!(nv_rd32(i2c->dev, i2c->rd) & 0x10);
+}
+
+static int
+nvd0_i2c_getsda(void *data)
+{
+	struct nouveau_i2c_chan *i2c = data;
+	return !!(nv_rd32(i2c->dev, i2c->rd) & 0x20);
+}
 
 int
 nouveau_i2c_init(struct drm_device *dev, struct dcb_i2c_entry *entry, int index)
@@ -163,7 +174,8 @@ nouveau_i2c_init(struct drm_device *dev, struct dcb_i2c_entry *entry, int index)
 	if (entry->chan)
 		return -EEXIST;
 
-	if (dev_priv->card_type >= NV_50 && entry->read >= NV50_I2C_PORTS) {
+	if (dev_priv->card_type >= NV_50 &&
+	    dev_priv->card_type <= NV_C0 && entry->read >= NV50_I2C_PORTS) {
 		NV_ERROR(dev, "unknown i2c port %d\n", entry->read);
 		return -EINVAL;
 	}
@@ -192,10 +204,17 @@ nouveau_i2c_init(struct drm_device *dev, struct dcb_i2c_entry *entry, int index)
 	case 5:
 		i2c->bit.setsda = nv50_i2c_setsda;
 		i2c->bit.setscl = nv50_i2c_setscl;
-		i2c->bit.getsda = nv50_i2c_getsda;
-		i2c->bit.getscl = nv50_i2c_getscl;
-		i2c->rd = nv50_i2c_port[entry->read];
-		i2c->wr = i2c->rd;
+		if (dev_priv->card_type < NV_D0) {
+			i2c->bit.getsda = nv50_i2c_getsda;
+			i2c->bit.getscl = nv50_i2c_getscl;
+			i2c->rd = nv50_i2c_port[entry->read];
+			i2c->wr = i2c->rd;
+		} else {
+			i2c->bit.getsda = nvd0_i2c_getsda;
+			i2c->bit.getscl = nvd0_i2c_getscl;
+			i2c->rd = 0x00d014 + (entry->read * 0x20);
+			i2c->wr = i2c->rd;
+		}
 		break;
 	case 6:
 		i2c->rd = entry->read;
@@ -267,7 +286,10 @@ nouveau_i2c_find(struct drm_device *dev, int index)
 			val  = 0xe001;
 		}
 
-		nv_wr32(dev, reg, (nv_rd32(dev, reg) & ~0xf003) | val);
+		/* nfi, but neither auxch or i2c work if it's 1 */
+		nv_mask(dev, reg + 0x0c, 0x00000001, 0x00000000);
+		/* nfi, but switches auxch vs normal i2c */
+		nv_mask(dev, reg + 0x00, 0x0000f003, val);
 	}
 
 	if (!i2c->chan && nouveau_i2c_init(dev, i2c, index))
diff --git a/drivers/gpu/drm/nouveau/nouveau_mem.c b/drivers/gpu/drm/nouveau/nouveau_mem.c
index f9ae2fc3d6f..36bec480770 100644
--- a/drivers/gpu/drm/nouveau/nouveau_mem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_mem.c
@@ -408,8 +408,6 @@ nouveau_mem_vram_init(struct drm_device *dev)
 	if (ret)
 		return ret;
 
-	dev_priv->fb_phys = pci_resource_start(dev->pdev, 1);
-
 	ret = nouveau_ttm_global_init(dev_priv);
 	if (ret)
 		return ret;
@@ -504,35 +502,146 @@ nouveau_mem_gart_init(struct drm_device *dev)
 	return 0;
 }
 
+/* XXX: For now a dummy. More samples required, possibly even a card
+ * Called from nouveau_perf.c */
+void nv30_mem_timing_entry(struct drm_device *dev, struct nouveau_pm_tbl_header *hdr,
+							struct nouveau_pm_tbl_entry *e, uint8_t magic_number,
+							struct nouveau_pm_memtiming *timing) {
+
+	NV_DEBUG(dev,"Timing entry format unknown, please contact nouveau developers");
+}
+
+void nv40_mem_timing_entry(struct drm_device *dev, struct nouveau_pm_tbl_header *hdr,
+							struct nouveau_pm_tbl_entry *e, uint8_t magic_number,
+							struct nouveau_pm_memtiming *timing) {
+
+	timing->reg_0 = (e->tRC << 24 | e->tRFC << 16 | e->tRAS << 8 | e->tRP);
+
+	/* XXX: I don't trust the -1's and +1's... they must come
+	 *      from somewhere! */
+	timing->reg_1 = (e->tWR + 2 + magic_number) << 24 |
+				  1 << 16 |
+				  (e->tUNK_1 + 2 + magic_number) << 8 |
+				  (e->tCL + 2 - magic_number);
+	timing->reg_2 = (magic_number << 24 | e->tUNK_12 << 16 | e->tUNK_11 << 8 | e->tUNK_10);
+	timing->reg_2 |= 0x20200000;
+
+	NV_DEBUG(dev, "Entry %d: 220: %08x %08x %08x\n", timing->id,
+		 timing->reg_0, timing->reg_1,timing->reg_2);
+}
+
+void nv50_mem_timing_entry(struct drm_device *dev, struct bit_entry *P, struct nouveau_pm_tbl_header *hdr,
+							struct nouveau_pm_tbl_entry *e, uint8_t magic_number,struct nouveau_pm_memtiming *timing) {
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+
+	uint8_t unk18 = 1,
+		unk19 = 1,
+		unk20 = 0,
+		unk21 = 0;
+
+	switch (min(hdr->entry_len, (u8) 22)) {
+	case 22:
+		unk21 = e->tUNK_21;
+	case 21:
+		unk20 = e->tUNK_20;
+	case 20:
+		unk19 = e->tUNK_19;
+	case 19:
+		unk18 = e->tUNK_18;
+		break;
+	}
+
+	timing->reg_0 = (e->tRC << 24 | e->tRFC << 16 | e->tRAS << 8 | e->tRP);
+
+	/* XXX: I don't trust the -1's and +1's... they must come
+	 *      from somewhere! */
+	timing->reg_1 = (e->tWR + unk19 + 1 + magic_number) << 24 |
+				  max(unk18, (u8) 1) << 16 |
+				  (e->tUNK_1 + unk19 + 1 + magic_number) << 8;
+	if (dev_priv->chipset == 0xa8) {
+		timing->reg_1 |= (e->tCL - 1);
+	} else {
+		timing->reg_1 |= (e->tCL + 2 - magic_number);
+	}
+	timing->reg_2 = (e->tUNK_12 << 16 | e->tUNK_11 << 8 | e->tUNK_10);
+
+	timing->reg_5 = (e->tRAS << 24 | e->tRC);
+	timing->reg_5 += max(e->tUNK_10, e->tUNK_11) << 16;
+
+	if (P->version == 1) {
+		timing->reg_2 |= magic_number << 24;
+		timing->reg_3 = (0x14 + e->tCL) << 24 |
+						0x16 << 16 |
+						(e->tCL - 1) << 8 |
+						(e->tCL - 1);
+		timing->reg_4 = (nv_rd32(dev,0x10022c) & 0xffff0000) | e->tUNK_13 << 8  | e->tUNK_13;
+		timing->reg_5 |= (e->tCL + 2) << 8;
+		timing->reg_7 = 0x4000202 | (e->tCL - 1) << 16;
+	} else {
+		timing->reg_2 |= (unk19 - 1) << 24;
+		/* XXX: reg_10022c for recentish cards pretty much unknown*/
+		timing->reg_3 = e->tCL - 1;
+		timing->reg_4 = (unk20 << 24 | unk21 << 16 |
+							e->tUNK_13 << 8  | e->tUNK_13);
+		/* XXX: +6? */
+		timing->reg_5 |= (unk19 + 6) << 8;
+
+		/* XXX: reg_10023c currently unknown
+		 * 10023c seen as 06xxxxxx, 0bxxxxxx or 0fxxxxxx */
+		timing->reg_7 = 0x202;
+	}
+
+	NV_DEBUG(dev, "Entry %d: 220: %08x %08x %08x %08x\n", timing->id,
+		 timing->reg_0, timing->reg_1,
+		 timing->reg_2, timing->reg_3);
+	NV_DEBUG(dev, "         230: %08x %08x %08x %08x\n",
+		 timing->reg_4, timing->reg_5,
+		 timing->reg_6, timing->reg_7);
+	NV_DEBUG(dev, "         240: %08x\n", timing->reg_8);
+}
+
+void nvc0_mem_timing_entry(struct drm_device *dev, struct nouveau_pm_tbl_header *hdr,
+							struct nouveau_pm_tbl_entry *e, struct nouveau_pm_memtiming *timing) {
+	timing->reg_0 = (e->tRC << 24 | (e->tRFC & 0x7f) << 17 | e->tRAS << 8 | e->tRP);
+	timing->reg_1 = (nv_rd32(dev,0x10f294) & 0xff000000) | (e->tUNK_11&0x0f) << 20 | (e->tUNK_19 << 7) | (e->tCL & 0x0f);
+	timing->reg_2 = (nv_rd32(dev,0x10f298) & 0xff0000ff) | e->tWR << 16 | e->tUNK_1 << 8;
+	timing->reg_3 = e->tUNK_20 << 9 | e->tUNK_13;
+	timing->reg_4 = (nv_rd32(dev,0x10f2a0) & 0xfff000ff) | e->tUNK_12 << 15;
+	NV_DEBUG(dev, "Entry %d: 290: %08x %08x %08x %08x\n", timing->id,
+		 timing->reg_0, timing->reg_1,
+		 timing->reg_2, timing->reg_3);
+	NV_DEBUG(dev, "         2a0: %08x %08x %08x %08x\n",
+		 timing->reg_4, timing->reg_5,
+		 timing->reg_6, timing->reg_7);
+}
+
+/**
+ * Processes the Memory Timing BIOS table, stores generated
+ * register values
+ * @pre init scripts were run, memtiming regs are initialized
+ */
 void
 nouveau_mem_timing_init(struct drm_device *dev)
 {
-	/* cards < NVC0 only */
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
 	struct nouveau_pm_engine *pm = &dev_priv->engine.pm;
 	struct nouveau_pm_memtimings *memtimings = &pm->memtimings;
 	struct nvbios *bios = &dev_priv->vbios;
 	struct bit_entry P;
-	u8 tUNK_0, tUNK_1, tUNK_2;
-	u8 tRP;		/* Byte 3 */
-	u8 tRAS;	/* Byte 5 */
-	u8 tRFC;	/* Byte 7 */
-	u8 tRC;		/* Byte 9 */
-	u8 tUNK_10, tUNK_11, tUNK_12, tUNK_13, tUNK_14;
-	u8 tUNK_18, tUNK_19, tUNK_20, tUNK_21;
-	u8 magic_number = 0; /* Yeah... sorry*/
-	u8 *mem = NULL, *entry;
-	int i, recordlen, entries;
+	struct nouveau_pm_tbl_header *hdr = NULL;
+	uint8_t magic_number;
+	u8 *entry;
+	int i;
 
 	if (bios->type == NVBIOS_BIT) {
 		if (bit_table(dev, 'P', &P))
 			return;
 
 		if (P.version == 1)
-			mem = ROMPTR(bios, P.data[4]);
+			hdr = (struct nouveau_pm_tbl_header *) ROMPTR(bios, P.data[4]);
 		else
 		if (P.version == 2)
-			mem = ROMPTR(bios, P.data[8]);
+			hdr = (struct nouveau_pm_tbl_header *) ROMPTR(bios, P.data[8]);
 		else {
 			NV_WARN(dev, "unknown mem for BIT P %d\n", P.version);
 		}
@@ -541,150 +650,56 @@ nouveau_mem_timing_init(struct drm_device *dev)
 		return;
 	}
 
-	if (!mem) {
+	if (!hdr) {
 		NV_DEBUG(dev, "memory timing table pointer invalid\n");
 		return;
 	}
 
-	if (mem[0] != 0x10) {
-		NV_WARN(dev, "memory timing table 0x%02x unknown\n", mem[0]);
+	if (hdr->version != 0x10) {
+		NV_WARN(dev, "memory timing table 0x%02x unknown\n", hdr->version);
 		return;
 	}
 
 	/* validate record length */
-	entries   = mem[2];
-	recordlen = mem[3];
-	if (recordlen < 15) {
-		NV_ERROR(dev, "mem timing table length unknown: %d\n", mem[3]);
+	if (hdr->entry_len < 15) {
+		NV_ERROR(dev, "mem timing table length unknown: %d\n", hdr->entry_len);
 		return;
 	}
 
 	/* parse vbios entries into common format */
 	memtimings->timing =
-		kcalloc(entries, sizeof(*memtimings->timing), GFP_KERNEL);
+		kcalloc(hdr->entry_cnt, sizeof(*memtimings->timing), GFP_KERNEL);
 	if (!memtimings->timing)
 		return;
 
 	/* Get "some number" from the timing reg for NV_40 and NV_50
-	 * Used in calculations later */
-	if (dev_priv->card_type >= NV_40 && dev_priv->chipset < 0x98) {
+	 * Used in calculations later... source unknown */
+	magic_number = 0;
+	if (P.version == 1) {
 		magic_number = (nv_rd32(dev, 0x100228) & 0x0f000000) >> 24;
 	}
 
-	entry = mem + mem[1];
-	for (i = 0; i < entries; i++, entry += recordlen) {
+	entry = (u8*) hdr + hdr->header_len;
+	for (i = 0; i < hdr->entry_cnt; i++, entry += hdr->entry_len) {
 		struct nouveau_pm_memtiming *timing = &pm->memtimings.timing[i];
 		if (entry[0] == 0)
 			continue;
 
-		tUNK_18 = 1;
-		tUNK_19 = 1;
-		tUNK_20 = 0;
-		tUNK_21 = 0;
-		switch (min(recordlen, 22)) {
-		case 22:
-			tUNK_21 = entry[21];
-		case 21:
-			tUNK_20 = entry[20];
-		case 20:
-			tUNK_19 = entry[19];
-		case 19:
-			tUNK_18 = entry[18];
-		default:
-			tUNK_0  = entry[0];
-			tUNK_1  = entry[1];
-			tUNK_2  = entry[2];
-			tRP     = entry[3];
-			tRAS    = entry[5];
-			tRFC    = entry[7];
-			tRC     = entry[9];
-			tUNK_10 = entry[10];
-			tUNK_11 = entry[11];
-			tUNK_12 = entry[12];
-			tUNK_13 = entry[13];
-			tUNK_14 = entry[14];
-			break;
-		}
-
-		timing->reg_100220 = (tRC << 24 | tRFC << 16 | tRAS << 8 | tRP);
-
-		/* XXX: I don't trust the -1's and +1's... they must come
-		 *      from somewhere! */
-		timing->reg_100224 = (tUNK_0 + tUNK_19 + 1 + magic_number) << 24 |
-				      max(tUNK_18, (u8) 1) << 16 |
-				      (tUNK_1 + tUNK_19 + 1 + magic_number) << 8;
-		if (dev_priv->chipset == 0xa8) {
-			timing->reg_100224 |= (tUNK_2 - 1);
-		} else {
-			timing->reg_100224 |= (tUNK_2 + 2 - magic_number);
-		}
-
-		timing->reg_100228 = (tUNK_12 << 16 | tUNK_11 << 8 | tUNK_10);
-		if (dev_priv->chipset >= 0xa3 && dev_priv->chipset < 0xaa)
-			timing->reg_100228 |= (tUNK_19 - 1) << 24;
-		else
-			timing->reg_100228 |= magic_number << 24;
-
-		if (dev_priv->card_type == NV_40) {
-			/* NV40: don't know what the rest of the regs are..
-			 * And don't need to know either */
-			timing->reg_100228 |= 0x20200000;
-		} else if (dev_priv->card_type >= NV_50) {
-			if (dev_priv->chipset < 0x98 ||
-			    (dev_priv->chipset == 0x98 &&
-			     dev_priv->stepping <= 0xa1)) {
-				timing->reg_10022c = (0x14 + tUNK_2) << 24 |
-						     0x16 << 16 |
-						     (tUNK_2 - 1) << 8 |
-						     (tUNK_2 - 1);
-			} else {
-				/* XXX: reg_10022c for recentish cards */
-				timing->reg_10022c = tUNK_2 - 1;
-			}
-
-			timing->reg_100230 = (tUNK_20 << 24 | tUNK_21 << 16 |
-						  tUNK_13 << 8  | tUNK_13);
-
-			timing->reg_100234 = (tRAS << 24 | tRC);
-			timing->reg_100234 += max(tUNK_10, tUNK_11) << 16;
-
-			if (dev_priv->chipset < 0x98 ||
-			    (dev_priv->chipset == 0x98 &&
-			     dev_priv->stepping <= 0xa1)) {
-				timing->reg_100234 |= (tUNK_2 + 2) << 8;
-			} else {
-				/* XXX: +6? */
-				timing->reg_100234 |= (tUNK_19 + 6) << 8;
-			}
-
-			/* XXX; reg_100238
-			 * reg_100238: 0x00?????? */
-			timing->reg_10023c = 0x202;
-			if (dev_priv->chipset < 0x98 ||
-			    (dev_priv->chipset == 0x98 &&
-			     dev_priv->stepping <= 0xa1)) {
-				timing->reg_10023c |= 0x4000000 | (tUNK_2 - 1) << 16;
-			} else {
-				/* XXX: reg_10023c
-				 * currently unknown
-				 * 10023c seen as 06xxxxxx, 0bxxxxxx or 0fxxxxxx */
-			}
-
-			/* XXX: reg_100240? */
-		}
 		timing->id = i;
-
-		NV_DEBUG(dev, "Entry %d: 220: %08x %08x %08x %08x\n", i,
-			 timing->reg_100220, timing->reg_100224,
-			 timing->reg_100228, timing->reg_10022c);
-		NV_DEBUG(dev, "         230: %08x %08x %08x %08x\n",
-			 timing->reg_100230, timing->reg_100234,
-			 timing->reg_100238, timing->reg_10023c);
-		NV_DEBUG(dev, "         240: %08x\n", timing->reg_100240);
+		timing->WR = entry[0];
+		timing->CL = entry[2];
+
+		if(dev_priv->card_type <= NV_40) {
+			nv40_mem_timing_entry(dev,hdr,(struct nouveau_pm_tbl_entry*) entry,magic_number,&pm->memtimings.timing[i]);
+		} else if(dev_priv->card_type == NV_50){
+			nv50_mem_timing_entry(dev,&P,hdr,(struct nouveau_pm_tbl_entry*) entry,magic_number,&pm->memtimings.timing[i]);
+		} else if(dev_priv->card_type == NV_C0) {
+			nvc0_mem_timing_entry(dev,hdr,(struct nouveau_pm_tbl_entry*) entry,&pm->memtimings.timing[i]);
+		}
 	}
 
-	memtimings->nr_timing = entries;
-	memtimings->supported = (dev_priv->chipset <= 0x98);
+	memtimings->nr_timing = hdr->entry_cnt;
+	memtimings->supported = P.version == 1;
 }
 
 void
@@ -693,7 +708,10 @@ nouveau_mem_timing_fini(struct drm_device *dev)
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
 	struct nouveau_pm_memtimings *mem = &dev_priv->engine.pm.memtimings;
 
-	kfree(mem->timing);
+	if(mem->timing) {
+		kfree(mem->timing);
+		mem->timing = NULL;
+	}
 }
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nouveau_mm.c b/drivers/gpu/drm/nouveau/nouveau_mm.c
index 1640dec3b82..b29ffb3d140 100644
--- a/drivers/gpu/drm/nouveau/nouveau_mm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_mm.c
@@ -27,7 +27,7 @@
 #include "nouveau_mm.h"
 
 static inline void
-region_put(struct nouveau_mm *rmm, struct nouveau_mm_node *a)
+region_put(struct nouveau_mm *mm, struct nouveau_mm_node *a)
 {
 	list_del(&a->nl_entry);
 	list_del(&a->fl_entry);
@@ -35,7 +35,7 @@ region_put(struct nouveau_mm *rmm, struct nouveau_mm_node *a)
 }
 
 static struct nouveau_mm_node *
-region_split(struct nouveau_mm *rmm, struct nouveau_mm_node *a, u32 size)
+region_split(struct nouveau_mm *mm, struct nouveau_mm_node *a, u32 size)
 {
 	struct nouveau_mm_node *b;
 
@@ -57,33 +57,33 @@ region_split(struct nouveau_mm *rmm, struct nouveau_mm_node *a, u32 size)
 	return b;
 }
 
-#define node(root, dir) ((root)->nl_entry.dir == &rmm->nodes) ? NULL : \
+#define node(root, dir) ((root)->nl_entry.dir == &mm->nodes) ? NULL : \
 	list_entry((root)->nl_entry.dir, struct nouveau_mm_node, nl_entry)
 
 void
-nouveau_mm_put(struct nouveau_mm *rmm, struct nouveau_mm_node *this)
+nouveau_mm_put(struct nouveau_mm *mm, struct nouveau_mm_node *this)
 {
 	struct nouveau_mm_node *prev = node(this, prev);
 	struct nouveau_mm_node *next = node(this, next);
 
-	list_add(&this->fl_entry, &rmm->free);
+	list_add(&this->fl_entry, &mm->free);
 	this->type = 0;
 
 	if (prev && prev->type == 0) {
 		prev->length += this->length;
-		region_put(rmm, this);
+		region_put(mm, this);
 		this = prev;
 	}
 
 	if (next && next->type == 0) {
 		next->offset  = this->offset;
 		next->length += this->length;
-		region_put(rmm, this);
+		region_put(mm, this);
 	}
 }
 
 int
-nouveau_mm_get(struct nouveau_mm *rmm, int type, u32 size, u32 size_nc,
+nouveau_mm_get(struct nouveau_mm *mm, int type, u32 size, u32 size_nc,
 	       u32 align, struct nouveau_mm_node **pnode)
 {
 	struct nouveau_mm_node *prev, *this, *next;
@@ -92,17 +92,17 @@ nouveau_mm_get(struct nouveau_mm *rmm, int type, u32 size, u32 size_nc,
 	u32 splitoff;
 	u32 s, e;
 
-	list_for_each_entry(this, &rmm->free, fl_entry) {
+	list_for_each_entry(this, &mm->free, fl_entry) {
 		e = this->offset + this->length;
 		s = this->offset;
 
 		prev = node(this, prev);
 		if (prev && prev->type != type)
-			s = roundup(s, rmm->block_size);
+			s = roundup(s, mm->block_size);
 
 		next = node(this, next);
 		if (next && next->type != type)
-			e = rounddown(e, rmm->block_size);
+			e = rounddown(e, mm->block_size);
 
 		s  = (s + align_mask) & ~align_mask;
 		e &= ~align_mask;
@@ -110,10 +110,10 @@ nouveau_mm_get(struct nouveau_mm *rmm, int type, u32 size, u32 size_nc,
 			continue;
 
 		splitoff = s - this->offset;
-		if (splitoff && !region_split(rmm, this, splitoff))
+		if (splitoff && !region_split(mm, this, splitoff))
 			return -ENOMEM;
 
-		this = region_split(rmm, this, min(size, e - s));
+		this = region_split(mm, this, min(size, e - s));
 		if (!this)
 			return -ENOMEM;
 
@@ -127,52 +127,49 @@ nouveau_mm_get(struct nouveau_mm *rmm, int type, u32 size, u32 size_nc,
 }
 
 int
-nouveau_mm_init(struct nouveau_mm **prmm, u32 offset, u32 length, u32 block)
+nouveau_mm_init(struct nouveau_mm *mm, u32 offset, u32 length, u32 block)
 {
-	struct nouveau_mm *rmm;
-	struct nouveau_mm_node *heap;
+	struct nouveau_mm_node *node;
+
+	if (block) {
+		mutex_init(&mm->mutex);
+		INIT_LIST_HEAD(&mm->nodes);
+		INIT_LIST_HEAD(&mm->free);
+		mm->block_size = block;
+		mm->heap_nodes = 0;
+	}
 
-	heap = kzalloc(sizeof(*heap), GFP_KERNEL);
-	if (!heap)
+	node = kzalloc(sizeof(*node), GFP_KERNEL);
+	if (!node)
 		return -ENOMEM;
-	heap->offset = roundup(offset, block);
-	heap->length = rounddown(offset + length, block) - heap->offset;
+	node->offset = roundup(offset, mm->block_size);
+	node->length = rounddown(offset + length, mm->block_size) - node->offset;
 
-	rmm = kzalloc(sizeof(*rmm), GFP_KERNEL);
-	if (!rmm) {
-		kfree(heap);
-		return -ENOMEM;
-	}
-	rmm->block_size = block;
-	mutex_init(&rmm->mutex);
-	INIT_LIST_HEAD(&rmm->nodes);
-	INIT_LIST_HEAD(&rmm->free);
-	list_add(&heap->nl_entry, &rmm->nodes);
-	list_add(&heap->fl_entry, &rmm->free);
-
-	*prmm = rmm;
+	list_add_tail(&node->nl_entry, &mm->nodes);
+	list_add_tail(&node->fl_entry, &mm->free);
+	mm->heap_nodes++;
 	return 0;
 }
 
 int
-nouveau_mm_fini(struct nouveau_mm **prmm)
+nouveau_mm_fini(struct nouveau_mm *mm)
 {
-	struct nouveau_mm *rmm = *prmm;
 	struct nouveau_mm_node *node, *heap =
-		list_first_entry(&rmm->nodes, struct nouveau_mm_node, nl_entry);
-
-	if (!list_is_singular(&rmm->nodes)) {
-		printk(KERN_ERR "nouveau_mm not empty at destroy time!\n");
-		list_for_each_entry(node, &rmm->nodes, nl_entry) {
-			printk(KERN_ERR "0x%02x: 0x%08x 0x%08x\n",
-			       node->type, node->offset, node->length);
+		list_first_entry(&mm->nodes, struct nouveau_mm_node, nl_entry);
+	int nodes = 0;
+
+	list_for_each_entry(node, &mm->nodes, nl_entry) {
+		if (nodes++ == mm->heap_nodes) {
+			printk(KERN_ERR "nouveau_mm in use at destroy time!\n");
+			list_for_each_entry(node, &mm->nodes, nl_entry) {
+				printk(KERN_ERR "0x%02x: 0x%08x 0x%08x\n",
+				       node->type, node->offset, node->length);
+			}
+			WARN_ON(1);
+			return -EBUSY;
 		}
-		WARN_ON(1);
-		return -EBUSY;
 	}
 
 	kfree(heap);
-	kfree(rmm);
-	*prmm = NULL;
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nouveau_mm.h b/drivers/gpu/drm/nouveau/nouveau_mm.h
index b9c016d2155..57a600c35c9 100644
--- a/drivers/gpu/drm/nouveau/nouveau_mm.h
+++ b/drivers/gpu/drm/nouveau/nouveau_mm.h
@@ -42,10 +42,11 @@ struct nouveau_mm {
 	struct mutex mutex;
 
 	u32 block_size;
+	int heap_nodes;
 };
 
-int  nouveau_mm_init(struct nouveau_mm **, u32 offset, u32 length, u32 block);
-int  nouveau_mm_fini(struct nouveau_mm **);
+int  nouveau_mm_init(struct nouveau_mm *, u32 offset, u32 length, u32 block);
+int  nouveau_mm_fini(struct nouveau_mm *);
 int  nouveau_mm_pre(struct nouveau_mm *);
 int  nouveau_mm_get(struct nouveau_mm *, int type, u32 size, u32 size_nc,
 		    u32 align, struct nouveau_mm_node **);
diff --git a/drivers/gpu/drm/nouveau/nouveau_object.c b/drivers/gpu/drm/nouveau/nouveau_object.c
index 159b7c437d3..02222c540ae 100644
--- a/drivers/gpu/drm/nouveau/nouveau_object.c
+++ b/drivers/gpu/drm/nouveau/nouveau_object.c
@@ -693,6 +693,7 @@ nouveau_gpuobj_channel_init_pramin(struct nouveau_channel *chan)
 static int
 nvc0_gpuobj_channel_init(struct nouveau_channel *chan, struct nouveau_vm *vm)
 {
+	struct drm_nouveau_private *dev_priv = chan->dev->dev_private;
 	struct drm_device *dev = chan->dev;
 	struct nouveau_gpuobj *pgd = NULL;
 	struct nouveau_vm_pgd *vpgd;
@@ -722,6 +723,9 @@ nvc0_gpuobj_channel_init(struct nouveau_channel *chan, struct nouveau_vm *vm)
 	nv_wo32(chan->ramin, 0x020c, 0x000000ff);
 
 	/* map display semaphore buffers into channel's vm */
+	if (dev_priv->card_type >= NV_D0)
+		return 0;
+
 	for (i = 0; i < 2; i++) {
 		struct nv50_display_crtc *dispc = &nv50_display(dev)->crtc[i];
 
@@ -746,7 +750,7 @@ nouveau_gpuobj_channel_init(struct nouveau_channel *chan,
 	int ret, i;
 
 	NV_DEBUG(dev, "ch%d vram=0x%08x tt=0x%08x\n", chan->id, vram_h, tt_h);
-	if (dev_priv->card_type == NV_C0)
+	if (dev_priv->card_type >= NV_C0)
 		return nvc0_gpuobj_channel_init(chan, vm);
 
 	/* Allocate a chunk of memory for per-channel object storage */
@@ -793,7 +797,7 @@ nouveau_gpuobj_channel_init(struct nouveau_channel *chan,
 			return ret;
 
 		/* dma objects for display sync channel semaphore blocks */
-		for (i = 0; i < 2; i++) {
+		for (i = 0; i < dev->mode_config.num_crtc; i++) {
 			struct nouveau_gpuobj *sem = NULL;
 			struct nv50_display_crtc *dispc =
 				&nv50_display(dev)->crtc[i];
@@ -875,18 +879,18 @@ nouveau_gpuobj_channel_takedown(struct nouveau_channel *chan)
 
 	NV_DEBUG(dev, "ch%d\n", chan->id);
 
-	if (dev_priv->card_type >= NV_50) {
+	if (dev_priv->card_type >= NV_50 && dev_priv->card_type <= NV_C0) {
 		struct nv50_display *disp = nv50_display(dev);
 
-		for (i = 0; i < 2; i++) {
+		for (i = 0; i < dev->mode_config.num_crtc; i++) {
 			struct nv50_display_crtc *dispc = &disp->crtc[i];
 			nouveau_bo_vma_del(dispc->sem.bo, &chan->dispc_vma[i]);
 		}
-
-		nouveau_vm_ref(NULL, &chan->vm, chan->vm_pd);
-		nouveau_gpuobj_ref(NULL, &chan->vm_pd);
 	}
 
+	nouveau_vm_ref(NULL, &chan->vm, chan->vm_pd);
+	nouveau_gpuobj_ref(NULL, &chan->vm_pd);
+
 	if (drm_mm_initialized(&chan->ramin_heap))
 		drm_mm_takedown(&chan->ramin_heap);
 	nouveau_gpuobj_ref(NULL, &chan->ramin);
diff --git a/drivers/gpu/drm/nouveau/nouveau_perf.c b/drivers/gpu/drm/nouveau/nouveau_perf.c
index ef9dec0e6f8..9f178aa9416 100644
--- a/drivers/gpu/drm/nouveau/nouveau_perf.c
+++ b/drivers/gpu/drm/nouveau/nouveau_perf.c
@@ -127,13 +127,57 @@ nouveau_perf_timing(struct drm_device *dev, struct bit_entry *P,
 
 	entry += ramcfg * recordlen;
 	if (entry[1] >= pm->memtimings.nr_timing) {
-		NV_WARN(dev, "timingset %d does not exist\n", entry[1]);
+		if (entry[1] != 0xff)
+			NV_WARN(dev, "timingset %d does not exist\n", entry[1]);
 		return NULL;
 	}
 
 	return &pm->memtimings.timing[entry[1]];
 }
 
+static void
+nouveau_perf_voltage(struct drm_device *dev, struct bit_entry *P,
+		     struct nouveau_pm_level *perflvl)
+{
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct nvbios *bios = &dev_priv->vbios;
+	u8 *vmap;
+	int id;
+
+	id = perflvl->volt_min;
+	perflvl->volt_min = 0;
+
+	/* boards using voltage table version <0x40 store the voltage
+	 * level directly in the perflvl entry as a multiple of 10mV
+	 */
+	if (dev_priv->engine.pm.voltage.version < 0x40) {
+		perflvl->volt_min = id * 10000;
+		perflvl->volt_max = perflvl->volt_min;
+		return;
+	}
+
+	/* on newer ones, the perflvl stores an index into yet another
+	 * vbios table containing a min/max voltage value for the perflvl
+	 */
+	if (P->version != 2 || P->length < 34) {
+		NV_DEBUG(dev, "where's our volt map table ptr? %d %d\n",
+			 P->version, P->length);
+		return;
+	}
+
+	vmap = ROMPTR(bios, P->data[32]);
+	if (!vmap) {
+		NV_DEBUG(dev, "volt map table pointer invalid\n");
+		return;
+	}
+
+	if (id < vmap[3]) {
+		vmap += vmap[1] + (vmap[2] * id);
+		perflvl->volt_min = ROM32(vmap[0]);
+		perflvl->volt_max = ROM32(vmap[4]);
+	}
+}
+
 void
 nouveau_perf_init(struct drm_device *dev)
 {
@@ -141,6 +185,8 @@ nouveau_perf_init(struct drm_device *dev)
 	struct nouveau_pm_engine *pm = &dev_priv->engine.pm;
 	struct nvbios *bios = &dev_priv->vbios;
 	struct bit_entry P;
+	struct nouveau_pm_memtimings *memtimings = &pm->memtimings;
+	struct nouveau_pm_tbl_header mt_hdr;
 	u8 version, headerlen, recordlen, entries;
 	u8 *perf, *entry;
 	int vid, i;
@@ -188,6 +234,22 @@ nouveau_perf_init(struct drm_device *dev)
 	}
 
 	entry = perf + headerlen;
+
+	/* For version 0x15, initialize memtiming table */
+	if(version == 0x15) {
+		memtimings->timing =
+				kcalloc(entries, sizeof(*memtimings->timing), GFP_KERNEL);
+		if(!memtimings) {
+			NV_WARN(dev,"Could not allocate memtiming table\n");
+			return;
+		}
+
+		mt_hdr.entry_cnt = entries;
+		mt_hdr.entry_len = 14;
+		mt_hdr.version = version;
+		mt_hdr.header_len = 4;
+	}
+
 	for (i = 0; i < entries; i++) {
 		struct nouveau_pm_level *perflvl = &pm->perflvl[pm->nr_perflvl];
 
@@ -203,7 +265,8 @@ nouveau_perf_init(struct drm_device *dev)
 		case 0x13:
 		case 0x15:
 			perflvl->fanspeed = entry[55];
-			perflvl->voltage = (recordlen > 56) ? entry[56] : 0;
+			if (recordlen > 56)
+				perflvl->volt_min = entry[56];
 			perflvl->core = ROM32(entry[1]) * 10;
 			perflvl->memory = ROM32(entry[5]) * 20;
 			break;
@@ -211,9 +274,10 @@ nouveau_perf_init(struct drm_device *dev)
 		case 0x23:
 		case 0x24:
 			perflvl->fanspeed = entry[4];
-			perflvl->voltage = entry[5];
-			perflvl->core = ROM16(entry[6]) * 1000;
-
+			perflvl->volt_min = entry[5];
+			perflvl->shader = ROM16(entry[6]) * 1000;
+			perflvl->core = perflvl->shader;
+			perflvl->core += (signed char)entry[8] * 1000;
 			if (dev_priv->chipset == 0x49 ||
 			    dev_priv->chipset == 0x4b)
 				perflvl->memory = ROM16(entry[11]) * 1000;
@@ -223,7 +287,7 @@ nouveau_perf_init(struct drm_device *dev)
 			break;
 		case 0x25:
 			perflvl->fanspeed = entry[4];
-			perflvl->voltage = entry[5];
+			perflvl->volt_min = entry[5];
 			perflvl->core = ROM16(entry[6]) * 1000;
 			perflvl->shader = ROM16(entry[10]) * 1000;
 			perflvl->memory = ROM16(entry[12]) * 1000;
@@ -232,7 +296,7 @@ nouveau_perf_init(struct drm_device *dev)
 			perflvl->memscript = ROM16(entry[2]);
 		case 0x35:
 			perflvl->fanspeed = entry[6];
-			perflvl->voltage = entry[7];
+			perflvl->volt_min = entry[7];
 			perflvl->core = ROM16(entry[8]) * 1000;
 			perflvl->shader = ROM16(entry[10]) * 1000;
 			perflvl->memory = ROM16(entry[12]) * 1000;
@@ -240,30 +304,34 @@ nouveau_perf_init(struct drm_device *dev)
 			perflvl->unk05 = ROM16(entry[16]) * 1000;
 			break;
 		case 0x40:
-#define subent(n) entry[perf[2] + ((n) * perf[3])]
+#define subent(n) (ROM16(entry[perf[2] + ((n) * perf[3])]) & 0xfff) * 1000
 			perflvl->fanspeed = 0; /*XXX*/
-			perflvl->voltage = entry[2];
+			perflvl->volt_min = entry[2];
 			if (dev_priv->card_type == NV_50) {
-				perflvl->core = ROM16(subent(0)) & 0xfff;
-				perflvl->shader = ROM16(subent(1)) & 0xfff;
-				perflvl->memory = ROM16(subent(2)) & 0xfff;
+				perflvl->core   = subent(0);
+				perflvl->shader = subent(1);
+				perflvl->memory = subent(2);
+				perflvl->vdec   = subent(3);
+				perflvl->unka0  = subent(4);
 			} else {
-				perflvl->shader = ROM16(subent(3)) & 0xfff;
+				perflvl->hub06  = subent(0);
+				perflvl->hub01  = subent(1);
+				perflvl->copy   = subent(2);
+				perflvl->shader = subent(3);
+				perflvl->rop    = subent(4);
+				perflvl->memory = subent(5);
+				perflvl->vdec   = subent(6);
+				perflvl->daemon = subent(10);
+				perflvl->hub07  = subent(11);
 				perflvl->core   = perflvl->shader / 2;
-				perflvl->unk0a  = ROM16(subent(4)) & 0xfff;
-				perflvl->memory = ROM16(subent(5)) & 0xfff;
 			}
-
-			perflvl->core *= 1000;
-			perflvl->shader *= 1000;
-			perflvl->memory *= 1000;
-			perflvl->unk0a *= 1000;
 			break;
 		}
 
 		/* make sure vid is valid */
-		if (pm->voltage.supported && perflvl->voltage) {
-			vid = nouveau_volt_vid_lookup(dev, perflvl->voltage);
+		nouveau_perf_voltage(dev, &P, perflvl);
+		if (pm->voltage.supported && perflvl->volt_min) {
+			vid = nouveau_volt_vid_lookup(dev, perflvl->volt_min);
 			if (vid < 0) {
 				NV_DEBUG(dev, "drop perflvl %d, bad vid\n", i);
 				entry += recordlen;
@@ -272,7 +340,11 @@ nouveau_perf_init(struct drm_device *dev)
 		}
 
 		/* get the corresponding memory timings */
-		if (version > 0x15) {
+		if (version == 0x15) {
+			memtimings->timing[i].id = i;
+			nv30_mem_timing_entry(dev,&mt_hdr,(struct nouveau_pm_tbl_entry*) &entry[41],0,&memtimings->timing[i]);
+			perflvl->timing = &memtimings->timing[i];
+		} else if (version > 0x15) {
 			/* last 3 args are for < 0x40, ignored for >= 0x40 */
 			perflvl->timing =
 				nouveau_perf_timing(dev, &P,
diff --git a/drivers/gpu/drm/nouveau/nouveau_pm.c b/drivers/gpu/drm/nouveau/nouveau_pm.c
index da8d994d5e8..a539fd25792 100644
--- a/drivers/gpu/drm/nouveau/nouveau_pm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_pm.c
@@ -64,18 +64,26 @@ nouveau_pm_perflvl_set(struct drm_device *dev, struct nouveau_pm_level *perflvl)
 	if (perflvl == pm->cur)
 		return 0;
 
-	if (pm->voltage.supported && pm->voltage_set && perflvl->voltage) {
-		ret = pm->voltage_set(dev, perflvl->voltage);
+	if (pm->voltage.supported && pm->voltage_set && perflvl->volt_min) {
+		ret = pm->voltage_set(dev, perflvl->volt_min);
 		if (ret) {
 			NV_ERROR(dev, "voltage_set %d failed: %d\n",
-				 perflvl->voltage, ret);
+				 perflvl->volt_min, ret);
 		}
 	}
 
-	nouveau_pm_clock_set(dev, perflvl, PLL_CORE, perflvl->core);
-	nouveau_pm_clock_set(dev, perflvl, PLL_SHADER, perflvl->shader);
-	nouveau_pm_clock_set(dev, perflvl, PLL_MEMORY, perflvl->memory);
-	nouveau_pm_clock_set(dev, perflvl, PLL_UNK05, perflvl->unk05);
+	if (pm->clocks_pre) {
+		void *state = pm->clocks_pre(dev, perflvl);
+		if (IS_ERR(state))
+			return PTR_ERR(state);
+		pm->clocks_set(dev, state);
+	} else
+	if (pm->clock_set) {
+		nouveau_pm_clock_set(dev, perflvl, PLL_CORE, perflvl->core);
+		nouveau_pm_clock_set(dev, perflvl, PLL_SHADER, perflvl->shader);
+		nouveau_pm_clock_set(dev, perflvl, PLL_MEMORY, perflvl->memory);
+		nouveau_pm_clock_set(dev, perflvl, PLL_UNK05, perflvl->unk05);
+	}
 
 	pm->cur = perflvl;
 	return 0;
@@ -92,9 +100,6 @@ nouveau_pm_profile_set(struct drm_device *dev, const char *profile)
 	if (nouveau_perflvl_wr != 7777)
 		return -EPERM;
 
-	if (!pm->clock_set)
-		return -EINVAL;
-
 	if (!strncmp(profile, "boot", 4))
 		perflvl = &pm->boot;
 	else {
@@ -123,31 +128,37 @@ nouveau_pm_perflvl_get(struct drm_device *dev, struct nouveau_pm_level *perflvl)
 	struct nouveau_pm_engine *pm = &dev_priv->engine.pm;
 	int ret;
 
-	if (!pm->clock_get)
-		return -EINVAL;
-
 	memset(perflvl, 0, sizeof(*perflvl));
 
-	ret = pm->clock_get(dev, PLL_CORE);
-	if (ret > 0)
-		perflvl->core = ret;
+	if (pm->clocks_get) {
+		ret = pm->clocks_get(dev, perflvl);
+		if (ret)
+			return ret;
+	} else
+	if (pm->clock_get) {
+		ret = pm->clock_get(dev, PLL_CORE);
+		if (ret > 0)
+			perflvl->core = ret;
 
-	ret = pm->clock_get(dev, PLL_MEMORY);
-	if (ret > 0)
-		perflvl->memory = ret;
+		ret = pm->clock_get(dev, PLL_MEMORY);
+		if (ret > 0)
+			perflvl->memory = ret;
 
-	ret = pm->clock_get(dev, PLL_SHADER);
-	if (ret > 0)
-		perflvl->shader = ret;
+		ret = pm->clock_get(dev, PLL_SHADER);
+		if (ret > 0)
+			perflvl->shader = ret;
 
-	ret = pm->clock_get(dev, PLL_UNK05);
-	if (ret > 0)
-		perflvl->unk05 = ret;
+		ret = pm->clock_get(dev, PLL_UNK05);
+		if (ret > 0)
+			perflvl->unk05 = ret;
+	}
 
 	if (pm->voltage.supported && pm->voltage_get) {
 		ret = pm->voltage_get(dev);
-		if (ret > 0)
-			perflvl->voltage = ret;
+		if (ret > 0) {
+			perflvl->volt_min = ret;
+			perflvl->volt_max = ret;
+		}
 	}
 
 	return 0;
@@ -156,7 +167,7 @@ nouveau_pm_perflvl_get(struct drm_device *dev, struct nouveau_pm_level *perflvl)
 static void
 nouveau_pm_perflvl_info(struct nouveau_pm_level *perflvl, char *ptr, int len)
 {
-	char c[16], s[16], v[16], f[16], t[16];
+	char c[16], s[16], v[32], f[16], t[16], m[16];
 
 	c[0] = '\0';
 	if (perflvl->core)
@@ -166,9 +177,19 @@ nouveau_pm_perflvl_info(struct nouveau_pm_level *perflvl, char *ptr, int len)
 	if (perflvl->shader)
 		snprintf(s, sizeof(s), " shader %dMHz", perflvl->shader / 1000);
 
+	m[0] = '\0';
+	if (perflvl->memory)
+		snprintf(m, sizeof(m), " memory %dMHz", perflvl->memory / 1000);
+
 	v[0] = '\0';
-	if (perflvl->voltage)
-		snprintf(v, sizeof(v), " voltage %dmV", perflvl->voltage * 10);
+	if (perflvl->volt_min && perflvl->volt_min != perflvl->volt_max) {
+		snprintf(v, sizeof(v), " voltage %dmV-%dmV",
+			 perflvl->volt_min / 1000, perflvl->volt_max / 1000);
+	} else
+	if (perflvl->volt_min) {
+		snprintf(v, sizeof(v), " voltage %dmV",
+			 perflvl->volt_min / 1000);
+	}
 
 	f[0] = '\0';
 	if (perflvl->fanspeed)
@@ -178,8 +199,7 @@ nouveau_pm_perflvl_info(struct nouveau_pm_level *perflvl, char *ptr, int len)
 	if (perflvl->timing)
 		snprintf(t, sizeof(t), " timing %d", perflvl->timing->id);
 
-	snprintf(ptr, len, "memory %dMHz%s%s%s%s%s\n", perflvl->memory / 1000,
-		 c, s, v, f, t);
+	snprintf(ptr, len, "%s%s%s%s%s%s\n", c, s, m, t, v, f);
 }
 
 static ssize_t
@@ -190,7 +210,7 @@ nouveau_pm_get_perflvl_info(struct device *d,
 	char *ptr = buf;
 	int len = PAGE_SIZE;
 
-	snprintf(ptr, len, "%d: ", perflvl->id);
+	snprintf(ptr, len, "%d:", perflvl->id);
 	ptr += strlen(buf);
 	len -= strlen(buf);
 
@@ -211,9 +231,9 @@ nouveau_pm_get_perflvl(struct device *d, struct device_attribute *a, char *buf)
 	if (!pm->cur)
 		snprintf(ptr, len, "setting: boot\n");
 	else if (pm->cur == &pm->boot)
-		snprintf(ptr, len, "setting: boot\nc: ");
+		snprintf(ptr, len, "setting: boot\nc:");
 	else
-		snprintf(ptr, len, "setting: static %d\nc: ", pm->cur->id);
+		snprintf(ptr, len, "setting: static %d\nc:", pm->cur->id);
 	ptr += strlen(buf);
 	len -= strlen(buf);
 
@@ -292,7 +312,7 @@ nouveau_sysfs_fini(struct drm_device *dev)
 	}
 }
 
-#ifdef CONFIG_HWMON
+#if defined(CONFIG_HWMON) || (defined(MODULE) && defined(CONFIG_HWMON_MODULE))
 static ssize_t
 nouveau_hwmon_show_temp(struct device *d, struct device_attribute *a, char *buf)
 {
@@ -409,7 +429,7 @@ static const struct attribute_group hwmon_attrgroup = {
 static int
 nouveau_hwmon_init(struct drm_device *dev)
 {
-#ifdef CONFIG_HWMON
+#if defined(CONFIG_HWMON) || (defined(MODULE) && defined(CONFIG_HWMON_MODULE))
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
 	struct nouveau_pm_engine *pm = &dev_priv->engine.pm;
 	struct device *hwmon_dev;
@@ -442,7 +462,7 @@ nouveau_hwmon_init(struct drm_device *dev)
 static void
 nouveau_hwmon_fini(struct drm_device *dev)
 {
-#ifdef CONFIG_HWMON
+#if defined(CONFIG_HWMON) || (defined(MODULE) && defined(CONFIG_HWMON_MODULE))
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
 	struct nouveau_pm_engine *pm = &dev_priv->engine.pm;
 
@@ -488,7 +508,7 @@ nouveau_pm_init(struct drm_device *dev)
 	NV_INFO(dev, "%d available performance level(s)\n", pm->nr_perflvl);
 	for (i = 0; i < pm->nr_perflvl; i++) {
 		nouveau_pm_perflvl_info(&pm->perflvl[i], info, sizeof(info));
-		NV_INFO(dev, "%d: %s", pm->perflvl[i].id, info);
+		NV_INFO(dev, "%d:%s", pm->perflvl[i].id, info);
 	}
 
 	/* determine current ("boot") performance level */
@@ -498,7 +518,7 @@ nouveau_pm_init(struct drm_device *dev)
 		pm->cur = &pm->boot;
 
 		nouveau_pm_perflvl_info(&pm->boot, info, sizeof(info));
-		NV_INFO(dev, "c: %s", info);
+		NV_INFO(dev, "c:%s", info);
 	}
 
 	/* switch performance levels now if requested */
diff --git a/drivers/gpu/drm/nouveau/nouveau_pm.h b/drivers/gpu/drm/nouveau/nouveau_pm.h
index 4a9838ddace..8ac02cdd03a 100644
--- a/drivers/gpu/drm/nouveau/nouveau_pm.h
+++ b/drivers/gpu/drm/nouveau/nouveau_pm.h
@@ -52,6 +52,11 @@ void *nv04_pm_clock_pre(struct drm_device *, struct nouveau_pm_level *,
 			u32 id, int khz);
 void nv04_pm_clock_set(struct drm_device *, void *);
 
+/* nv40_pm.c */
+int nv40_pm_clocks_get(struct drm_device *, struct nouveau_pm_level *);
+void *nv40_pm_clocks_pre(struct drm_device *, struct nouveau_pm_level *);
+void nv40_pm_clocks_set(struct drm_device *, void *);
+
 /* nv50_pm.c */
 int nv50_pm_clock_get(struct drm_device *, u32 id);
 void *nv50_pm_clock_pre(struct drm_device *, struct nouveau_pm_level *,
@@ -59,10 +64,12 @@ void *nv50_pm_clock_pre(struct drm_device *, struct nouveau_pm_level *,
 void nv50_pm_clock_set(struct drm_device *, void *);
 
 /* nva3_pm.c */
-int nva3_pm_clock_get(struct drm_device *, u32 id);
-void *nva3_pm_clock_pre(struct drm_device *, struct nouveau_pm_level *,
-			u32 id, int khz);
-void nva3_pm_clock_set(struct drm_device *, void *);
+int nva3_pm_clocks_get(struct drm_device *, struct nouveau_pm_level *);
+void *nva3_pm_clocks_pre(struct drm_device *, struct nouveau_pm_level *);
+void nva3_pm_clocks_set(struct drm_device *, void *);
+
+/* nvc0_pm.c */
+int nvc0_pm_clocks_get(struct drm_device *, struct nouveau_pm_level *);
 
 /* nouveau_temp.c */
 void nouveau_temp_init(struct drm_device *dev);
diff --git a/drivers/gpu/drm/nouveau/nouveau_reg.h b/drivers/gpu/drm/nouveau/nouveau_reg.h
index f18cdfc3400..43a96b99e18 100644
--- a/drivers/gpu/drm/nouveau/nouveau_reg.h
+++ b/drivers/gpu/drm/nouveau/nouveau_reg.h
@@ -826,9 +826,12 @@
 #define NV50_PDISPLAY_SOR_DPMS_STATE_ACTIVE                          0x00030000
 #define NV50_PDISPLAY_SOR_DPMS_STATE_BLANKED                         0x00080000
 #define NV50_PDISPLAY_SOR_DPMS_STATE_WAIT                            0x10000000
-#define NV50_PDISPLAY_SOR_BACKLIGHT                                  0x0061c084
-#define NV50_PDISPLAY_SOR_BACKLIGHT_ENABLE                           0x80000000
-#define NV50_PDISPLAY_SOR_BACKLIGHT_LEVEL                            0x00000fff
+#define NV50_PDISP_SOR_PWM_DIV(i)                     (0x0061c080 + (i) * 0x800)
+#define NV50_PDISP_SOR_PWM_CTL(i)                     (0x0061c084 + (i) * 0x800)
+#define NV50_PDISP_SOR_PWM_CTL_NEW                                   0x80000000
+#define NVA3_PDISP_SOR_PWM_CTL_UNK                                   0x40000000
+#define NV50_PDISP_SOR_PWM_CTL_VAL                                   0x000007ff
+#define NVA3_PDISP_SOR_PWM_CTL_VAL                                   0x00ffffff
 #define NV50_SOR_DP_CTRL(i, l)           (0x0061c10c + (i) * 0x800 + (l) * 0x80)
 #define NV50_SOR_DP_CTRL_ENABLED                                     0x00000001
 #define NV50_SOR_DP_CTRL_ENHANCED_FRAME_ENABLED                      0x00004000
@@ -843,7 +846,7 @@
 #define NV50_SOR_DP_CTRL_TRAINING_PATTERN_2                          0x02000000
 #define NV50_SOR_DP_UNK118(i, l)         (0x0061c118 + (i) * 0x800 + (l) * 0x80)
 #define NV50_SOR_DP_UNK120(i, l)         (0x0061c120 + (i) * 0x800 + (l) * 0x80)
-#define NV50_SOR_DP_UNK128(i, l)         (0x0061c128 + (i) * 0x800 + (l) * 0x80)
+#define NV50_SOR_DP_SCFG(i, l)           (0x0061c128 + (i) * 0x800 + (l) * 0x80)
 #define NV50_SOR_DP_UNK130(i, l)         (0x0061c130 + (i) * 0x800 + (l) * 0x80)
 
 #define NV50_PDISPLAY_USER(i)                        ((i) * 0x1000 + 0x00640000)
diff --git a/drivers/gpu/drm/nouveau/nouveau_sgdma.c b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
index c444cadbf84..b75258a9fe4 100644
--- a/drivers/gpu/drm/nouveau/nouveau_sgdma.c
+++ b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
@@ -12,8 +12,8 @@ struct nouveau_sgdma_be {
 	struct drm_device *dev;
 
 	dma_addr_t *pages;
-	bool *ttm_alloced;
 	unsigned nr_pages;
+	bool unmap_pages;
 
 	u64 offset;
 	bool bound;
@@ -26,40 +26,28 @@ nouveau_sgdma_populate(struct ttm_backend *be, unsigned long num_pages,
 {
 	struct nouveau_sgdma_be *nvbe = (struct nouveau_sgdma_be *)be;
 	struct drm_device *dev = nvbe->dev;
+	int i;
 
 	NV_DEBUG(nvbe->dev, "num_pages = %ld\n", num_pages);
 
-	if (nvbe->pages)
-		return -EINVAL;
-
-	nvbe->pages = kmalloc(sizeof(dma_addr_t) * num_pages, GFP_KERNEL);
-	if (!nvbe->pages)
-		return -ENOMEM;
+	nvbe->pages = dma_addrs;
+	nvbe->nr_pages = num_pages;
+	nvbe->unmap_pages = true;
 
-	nvbe->ttm_alloced = kmalloc(sizeof(bool) * num_pages, GFP_KERNEL);
-	if (!nvbe->ttm_alloced)
-		return -ENOMEM;
+	/* this code path isn't called and is incorrect anyways */
+	if (0) { /* dma_addrs[0] != DMA_ERROR_CODE) { */
+		nvbe->unmap_pages = false;
+		return 0;
+	}
 
-	nvbe->nr_pages = 0;
-	while (num_pages--) {
-		/* this code path isn't called and is incorrect anyways */
-		if (0) { /*dma_addrs[nvbe->nr_pages] != DMA_ERROR_CODE)*/
-			nvbe->pages[nvbe->nr_pages] =
-					dma_addrs[nvbe->nr_pages];
-		 	nvbe->ttm_alloced[nvbe->nr_pages] = true;
-		} else {
-			nvbe->pages[nvbe->nr_pages] =
-				pci_map_page(dev->pdev, pages[nvbe->nr_pages], 0,
-				     PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
-			if (pci_dma_mapping_error(dev->pdev,
-						  nvbe->pages[nvbe->nr_pages])) {
-				be->func->clear(be);
-				return -EFAULT;
-			}
-			nvbe->ttm_alloced[nvbe->nr_pages] = false;
+	for (i = 0; i < num_pages; i++) {
+		nvbe->pages[i] = pci_map_page(dev->pdev, pages[i], 0,
+					      PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
+		if (pci_dma_mapping_error(dev->pdev, nvbe->pages[i])) {
+			nvbe->nr_pages = --i;
+			be->func->clear(be);
+			return -EFAULT;
 		}
-
-		nvbe->nr_pages++;
 	}
 
 	return 0;
@@ -69,25 +57,16 @@ static void
 nouveau_sgdma_clear(struct ttm_backend *be)
 {
 	struct nouveau_sgdma_be *nvbe = (struct nouveau_sgdma_be *)be;
-	struct drm_device *dev;
-
-	if (nvbe && nvbe->pages) {
-		dev = nvbe->dev;
-		NV_DEBUG(dev, "\n");
+	struct drm_device *dev = nvbe->dev;
 
-		if (nvbe->bound)
-			be->func->unbind(be);
+	if (nvbe->bound)
+		be->func->unbind(be);
 
+	if (nvbe->unmap_pages) {
 		while (nvbe->nr_pages--) {
-			if (!nvbe->ttm_alloced[nvbe->nr_pages])
-				pci_unmap_page(dev->pdev, nvbe->pages[nvbe->nr_pages],
+			pci_unmap_page(dev->pdev, nvbe->pages[nvbe->nr_pages],
 				       PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
 		}
-		kfree(nvbe->pages);
-		kfree(nvbe->ttm_alloced);
-		nvbe->pages = NULL;
-		nvbe->ttm_alloced = NULL;
-		nvbe->nr_pages = 0;
 	}
 }
 
@@ -126,7 +105,7 @@ nv04_sgdma_bind(struct ttm_backend *be, struct ttm_mem_reg *mem)
 
 		for (j = 0; j < PAGE_SIZE / NV_CTXDMA_PAGE_SIZE; j++, pte++) {
 			nv_wo32(gpuobj, (pte * 4) + 0, offset_l | 3);
-			dma_offset += NV_CTXDMA_PAGE_SIZE;
+			offset_l += NV_CTXDMA_PAGE_SIZE;
 		}
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_state.c b/drivers/gpu/drm/nouveau/nouveau_state.c
index 10656e430b4..82478e0998e 100644
--- a/drivers/gpu/drm/nouveau/nouveau_state.c
+++ b/drivers/gpu/drm/nouveau/nouveau_state.c
@@ -286,9 +286,9 @@ static int nouveau_init_engine_ptrs(struct drm_device *dev)
 		engine->gpio.get		= nv10_gpio_get;
 		engine->gpio.set		= nv10_gpio_set;
 		engine->gpio.irq_enable		= NULL;
-		engine->pm.clock_get		= nv04_pm_clock_get;
-		engine->pm.clock_pre		= nv04_pm_clock_pre;
-		engine->pm.clock_set		= nv04_pm_clock_set;
+		engine->pm.clocks_get		= nv40_pm_clocks_get;
+		engine->pm.clocks_pre		= nv40_pm_clocks_pre;
+		engine->pm.clocks_set		= nv40_pm_clocks_set;
 		engine->pm.voltage_get		= nouveau_voltage_gpio_get;
 		engine->pm.voltage_set		= nouveau_voltage_gpio_set;
 		engine->pm.temp_get		= nv40_temp_get;
@@ -299,7 +299,7 @@ static int nouveau_init_engine_ptrs(struct drm_device *dev)
 	case 0x50:
 	case 0x80: /* gotta love NVIDIA's consistency.. */
 	case 0x90:
-	case 0xA0:
+	case 0xa0:
 		engine->instmem.init		= nv50_instmem_init;
 		engine->instmem.takedown	= nv50_instmem_takedown;
 		engine->instmem.suspend		= nv50_instmem_suspend;
@@ -359,9 +359,9 @@ static int nouveau_init_engine_ptrs(struct drm_device *dev)
 			engine->pm.clock_set	= nv50_pm_clock_set;
 			break;
 		default:
-			engine->pm.clock_get	= nva3_pm_clock_get;
-			engine->pm.clock_pre	= nva3_pm_clock_pre;
-			engine->pm.clock_set	= nva3_pm_clock_set;
+			engine->pm.clocks_get	= nva3_pm_clocks_get;
+			engine->pm.clocks_pre	= nva3_pm_clocks_pre;
+			engine->pm.clocks_set	= nva3_pm_clocks_set;
 			break;
 		}
 		engine->pm.voltage_get		= nouveau_voltage_gpio_get;
@@ -376,7 +376,7 @@ static int nouveau_init_engine_ptrs(struct drm_device *dev)
 		engine->vram.put		= nv50_vram_del;
 		engine->vram.flags_valid	= nv50_vram_flags_valid;
 		break;
-	case 0xC0:
+	case 0xc0:
 		engine->instmem.init		= nvc0_instmem_init;
 		engine->instmem.takedown	= nvc0_instmem_takedown;
 		engine->instmem.suspend		= nvc0_instmem_suspend;
@@ -422,12 +422,73 @@ static int nouveau_init_engine_ptrs(struct drm_device *dev)
 		engine->vram.put		= nv50_vram_del;
 		engine->vram.flags_valid	= nvc0_vram_flags_valid;
 		engine->pm.temp_get		= nv84_temp_get;
+		engine->pm.clocks_get		= nvc0_pm_clocks_get;
+		engine->pm.voltage_get		= nouveau_voltage_gpio_get;
+		engine->pm.voltage_set		= nouveau_voltage_gpio_set;
+		break;
+	case 0xd0:
+		engine->instmem.init		= nvc0_instmem_init;
+		engine->instmem.takedown	= nvc0_instmem_takedown;
+		engine->instmem.suspend		= nvc0_instmem_suspend;
+		engine->instmem.resume		= nvc0_instmem_resume;
+		engine->instmem.get		= nv50_instmem_get;
+		engine->instmem.put		= nv50_instmem_put;
+		engine->instmem.map		= nv50_instmem_map;
+		engine->instmem.unmap		= nv50_instmem_unmap;
+		engine->instmem.flush		= nv84_instmem_flush;
+		engine->mc.init			= nv50_mc_init;
+		engine->mc.takedown		= nv50_mc_takedown;
+		engine->timer.init		= nv04_timer_init;
+		engine->timer.read		= nv04_timer_read;
+		engine->timer.takedown		= nv04_timer_takedown;
+		engine->fb.init			= nvc0_fb_init;
+		engine->fb.takedown		= nvc0_fb_takedown;
+		engine->fifo.channels		= 128;
+		engine->fifo.init		= nvc0_fifo_init;
+		engine->fifo.takedown		= nvc0_fifo_takedown;
+		engine->fifo.disable		= nvc0_fifo_disable;
+		engine->fifo.enable		= nvc0_fifo_enable;
+		engine->fifo.reassign		= nvc0_fifo_reassign;
+		engine->fifo.channel_id		= nvc0_fifo_channel_id;
+		engine->fifo.create_context	= nvc0_fifo_create_context;
+		engine->fifo.destroy_context	= nvc0_fifo_destroy_context;
+		engine->fifo.load_context	= nvc0_fifo_load_context;
+		engine->fifo.unload_context	= nvc0_fifo_unload_context;
+		engine->display.early_init	= nouveau_stub_init;
+		engine->display.late_takedown	= nouveau_stub_takedown;
+		engine->display.create		= nvd0_display_create;
+		engine->display.init		= nvd0_display_init;
+		engine->display.destroy		= nvd0_display_destroy;
+		engine->gpio.init		= nv50_gpio_init;
+		engine->gpio.takedown		= nouveau_stub_takedown;
+		engine->gpio.get		= nvd0_gpio_get;
+		engine->gpio.set		= nvd0_gpio_set;
+		engine->gpio.irq_register	= nv50_gpio_irq_register;
+		engine->gpio.irq_unregister	= nv50_gpio_irq_unregister;
+		engine->gpio.irq_enable		= nv50_gpio_irq_enable;
+		engine->vram.init		= nvc0_vram_init;
+		engine->vram.takedown		= nv50_vram_fini;
+		engine->vram.get		= nvc0_vram_new;
+		engine->vram.put		= nv50_vram_del;
+		engine->vram.flags_valid	= nvc0_vram_flags_valid;
+		engine->pm.clocks_get		= nvc0_pm_clocks_get;
+		engine->pm.voltage_get		= nouveau_voltage_gpio_get;
+		engine->pm.voltage_set		= nouveau_voltage_gpio_set;
 		break;
 	default:
 		NV_ERROR(dev, "NV%02x unsupported\n", dev_priv->chipset);
 		return 1;
 	}
 
+	/* headless mode */
+	if (nouveau_modeset == 2) {
+		engine->display.early_init = nouveau_stub_init;
+		engine->display.late_takedown = nouveau_stub_takedown;
+		engine->display.create = nouveau_stub_init;
+		engine->display.init = nouveau_stub_init;
+		engine->display.destroy = nouveau_stub_takedown;
+	}
+
 	return 0;
 }
 
@@ -449,21 +510,6 @@ nouveau_vga_set_decode(void *priv, bool state)
 		return VGA_RSRC_NORMAL_IO | VGA_RSRC_NORMAL_MEM;
 }
 
-static int
-nouveau_card_init_channel(struct drm_device *dev)
-{
-	struct drm_nouveau_private *dev_priv = dev->dev_private;
-	int ret;
-
-	ret = nouveau_channel_alloc(dev, &dev_priv->channel, NULL,
-				    NvDmaFB, NvDmaTT);
-	if (ret)
-		return ret;
-
-	mutex_unlock(&dev_priv->channel->mutex);
-	return 0;
-}
-
 static void nouveau_switcheroo_set_state(struct pci_dev *pdev,
 					 enum vga_switcheroo_state state)
 {
@@ -630,8 +676,11 @@ nouveau_card_init(struct drm_device *dev)
 			break;
 		}
 
-		if (dev_priv->card_type == NV_40)
-			nv40_mpeg_create(dev);
+		if (dev_priv->card_type == NV_40 ||
+		    dev_priv->chipset == 0x31 ||
+		    dev_priv->chipset == 0x34 ||
+		    dev_priv->chipset == 0x36)
+			nv31_mpeg_create(dev);
 		else
 		if (dev_priv->card_type == NV_50 &&
 		    (dev_priv->chipset < 0x98 || dev_priv->chipset == 0xa0))
@@ -651,41 +700,69 @@ nouveau_card_init(struct drm_device *dev)
 			goto out_engine;
 	}
 
-	ret = engine->display.create(dev);
+	ret = nouveau_irq_init(dev);
 	if (ret)
 		goto out_fifo;
 
-	ret = drm_vblank_init(dev, nv_two_heads(dev) ? 2 : 1);
-	if (ret)
-		goto out_vblank;
+	/* initialise general modesetting */
+	drm_mode_config_init(dev);
+	drm_mode_create_scaling_mode_property(dev);
+	drm_mode_create_dithering_property(dev);
+	dev->mode_config.funcs = (void *)&nouveau_mode_config_funcs;
+	dev->mode_config.fb_base = pci_resource_start(dev->pdev, 1);
+	dev->mode_config.min_width = 0;
+	dev->mode_config.min_height = 0;
+	if (dev_priv->card_type < NV_10) {
+		dev->mode_config.max_width = 2048;
+		dev->mode_config.max_height = 2048;
+	} else
+	if (dev_priv->card_type < NV_50) {
+		dev->mode_config.max_width = 4096;
+		dev->mode_config.max_height = 4096;
+	} else {
+		dev->mode_config.max_width = 8192;
+		dev->mode_config.max_height = 8192;
+	}
 
-	ret = nouveau_irq_init(dev);
+	ret = engine->display.create(dev);
 	if (ret)
-		goto out_vblank;
+		goto out_irq;
 
-	/* what about PVIDEO/PCRTC/PRAMDAC etc? */
+	nouveau_backlight_init(dev);
 
 	if (dev_priv->eng[NVOBJ_ENGINE_GR]) {
 		ret = nouveau_fence_init(dev);
 		if (ret)
-			goto out_irq;
+			goto out_disp;
 
-		ret = nouveau_card_init_channel(dev);
+		ret = nouveau_channel_alloc(dev, &dev_priv->channel, NULL,
+					    NvDmaFB, NvDmaTT);
 		if (ret)
 			goto out_fence;
+
+		mutex_unlock(&dev_priv->channel->mutex);
+	}
+
+	if (dev->mode_config.num_crtc) {
+		ret = drm_vblank_init(dev, dev->mode_config.num_crtc);
+		if (ret)
+			goto out_chan;
+
+		nouveau_fbcon_init(dev);
+		drm_kms_helper_poll_init(dev);
 	}
 
-	nouveau_fbcon_init(dev);
-	drm_kms_helper_poll_init(dev);
 	return 0;
 
+out_chan:
+	nouveau_channel_put_unlocked(&dev_priv->channel);
 out_fence:
 	nouveau_fence_fini(dev);
+out_disp:
+	nouveau_backlight_exit(dev);
+	engine->display.destroy(dev);
 out_irq:
 	nouveau_irq_fini(dev);
-out_vblank:
-	drm_vblank_cleanup(dev);
-	engine->display.destroy(dev);
 out_fifo:
 	if (!dev_priv->noaccel)
 		engine->fifo.takedown(dev);
@@ -732,15 +809,20 @@ static void nouveau_card_takedown(struct drm_device *dev)
 	struct nouveau_engine *engine = &dev_priv->engine;
 	int e;
 
-	drm_kms_helper_poll_fini(dev);
-	nouveau_fbcon_fini(dev);
+	if (dev->mode_config.num_crtc) {
+		drm_kms_helper_poll_fini(dev);
+		nouveau_fbcon_fini(dev);
+		drm_vblank_cleanup(dev);
+	}
 
 	if (dev_priv->channel) {
 		nouveau_channel_put_unlocked(&dev_priv->channel);
 		nouveau_fence_fini(dev);
 	}
 
+	nouveau_backlight_exit(dev);
 	engine->display.destroy(dev);
+	drm_mode_config_cleanup(dev);
 
 	if (!dev_priv->noaccel) {
 		engine->fifo.takedown(dev);
@@ -774,7 +856,6 @@ static void nouveau_card_takedown(struct drm_device *dev)
 	engine->vram.takedown(dev);
 
 	nouveau_irq_fini(dev);
-	drm_vblank_cleanup(dev);
 
 	nouveau_pm_fini(dev);
 	nouveau_bios_takedown(dev);
@@ -907,7 +988,7 @@ static int nouveau_remove_conflicting_drivers(struct drm_device *dev)
 int nouveau_load(struct drm_device *dev, unsigned long flags)
 {
 	struct drm_nouveau_private *dev_priv;
-	uint32_t reg0;
+	uint32_t reg0, strap;
 	resource_size_t mmio_start_offs;
 	int ret;
 
@@ -951,13 +1032,11 @@ int nouveau_load(struct drm_device *dev, unsigned long flags)
 
 	/* Time to determine the card architecture */
 	reg0 = nv_rd32(dev, NV03_PMC_BOOT_0);
-	dev_priv->stepping = 0; /* XXX: add stepping for pre-NV10? */
 
 	/* We're dealing with >=NV10 */
 	if ((reg0 & 0x0f000000) > 0) {
 		/* Bit 27-20 contain the architecture in hex */
 		dev_priv->chipset = (reg0 & 0xff00000) >> 20;
-		dev_priv->stepping = (reg0 & 0xff);
 	/* NV04 or NV05 */
 	} else if ((reg0 & 0xff00fff0) == 0x20004000) {
 		if (reg0 & 0x00f00000)
@@ -987,6 +1066,9 @@ int nouveau_load(struct drm_device *dev, unsigned long flags)
 	case 0xc0:
 		dev_priv->card_type = NV_C0;
 		break;
+	case 0xd0:
+		dev_priv->card_type = NV_D0;
+		break;
 	default:
 		NV_INFO(dev, "Unsupported chipset 0x%08x\n", reg0);
 		ret = -EINVAL;
@@ -996,6 +1078,23 @@ int nouveau_load(struct drm_device *dev, unsigned long flags)
 	NV_INFO(dev, "Detected an NV%2x generation card (0x%08x)\n",
 		dev_priv->card_type, reg0);
 
+	/* determine frequency of timing crystal */
+	strap = nv_rd32(dev, 0x101000);
+	if ( dev_priv->chipset < 0x17 ||
+	    (dev_priv->chipset >= 0x20 && dev_priv->chipset <= 0x25))
+		strap &= 0x00000040;
+	else
+		strap &= 0x00400040;
+
+	switch (strap) {
+	case 0x00000000: dev_priv->crystal = 13500; break;
+	case 0x00000040: dev_priv->crystal = 14318; break;
+	case 0x00400000: dev_priv->crystal = 27000; break;
+	case 0x00400040: dev_priv->crystal = 25000; break;
+	}
+
+	NV_DEBUG(dev, "crystal freq: %dKHz\n", dev_priv->crystal);
+
 	/* Determine whether we'll attempt acceleration or not, some
 	 * cards are disabled by default here due to them being known
 	 * non-functional, or never been tested due to lack of hw.
@@ -1030,7 +1129,7 @@ int nouveau_load(struct drm_device *dev, unsigned long flags)
 			ioremap(pci_resource_start(dev->pdev, ramin_bar),
 				dev_priv->ramin_size);
 		if (!dev_priv->ramin) {
-			NV_ERROR(dev, "Failed to PRAMIN BAR");
+			NV_ERROR(dev, "Failed to map PRAMIN BAR\n");
 			ret = -ENOMEM;
 			goto err_mmio;
 		}
@@ -1130,7 +1229,7 @@ int nouveau_ioctl_getparam(struct drm_device *dev, void *data,
 		getparam->value = 1;
 		break;
 	case NOUVEAU_GETPARAM_HAS_PAGEFLIP:
-		getparam->value = 1;
+		getparam->value = dev_priv->card_type < NV_D0;
 		break;
 	case NOUVEAU_GETPARAM_GRAPH_UNITS:
 		/* NV40 and NV50 versions are quite different, but register
@@ -1198,6 +1297,23 @@ nouveau_wait_ne(struct drm_device *dev, uint64_t timeout,
 	return false;
 }
 
+/* Wait until cond(data) == true, up until timeout has hit */
+bool
+nouveau_wait_cb(struct drm_device *dev, u64 timeout,
+		bool (*cond)(void *), void *data)
+{
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct nouveau_timer_engine *ptimer = &dev_priv->engine.timer;
+	u64 start = ptimer->read(dev);
+
+	do {
+		if (cond(data) == true)
+			return true;
+	} while (ptimer->read(dev) - start < timeout);
+
+	return false;
+}
+
 /* Waits for PGRAPH to go completely idle */
 bool nouveau_wait_for_idle(struct drm_device *dev)
 {
diff --git a/drivers/gpu/drm/nouveau/nouveau_vm.c b/drivers/gpu/drm/nouveau/nouveau_vm.c
index 244fd38fdb8..ef0832b29ad 100644
--- a/drivers/gpu/drm/nouveau/nouveau_vm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_vm.c
@@ -172,9 +172,9 @@ nouveau_vm_unmap_pgt(struct nouveau_vm *vm, int big, u32 fpde, u32 lpde)
 			vm->map_pgt(vpgd->obj, pde, vpgt->obj);
 		}
 
-		mutex_unlock(&vm->mm->mutex);
+		mutex_unlock(&vm->mm.mutex);
 		nouveau_gpuobj_ref(NULL, &pgt);
-		mutex_lock(&vm->mm->mutex);
+		mutex_lock(&vm->mm.mutex);
 	}
 }
 
@@ -191,18 +191,18 @@ nouveau_vm_map_pgt(struct nouveau_vm *vm, u32 pde, u32 type)
 	pgt_size  = (1 << (vm->pgt_bits + 12)) >> type;
 	pgt_size *= 8;
 
-	mutex_unlock(&vm->mm->mutex);
+	mutex_unlock(&vm->mm.mutex);
 	ret = nouveau_gpuobj_new(vm->dev, NULL, pgt_size, 0x1000,
 				 NVOBJ_FLAG_ZERO_ALLOC, &pgt);
-	mutex_lock(&vm->mm->mutex);
+	mutex_lock(&vm->mm.mutex);
 	if (unlikely(ret))
 		return ret;
 
 	/* someone beat us to filling the PDE while we didn't have the lock */
 	if (unlikely(vpgt->refcount[big]++)) {
-		mutex_unlock(&vm->mm->mutex);
+		mutex_unlock(&vm->mm.mutex);
 		nouveau_gpuobj_ref(NULL, &pgt);
-		mutex_lock(&vm->mm->mutex);
+		mutex_lock(&vm->mm.mutex);
 		return 0;
 	}
 
@@ -223,10 +223,10 @@ nouveau_vm_get(struct nouveau_vm *vm, u64 size, u32 page_shift,
 	u32 fpde, lpde, pde;
 	int ret;
 
-	mutex_lock(&vm->mm->mutex);
-	ret = nouveau_mm_get(vm->mm, page_shift, msize, 0, align, &vma->node);
+	mutex_lock(&vm->mm.mutex);
+	ret = nouveau_mm_get(&vm->mm, page_shift, msize, 0, align, &vma->node);
 	if (unlikely(ret != 0)) {
-		mutex_unlock(&vm->mm->mutex);
+		mutex_unlock(&vm->mm.mutex);
 		return ret;
 	}
 
@@ -245,13 +245,13 @@ nouveau_vm_get(struct nouveau_vm *vm, u64 size, u32 page_shift,
 		if (ret) {
 			if (pde != fpde)
 				nouveau_vm_unmap_pgt(vm, big, fpde, pde - 1);
-			nouveau_mm_put(vm->mm, vma->node);
-			mutex_unlock(&vm->mm->mutex);
+			nouveau_mm_put(&vm->mm, vma->node);
+			mutex_unlock(&vm->mm.mutex);
 			vma->node = NULL;
 			return ret;
 		}
 	}
-	mutex_unlock(&vm->mm->mutex);
+	mutex_unlock(&vm->mm.mutex);
 
 	vma->vm     = vm;
 	vma->offset = (u64)vma->node->offset << 12;
@@ -270,11 +270,11 @@ nouveau_vm_put(struct nouveau_vma *vma)
 	fpde = (vma->node->offset >> vm->pgt_bits);
 	lpde = (vma->node->offset + vma->node->length - 1) >> vm->pgt_bits;
 
-	mutex_lock(&vm->mm->mutex);
+	mutex_lock(&vm->mm.mutex);
 	nouveau_vm_unmap_pgt(vm, vma->node->type != vm->spg_shift, fpde, lpde);
-	nouveau_mm_put(vm->mm, vma->node);
+	nouveau_mm_put(&vm->mm, vma->node);
 	vma->node = NULL;
-	mutex_unlock(&vm->mm->mutex);
+	mutex_unlock(&vm->mm.mutex);
 }
 
 int
@@ -306,7 +306,7 @@ nouveau_vm_new(struct drm_device *dev, u64 offset, u64 length, u64 mm_offset,
 			block = length;
 
 	} else
-	if (dev_priv->card_type == NV_C0) {
+	if (dev_priv->card_type >= NV_C0) {
 		vm->map_pgt = nvc0_vm_map_pgt;
 		vm->map = nvc0_vm_map;
 		vm->map_sg = nvc0_vm_map_sg;
@@ -360,11 +360,11 @@ nouveau_vm_link(struct nouveau_vm *vm, struct nouveau_gpuobj *pgd)
 
 	nouveau_gpuobj_ref(pgd, &vpgd->obj);
 
-	mutex_lock(&vm->mm->mutex);
+	mutex_lock(&vm->mm.mutex);
 	for (i = vm->fpde; i <= vm->lpde; i++)
 		vm->map_pgt(pgd, i, vm->pgt[i - vm->fpde].obj);
 	list_add(&vpgd->head, &vm->pgd_list);
-	mutex_unlock(&vm->mm->mutex);
+	mutex_unlock(&vm->mm.mutex);
 	return 0;
 }
 
@@ -377,7 +377,7 @@ nouveau_vm_unlink(struct nouveau_vm *vm, struct nouveau_gpuobj *mpgd)
 	if (!mpgd)
 		return;
 
-	mutex_lock(&vm->mm->mutex);
+	mutex_lock(&vm->mm.mutex);
 	list_for_each_entry_safe(vpgd, tmp, &vm->pgd_list, head) {
 		if (vpgd->obj == mpgd) {
 			pgd = vpgd->obj;
@@ -386,7 +386,7 @@ nouveau_vm_unlink(struct nouveau_vm *vm, struct nouveau_gpuobj *mpgd)
 			break;
 		}
 	}
-	mutex_unlock(&vm->mm->mutex);
+	mutex_unlock(&vm->mm.mutex);
 
 	nouveau_gpuobj_ref(NULL, &pgd);
 }
diff --git a/drivers/gpu/drm/nouveau/nouveau_vm.h b/drivers/gpu/drm/nouveau/nouveau_vm.h
index 579ca8cc223..6ce995f7797 100644
--- a/drivers/gpu/drm/nouveau/nouveau_vm.h
+++ b/drivers/gpu/drm/nouveau/nouveau_vm.h
@@ -51,7 +51,7 @@ struct nouveau_vma {
 
 struct nouveau_vm {
 	struct drm_device *dev;
-	struct nouveau_mm *mm;
+	struct nouveau_mm mm;
 	int refcount;
 
 	struct list_head pgd_list;
diff --git a/drivers/gpu/drm/nouveau/nouveau_volt.c b/drivers/gpu/drm/nouveau/nouveau_volt.c
index 75e872741d9..86d03e15735 100644
--- a/drivers/gpu/drm/nouveau/nouveau_volt.c
+++ b/drivers/gpu/drm/nouveau/nouveau_volt.c
@@ -27,7 +27,7 @@
 #include "nouveau_drv.h"
 #include "nouveau_pm.h"
 
-static const enum dcb_gpio_tag vidtag[] = { 0x04, 0x05, 0x06, 0x1a };
+static const enum dcb_gpio_tag vidtag[] = { 0x04, 0x05, 0x06, 0x1a, 0x73 };
 static int nr_vidtag = sizeof(vidtag) / sizeof(vidtag[0]);
 
 int
@@ -170,6 +170,13 @@ nouveau_volt_init(struct drm_device *dev)
 		 */
 		vidshift  = 2;
 		break;
+	case 0x40:
+		headerlen = volt[1];
+		recordlen = volt[2];
+		entries   = volt[3]; /* not a clue what the entries are for.. */
+		vidmask   = volt[11]; /* guess.. */
+		vidshift  = 0;
+		break;
 	default:
 		NV_WARN(dev, "voltage table 0x%02x unknown\n", volt[0]);
 		return;
@@ -197,16 +204,37 @@ nouveau_volt_init(struct drm_device *dev)
 	}
 
 	/* parse vbios entries into common format */
-	voltage->level = kcalloc(entries, sizeof(*voltage->level), GFP_KERNEL);
-	if (!voltage->level)
-		return;
+	voltage->version = volt[0];
+	if (voltage->version < 0x40) {
+		voltage->nr_level = entries;
+		voltage->level =
+			kcalloc(entries, sizeof(*voltage->level), GFP_KERNEL);
+		if (!voltage->level)
+			return;
 
-	entry = volt + headerlen;
-	for (i = 0; i < entries; i++, entry += recordlen) {
-		voltage->level[i].voltage = entry[0];
-		voltage->level[i].vid     = entry[1] >> vidshift;
+		entry = volt + headerlen;
+		for (i = 0; i < entries; i++, entry += recordlen) {
+			voltage->level[i].voltage = entry[0] * 10000;
+			voltage->level[i].vid     = entry[1] >> vidshift;
+		}
+	} else {
+		u32 volt_uv = ROM32(volt[4]);
+		s16 step_uv = ROM16(volt[8]);
+		u8 vid;
+
+		voltage->nr_level = voltage->vid_mask + 1;
+		voltage->level = kcalloc(voltage->nr_level,
+					 sizeof(*voltage->level), GFP_KERNEL);
+		if (!voltage->level)
+			return;
+
+		for (vid = 0; vid <= voltage->vid_mask; vid++) {
+			voltage->level[vid].voltage = volt_uv;
+			voltage->level[vid].vid = vid;
+			volt_uv += step_uv;
+		}
 	}
-	voltage->nr_level  = entries;
+
 	voltage->supported = true;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nv04_crtc.c b/drivers/gpu/drm/nouveau/nv04_crtc.c
index 118261d4927..5e45398a9e2 100644
--- a/drivers/gpu/drm/nouveau/nv04_crtc.c
+++ b/drivers/gpu/drm/nouveau/nv04_crtc.c
@@ -781,11 +781,20 @@ nv04_crtc_do_mode_set_base(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
 	struct nv04_crtc_reg *regp = &dev_priv->mode_reg.crtc_reg[nv_crtc->index];
-	struct drm_framebuffer *drm_fb = nv_crtc->base.fb;
-	struct nouveau_framebuffer *fb = nouveau_framebuffer(drm_fb);
+	struct drm_framebuffer *drm_fb;
+	struct nouveau_framebuffer *fb;
 	int arb_burst, arb_lwm;
 	int ret;
 
+	NV_DEBUG_KMS(dev, "index %d\n", nv_crtc->index);
+
+	/* no fb bound */
+	if (!atomic && !crtc->fb) {
+		NV_DEBUG_KMS(dev, "No FB bound\n");
+		return 0;
+	}
+
+
 	/* If atomic, we want to switch to the fb we were passed, so
 	 * now we update pointers to do that.  (We don't pin; just
 	 * assume we're already pinned and update the base address.)
@@ -794,6 +803,8 @@ nv04_crtc_do_mode_set_base(struct drm_crtc *crtc,
 		drm_fb = passed_fb;
 		fb = nouveau_framebuffer(passed_fb);
 	} else {
+		drm_fb = crtc->fb;
+		fb = nouveau_framebuffer(crtc->fb);
 		/* If not atomic, we can go ahead and pin, and unpin the
 		 * old fb we were passed.
 		 */
diff --git a/drivers/gpu/drm/nouveau/nv04_display.c b/drivers/gpu/drm/nouveau/nv04_display.c
index 1715e1464b7..6bd8518d7b2 100644
--- a/drivers/gpu/drm/nouveau/nv04_display.c
+++ b/drivers/gpu/drm/nouveau/nv04_display.c
@@ -126,27 +126,6 @@ nv04_display_create(struct drm_device *dev)
 
 	nouveau_hw_save_vga_fonts(dev, 1);
 
-	drm_mode_config_init(dev);
-	drm_mode_create_scaling_mode_property(dev);
-	drm_mode_create_dithering_property(dev);
-
-	dev->mode_config.funcs = (void *)&nouveau_mode_config_funcs;
-
-	dev->mode_config.min_width = 0;
-	dev->mode_config.min_height = 0;
-	switch (dev_priv->card_type) {
-	case NV_04:
-		dev->mode_config.max_width = 2048;
-		dev->mode_config.max_height = 2048;
-		break;
-	default:
-		dev->mode_config.max_width = 4096;
-		dev->mode_config.max_height = 4096;
-		break;
-	}
-
-	dev->mode_config.fb_base = dev_priv->fb_phys;
-
 	nv04_crtc_create(dev, 0);
 	if (nv_two_heads(dev))
 		nv04_crtc_create(dev, 1);
@@ -235,8 +214,6 @@ nv04_display_destroy(struct drm_device *dev)
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
 		crtc->funcs->restore(crtc);
 
-	drm_mode_config_cleanup(dev);
-
 	nouveau_hw_save_vga_fonts(dev, 0);
 }
 
diff --git a/drivers/gpu/drm/nouveau/nv04_pm.c b/drivers/gpu/drm/nouveau/nv04_pm.c
index eb1c70dd82e..9ae92a87b8c 100644
--- a/drivers/gpu/drm/nouveau/nv04_pm.c
+++ b/drivers/gpu/drm/nouveau/nv04_pm.c
@@ -68,6 +68,7 @@ void
 nv04_pm_clock_set(struct drm_device *dev, void *pre_state)
 {
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct nouveau_timer_engine *ptimer = &dev_priv->engine.timer;
 	struct nv04_pm_state *state = pre_state;
 	u32 reg = state->pll.reg;
 
@@ -85,6 +86,9 @@ nv04_pm_clock_set(struct drm_device *dev, void *pre_state)
 		nv_mask(dev, 0x1002c0, 0, 1 << 8);
 	}
 
+	if (reg == NV_PRAMDAC_NVPLL_COEFF)
+		ptimer->init(dev);
+
 	kfree(state);
 }
 
diff --git a/drivers/gpu/drm/nouveau/nv04_timer.c b/drivers/gpu/drm/nouveau/nv04_timer.c
index 1d09ddd5739..263301b809d 100644
--- a/drivers/gpu/drm/nouveau/nv04_timer.c
+++ b/drivers/gpu/drm/nouveau/nv04_timer.c
@@ -6,43 +6,75 @@
 int
 nv04_timer_init(struct drm_device *dev)
 {
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	u32 m, n, d;
+
 	nv_wr32(dev, NV04_PTIMER_INTR_EN_0, 0x00000000);
 	nv_wr32(dev, NV04_PTIMER_INTR_0, 0xFFFFFFFF);
 
-	/* Just use the pre-existing values when possible for now; these regs
-	 * are not written in nv (driver writer missed a /4 on the address), and
-	 * writing 8 and 3 to the correct regs breaks the timings on the LVDS
-	 * hardware sequencing microcode.
-	 * A correct solution (involving calculations with the GPU PLL) can
-	 * be done when kernel modesetting lands
-	 */
-	if (!nv_rd32(dev, NV04_PTIMER_NUMERATOR) ||
-				!nv_rd32(dev, NV04_PTIMER_DENOMINATOR)) {
-		nv_wr32(dev, NV04_PTIMER_NUMERATOR, 0x00000008);
-		nv_wr32(dev, NV04_PTIMER_DENOMINATOR, 0x00000003);
+	/* aim for 31.25MHz, which gives us nanosecond timestamps */
+	d = 1000000 / 32;
+
+	/* determine base clock for timer source */
+	if (dev_priv->chipset < 0x40) {
+		n = dev_priv->engine.pm.clock_get(dev, PLL_CORE);
+	} else
+	if (dev_priv->chipset == 0x40) {
+		/*XXX: figure this out */
+		n = 0;
+	} else {
+		n = dev_priv->crystal;
+		m = 1;
+		while (n < (d * 2)) {
+			n += (n / m);
+			m++;
+		}
+
+		nv_wr32(dev, 0x009220, m - 1);
+	}
+
+	if (!n) {
+		NV_WARN(dev, "PTIMER: unknown input clock freq\n");
+		if (!nv_rd32(dev, NV04_PTIMER_NUMERATOR) ||
+		    !nv_rd32(dev, NV04_PTIMER_DENOMINATOR)) {
+			nv_wr32(dev, NV04_PTIMER_NUMERATOR, 1);
+			nv_wr32(dev, NV04_PTIMER_DENOMINATOR, 1);
+		}
+		return 0;
+	}
+
+	/* reduce ratio to acceptable values */
+	while (((n % 5) == 0) && ((d % 5) == 0)) {
+		n /= 5;
+		d /= 5;
 	}
 
+	while (((n % 2) == 0) && ((d % 2) == 0)) {
+		n /= 2;
+		d /= 2;
+	}
+
+	while (n > 0xffff || d > 0xffff) {
+		n >>= 1;
+		d >>= 1;
+	}
+
+	nv_wr32(dev, NV04_PTIMER_NUMERATOR, n);
+	nv_wr32(dev, NV04_PTIMER_DENOMINATOR, d);
 	return 0;
 }
 
-uint64_t
+u64
 nv04_timer_read(struct drm_device *dev)
 {
-	uint32_t low;
-	/* From kmmio dumps on nv28 this looks like how the blob does this.
-	 * It reads the high dword twice, before and after.
-	 * The only explanation seems to be that the 64-bit timer counter
-	 * advances between high and low dword reads and may corrupt the
-	 * result. Not confirmed.
-	 */
-	uint32_t high2 = nv_rd32(dev, NV04_PTIMER_TIME_1);
-	uint32_t high1;
+	u32 hi, lo;
+
 	do {
-		high1 = high2;
-		low = nv_rd32(dev, NV04_PTIMER_TIME_0);
-		high2 = nv_rd32(dev, NV04_PTIMER_TIME_1);
-	} while (high1 != high2);
-	return (((uint64_t)high2) << 32) | (uint64_t)low;
+		hi = nv_rd32(dev, NV04_PTIMER_TIME_1);
+		lo = nv_rd32(dev, NV04_PTIMER_TIME_0);
+	} while (hi != nv_rd32(dev, NV04_PTIMER_TIME_1));
+
+	return ((u64)hi << 32 | lo);
 }
 
 void
diff --git a/drivers/gpu/drm/nouveau/nv40_mpeg.c b/drivers/gpu/drm/nouveau/nv31_mpeg.c
index ad03a0e1fc7..6f06a0713f0 100644
--- a/drivers/gpu/drm/nouveau/nv40_mpeg.c
+++ b/drivers/gpu/drm/nouveau/nv31_mpeg.c
@@ -26,10 +26,32 @@
 #include "nouveau_drv.h"
 #include "nouveau_ramht.h"
 
-struct nv40_mpeg_engine {
+struct nv31_mpeg_engine {
 	struct nouveau_exec_engine base;
+	atomic_t refcount;
 };
 
+
+static int
+nv31_mpeg_context_new(struct nouveau_channel *chan, int engine)
+{
+	struct nv31_mpeg_engine *pmpeg = nv_engine(chan->dev, engine);
+
+	if (!atomic_add_unless(&pmpeg->refcount, 1, 1))
+		return -EBUSY;
+
+	chan->engctx[engine] = (void *)0xdeadcafe;
+	return 0;
+}
+
+static void
+nv31_mpeg_context_del(struct nouveau_channel *chan, int engine)
+{
+	struct nv31_mpeg_engine *pmpeg = nv_engine(chan->dev, engine);
+	atomic_dec(&pmpeg->refcount);
+	chan->engctx[engine] = NULL;
+}
+
 static int
 nv40_mpeg_context_new(struct nouveau_channel *chan, int engine)
 {
@@ -81,7 +103,7 @@ nv40_mpeg_context_del(struct nouveau_channel *chan, int engine)
 }
 
 static int
-nv40_mpeg_object_new(struct nouveau_channel *chan, int engine,
+nv31_mpeg_object_new(struct nouveau_channel *chan, int engine,
 		      u32 handle, u16 class)
 {
 	struct drm_device *dev = chan->dev;
@@ -103,10 +125,10 @@ nv40_mpeg_object_new(struct nouveau_channel *chan, int engine,
 }
 
 static int
-nv40_mpeg_init(struct drm_device *dev, int engine)
+nv31_mpeg_init(struct drm_device *dev, int engine)
 {
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
-	struct nv40_mpeg_engine *pmpeg = nv_engine(dev, engine);
+	struct nv31_mpeg_engine *pmpeg = nv_engine(dev, engine);
 	int i;
 
 	/* VPE init */
@@ -121,7 +143,7 @@ nv40_mpeg_init(struct drm_device *dev, int engine)
 	/* PMPEG init */
 	nv_wr32(dev, 0x00b32c, 0x00000000);
 	nv_wr32(dev, 0x00b314, 0x00000100);
-	nv_wr32(dev, 0x00b220, 0x00000044);
+	nv_wr32(dev, 0x00b220, nv44_graph_class(dev) ? 0x00000044 : 0x00000031);
 	nv_wr32(dev, 0x00b300, 0x02001ec1);
 	nv_mask(dev, 0x00b32c, 0x00000001, 0x00000001);
 
@@ -137,7 +159,7 @@ nv40_mpeg_init(struct drm_device *dev, int engine)
 }
 
 static int
-nv40_mpeg_fini(struct drm_device *dev, int engine, bool suspend)
+nv31_mpeg_fini(struct drm_device *dev, int engine, bool suspend)
 {
 	/*XXX: context save? */
 	nv_mask(dev, 0x00b32c, 0x00000001, 0x00000000);
@@ -146,7 +168,7 @@ nv40_mpeg_fini(struct drm_device *dev, int engine, bool suspend)
 }
 
 static int
-nv40_mpeg_mthd_dma(struct nouveau_channel *chan, u32 class, u32 mthd, u32 data)
+nv31_mpeg_mthd_dma(struct nouveau_channel *chan, u32 class, u32 mthd, u32 data)
 {
 	struct drm_device *dev = chan->dev;
 	u32 inst = data << 4;
@@ -184,13 +206,17 @@ nv40_mpeg_mthd_dma(struct nouveau_channel *chan, u32 class, u32 mthd, u32 data)
 }
 
 static int
-nv40_mpeg_isr_chid(struct drm_device *dev, u32 inst)
+nv31_mpeg_isr_chid(struct drm_device *dev, u32 inst)
 {
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
 	struct nouveau_gpuobj *ctx;
 	unsigned long flags;
 	int i;
 
+	/* hardcode drm channel id on nv3x, so swmthd lookup works */
+	if (dev_priv->card_type < NV_40)
+		return 0;
+
 	spin_lock_irqsave(&dev_priv->channels.lock, flags);
 	for (i = 0; i < dev_priv->engine.fifo.channels; i++) {
 		if (!dev_priv->channels.ptr[i])
@@ -205,7 +231,7 @@ nv40_mpeg_isr_chid(struct drm_device *dev, u32 inst)
 }
 
 static void
-nv40_vpe_set_tile_region(struct drm_device *dev, int i)
+nv31_vpe_set_tile_region(struct drm_device *dev, int i)
 {
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
 	struct nouveau_tile_reg *tile = &dev_priv->tile.reg[i];
@@ -216,10 +242,10 @@ nv40_vpe_set_tile_region(struct drm_device *dev, int i)
 }
 
 static void
-nv40_mpeg_isr(struct drm_device *dev)
+nv31_mpeg_isr(struct drm_device *dev)
 {
 	u32 inst = (nv_rd32(dev, 0x00b318) & 0x000fffff) << 4;
-	u32 chid = nv40_mpeg_isr_chid(dev, inst);
+	u32 chid = nv31_mpeg_isr_chid(dev, inst);
 	u32 stat = nv_rd32(dev, 0x00b100);
 	u32 type = nv_rd32(dev, 0x00b230);
 	u32 mthd = nv_rd32(dev, 0x00b234);
@@ -249,10 +275,10 @@ nv40_mpeg_isr(struct drm_device *dev)
 }
 
 static void
-nv40_vpe_isr(struct drm_device *dev)
+nv31_vpe_isr(struct drm_device *dev)
 {
 	if (nv_rd32(dev, 0x00b100))
-		nv40_mpeg_isr(dev);
+		nv31_mpeg_isr(dev);
 
 	if (nv_rd32(dev, 0x00b800)) {
 		u32 stat = nv_rd32(dev, 0x00b800);
@@ -262,9 +288,9 @@ nv40_vpe_isr(struct drm_device *dev)
 }
 
 static void
-nv40_mpeg_destroy(struct drm_device *dev, int engine)
+nv31_mpeg_destroy(struct drm_device *dev, int engine)
 {
-	struct nv40_mpeg_engine *pmpeg = nv_engine(dev, engine);
+	struct nv31_mpeg_engine *pmpeg = nv_engine(dev, engine);
 
 	nouveau_irq_unregister(dev, 0);
 
@@ -273,34 +299,41 @@ nv40_mpeg_destroy(struct drm_device *dev, int engine)
 }
 
 int
-nv40_mpeg_create(struct drm_device *dev)
+nv31_mpeg_create(struct drm_device *dev)
 {
-	struct nv40_mpeg_engine *pmpeg;
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct nv31_mpeg_engine *pmpeg;
 
 	pmpeg = kzalloc(sizeof(*pmpeg), GFP_KERNEL);
 	if (!pmpeg)
 		return -ENOMEM;
-
-	pmpeg->base.destroy = nv40_mpeg_destroy;
-	pmpeg->base.init = nv40_mpeg_init;
-	pmpeg->base.fini = nv40_mpeg_fini;
-	pmpeg->base.context_new = nv40_mpeg_context_new;
-	pmpeg->base.context_del = nv40_mpeg_context_del;
-	pmpeg->base.object_new = nv40_mpeg_object_new;
+	atomic_set(&pmpeg->refcount, 0);
+
+	pmpeg->base.destroy = nv31_mpeg_destroy;
+	pmpeg->base.init = nv31_mpeg_init;
+	pmpeg->base.fini = nv31_mpeg_fini;
+	if (dev_priv->card_type < NV_40) {
+		pmpeg->base.context_new = nv31_mpeg_context_new;
+		pmpeg->base.context_del = nv31_mpeg_context_del;
+	} else {
+		pmpeg->base.context_new = nv40_mpeg_context_new;
+		pmpeg->base.context_del = nv40_mpeg_context_del;
+	}
+	pmpeg->base.object_new = nv31_mpeg_object_new;
 
 	/* ISR vector, PMC_ENABLE bit,  and TILE regs are shared between
 	 * all VPE engines, for this driver's purposes the PMPEG engine
 	 * will be treated as the "master" and handle the global VPE
 	 * bits too
 	 */
-	pmpeg->base.set_tile_region = nv40_vpe_set_tile_region;
-	nouveau_irq_register(dev, 0, nv40_vpe_isr);
+	pmpeg->base.set_tile_region = nv31_vpe_set_tile_region;
+	nouveau_irq_register(dev, 0, nv31_vpe_isr);
 
 	NVOBJ_ENGINE_ADD(dev, MPEG, &pmpeg->base);
 	NVOBJ_CLASS(dev, 0x3174, MPEG);
-	NVOBJ_MTHD (dev, 0x3174, 0x0190, nv40_mpeg_mthd_dma);
-	NVOBJ_MTHD (dev, 0x3174, 0x01a0, nv40_mpeg_mthd_dma);
-	NVOBJ_MTHD (dev, 0x3174, 0x01b0, nv40_mpeg_mthd_dma);
+	NVOBJ_MTHD (dev, 0x3174, 0x0190, nv31_mpeg_mthd_dma);
+	NVOBJ_MTHD (dev, 0x3174, 0x01a0, nv31_mpeg_mthd_dma);
+	NVOBJ_MTHD (dev, 0x3174, 0x01b0, nv31_mpeg_mthd_dma);
 
 #if 0
 	NVOBJ_ENGINE_ADD(dev, ME, &pme->base);
diff --git a/drivers/gpu/drm/nouveau/nv40_pm.c b/drivers/gpu/drm/nouveau/nv40_pm.c
new file mode 100644
index 00000000000..bbc0b9c7e1f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nv40_pm.c
@@ -0,0 +1,338 @@
+/*
+ * Copyright 2011 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "drmP.h"
+#include "nouveau_drv.h"
+#include "nouveau_bios.h"
+#include "nouveau_pm.h"
+#include "nouveau_hw.h"
+
+#define min2(a,b) ((a) < (b) ? (a) : (b))
+
+static u32
+read_pll_1(struct drm_device *dev, u32 reg)
+{
+	u32 ctrl = nv_rd32(dev, reg + 0x00);
+	int P = (ctrl & 0x00070000) >> 16;
+	int N = (ctrl & 0x0000ff00) >> 8;
+	int M = (ctrl & 0x000000ff) >> 0;
+	u32 ref = 27000, clk = 0;
+
+	if (ctrl & 0x80000000)
+		clk = ref * N / M;
+
+	return clk >> P;
+}
+
+static u32
+read_pll_2(struct drm_device *dev, u32 reg)
+{
+	u32 ctrl = nv_rd32(dev, reg + 0x00);
+	u32 coef = nv_rd32(dev, reg + 0x04);
+	int N2 = (coef & 0xff000000) >> 24;
+	int M2 = (coef & 0x00ff0000) >> 16;
+	int N1 = (coef & 0x0000ff00) >> 8;
+	int M1 = (coef & 0x000000ff) >> 0;
+	int P = (ctrl & 0x00070000) >> 16;
+	u32 ref = 27000, clk = 0;
+
+	if (ctrl & 0x80000000)
+		clk = ref * N1 / M1;
+
+	if (!(ctrl & 0x00000100)) {
+		if (ctrl & 0x40000000)
+			clk = clk * N2 / M2;
+	}
+
+	return clk >> P;
+}
+
+static u32
+read_clk(struct drm_device *dev, u32 src)
+{
+	switch (src) {
+	case 3:
+		return read_pll_2(dev, 0x004000);
+	case 2:
+		return read_pll_1(dev, 0x004008);
+	default:
+		break;
+	}
+
+	return 0;
+}
+
+int
+nv40_pm_clocks_get(struct drm_device *dev, struct nouveau_pm_level *perflvl)
+{
+	u32 ctrl = nv_rd32(dev, 0x00c040);
+
+	perflvl->core   = read_clk(dev, (ctrl & 0x00000003) >> 0);
+	perflvl->shader = read_clk(dev, (ctrl & 0x00000030) >> 4);
+	perflvl->memory = read_pll_2(dev, 0x4020);
+	return 0;
+}
+
+struct nv40_pm_state {
+	u32 ctrl;
+	u32 npll_ctrl;
+	u32 npll_coef;
+	u32 spll;
+	u32 mpll_ctrl;
+	u32 mpll_coef;
+};
+
+static int
+nv40_calc_pll(struct drm_device *dev, u32 reg, struct pll_lims *pll,
+	      u32 clk, int *N1, int *M1, int *N2, int *M2, int *log2P)
+{
+	struct nouveau_pll_vals coef;
+	int ret;
+
+	ret = get_pll_limits(dev, reg, pll);
+	if (ret)
+		return ret;
+
+	if (clk < pll->vco1.maxfreq)
+		pll->vco2.maxfreq = 0;
+
+	ret = nouveau_calc_pll_mnp(dev, pll, clk, &coef);
+	if (ret == 0)
+		return -ERANGE;
+
+	*N1 = coef.N1;
+	*M1 = coef.M1;
+	if (N2 && M2) {
+		if (pll->vco2.maxfreq) {
+			*N2 = coef.N2;
+			*M2 = coef.M2;
+		} else {
+			*N2 = 1;
+			*M2 = 1;
+		}
+	}
+	*log2P = coef.log2P;
+	return 0;
+}
+
+void *
+nv40_pm_clocks_pre(struct drm_device *dev, struct nouveau_pm_level *perflvl)
+{
+	struct nv40_pm_state *info;
+	struct pll_lims pll;
+	int N1, N2, M1, M2, log2P;
+	int ret;
+
+	info = kmalloc(sizeof(*info), GFP_KERNEL);
+	if (!info)
+		return ERR_PTR(-ENOMEM);
+
+	/* core/geometric clock */
+	ret = nv40_calc_pll(dev, 0x004000, &pll, perflvl->core,
+			    &N1, &M1, &N2, &M2, &log2P);
+	if (ret < 0)
+		goto out;
+
+	if (N2 == M2) {
+		info->npll_ctrl = 0x80000100 | (log2P << 16);
+		info->npll_coef = (N1 << 8) | M1;
+	} else {
+		info->npll_ctrl = 0xc0000000 | (log2P << 16);
+		info->npll_coef = (N2 << 24) | (M2 << 16) | (N1 << 8) | M1;
+	}
+
+	/* use the second PLL for shader/rop clock, if it differs from core */
+	if (perflvl->shader && perflvl->shader != perflvl->core) {
+		ret = nv40_calc_pll(dev, 0x004008, &pll, perflvl->shader,
+				    &N1, &M1, NULL, NULL, &log2P);
+		if (ret < 0)
+			goto out;
+
+		info->spll = 0xc0000000 | (log2P << 16) | (N1 << 8) | M1;
+		info->ctrl = 0x00000223;
+	} else {
+		info->spll = 0x00000000;
+		info->ctrl = 0x00000333;
+	}
+
+	/* memory clock */
+	ret = nv40_calc_pll(dev, 0x004020, &pll, perflvl->memory,
+			    &N1, &M1, &N2, &M2, &log2P);
+	if (ret < 0)
+		goto out;
+
+	info->mpll_ctrl  = 0x80000000 | (log2P << 16);
+	info->mpll_ctrl |= min2(pll.log2p_bias + log2P, pll.max_log2p) << 20;
+	if (N2 == M2) {
+		info->mpll_ctrl |= 0x00000100;
+		info->mpll_coef  = (N1 << 8) | M1;
+	} else {
+		info->mpll_ctrl |= 0x40000000;
+		info->mpll_coef  = (N2 << 24) | (M2 << 16) | (N1 << 8) | M1;
+	}
+
+out:
+	if (ret < 0) {
+		kfree(info);
+		info = ERR_PTR(ret);
+	}
+	return info;
+}
+
+static bool
+nv40_pm_gr_idle(void *data)
+{
+	struct drm_device *dev = data;
+
+	if ((nv_rd32(dev, 0x400760) & 0x000000f0) >> 4 !=
+	    (nv_rd32(dev, 0x400760) & 0x0000000f))
+		return false;
+
+	if (nv_rd32(dev, 0x400700))
+		return false;
+
+	return true;
+}
+
+void
+nv40_pm_clocks_set(struct drm_device *dev, void *pre_state)
+{
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct nv40_pm_state *info = pre_state;
+	unsigned long flags;
+	struct bit_entry M;
+	u32 crtc_mask = 0;
+	u8 sr1[2];
+	int i;
+
+	/* determine which CRTCs are active, fetch VGA_SR1 for each */
+	for (i = 0; i < 2; i++) {
+		u32 vbl = nv_rd32(dev, 0x600808 + (i * 0x2000));
+		u32 cnt = 0;
+		do {
+			if (vbl != nv_rd32(dev, 0x600808 + (i * 0x2000))) {
+				nv_wr08(dev, 0x0c03c4 + (i * 0x2000), 0x01);
+				sr1[i] = nv_rd08(dev, 0x0c03c5 + (i * 0x2000));
+				if (!(sr1[i] & 0x20))
+					crtc_mask |= (1 << i);
+				break;
+			}
+			udelay(1);
+		} while (cnt++ < 32);
+	}
+
+	/* halt and idle engines */
+	spin_lock_irqsave(&dev_priv->context_switch_lock, flags);
+	nv_mask(dev, 0x002500, 0x00000001, 0x00000000);
+	if (!nv_wait(dev, 0x002500, 0x00000010, 0x00000000))
+		goto resume;
+	nv_mask(dev, 0x003220, 0x00000001, 0x00000000);
+	if (!nv_wait(dev, 0x003220, 0x00000010, 0x00000000))
+		goto resume;
+	nv_mask(dev, 0x003200, 0x00000001, 0x00000000);
+	nv04_fifo_cache_pull(dev, false);
+
+	if (!nv_wait_cb(dev, nv40_pm_gr_idle, dev))
+		goto resume;
+
+	/* set engine clocks */
+	nv_mask(dev, 0x00c040, 0x00000333, 0x00000000);
+	nv_wr32(dev, 0x004004, info->npll_coef);
+	nv_mask(dev, 0x004000, 0xc0070100, info->npll_ctrl);
+	nv_mask(dev, 0x004008, 0xc007ffff, info->spll);
+	mdelay(5);
+	nv_mask(dev, 0x00c040, 0x00000333, info->ctrl);
+
+	/* wait for vblank start on active crtcs, disable memory access */
+	for (i = 0; i < 2; i++) {
+		if (!(crtc_mask & (1 << i)))
+			continue;
+		nv_wait(dev, 0x600808 + (i * 0x2000), 0x00010000, 0x00000000);
+		nv_wait(dev, 0x600808 + (i * 0x2000), 0x00010000, 0x00010000);
+		nv_wr08(dev, 0x0c03c4 + (i * 0x2000), 0x01);
+		nv_wr08(dev, 0x0c03c5 + (i * 0x2000), sr1[i] | 0x20);
+	}
+
+	/* prepare ram for reclocking */
+	nv_wr32(dev, 0x1002d4, 0x00000001); /* precharge */
+	nv_wr32(dev, 0x1002d0, 0x00000001); /* refresh */
+	nv_wr32(dev, 0x1002d0, 0x00000001); /* refresh */
+	nv_mask(dev, 0x100210, 0x80000000, 0x00000000); /* no auto refresh */
+	nv_wr32(dev, 0x1002dc, 0x00000001); /* enable self-refresh */
+
+	/* change the PLL of each memory partition */
+	nv_mask(dev, 0x00c040, 0x0000c000, 0x00000000);
+	switch (dev_priv->chipset) {
+	case 0x40:
+	case 0x45:
+	case 0x41:
+	case 0x42:
+	case 0x47:
+		nv_mask(dev, 0x004044, 0xc0771100, info->mpll_ctrl);
+		nv_mask(dev, 0x00402c, 0xc0771100, info->mpll_ctrl);
+		nv_wr32(dev, 0x004048, info->mpll_coef);
+		nv_wr32(dev, 0x004030, info->mpll_coef);
+	case 0x43:
+	case 0x49:
+	case 0x4b:
+		nv_mask(dev, 0x004038, 0xc0771100, info->mpll_ctrl);
+		nv_wr32(dev, 0x00403c, info->mpll_coef);
+	default:
+		nv_mask(dev, 0x004020, 0xc0771100, info->mpll_ctrl);
+		nv_wr32(dev, 0x004024, info->mpll_coef);
+		break;
+	}
+	udelay(100);
+	nv_mask(dev, 0x00c040, 0x0000c000, 0x0000c000);
+
+	/* re-enable normal operation of memory controller */
+	nv_wr32(dev, 0x1002dc, 0x00000000);
+	nv_mask(dev, 0x100210, 0x80000000, 0x80000000);
+	udelay(100);
+
+	/* execute memory reset script from vbios */
+	if (!bit_table(dev, 'M', &M))
+		nouveau_bios_init_exec(dev, ROM16(M.data[0]));
+
+	/* make sure we're in vblank (hopefully the same one as before), and
+	 * then re-enable crtc memory access
+	 */
+	for (i = 0; i < 2; i++) {
+		if (!(crtc_mask & (1 << i)))
+			continue;
+		nv_wait(dev, 0x600808 + (i * 0x2000), 0x00010000, 0x00010000);
+		nv_wr08(dev, 0x0c03c4 + (i * 0x2000), 0x01);
+		nv_wr08(dev, 0x0c03c5 + (i * 0x2000), sr1[i]);
+	}
+
+	/* resume engines */
+resume:
+	nv_wr32(dev, 0x003250, 0x00000001);
+	nv_mask(dev, 0x003220, 0x00000001, 0x00000001);
+	nv_wr32(dev, 0x003200, 0x00000001);
+	nv_wr32(dev, 0x002500, 0x00000001);
+	spin_unlock_irqrestore(&dev_priv->context_switch_lock, flags);
+
+	kfree(info);
+}
diff --git a/drivers/gpu/drm/nouveau/nv50_crtc.c b/drivers/gpu/drm/nouveau/nv50_crtc.c
index 46ad59ea218..882080e0b4f 100644
--- a/drivers/gpu/drm/nouveau/nv50_crtc.c
+++ b/drivers/gpu/drm/nouveau/nv50_crtc.c
@@ -329,8 +329,6 @@ nv50_crtc_destroy(struct drm_crtc *crtc)
 
 	drm_crtc_cleanup(&nv_crtc->base);
 
-	nv50_cursor_fini(nv_crtc);
-
 	nouveau_bo_unmap(nv_crtc->lut.nvbo);
 	nouveau_bo_ref(NULL, &nv_crtc->lut.nvbo);
 	nouveau_bo_unmap(nv_crtc->cursor.nvbo);
@@ -519,12 +517,18 @@ nv50_crtc_do_mode_set_base(struct drm_crtc *crtc,
 	struct drm_device *dev = nv_crtc->base.dev;
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
 	struct nouveau_channel *evo = nv50_display(dev)->master;
-	struct drm_framebuffer *drm_fb = nv_crtc->base.fb;
-	struct nouveau_framebuffer *fb = nouveau_framebuffer(drm_fb);
+	struct drm_framebuffer *drm_fb;
+	struct nouveau_framebuffer *fb;
 	int ret;
 
 	NV_DEBUG_KMS(dev, "index %d\n", nv_crtc->index);
 
+	/* no fb bound */
+	if (!atomic && !crtc->fb) {
+		NV_DEBUG_KMS(dev, "No FB bound\n");
+		return 0;
+	}
+
 	/* If atomic, we want to switch to the fb we were passed, so
 	 * now we update pointers to do that.  (We don't pin; just
 	 * assume we're already pinned and update the base address.)
@@ -533,6 +537,8 @@ nv50_crtc_do_mode_set_base(struct drm_crtc *crtc,
 		drm_fb = passed_fb;
 		fb = nouveau_framebuffer(passed_fb);
 	} else {
+		drm_fb = crtc->fb;
+		fb = nouveau_framebuffer(crtc->fb);
 		/* If not atomic, we can go ahead and pin, and unpin the
 		 * old fb we were passed.
 		 */
diff --git a/drivers/gpu/drm/nouveau/nv50_cursor.c b/drivers/gpu/drm/nouveau/nv50_cursor.c
index 9752c35bb84..adfc9b607a5 100644
--- a/drivers/gpu/drm/nouveau/nv50_cursor.c
+++ b/drivers/gpu/drm/nouveau/nv50_cursor.c
@@ -137,21 +137,3 @@ nv50_cursor_init(struct nouveau_crtc *nv_crtc)
 	nv_crtc->cursor.show = nv50_cursor_show;
 	return 0;
 }
-
-void
-nv50_cursor_fini(struct nouveau_crtc *nv_crtc)
-{
-	struct drm_device *dev = nv_crtc->base.dev;
-	int idx = nv_crtc->index;
-
-	NV_DEBUG_KMS(dev, "\n");
-
-	nv_wr32(dev, NV50_PDISPLAY_CURSOR_CURSOR_CTRL2(idx), 0);
-	if (!nv_wait(dev, NV50_PDISPLAY_CURSOR_CURSOR_CTRL2(idx),
-		     NV50_PDISPLAY_CURSOR_CURSOR_CTRL2_STATUS, 0)) {
-		NV_ERROR(dev, "timeout: CURSOR_CTRL2_STATUS == 0\n");
-		NV_ERROR(dev, "CURSOR_CTRL2 = 0x%08x\n",
-			 nv_rd32(dev, NV50_PDISPLAY_CURSOR_CURSOR_CTRL2(idx)));
-	}
-}
-
diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
index db1a5f4b711..d23ca00e7d6 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ b/drivers/gpu/drm/nouveau/nv50_display.c
@@ -247,6 +247,16 @@ static int nv50_display_disable(struct drm_device *dev)
 		}
 	}
 
+	for (i = 0; i < 2; i++) {
+		nv_wr32(dev, NV50_PDISPLAY_CURSOR_CURSOR_CTRL2(i), 0);
+		if (!nv_wait(dev, NV50_PDISPLAY_CURSOR_CURSOR_CTRL2(i),
+			     NV50_PDISPLAY_CURSOR_CURSOR_CTRL2_STATUS, 0)) {
+			NV_ERROR(dev, "timeout: CURSOR_CTRL2_STATUS == 0\n");
+			NV_ERROR(dev, "CURSOR_CTRL2 = 0x%08x\n",
+				 nv_rd32(dev, NV50_PDISPLAY_CURSOR_CURSOR_CTRL2(i)));
+		}
+	}
+
 	nv50_evo_fini(dev);
 
 	for (i = 0; i < 3; i++) {
@@ -286,23 +296,6 @@ int nv50_display_create(struct drm_device *dev)
 		return -ENOMEM;
 	dev_priv->engine.display.priv = priv;
 
-	/* init basic kernel modesetting */
-	drm_mode_config_init(dev);
-
-	/* Initialise some optional connector properties. */
-	drm_mode_create_scaling_mode_property(dev);
-	drm_mode_create_dithering_property(dev);
-
-	dev->mode_config.min_width = 0;
-	dev->mode_config.min_height = 0;
-
-	dev->mode_config.funcs = (void *)&nouveau_mode_config_funcs;
-
-	dev->mode_config.max_width = 8192;
-	dev->mode_config.max_height = 8192;
-
-	dev->mode_config.fb_base = dev_priv->fb_phys;
-
 	/* Create CRTC objects */
 	for (i = 0; i < 2; i++)
 		nv50_crtc_create(dev, i);
@@ -364,8 +357,6 @@ nv50_display_destroy(struct drm_device *dev)
 
 	NV_DEBUG_KMS(dev, "\n");
 
-	drm_mode_config_cleanup(dev);
-
 	nv50_display_disable(dev);
 	nouveau_irq_unregister(dev, 26);
 	kfree(disp);
@@ -698,7 +689,7 @@ nv50_display_unk10_handler(struct drm_device *dev)
 		struct dcb_entry *dcb = &dev_priv->vbios.dcb.entry[i];
 
 		if (dcb->type == type && (dcb->or & (1 << or))) {
-			nouveau_bios_run_display_table(dev, dcb, 0, -1);
+			nouveau_bios_run_display_table(dev, 0, -1, dcb, -1);
 			disp->irq.dcb = dcb;
 			goto ack;
 		}
@@ -711,37 +702,6 @@ ack:
 }
 
 static void
-nv50_display_unk20_dp_hack(struct drm_device *dev, struct dcb_entry *dcb)
-{
-	int or = ffs(dcb->or) - 1, link = !(dcb->dpconf.sor.link & 1);
-	struct drm_encoder *encoder;
-	uint32_t tmp, unk0 = 0, unk1 = 0;
-
-	if (dcb->type != OUTPUT_DP)
-		return;
-
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
-		struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-
-		if (nv_encoder->dcb == dcb) {
-			unk0 = nv_encoder->dp.unk0;
-			unk1 = nv_encoder->dp.unk1;
-			break;
-		}
-	}
-
-	if (unk0 || unk1) {
-		tmp  = nv_rd32(dev, NV50_SOR_DP_CTRL(or, link));
-		tmp &= 0xfffffe03;
-		nv_wr32(dev, NV50_SOR_DP_CTRL(or, link), tmp | unk0);
-
-		tmp  = nv_rd32(dev, NV50_SOR_DP_UNK128(or, link));
-		tmp &= 0xfef080c0;
-		nv_wr32(dev, NV50_SOR_DP_UNK128(or, link), tmp | unk1);
-	}
-}
-
-static void
 nv50_display_unk20_handler(struct drm_device *dev)
 {
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
@@ -753,7 +713,7 @@ nv50_display_unk20_handler(struct drm_device *dev)
 	NV_DEBUG_KMS(dev, "0x610030: 0x%08x\n", unk30);
 	dcb = disp->irq.dcb;
 	if (dcb) {
-		nouveau_bios_run_display_table(dev, dcb, 0, -2);
+		nouveau_bios_run_display_table(dev, 0, -2, dcb, -1);
 		disp->irq.dcb = NULL;
 	}
 
@@ -837,9 +797,15 @@ nv50_display_unk20_handler(struct drm_device *dev)
 	}
 
 	script = nv50_display_script_select(dev, dcb, mc, pclk);
-	nouveau_bios_run_display_table(dev, dcb, script, pclk);
+	nouveau_bios_run_display_table(dev, script, pclk, dcb, -1);
 
-	nv50_display_unk20_dp_hack(dev, dcb);
+	if (type == OUTPUT_DP) {
+		int link = !(dcb->dpconf.sor.link & 1);
+		if ((mc & 0x000f0000) == 0x00020000)
+			nouveau_dp_tu_update(dev, or, link, pclk, 18);
+		else
+			nouveau_dp_tu_update(dev, or, link, pclk, 24);
+	}
 
 	if (dcb->type != OUTPUT_ANALOG) {
 		tmp = nv_rd32(dev, NV50_PDISPLAY_SOR_CLK_CTRL2(or));
@@ -904,7 +870,7 @@ nv50_display_unk40_handler(struct drm_device *dev)
 	if (!dcb)
 		goto ack;
 
-	nouveau_bios_run_display_table(dev, dcb, script, -pclk);
+	nouveau_bios_run_display_table(dev, script, -pclk, dcb, -1);
 	nv50_display_unk40_dp_set_tmds(dev, dcb);
 
 ack:
diff --git a/drivers/gpu/drm/nouveau/nv50_gpio.c b/drivers/gpu/drm/nouveau/nv50_gpio.c
index d4f4206dad7..793a5ccca12 100644
--- a/drivers/gpu/drm/nouveau/nv50_gpio.c
+++ b/drivers/gpu/drm/nouveau/nv50_gpio.c
@@ -98,6 +98,37 @@ nv50_gpio_set(struct drm_device *dev, enum dcb_gpio_tag tag, int state)
 }
 
 int
+nvd0_gpio_get(struct drm_device *dev, enum dcb_gpio_tag tag)
+{
+	struct dcb_gpio_entry *gpio;
+	u32 v;
+
+	gpio = nouveau_bios_gpio_entry(dev, tag);
+	if (!gpio)
+		return -ENOENT;
+
+	v  = nv_rd32(dev, 0x00d610 + (gpio->line * 4));
+	v &= 0x00004000;
+	return (!!v == (gpio->state[1] & 1));
+}
+
+int
+nvd0_gpio_set(struct drm_device *dev, enum dcb_gpio_tag tag, int state)
+{
+	struct dcb_gpio_entry *gpio;
+	u32 v;
+
+	gpio = nouveau_bios_gpio_entry(dev, tag);
+	if (!gpio)
+		return -ENOENT;
+
+	v = gpio->state[state] ^ 2;
+
+	nv_mask(dev, 0x00d610 + (gpio->line * 4), 0x00003000, v << 12);
+	return 0;
+}
+
+int
 nv50_gpio_irq_register(struct drm_device *dev, enum dcb_gpio_tag tag,
 		       void (*handler)(void *, int), void *data)
 {
diff --git a/drivers/gpu/drm/nouveau/nv50_graph.c b/drivers/gpu/drm/nouveau/nv50_graph.c
index d43c46caa76..8c979b31ff6 100644
--- a/drivers/gpu/drm/nouveau/nv50_graph.c
+++ b/drivers/gpu/drm/nouveau/nv50_graph.c
@@ -120,70 +120,62 @@ nv50_graph_unload_context(struct drm_device *dev)
 	return 0;
 }
 
-static void
-nv50_graph_init_reset(struct drm_device *dev)
-{
-	uint32_t pmc_e = NV_PMC_ENABLE_PGRAPH | (1 << 21);
-	NV_DEBUG(dev, "\n");
-
-	nv_wr32(dev, NV03_PMC_ENABLE, nv_rd32(dev, NV03_PMC_ENABLE) & ~pmc_e);
-	nv_wr32(dev, NV03_PMC_ENABLE, nv_rd32(dev, NV03_PMC_ENABLE) |  pmc_e);
-}
-
-static void
-nv50_graph_init_intr(struct drm_device *dev)
-{
-	NV_DEBUG(dev, "\n");
-
-	nv_wr32(dev, NV03_PGRAPH_INTR, 0xffffffff);
-	nv_wr32(dev, 0x400138, 0xffffffff);
-	nv_wr32(dev, NV40_PGRAPH_INTR_EN, 0xffffffff);
-}
-
-static void
-nv50_graph_init_regs__nv(struct drm_device *dev)
+static int
+nv50_graph_init(struct drm_device *dev, int engine)
 {
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
-	uint32_t units = nv_rd32(dev, 0x1540);
+	struct nv50_graph_engine *pgraph = nv_engine(dev, engine);
+	u32 units = nv_rd32(dev, 0x001540);
 	int i;
 
 	NV_DEBUG(dev, "\n");
 
+	/* master reset */
+	nv_mask(dev, 0x000200, 0x00200100, 0x00000000);
+	nv_mask(dev, 0x000200, 0x00200100, 0x00200100);
+	nv_wr32(dev, 0x40008c, 0x00000004); /* HW_CTX_SWITCH_ENABLED */
+
+	/* reset/enable traps and interrupts */
 	nv_wr32(dev, 0x400804, 0xc0000000);
 	nv_wr32(dev, 0x406800, 0xc0000000);
 	nv_wr32(dev, 0x400c04, 0xc0000000);
 	nv_wr32(dev, 0x401800, 0xc0000000);
 	nv_wr32(dev, 0x405018, 0xc0000000);
 	nv_wr32(dev, 0x402000, 0xc0000000);
-
 	for (i = 0; i < 16; i++) {
-		if (units & 1 << i) {
-			if (dev_priv->chipset < 0xa0) {
-				nv_wr32(dev, 0x408900 + (i << 12), 0xc0000000);
-				nv_wr32(dev, 0x408e08 + (i << 12), 0xc0000000);
-				nv_wr32(dev, 0x408314 + (i << 12), 0xc0000000);
-			} else {
-				nv_wr32(dev, 0x408600 + (i << 11), 0xc0000000);
-				nv_wr32(dev, 0x408708 + (i << 11), 0xc0000000);
-				nv_wr32(dev, 0x40831c + (i << 11), 0xc0000000);
-			}
+		if (!(units & (1 << i)))
+			continue;
+
+		if (dev_priv->chipset < 0xa0) {
+			nv_wr32(dev, 0x408900 + (i << 12), 0xc0000000);
+			nv_wr32(dev, 0x408e08 + (i << 12), 0xc0000000);
+			nv_wr32(dev, 0x408314 + (i << 12), 0xc0000000);
+		} else {
+			nv_wr32(dev, 0x408600 + (i << 11), 0xc0000000);
+			nv_wr32(dev, 0x408708 + (i << 11), 0xc0000000);
+			nv_wr32(dev, 0x40831c + (i << 11), 0xc0000000);
 		}
 	}
 
 	nv_wr32(dev, 0x400108, 0xffffffff);
-
-	nv_wr32(dev, 0x400824, 0x00004000);
+	nv_wr32(dev, 0x400138, 0xffffffff);
+	nv_wr32(dev, 0x400100, 0xffffffff);
+	nv_wr32(dev, 0x40013c, 0xffffffff);
 	nv_wr32(dev, 0x400500, 0x00010001);
-}
-
-static void
-nv50_graph_init_zcull(struct drm_device *dev)
-{
-	struct drm_nouveau_private *dev_priv = dev->dev_private;
-	int i;
-
-	NV_DEBUG(dev, "\n");
 
+	/* upload context program, initialise ctxctl defaults */
+	nv_wr32(dev, 0x400324, 0x00000000);
+	for (i = 0; i < pgraph->ctxprog_size; i++)
+		nv_wr32(dev, 0x400328, pgraph->ctxprog[i]);
+	nv_wr32(dev, 0x400824, 0x00000000);
+	nv_wr32(dev, 0x400828, 0x00000000);
+	nv_wr32(dev, 0x40082c, 0x00000000);
+	nv_wr32(dev, 0x400830, 0x00000000);
+	nv_wr32(dev, 0x400724, 0x00000000);
+	nv_wr32(dev, 0x40032c, 0x00000000);
+	nv_wr32(dev, 0x400320, 4);	/* CTXCTL_CMD = NEWCTXDMA */
+
+	/* some unknown zcull magic */
 	switch (dev_priv->chipset & 0xf0) {
 	case 0x50:
 	case 0x80:
@@ -212,43 +204,7 @@ nv50_graph_init_zcull(struct drm_device *dev)
 		nv_wr32(dev, 0x402c28 + (i * 8), 0x00000000);
 		nv_wr32(dev, 0x402c2c + (i * 8), 0x00000000);
 	}
-}
-
-static int
-nv50_graph_init_ctxctl(struct drm_device *dev)
-{
-	struct nv50_graph_engine *pgraph = nv_engine(dev, NVOBJ_ENGINE_GR);
-	int i;
-
-	NV_DEBUG(dev, "\n");
-
-	nv_wr32(dev, NV40_PGRAPH_CTXCTL_UCODE_INDEX, 0);
-	for (i = 0; i < pgraph->ctxprog_size; i++)
-		nv_wr32(dev, NV40_PGRAPH_CTXCTL_UCODE_DATA, pgraph->ctxprog[i]);
-
-	nv_wr32(dev, 0x40008c, 0x00000004); /* HW_CTX_SWITCH_ENABLED */
-	nv_wr32(dev, 0x400320, 4);
-	nv_wr32(dev, NV40_PGRAPH_CTXCTL_CUR, 0);
-	nv_wr32(dev, NV20_PGRAPH_CHANNEL_CTX_POINTER, 0);
-	return 0;
-}
-
-static int
-nv50_graph_init(struct drm_device *dev, int engine)
-{
-	int ret;
-
-	NV_DEBUG(dev, "\n");
-
-	nv50_graph_init_reset(dev);
-	nv50_graph_init_regs__nv(dev);
-	nv50_graph_init_zcull(dev);
-
-	ret = nv50_graph_init_ctxctl(dev);
-	if (ret)
-		return ret;
 
-	nv50_graph_init_intr(dev);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nv50_grctx.c b/drivers/gpu/drm/nouveau/nv50_grctx.c
index de9abff12b9..d05c2c3b244 100644
--- a/drivers/gpu/drm/nouveau/nv50_grctx.c
+++ b/drivers/gpu/drm/nouveau/nv50_grctx.c
@@ -40,6 +40,12 @@
 #define CP_FLAG_UNK0B                 ((0 * 32) + 0xb)
 #define CP_FLAG_UNK0B_CLEAR           0
 #define CP_FLAG_UNK0B_SET             1
+#define CP_FLAG_XFER_SWITCH           ((0 * 32) + 0xe)
+#define CP_FLAG_XFER_SWITCH_DISABLE   0
+#define CP_FLAG_XFER_SWITCH_ENABLE    1
+#define CP_FLAG_STATE                 ((0 * 32) + 0x1c)
+#define CP_FLAG_STATE_STOPPED         0
+#define CP_FLAG_STATE_RUNNING         1
 #define CP_FLAG_UNK1D                 ((0 * 32) + 0x1d)
 #define CP_FLAG_UNK1D_CLEAR           0
 #define CP_FLAG_UNK1D_SET             1
@@ -194,6 +200,9 @@ nv50_grctx_init(struct nouveau_grctx *ctx)
 				   "the devs.\n");
 		return -ENOSYS;
 	}
+
+	cp_set (ctx, STATE, RUNNING);
+	cp_set (ctx, XFER_SWITCH, ENABLE);
 	/* decide whether we're loading/unloading the context */
 	cp_bra (ctx, AUTO_SAVE, PENDING, cp_setup_save);
 	cp_bra (ctx, USER_SAVE, PENDING, cp_setup_save);
@@ -260,6 +269,8 @@ nv50_grctx_init(struct nouveau_grctx *ctx)
 	cp_name(ctx, cp_exit);
 	cp_set (ctx, USER_SAVE, NOT_PENDING);
 	cp_set (ctx, USER_LOAD, NOT_PENDING);
+	cp_set (ctx, XFER_SWITCH, DISABLE);
+	cp_set (ctx, STATE, STOPPED);
 	cp_out (ctx, CP_END);
 	ctx->ctxvals_pos += 0x400; /* padding... no idea why you need it */
 
diff --git a/drivers/gpu/drm/nouveau/nv50_pm.c b/drivers/gpu/drm/nouveau/nv50_pm.c
index 8a2810011bd..3d5a86b9828 100644
--- a/drivers/gpu/drm/nouveau/nv50_pm.c
+++ b/drivers/gpu/drm/nouveau/nv50_pm.c
@@ -115,15 +115,15 @@ nv50_pm_clock_set(struct drm_device *dev, void *pre_state)
 	    BIT_M.version == 1 && BIT_M.length >= 0x0b) {
 		script = ROM16(BIT_M.data[0x05]);
 		if (script)
-			nouveau_bios_run_init_table(dev, script, NULL);
+			nouveau_bios_run_init_table(dev, script, NULL, -1);
 		script = ROM16(BIT_M.data[0x07]);
 		if (script)
-			nouveau_bios_run_init_table(dev, script, NULL);
+			nouveau_bios_run_init_table(dev, script, NULL, -1);
 		script = ROM16(BIT_M.data[0x09]);
 		if (script)
-			nouveau_bios_run_init_table(dev, script, NULL);
+			nouveau_bios_run_init_table(dev, script, NULL, -1);
 
-		nouveau_bios_run_init_table(dev, perflvl->memscript, NULL);
+		nouveau_bios_run_init_table(dev, perflvl->memscript, NULL, -1);
 	}
 
 	if (state->type == PLL_MEMORY) {
diff --git a/drivers/gpu/drm/nouveau/nv50_sor.c b/drivers/gpu/drm/nouveau/nv50_sor.c
index ffe8b483b7b..2633aa8554e 100644
--- a/drivers/gpu/drm/nouveau/nv50_sor.c
+++ b/drivers/gpu/drm/nouveau/nv50_sor.c
@@ -124,7 +124,7 @@ nv50_sor_dpms(struct drm_encoder *encoder, int mode)
 		if (mode == DRM_MODE_DPMS_ON) {
 			u8 status = DP_SET_POWER_D0;
 			nouveau_dp_auxch(auxch, 8, DP_SET_POWER, &status, 1);
-			nouveau_dp_link_train(encoder);
+			nouveau_dp_link_train(encoder, nv_encoder->dp.datarate);
 		} else {
 			u8 status = DP_SET_POWER_D3;
 			nouveau_dp_auxch(auxch, 8, DP_SET_POWER, &status, 1);
@@ -187,14 +187,13 @@ nv50_sor_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode,
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
 	struct drm_device *dev = encoder->dev;
 	struct nouveau_crtc *crtc = nouveau_crtc(encoder->crtc);
+	struct nouveau_connector *nv_connector;
 	uint32_t mode_ctl = 0;
 	int ret;
 
 	NV_DEBUG_KMS(dev, "or %d type %d -> crtc %d\n",
 		     nv_encoder->or, nv_encoder->dcb->type, crtc->index);
 
-	nv50_sor_dpms(encoder, DRM_MODE_DPMS_ON);
-
 	switch (nv_encoder->dcb->type) {
 	case OUTPUT_TMDS:
 		if (nv_encoder->dcb->sorconf.link & 1) {
@@ -206,7 +205,15 @@ nv50_sor_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode,
 			mode_ctl = 0x0200;
 		break;
 	case OUTPUT_DP:
-		mode_ctl |= (nv_encoder->dp.mc_unknown << 16);
+		nv_connector = nouveau_encoder_connector_get(nv_encoder);
+		if (nv_connector && nv_connector->base.display_info.bpc == 6) {
+			nv_encoder->dp.datarate = crtc->mode->clock * 18 / 8;
+			mode_ctl |= 0x00020000;
+		} else {
+			nv_encoder->dp.datarate = crtc->mode->clock * 24 / 8;
+			mode_ctl |= 0x00050000;
+		}
+
 		if (nv_encoder->dcb->sorconf.link & 1)
 			mode_ctl |= 0x00000800;
 		else
@@ -227,6 +234,8 @@ nv50_sor_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode,
 	if (adjusted_mode->flags & DRM_MODE_FLAG_NVSYNC)
 		mode_ctl |= NV50_EVO_SOR_MODE_CTRL_NVSYNC;
 
+	nv50_sor_dpms(encoder, DRM_MODE_DPMS_ON);
+
 	ret = RING_SPACE(evo, 2);
 	if (ret) {
 		NV_ERROR(dev, "no space while connecting SOR\n");
@@ -313,31 +322,6 @@ nv50_sor_create(struct drm_connector *connector, struct dcb_entry *entry)
 	encoder->possible_crtcs = entry->heads;
 	encoder->possible_clones = 0;
 
-	if (nv_encoder->dcb->type == OUTPUT_DP) {
-		int or = nv_encoder->or, link = !(entry->dpconf.sor.link & 1);
-		uint32_t tmp;
-
-		tmp = nv_rd32(dev, 0x61c700 + (or * 0x800));
-		if (!tmp)
-			tmp = nv_rd32(dev, 0x610798 + (or * 8));
-
-		switch ((tmp & 0x00000f00) >> 8) {
-		case 8:
-		case 9:
-			nv_encoder->dp.mc_unknown = (tmp & 0x000f0000) >> 16;
-			tmp = nv_rd32(dev, NV50_SOR_DP_CTRL(or, link));
-			nv_encoder->dp.unk0 = tmp & 0x000001fc;
-			tmp = nv_rd32(dev, NV50_SOR_DP_UNK128(or, link));
-			nv_encoder->dp.unk1 = tmp & 0x010f7f3f;
-			break;
-		default:
-			break;
-		}
-
-		if (!nv_encoder->dp.mc_unknown)
-			nv_encoder->dp.mc_unknown = 5;
-	}
-
 	drm_mode_connector_attach_encoder(connector, encoder);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nv50_vram.c b/drivers/gpu/drm/nouveau/nv50_vram.c
index af32daecd1e..9da23838e63 100644
--- a/drivers/gpu/drm/nouveau/nv50_vram.c
+++ b/drivers/gpu/drm/nouveau/nv50_vram.c
@@ -51,7 +51,7 @@ void
 nv50_vram_del(struct drm_device *dev, struct nouveau_mem **pmem)
 {
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
-	struct nouveau_mm *mm = dev_priv->engine.vram.mm;
+	struct nouveau_mm *mm = &dev_priv->engine.vram.mm;
 	struct nouveau_mm_node *this;
 	struct nouveau_mem *mem;
 
@@ -82,7 +82,7 @@ nv50_vram_new(struct drm_device *dev, u64 size, u32 align, u32 size_nc,
 	      u32 memtype, struct nouveau_mem **pmem)
 {
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
-	struct nouveau_mm *mm = dev_priv->engine.vram.mm;
+	struct nouveau_mm *mm = &dev_priv->engine.vram.mm;
 	struct nouveau_mm_node *r;
 	struct nouveau_mem *mem;
 	int comp = (memtype & 0x300) >> 8;
diff --git a/drivers/gpu/drm/nouveau/nva3_pm.c b/drivers/gpu/drm/nouveau/nva3_pm.c
index e4b2b9e934b..618c144b7a3 100644
--- a/drivers/gpu/drm/nouveau/nva3_pm.c
+++ b/drivers/gpu/drm/nouveau/nva3_pm.c
@@ -27,178 +27,316 @@
 #include "nouveau_bios.h"
 #include "nouveau_pm.h"
 
-/* This is actually a lot more complex than it appears here, but hopefully
- * this should be able to deal with what the VBIOS leaves for us..
- *
- * If not, well, I'll jump off that bridge when I come to it.
- */
+static u32 read_clk(struct drm_device *, int, bool);
+static u32 read_pll(struct drm_device *, int, u32);
 
-struct nva3_pm_state {
-	enum pll_types type;
-	u32 src0;
-	u32 src1;
-	u32 ctrl;
-	u32 coef;
-	u32 old_pnm;
-	u32 new_pnm;
-	u32 new_div;
-};
+static u32
+read_vco(struct drm_device *dev, int clk)
+{
+	u32 sctl = nv_rd32(dev, 0x4120 + (clk * 4));
+	if ((sctl & 0x00000030) != 0x00000030)
+		return read_pll(dev, 0x41, 0x00e820);
+	return read_pll(dev, 0x42, 0x00e8a0);
+}
 
-static int
-nva3_pm_pll_offset(u32 id)
+static u32
+read_clk(struct drm_device *dev, int clk, bool ignore_en)
 {
-	static const u32 pll_map[] = {
-		0x00, PLL_CORE,
-		0x01, PLL_SHADER,
-		0x02, PLL_MEMORY,
-		0x00, 0x00
-	};
-	const u32 *map = pll_map;
-
-	while (map[1]) {
-		if (id == map[1])
-			return map[0];
-		map += 2;
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	u32 sctl, sdiv, sclk;
+
+	/* refclk for the 0xe8xx plls is a fixed frequency */
+	if (clk >= 0x40) {
+		if (dev_priv->chipset == 0xaf) {
+			/* no joke.. seriously.. sigh.. */
+			return nv_rd32(dev, 0x00471c) * 1000;
+		}
+
+		return dev_priv->crystal;
 	}
 
-	return -ENOENT;
+	sctl = nv_rd32(dev, 0x4120 + (clk * 4));
+	if (!ignore_en && !(sctl & 0x00000100))
+		return 0;
+
+	switch (sctl & 0x00003000) {
+	case 0x00000000:
+		return dev_priv->crystal;
+	case 0x00002000:
+		if (sctl & 0x00000040)
+			return 108000;
+		return 100000;
+	case 0x00003000:
+		sclk = read_vco(dev, clk);
+		sdiv = ((sctl & 0x003f0000) >> 16) + 2;
+		return (sclk * 2) / sdiv;
+	default:
+		return 0;
+	}
 }
 
-int
-nva3_pm_clock_get(struct drm_device *dev, u32 id)
+static u32
+read_pll(struct drm_device *dev, int clk, u32 pll)
+{
+	u32 ctrl = nv_rd32(dev, pll + 0);
+	u32 sclk = 0, P = 1, N = 1, M = 1;
+
+	if (!(ctrl & 0x00000008)) {
+		if (ctrl & 0x00000001) {
+			u32 coef = nv_rd32(dev, pll + 4);
+			M = (coef & 0x000000ff) >> 0;
+			N = (coef & 0x0000ff00) >> 8;
+			P = (coef & 0x003f0000) >> 16;
+
+			/* no post-divider on these.. */
+			if ((pll & 0x00ff00) == 0x00e800)
+				P = 1;
+
+			sclk = read_clk(dev, 0x00 + clk, false);
+		}
+	} else {
+		sclk = read_clk(dev, 0x10 + clk, false);
+	}
+
+	return sclk * N / (M * P);
+}
+
+struct creg {
+	u32 clk;
+	u32 pll;
+};
+
+static int
+calc_clk(struct drm_device *dev, int clk, u32 pll, u32 khz, struct creg *reg)
 {
-	u32 src0, src1, ctrl, coef;
-	struct pll_lims pll;
-	int ret, off;
-	int P, N, M;
+	struct pll_lims limits;
+	u32 oclk, sclk, sdiv;
+	int P, N, M, diff;
+	int ret;
+
+	reg->pll = 0;
+	reg->clk = 0;
+	if (!khz) {
+		NV_DEBUG(dev, "no clock for 0x%04x/0x%02x\n", pll, clk);
+		return 0;
+	}
 
-	ret = get_pll_limits(dev, id, &pll);
+	switch (khz) {
+	case 27000:
+		reg->clk = 0x00000100;
+		return khz;
+	case 100000:
+		reg->clk = 0x00002100;
+		return khz;
+	case 108000:
+		reg->clk = 0x00002140;
+		return khz;
+	default:
+		sclk = read_vco(dev, clk);
+		sdiv = min((sclk * 2) / (khz - 2999), (u32)65);
+		/* if the clock has a PLL attached, and we can get a within
+		 * [-2, 3) MHz of a divider, we'll disable the PLL and use
+		 * the divider instead.
+		 *
+		 * divider can go as low as 2, limited here because NVIDIA
+		 * and the VBIOS on my NVA8 seem to prefer using the PLL
+		 * for 810MHz - is there a good reason?
+		 */
+		if (sdiv > 4) {
+			oclk = (sclk * 2) / sdiv;
+			diff = khz - oclk;
+			if (!pll || (diff >= -2000 && diff < 3000)) {
+				reg->clk = (((sdiv - 2) << 16) | 0x00003100);
+				return oclk;
+			}
+		}
+
+		if (!pll) {
+			NV_ERROR(dev, "bad freq %02x: %d %d\n", clk, khz, sclk);
+			return -ERANGE;
+		}
+
+		break;
+	}
+
+	ret = get_pll_limits(dev, pll, &limits);
 	if (ret)
 		return ret;
 
-	off = nva3_pm_pll_offset(id);
-	if (off < 0)
-		return off;
+	limits.refclk = read_clk(dev, clk - 0x10, true);
+	if (!limits.refclk)
+		return -EINVAL;
+
+	ret = nva3_calc_pll(dev, &limits, khz, &N, NULL, &M, &P);
+	if (ret >= 0) {
+		reg->clk = nv_rd32(dev, 0x4120 + (clk * 4));
+		reg->pll = (P << 16) | (N << 8) | M;
+	}
+	return ret;
+}
+
+static void
+prog_pll(struct drm_device *dev, int clk, u32 pll, struct creg *reg)
+{
+	const u32 src0 = 0x004120 + (clk * 4);
+	const u32 src1 = 0x004160 + (clk * 4);
+	const u32 ctrl = pll + 0;
+	const u32 coef = pll + 4;
+	u32 cntl;
+
+	if (!reg->clk && !reg->pll) {
+		NV_DEBUG(dev, "no clock for %02x\n", clk);
+		return;
+	}
 
-	src0 = nv_rd32(dev, 0x4120 + (off * 4));
-	src1 = nv_rd32(dev, 0x4160 + (off * 4));
-	ctrl = nv_rd32(dev, pll.reg + 0);
-	coef = nv_rd32(dev, pll.reg + 4);
-	NV_DEBUG(dev, "PLL %02x: 0x%08x 0x%08x 0x%08x 0x%08x\n",
-		      id, src0, src1, ctrl, coef);
+	cntl = nv_rd32(dev, ctrl) & 0xfffffff2;
+	if (reg->pll) {
+		nv_mask(dev, src0, 0x00000101, 0x00000101);
+		nv_wr32(dev, coef, reg->pll);
+		nv_wr32(dev, ctrl, cntl | 0x00000015);
+		nv_mask(dev, src1, 0x00000100, 0x00000000);
+		nv_mask(dev, src1, 0x00000001, 0x00000000);
+	} else {
+		nv_mask(dev, src1, 0x003f3141, 0x00000101 | reg->clk);
+		nv_wr32(dev, ctrl, cntl | 0x0000001d);
+		nv_mask(dev, ctrl, 0x00000001, 0x00000000);
+		nv_mask(dev, src0, 0x00000100, 0x00000000);
+		nv_mask(dev, src0, 0x00000001, 0x00000000);
+	}
+}
 
-	if (ctrl & 0x00000008) {
-		u32 div = ((src1 & 0x003c0000) >> 18) + 1;
-		return (pll.refclk * 2) / div;
+static void
+prog_clk(struct drm_device *dev, int clk, struct creg *reg)
+{
+	if (!reg->clk) {
+		NV_DEBUG(dev, "no clock for %02x\n", clk);
+		return;
 	}
 
-	P = (coef & 0x003f0000) >> 16;
-	N = (coef & 0x0000ff00) >> 8;
-	M = (coef & 0x000000ff);
-	return pll.refclk * N / M / P;
+	nv_mask(dev, 0x004120 + (clk * 4), 0x003f3141, 0x00000101 | reg->clk);
+}
+
+int
+nva3_pm_clocks_get(struct drm_device *dev, struct nouveau_pm_level *perflvl)
+{
+	perflvl->core   = read_pll(dev, 0x00, 0x4200);
+	perflvl->shader = read_pll(dev, 0x01, 0x4220);
+	perflvl->memory = read_pll(dev, 0x02, 0x4000);
+	perflvl->unka0  = read_clk(dev, 0x20, false);
+	perflvl->vdec   = read_clk(dev, 0x21, false);
+	perflvl->daemon = read_clk(dev, 0x25, false);
+	perflvl->copy   = perflvl->core;
+	return 0;
 }
 
+struct nva3_pm_state {
+	struct creg nclk;
+	struct creg sclk;
+	struct creg mclk;
+	struct creg vdec;
+	struct creg unka0;
+};
+
 void *
-nva3_pm_clock_pre(struct drm_device *dev, struct nouveau_pm_level *perflvl,
-		  u32 id, int khz)
+nva3_pm_clocks_pre(struct drm_device *dev, struct nouveau_pm_level *perflvl)
 {
-	struct nva3_pm_state *pll;
-	struct pll_lims limits;
-	int N, M, P, diff;
-	int ret, off;
+	struct nva3_pm_state *info;
+	int ret;
 
-	ret = get_pll_limits(dev, id, &limits);
+	info = kzalloc(sizeof(*info), GFP_KERNEL);
+	if (!info)
+		return ERR_PTR(-ENOMEM);
+
+	ret = calc_clk(dev, 0x10, 0x4200, perflvl->core, &info->nclk);
 	if (ret < 0)
-		return (ret == -ENOENT) ? NULL : ERR_PTR(ret);
+		goto out;
 
-	off = nva3_pm_pll_offset(id);
-	if (id < 0)
-		return ERR_PTR(-EINVAL);
+	ret = calc_clk(dev, 0x11, 0x4220, perflvl->shader, &info->sclk);
+	if (ret < 0)
+		goto out;
 
+	ret = calc_clk(dev, 0x12, 0x4000, perflvl->memory, &info->mclk);
+	if (ret < 0)
+		goto out;
 
-	pll = kzalloc(sizeof(*pll), GFP_KERNEL);
-	if (!pll)
-		return ERR_PTR(-ENOMEM);
-	pll->type = id;
-	pll->src0 = 0x004120 + (off * 4);
-	pll->src1 = 0x004160 + (off * 4);
-	pll->ctrl = limits.reg + 0;
-	pll->coef = limits.reg + 4;
-
-	/* If target clock is within [-2, 3) MHz of a divisor, we'll
-	 * use that instead of calculating MNP values
-	 */
-	pll->new_div = min((limits.refclk * 2) / (khz - 2999), 16);
-	if (pll->new_div) {
-		diff = khz - ((limits.refclk * 2) / pll->new_div);
-		if (diff < -2000 || diff >= 3000)
-			pll->new_div = 0;
-	}
+	ret = calc_clk(dev, 0x20, 0x0000, perflvl->unka0, &info->unka0);
+	if (ret < 0)
+		goto out;
 
-	if (!pll->new_div) {
-		ret = nva3_calc_pll(dev, &limits, khz, &N, NULL, &M, &P);
-		if (ret < 0)
-			return ERR_PTR(ret);
+	ret = calc_clk(dev, 0x21, 0x0000, perflvl->vdec, &info->vdec);
+	if (ret < 0)
+		goto out;
 
-		pll->new_pnm = (P << 16) | (N << 8) | M;
-		pll->new_div = 2 - 1;
-	} else {
-		pll->new_pnm = 0;
-		pll->new_div--;
+out:
+	if (ret < 0) {
+		kfree(info);
+		info = ERR_PTR(ret);
 	}
+	return info;
+}
+
+static bool
+nva3_pm_grcp_idle(void *data)
+{
+	struct drm_device *dev = data;
 
-	if ((nv_rd32(dev, pll->src1) & 0x00000101) != 0x00000101)
-		pll->old_pnm = nv_rd32(dev, pll->coef);
-	return pll;
+	if (!(nv_rd32(dev, 0x400304) & 0x00000001))
+		return true;
+	if (nv_rd32(dev, 0x400308) == 0x0050001c)
+		return true;
+	return false;
 }
 
 void
-nva3_pm_clock_set(struct drm_device *dev, void *pre_state)
+nva3_pm_clocks_set(struct drm_device *dev, void *pre_state)
 {
-	struct nva3_pm_state *pll = pre_state;
-	u32 ctrl = 0;
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct nva3_pm_state *info = pre_state;
+	unsigned long flags;
 
-	/* For the memory clock, NVIDIA will build a "script" describing
-	 * the reclocking process and ask PDAEMON to execute it.
-	 */
-	if (pll->type == PLL_MEMORY) {
-		nv_wr32(dev, 0x100210, 0);
-		nv_wr32(dev, 0x1002dc, 1);
-		nv_wr32(dev, 0x004018, 0x00001000);
-		ctrl = 0x18000100;
+	/* prevent any new grctx switches from starting */
+	spin_lock_irqsave(&dev_priv->context_switch_lock, flags);
+	nv_wr32(dev, 0x400324, 0x00000000);
+	nv_wr32(dev, 0x400328, 0x0050001c); /* wait flag 0x1c */
+	/* wait for any pending grctx switches to complete */
+	if (!nv_wait_cb(dev, nva3_pm_grcp_idle, dev)) {
+		NV_ERROR(dev, "pm: ctxprog didn't go idle\n");
+		goto cleanup;
 	}
-
-	if (pll->old_pnm || !pll->new_pnm) {
-		nv_mask(dev, pll->src1, 0x003c0101, 0x00000101 |
-						    (pll->new_div << 18));
-		nv_wr32(dev, pll->ctrl, 0x0001001d | ctrl);
-		nv_mask(dev, pll->ctrl, 0x00000001, 0x00000000);
+	/* freeze PFIFO */
+	nv_mask(dev, 0x002504, 0x00000001, 0x00000001);
+	if (!nv_wait(dev, 0x002504, 0x00000010, 0x00000010)) {
+		NV_ERROR(dev, "pm: fifo didn't go idle\n");
+		goto cleanup;
 	}
 
-	if (pll->new_pnm) {
-		nv_mask(dev, pll->src0, 0x00000101, 0x00000101);
-		nv_wr32(dev, pll->coef, pll->new_pnm);
-		nv_wr32(dev, pll->ctrl, 0x0001001d | ctrl);
-		nv_mask(dev, pll->ctrl, 0x00000010, 0x00000000);
-		nv_mask(dev, pll->ctrl, 0x00020010, 0x00020010);
-		nv_wr32(dev, pll->ctrl, 0x00010015 | ctrl);
-		nv_mask(dev, pll->src1, 0x00000100, 0x00000000);
-		nv_mask(dev, pll->src1, 0x00000001, 0x00000000);
-		if (pll->type == PLL_MEMORY)
-			nv_wr32(dev, 0x4018, 0x10005000);
-	} else {
-		nv_mask(dev, pll->ctrl, 0x00000001, 0x00000000);
-		nv_mask(dev, pll->src0, 0x00000100, 0x00000000);
-		nv_mask(dev, pll->src0, 0x00000001, 0x00000000);
-		if (pll->type == PLL_MEMORY)
-			nv_wr32(dev, 0x4018, 0x1000d000);
-	}
+	prog_pll(dev, 0x00, 0x004200, &info->nclk);
+	prog_pll(dev, 0x01, 0x004220, &info->sclk);
+	prog_clk(dev, 0x20, &info->unka0);
+	prog_clk(dev, 0x21, &info->vdec);
 
-	if (pll->type == PLL_MEMORY) {
+	if (info->mclk.clk || info->mclk.pll) {
+		nv_wr32(dev, 0x100210, 0);
+		nv_wr32(dev, 0x1002dc, 1);
+		nv_wr32(dev, 0x004018, 0x00001000);
+		prog_pll(dev, 0x02, 0x004000, &info->mclk);
+		if (nv_rd32(dev, 0x4000) & 0x00000008)
+			nv_wr32(dev, 0x004018, 0x1000d000);
+		else
+			nv_wr32(dev, 0x004018, 0x10005000);
 		nv_wr32(dev, 0x1002dc, 0);
 		nv_wr32(dev, 0x100210, 0x80000000);
 	}
 
-	kfree(pll);
+cleanup:
+	/* unfreeze PFIFO */
+	nv_mask(dev, 0x002504, 0x00000001, 0x00000000);
+	/* restore ctxprog to normal */
+	nv_wr32(dev, 0x400324, 0x00000000);
+	nv_wr32(dev, 0x400328, 0x0070009c); /* set flag 0x1c */
+	/* unblock it if necessary */
+	if (nv_rd32(dev, 0x400308) == 0x0050001c)
+		nv_mask(dev, 0x400824, 0x10000000, 0x10000000);
+	spin_unlock_irqrestore(&dev_priv->context_switch_lock, flags);
+	kfree(info);
 }
-
diff --git a/drivers/gpu/drm/nouveau/nvc0_fb.c b/drivers/gpu/drm/nouveau/nvc0_fb.c
index 08e6b118f02..5bf55038fd9 100644
--- a/drivers/gpu/drm/nouveau/nvc0_fb.c
+++ b/drivers/gpu/drm/nouveau/nvc0_fb.c
@@ -32,6 +32,30 @@ struct nvc0_fb_priv {
 	dma_addr_t r100c10;
 };
 
+static inline void
+nvc0_mfb_subp_isr(struct drm_device *dev, int unit, int subp)
+{
+	u32 subp_base = 0x141000 + (unit * 0x2000) + (subp * 0x400);
+	u32 stat = nv_rd32(dev, subp_base + 0x020);
+
+	if (stat) {
+		NV_INFO(dev, "PMFB%d_SUBP%d: 0x%08x\n", unit, subp, stat);
+		nv_wr32(dev, subp_base + 0x020, stat);
+	}
+}
+
+static void
+nvc0_mfb_isr(struct drm_device *dev)
+{
+	u32 units = nv_rd32(dev, 0x00017c);
+	while (units) {
+		u32 subp, unit = ffs(units) - 1;
+		for (subp = 0; subp < 2; subp++)
+			nvc0_mfb_subp_isr(dev, unit, subp);
+		units &= ~(1 << unit);
+	}
+}
+
 static void
 nvc0_fb_destroy(struct drm_device *dev)
 {
@@ -39,6 +63,8 @@ nvc0_fb_destroy(struct drm_device *dev)
 	struct nouveau_fb_engine *pfb = &dev_priv->engine.fb;
 	struct nvc0_fb_priv *priv = pfb->priv;
 
+	nouveau_irq_unregister(dev, 25);
+
 	if (priv->r100c10_page) {
 		pci_unmap_page(dev->pdev, priv->r100c10, PAGE_SIZE,
 			       PCI_DMA_BIDIRECTIONAL);
@@ -74,6 +100,7 @@ nvc0_fb_create(struct drm_device *dev)
 		return -EFAULT;
 	}
 
+	nouveau_irq_register(dev, 25, nvc0_mfb_isr);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvc0_fifo.c b/drivers/gpu/drm/nouveau/nvc0_fifo.c
index 6f9f341c3e8..dcbe0d5d024 100644
--- a/drivers/gpu/drm/nouveau/nvc0_fifo.c
+++ b/drivers/gpu/drm/nouveau/nvc0_fifo.c
@@ -322,7 +322,7 @@ nvc0_fifo_init(struct drm_device *dev)
 	}
 
 	/* PSUBFIFO[n] */
-	for (i = 0; i < 3; i++) {
+	for (i = 0; i < priv->spoon_nr; i++) {
 		nv_mask(dev, 0x04013c + (i * 0x2000), 0x10000100, 0x00000000);
 		nv_wr32(dev, 0x040108 + (i * 0x2000), 0xffffffff); /* INTR */
 		nv_wr32(dev, 0x04010c + (i * 0x2000), 0xfffffeff); /* INTR_EN */
diff --git a/drivers/gpu/drm/nouveau/nvc0_graph.c b/drivers/gpu/drm/nouveau/nvc0_graph.c
index 5b2f6f42046..4b8d0b3f7d2 100644
--- a/drivers/gpu/drm/nouveau/nvc0_graph.c
+++ b/drivers/gpu/drm/nouveau/nvc0_graph.c
@@ -390,7 +390,7 @@ nvc0_graph_init_gpc_0(struct drm_device *dev)
 	}
 
 	nv_wr32(dev, GPC_BCAST(0x1bd4), magicgpc918);
-	nv_wr32(dev, GPC_BCAST(0x08ac), priv->rop_nr);
+	nv_wr32(dev, GPC_BCAST(0x08ac), nv_rd32(dev, 0x100800));
 }
 
 static void
@@ -700,22 +700,6 @@ nvc0_graph_isr(struct drm_device *dev)
 	nv_wr32(dev, 0x400500, 0x00010001);
 }
 
-static void
-nvc0_runk140_isr(struct drm_device *dev)
-{
-	u32 units = nv_rd32(dev, 0x00017c) & 0x1f;
-
-	while (units) {
-		u32 unit = ffs(units) - 1;
-		u32 reg = 0x140000 + unit * 0x2000;
-		u32 st0 = nv_mask(dev, reg + 0x1020, 0, 0);
-		u32 st1 = nv_mask(dev, reg + 0x1420, 0, 0);
-
-		NV_DEBUG(dev, "PRUNK140: %d 0x%08x 0x%08x\n", unit, st0, st1);
-		units &= ~(1 << unit);
-	}
-}
-
 static int
 nvc0_graph_create_fw(struct drm_device *dev, const char *fwname,
 		     struct nvc0_graph_fuc *fuc)
@@ -764,7 +748,6 @@ nvc0_graph_destroy(struct drm_device *dev, int engine)
 	}
 
 	nouveau_irq_unregister(dev, 12);
-	nouveau_irq_unregister(dev, 25);
 
 	nouveau_gpuobj_ref(NULL, &priv->unk4188b8);
 	nouveau_gpuobj_ref(NULL, &priv->unk4188b4);
@@ -803,7 +786,6 @@ nvc0_graph_create(struct drm_device *dev)
 
 	NVOBJ_ENGINE_ADD(dev, GR, &priv->base);
 	nouveau_irq_register(dev, 12, nvc0_graph_isr);
-	nouveau_irq_register(dev, 25, nvc0_runk140_isr);
 
 	if (nouveau_ctxfw) {
 		NV_INFO(dev, "PGRAPH: using external firmware\n");
@@ -864,6 +846,9 @@ nvc0_graph_create(struct drm_device *dev)
 	case 0xce: /* 4/4/0/0, 4 */
 		priv->magic_not_rop_nr = 0x03;
 		break;
+	case 0xcf: /* 4/0/0/0, 3 */
+		priv->magic_not_rop_nr = 0x03;
+		break;
 	}
 
 	if (!priv->magic_not_rop_nr) {
@@ -889,20 +874,3 @@ error:
 	nvc0_graph_destroy(dev, NVOBJ_ENGINE_GR);
 	return ret;
 }
-
-MODULE_FIRMWARE("nouveau/nvc0_fuc409c");
-MODULE_FIRMWARE("nouveau/nvc0_fuc409d");
-MODULE_FIRMWARE("nouveau/nvc0_fuc41ac");
-MODULE_FIRMWARE("nouveau/nvc0_fuc41ad");
-MODULE_FIRMWARE("nouveau/nvc3_fuc409c");
-MODULE_FIRMWARE("nouveau/nvc3_fuc409d");
-MODULE_FIRMWARE("nouveau/nvc3_fuc41ac");
-MODULE_FIRMWARE("nouveau/nvc3_fuc41ad");
-MODULE_FIRMWARE("nouveau/nvc4_fuc409c");
-MODULE_FIRMWARE("nouveau/nvc4_fuc409d");
-MODULE_FIRMWARE("nouveau/nvc4_fuc41ac");
-MODULE_FIRMWARE("nouveau/nvc4_fuc41ad");
-MODULE_FIRMWARE("nouveau/fuc409c");
-MODULE_FIRMWARE("nouveau/fuc409d");
-MODULE_FIRMWARE("nouveau/fuc41ac");
-MODULE_FIRMWARE("nouveau/fuc41ad");
diff --git a/drivers/gpu/drm/nouveau/nvc0_graph.h b/drivers/gpu/drm/nouveau/nvc0_graph.h
index 55689e99728..636fe9812f7 100644
--- a/drivers/gpu/drm/nouveau/nvc0_graph.h
+++ b/drivers/gpu/drm/nouveau/nvc0_graph.h
@@ -82,6 +82,7 @@ nvc0_graph_class(struct drm_device *dev)
 	case 0xc3:
 	case 0xc4:
 	case 0xce: /* guess, mmio trace shows only 0x9097 state */
+	case 0xcf: /* guess, mmio trace shows only 0x9097 state */
 		return 0x9097;
 	case 0xc1:
 		return 0x9197;
diff --git a/drivers/gpu/drm/nouveau/nvc0_grctx.c b/drivers/gpu/drm/nouveau/nvc0_grctx.c
index 31018eaf527..dd0e6a736b3 100644
--- a/drivers/gpu/drm/nouveau/nvc0_grctx.c
+++ b/drivers/gpu/drm/nouveau/nvc0_grctx.c
@@ -1678,7 +1678,10 @@ nvc0_grctx_generate_tp(struct drm_device *dev)
 	nv_wr32(dev, 0x419c04, 0x00000006);
 	nv_wr32(dev, 0x419c08, 0x00000002);
 	nv_wr32(dev, 0x419c20, 0x00000000);
-	nv_wr32(dev, 0x419cb0, 0x00060048); //XXX: 0xce 0x00020048
+	if (chipset == 0xce || chipset == 0xcf)
+		nv_wr32(dev, 0x419cb0, 0x00020048);
+	else
+		nv_wr32(dev, 0x419cb0, 0x00060048);
 	nv_wr32(dev, 0x419ce8, 0x00000000);
 	nv_wr32(dev, 0x419cf4, 0x00000183);
 	nv_wr32(dev, 0x419d20, chipset != 0xc1 ? 0x02180000 : 0x12180000);
@@ -1783,11 +1786,7 @@ nvc0_grctx_generate(struct nouveau_channel *chan)
 	nv_wr32(dev, 0x40587c, 0x00000000);
 
 	if (1) {
-		const u8 chipset_tp_max[] = { 16, 4, 0, 4, 8, 0, 0, 0,
-					      16, 0, 0, 0, 0, 0, 8, 0 };
-		u8 max = chipset_tp_max[dev_priv->chipset & 0x0f];
-		u8 tpnr[GPC_MAX];
-		u8 data[TP_MAX];
+		u8 tpnr[GPC_MAX], data[TP_MAX];
 
 		memcpy(tpnr, priv->tp_nr, sizeof(priv->tp_nr));
 		memset(data, 0x1f, sizeof(data));
@@ -1801,7 +1800,7 @@ nvc0_grctx_generate(struct nouveau_channel *chan)
 			data[tp] = gpc;
 		}
 
-		for (i = 0; i < max / 4; i++)
+		for (i = 0; i < 4; i++)
 			nv_wr32(dev, 0x4060a8 + (i * 4), ((u32 *)data)[i]);
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nvc0_grgpc.fuc b/drivers/gpu/drm/nouveau/nvc0_grgpc.fuc
index 0ec2add72a7..06f5e26d1e0 100644
--- a/drivers/gpu/drm/nouveau/nvc0_grgpc.fuc
+++ b/drivers/gpu/drm/nouveau/nvc0_grgpc.fuc
@@ -77,6 +77,11 @@ chipsets:
 .b16 nvc0_gpc_mmio_tail
 .b16 nvc0_tpc_mmio_head
 .b16 nvc3_tpc_mmio_tail
+.b8  0xcf 0 0 0
+.b16 nvc0_gpc_mmio_head
+.b16 nvc0_gpc_mmio_tail
+.b16 nvc0_tpc_mmio_head
+.b16 nvcf_tpc_mmio_tail
 .b8  0 0 0 0
 
 // GPC mmio lists
@@ -134,8 +139,9 @@ mmctx_data(0x000750, 2)
 nvc0_tpc_mmio_tail:
 mmctx_data(0x000758, 1)
 mmctx_data(0x0002c4, 1)
-mmctx_data(0x0004bc, 1)
 mmctx_data(0x0006e0, 1)
+nvcf_tpc_mmio_tail:
+mmctx_data(0x0004bc, 1)
 nvc3_tpc_mmio_tail:
 mmctx_data(0x000544, 1)
 nvc1_tpc_mmio_tail:
diff --git a/drivers/gpu/drm/nouveau/nvc0_grgpc.fuc.h b/drivers/gpu/drm/nouveau/nvc0_grgpc.fuc.h
index 1896c898f5b..6f820324480 100644
--- a/drivers/gpu/drm/nouveau/nvc0_grgpc.fuc.h
+++ b/drivers/gpu/drm/nouveau/nvc0_grgpc.fuc.h
@@ -25,23 +25,26 @@ uint32_t nvc0_grgpc_data[] = {
 	0x00000000,
 	0x00000000,
 	0x000000c0,
-	0x011000b0,
-	0x01640114,
+	0x011c00bc,
+	0x01700120,
 	0x000000c1,
-	0x011400b0,
-	0x01780114,
+	0x012000bc,
+	0x01840120,
 	0x000000c3,
-	0x011000b0,
-	0x01740114,
+	0x011c00bc,
+	0x01800120,
 	0x000000c4,
-	0x011000b0,
-	0x01740114,
+	0x011c00bc,
+	0x01800120,
 	0x000000c8,
-	0x011000b0,
-	0x01640114,
+	0x011c00bc,
+	0x01700120,
 	0x000000ce,
-	0x011000b0,
-	0x01740114,
+	0x011c00bc,
+	0x01800120,
+	0x000000cf,
+	0x011c00bc,
+	0x017c0120,
 	0x00000000,
 	0x00000380,
 	0x14000400,
@@ -90,8 +93,8 @@ uint32_t nvc0_grgpc_data[] = {
 	0x04000750,
 	0x00000758,
 	0x000002c4,
-	0x000004bc,
 	0x000006e0,
+	0x000004bc,
 	0x00000544,
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvc0_grhub.fuc b/drivers/gpu/drm/nouveau/nvc0_grhub.fuc
index a1a599124cf..e4f8c7e89dd 100644
--- a/drivers/gpu/drm/nouveau/nvc0_grhub.fuc
+++ b/drivers/gpu/drm/nouveau/nvc0_grhub.fuc
@@ -56,6 +56,9 @@ chipsets:
 .b8  0xce 0 0 0
 .b16 nvc0_hub_mmio_head
 .b16 nvc0_hub_mmio_tail
+.b8  0xcf 0 0 0
+.b16 nvc0_hub_mmio_head
+.b16 nvc0_hub_mmio_tail
 .b8  0 0 0 0
 
 nvc0_hub_mmio_head:
diff --git a/drivers/gpu/drm/nouveau/nvc0_grhub.fuc.h b/drivers/gpu/drm/nouveau/nvc0_grhub.fuc.h
index b3b541b6d04..241d3263f1e 100644
--- a/drivers/gpu/drm/nouveau/nvc0_grhub.fuc.h
+++ b/drivers/gpu/drm/nouveau/nvc0_grhub.fuc.h
@@ -23,17 +23,19 @@ uint32_t nvc0_grhub_data[] = {
 	0x00000000,
 	0x00000000,
 	0x000000c0,
-	0x012c0090,
+	0x01340098,
 	0x000000c1,
-	0x01300090,
+	0x01380098,
 	0x000000c3,
-	0x012c0090,
+	0x01340098,
 	0x000000c4,
-	0x012c0090,
+	0x01340098,
 	0x000000c8,
-	0x012c0090,
+	0x01340098,
 	0x000000ce,
-	0x012c0090,
+	0x01340098,
+	0x000000cf,
+	0x01340098,
 	0x00000000,
 	0x0417e91c,
 	0x04400204,
@@ -190,8 +192,6 @@ uint32_t nvc0_grhub_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x00000000,
-	0x00000000,
 };
 
 uint32_t nvc0_grhub_code[] = {
diff --git a/drivers/gpu/drm/nouveau/nvc0_pm.c b/drivers/gpu/drm/nouveau/nvc0_pm.c
new file mode 100644
index 00000000000..929aded35cb
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvc0_pm.c
@@ -0,0 +1,155 @@
+/*
+ * Copyright 2011 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "drmP.h"
+#include "nouveau_drv.h"
+#include "nouveau_bios.h"
+#include "nouveau_pm.h"
+
+static u32 read_div(struct drm_device *, int, u32, u32);
+static u32 read_pll(struct drm_device *, u32);
+
+static u32
+read_vco(struct drm_device *dev, u32 dsrc)
+{
+	u32 ssrc = nv_rd32(dev, dsrc);
+	if (!(ssrc & 0x00000100))
+		return read_pll(dev, 0x00e800);
+	return read_pll(dev, 0x00e820);
+}
+
+static u32
+read_pll(struct drm_device *dev, u32 pll)
+{
+	u32 ctrl = nv_rd32(dev, pll + 0);
+	u32 coef = nv_rd32(dev, pll + 4);
+	u32 P = (coef & 0x003f0000) >> 16;
+	u32 N = (coef & 0x0000ff00) >> 8;
+	u32 M = (coef & 0x000000ff) >> 0;
+	u32 sclk, doff;
+
+	if (!(ctrl & 0x00000001))
+		return 0;
+
+	switch (pll & 0xfff000) {
+	case 0x00e000:
+		sclk = 27000;
+		P = 1;
+		break;
+	case 0x137000:
+		doff = (pll - 0x137000) / 0x20;
+		sclk = read_div(dev, doff, 0x137120, 0x137140);
+		break;
+	case 0x132000:
+		switch (pll) {
+		case 0x132000:
+			sclk = read_pll(dev, 0x132020);
+			break;
+		case 0x132020:
+			sclk = read_div(dev, 0, 0x137320, 0x137330);
+			break;
+		default:
+			return 0;
+		}
+		break;
+	default:
+		return 0;
+	}
+
+	return sclk * N / M / P;
+}
+
+static u32
+read_div(struct drm_device *dev, int doff, u32 dsrc, u32 dctl)
+{
+	u32 ssrc = nv_rd32(dev, dsrc + (doff * 4));
+	u32 sctl = nv_rd32(dev, dctl + (doff * 4));
+
+	switch (ssrc & 0x00000003) {
+	case 0:
+		if ((ssrc & 0x00030000) != 0x00030000)
+			return 27000;
+		return 108000;
+	case 2:
+		return 100000;
+	case 3:
+		if (sctl & 0x80000000) {
+			u32 sclk = read_vco(dev, dsrc + (doff * 4));
+			u32 sdiv = (sctl & 0x0000003f) + 2;
+			return (sclk * 2) / sdiv;
+		}
+
+		return read_vco(dev, dsrc + (doff * 4));
+	default:
+		return 0;
+	}
+}
+
+static u32
+read_mem(struct drm_device *dev)
+{
+	u32 ssel = nv_rd32(dev, 0x1373f0);
+	if (ssel & 0x00000001)
+		return read_div(dev, 0, 0x137300, 0x137310);
+	return read_pll(dev, 0x132000);
+}
+
+static u32
+read_clk(struct drm_device *dev, int clk)
+{
+	u32 sctl = nv_rd32(dev, 0x137250 + (clk * 4));
+	u32 ssel = nv_rd32(dev, 0x137100);
+	u32 sclk, sdiv;
+
+	if (ssel & (1 << clk)) {
+		if (clk < 7)
+			sclk = read_pll(dev, 0x137000 + (clk * 0x20));
+		else
+			sclk = read_pll(dev, 0x1370e0);
+		sdiv = ((sctl & 0x00003f00) >> 8) + 2;
+	} else {
+		sclk = read_div(dev, clk, 0x137160, 0x1371d0);
+		sdiv = ((sctl & 0x0000003f) >> 0) + 2;
+	}
+
+	if (sctl & 0x80000000)
+		return (sclk * 2) / sdiv;
+	return sclk;
+}
+
+int
+nvc0_pm_clocks_get(struct drm_device *dev, struct nouveau_pm_level *perflvl)
+{
+	perflvl->shader = read_clk(dev, 0x00);
+	perflvl->core   = perflvl->shader / 2;
+	perflvl->memory = read_mem(dev);
+	perflvl->rop    = read_clk(dev, 0x01);
+	perflvl->hub07  = read_clk(dev, 0x02);
+	perflvl->hub06  = read_clk(dev, 0x07);
+	perflvl->hub01  = read_clk(dev, 0x08);
+	perflvl->copy   = read_clk(dev, 0x09);
+	perflvl->daemon = read_clk(dev, 0x0c);
+	perflvl->vdec   = read_clk(dev, 0x0e);
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/nvc0_vram.c b/drivers/gpu/drm/nouveau/nvc0_vram.c
index e45a24d84e9..edbfe9360ae 100644
--- a/drivers/gpu/drm/nouveau/nvc0_vram.c
+++ b/drivers/gpu/drm/nouveau/nvc0_vram.c
@@ -61,7 +61,7 @@ nvc0_vram_new(struct drm_device *dev, u64 size, u32 align, u32 ncmin,
 	      u32 type, struct nouveau_mem **pmem)
 {
 	struct drm_nouveau_private *dev_priv = dev->dev_private;
-	struct nouveau_mm *mm = dev_priv->engine.vram.mm;
+	struct nouveau_mm *mm = &dev_priv->engine.vram.mm;
 	struct nouveau_mm_node *r;
 	struct nouveau_mem *mem;
 	int ret;
@@ -106,12 +106,50 @@ nvc0_vram_init(struct drm_device *dev)
 	struct nouveau_vram_engine *vram = &dev_priv->engine.vram;
 	const u32 rsvd_head = ( 256 * 1024) >> 12; /* vga memory */
 	const u32 rsvd_tail = (1024 * 1024) >> 12; /* vbios etc */
-	u32 length;
+	u32 parts = nv_rd32(dev, 0x121c74);
+	u32 bsize = nv_rd32(dev, 0x10f20c);
+	u32 offset, length;
+	bool uniform = true;
+	int ret, i;
 
-	dev_priv->vram_size  = nv_rd32(dev, 0x10f20c) << 20;
-	dev_priv->vram_size *= nv_rd32(dev, 0x121c74);
+	NV_DEBUG(dev, "0x100800: 0x%08x\n", nv_rd32(dev, 0x100800));
+	NV_DEBUG(dev, "parts 0x%08x bcast_mem_amount 0x%08x\n", parts, bsize);
 
-	length = (dev_priv->vram_size >> 12) - rsvd_head - rsvd_tail;
+	/* read amount of vram attached to each memory controller */
+	for (i = 0; i < parts; i++) {
+		u32 psize = nv_rd32(dev, 0x11020c + (i * 0x1000));
+		if (psize != bsize) {
+			if (psize < bsize)
+				bsize = psize;
+			uniform = false;
+		}
+
+		NV_DEBUG(dev, "%d: mem_amount 0x%08x\n", i, psize);
+
+		dev_priv->vram_size += (u64)psize << 20;
+	}
+
+	/* if all controllers have the same amount attached, there's no holes */
+	if (uniform) {
+		offset = rsvd_head;
+		length = (dev_priv->vram_size >> 12) - rsvd_head - rsvd_tail;
+		return nouveau_mm_init(&vram->mm, offset, length, 1);
+	}
 
-	return nouveau_mm_init(&vram->mm, rsvd_head, length, 1);
+	/* otherwise, address lowest common amount from 0GiB */
+	ret = nouveau_mm_init(&vram->mm, rsvd_head, (bsize << 8) * parts, 1);
+	if (ret)
+		return ret;
+
+	/* and the rest starting from (8GiB + common_size) */
+	offset = (0x0200000000ULL >> 12) + (bsize << 8);
+	length = (dev_priv->vram_size >> 12) - (bsize << 8) - rsvd_tail;
+
+	ret = nouveau_mm_init(&vram->mm, offset, length, 0);
+	if (ret) {
+		nouveau_mm_fini(&vram->mm);
+		return ret;
+	}
+
+	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nvd0_display.c b/drivers/gpu/drm/nouveau/nvd0_display.c
new file mode 100644
index 00000000000..23d63b4b3d7
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvd0_display.c
@@ -0,0 +1,1473 @@
+/*
+ * Copyright 2011 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <linux/dma-mapping.h>
+
+#include "drmP.h"
+#include "drm_crtc_helper.h"
+
+#include "nouveau_drv.h"
+#include "nouveau_connector.h"
+#include "nouveau_encoder.h"
+#include "nouveau_crtc.h"
+#include "nouveau_dma.h"
+#include "nouveau_fb.h"
+#include "nv50_display.h"
+
+struct nvd0_display {
+	struct nouveau_gpuobj *mem;
+	struct {
+		dma_addr_t handle;
+		u32 *ptr;
+	} evo[1];
+
+	struct tasklet_struct tasklet;
+	u32 modeset;
+};
+
+static struct nvd0_display *
+nvd0_display(struct drm_device *dev)
+{
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	return dev_priv->engine.display.priv;
+}
+
+static inline int
+evo_icmd(struct drm_device *dev, int id, u32 mthd, u32 data)
+{
+	int ret = 0;
+	nv_mask(dev, 0x610700 + (id * 0x10), 0x00000001, 0x00000001);
+	nv_wr32(dev, 0x610704 + (id * 0x10), data);
+	nv_mask(dev, 0x610704 + (id * 0x10), 0x80000ffc, 0x80000000 | mthd);
+	if (!nv_wait(dev, 0x610704 + (id * 0x10), 0x80000000, 0x00000000))
+		ret = -EBUSY;
+	nv_mask(dev, 0x610700 + (id * 0x10), 0x00000001, 0x00000000);
+	return ret;
+}
+
+static u32 *
+evo_wait(struct drm_device *dev, int id, int nr)
+{
+	struct nvd0_display *disp = nvd0_display(dev);
+	u32 put = nv_rd32(dev, 0x640000 + (id * 0x1000)) / 4;
+
+	if (put + nr >= (PAGE_SIZE / 4)) {
+		disp->evo[id].ptr[put] = 0x20000000;
+
+		nv_wr32(dev, 0x640000 + (id * 0x1000), 0x00000000);
+		if (!nv_wait(dev, 0x640004 + (id * 0x1000), ~0, 0x00000000)) {
+			NV_ERROR(dev, "evo %d dma stalled\n", id);
+			return NULL;
+		}
+
+		put = 0;
+	}
+
+	return disp->evo[id].ptr + put;
+}
+
+static void
+evo_kick(u32 *push, struct drm_device *dev, int id)
+{
+	struct nvd0_display *disp = nvd0_display(dev);
+	nv_wr32(dev, 0x640000 + (id * 0x1000), (push - disp->evo[id].ptr) << 2);
+}
+
+#define evo_mthd(p,m,s) *((p)++) = (((s) << 18) | (m))
+#define evo_data(p,d)   *((p)++) = (d)
+
+static struct drm_crtc *
+nvd0_display_crtc_get(struct drm_encoder *encoder)
+{
+	return nouveau_encoder(encoder)->crtc;
+}
+
+/******************************************************************************
+ * CRTC
+ *****************************************************************************/
+static int
+nvd0_crtc_set_dither(struct nouveau_crtc *nv_crtc, bool on, bool update)
+{
+	struct drm_device *dev = nv_crtc->base.dev;
+	u32 *push, mode;
+
+	mode = 0x00000000;
+	if (on) {
+		/* 0x11: 6bpc dynamic 2x2
+		 * 0x13: 8bpc dynamic 2x2
+		 * 0x19: 6bpc static 2x2
+		 * 0x1b: 8bpc static 2x2
+		 * 0x21: 6bpc temporal
+		 * 0x23: 8bpc temporal
+		 */
+		mode = 0x00000011;
+	}
+
+	push = evo_wait(dev, 0, 4);
+	if (push) {
+		evo_mthd(push, 0x0490 + (nv_crtc->index * 0x300), 1);
+		evo_data(push, mode);
+		if (update) {
+			evo_mthd(push, 0x0080, 1);
+			evo_data(push, 0x00000000);
+		}
+		evo_kick(push, dev, 0);
+	}
+
+	return 0;
+}
+
+static int
+nvd0_crtc_set_scale(struct nouveau_crtc *nv_crtc, int type, bool update)
+{
+	struct drm_display_mode *mode = &nv_crtc->base.mode;
+	struct drm_device *dev = nv_crtc->base.dev;
+	struct nouveau_connector *nv_connector;
+	u32 *push, outX, outY;
+
+	outX = mode->hdisplay;
+	outY = mode->vdisplay;
+
+	nv_connector = nouveau_crtc_connector_get(nv_crtc);
+	if (nv_connector && nv_connector->native_mode) {
+		struct drm_display_mode *native = nv_connector->native_mode;
+		u32 xratio = (native->hdisplay << 19) / mode->hdisplay;
+		u32 yratio = (native->vdisplay << 19) / mode->vdisplay;
+
+		switch (type) {
+		case DRM_MODE_SCALE_ASPECT:
+			if (xratio > yratio) {
+				outX = (mode->hdisplay * yratio) >> 19;
+				outY = (mode->vdisplay * yratio) >> 19;
+			} else {
+				outX = (mode->hdisplay * xratio) >> 19;
+				outY = (mode->vdisplay * xratio) >> 19;
+			}
+			break;
+		case DRM_MODE_SCALE_FULLSCREEN:
+			outX = native->hdisplay;
+			outY = native->vdisplay;
+			break;
+		default:
+			break;
+		}
+	}
+
+	push = evo_wait(dev, 0, 16);
+	if (push) {
+		evo_mthd(push, 0x04c0 + (nv_crtc->index * 0x300), 3);
+		evo_data(push, (outY << 16) | outX);
+		evo_data(push, (outY << 16) | outX);
+		evo_data(push, (outY << 16) | outX);
+		evo_mthd(push, 0x0494 + (nv_crtc->index * 0x300), 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x04b8 + (nv_crtc->index * 0x300), 1);
+		evo_data(push, (mode->vdisplay << 16) | mode->hdisplay);
+		if (update) {
+			evo_mthd(push, 0x0080, 1);
+			evo_data(push, 0x00000000);
+		}
+		evo_kick(push, dev, 0);
+	}
+
+	return 0;
+}
+
+static int
+nvd0_crtc_set_image(struct nouveau_crtc *nv_crtc, struct drm_framebuffer *fb,
+		    int x, int y, bool update)
+{
+	struct nouveau_framebuffer *nvfb = nouveau_framebuffer(fb);
+	u32 *push;
+
+	push = evo_wait(fb->dev, 0, 16);
+	if (push) {
+		evo_mthd(push, 0x0460 + (nv_crtc->index * 0x300), 1);
+		evo_data(push, nvfb->nvbo->bo.offset >> 8);
+		evo_mthd(push, 0x0468 + (nv_crtc->index * 0x300), 4);
+		evo_data(push, (fb->height << 16) | fb->width);
+		evo_data(push, nvfb->r_pitch);
+		evo_data(push, nvfb->r_format);
+		evo_data(push, nvfb->r_dma);
+		evo_mthd(push, 0x04b0 + (nv_crtc->index * 0x300), 1);
+		evo_data(push, (y << 16) | x);
+		if (update) {
+			evo_mthd(push, 0x0080, 1);
+			evo_data(push, 0x00000000);
+		}
+		evo_kick(push, fb->dev, 0);
+	}
+
+	nv_crtc->fb.tile_flags = nvfb->r_dma;
+	return 0;
+}
+
+static void
+nvd0_crtc_cursor_show(struct nouveau_crtc *nv_crtc, bool show, bool update)
+{
+	struct drm_device *dev = nv_crtc->base.dev;
+	u32 *push = evo_wait(dev, 0, 16);
+	if (push) {
+		if (show) {
+			evo_mthd(push, 0x0480 + (nv_crtc->index * 0x300), 2);
+			evo_data(push, 0x85000000);
+			evo_data(push, nv_crtc->cursor.nvbo->bo.offset >> 8);
+			evo_mthd(push, 0x048c + (nv_crtc->index * 0x300), 1);
+			evo_data(push, NvEvoVRAM);
+		} else {
+			evo_mthd(push, 0x0480 + (nv_crtc->index * 0x300), 1);
+			evo_data(push, 0x05000000);
+			evo_mthd(push, 0x048c + (nv_crtc->index * 0x300), 1);
+			evo_data(push, 0x00000000);
+		}
+
+		if (update) {
+			evo_mthd(push, 0x0080, 1);
+			evo_data(push, 0x00000000);
+		}
+
+		evo_kick(push, dev, 0);
+	}
+}
+
+static void
+nvd0_crtc_dpms(struct drm_crtc *crtc, int mode)
+{
+}
+
+static void
+nvd0_crtc_prepare(struct drm_crtc *crtc)
+{
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+	u32 *push;
+
+	push = evo_wait(crtc->dev, 0, 2);
+	if (push) {
+		evo_mthd(push, 0x0474 + (nv_crtc->index * 0x300), 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x0440 + (nv_crtc->index * 0x300), 1);
+		evo_data(push, 0x03000000);
+		evo_mthd(push, 0x045c + (nv_crtc->index * 0x300), 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, crtc->dev, 0);
+	}
+
+	nvd0_crtc_cursor_show(nv_crtc, false, false);
+}
+
+static void
+nvd0_crtc_commit(struct drm_crtc *crtc)
+{
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+	u32 *push;
+
+	push = evo_wait(crtc->dev, 0, 32);
+	if (push) {
+		evo_mthd(push, 0x0474 + (nv_crtc->index * 0x300), 1);
+		evo_data(push, nv_crtc->fb.tile_flags);
+		evo_mthd(push, 0x0440 + (nv_crtc->index * 0x300), 4);
+		evo_data(push, 0x83000000);
+		evo_data(push, nv_crtc->lut.nvbo->bo.offset >> 8);
+		evo_data(push, 0x00000000);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x045c + (nv_crtc->index * 0x300), 1);
+		evo_data(push, NvEvoVRAM);
+		evo_mthd(push, 0x0430 + (nv_crtc->index * 0x300), 1);
+		evo_data(push, 0xffffff00);
+		evo_kick(push, crtc->dev, 0);
+	}
+
+	nvd0_crtc_cursor_show(nv_crtc, nv_crtc->cursor.visible, true);
+}
+
+static bool
+nvd0_crtc_mode_fixup(struct drm_crtc *crtc, struct drm_display_mode *mode,
+		     struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
+static int
+nvd0_crtc_swap_fbs(struct drm_crtc *crtc, struct drm_framebuffer *old_fb)
+{
+	struct nouveau_framebuffer *nvfb = nouveau_framebuffer(crtc->fb);
+	int ret;
+
+	ret = nouveau_bo_pin(nvfb->nvbo, TTM_PL_FLAG_VRAM);
+	if (ret)
+		return ret;
+
+	if (old_fb) {
+		nvfb = nouveau_framebuffer(old_fb);
+		nouveau_bo_unpin(nvfb->nvbo);
+	}
+
+	return 0;
+}
+
+static int
+nvd0_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *umode,
+		   struct drm_display_mode *mode, int x, int y,
+		   struct drm_framebuffer *old_fb)
+{
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+	struct nouveau_connector *nv_connector;
+	u32 htotal = mode->htotal;
+	u32 vtotal = mode->vtotal;
+	u32 hsyncw = mode->hsync_end - mode->hsync_start - 1;
+	u32 vsyncw = mode->vsync_end - mode->vsync_start - 1;
+	u32 hfrntp = mode->hsync_start - mode->hdisplay;
+	u32 vfrntp = mode->vsync_start - mode->vdisplay;
+	u32 hbackp = mode->htotal - mode->hsync_end;
+	u32 vbackp = mode->vtotal - mode->vsync_end;
+	u32 hss2be = hsyncw + hbackp;
+	u32 vss2be = vsyncw + vbackp;
+	u32 hss2de = htotal - hfrntp;
+	u32 vss2de = vtotal - vfrntp;
+	u32 syncs, *push;
+	int ret;
+
+	syncs = 0x00000001;
+	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+		syncs |= 0x00000008;
+	if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+		syncs |= 0x00000010;
+
+	ret = nvd0_crtc_swap_fbs(crtc, old_fb);
+	if (ret)
+		return ret;
+
+	push = evo_wait(crtc->dev, 0, 64);
+	if (push) {
+		evo_mthd(push, 0x0410 + (nv_crtc->index * 0x300), 5);
+		evo_data(push, 0x00000000);
+		evo_data(push, (vtotal << 16) | htotal);
+		evo_data(push, (vsyncw << 16) | hsyncw);
+		evo_data(push, (vss2be << 16) | hss2be);
+		evo_data(push, (vss2de << 16) | hss2de);
+		evo_mthd(push, 0x042c + (nv_crtc->index * 0x300), 1);
+		evo_data(push, 0x00000000); /* ??? */
+		evo_mthd(push, 0x0450 + (nv_crtc->index * 0x300), 3);
+		evo_data(push, mode->clock * 1000);
+		evo_data(push, 0x00200000); /* ??? */
+		evo_data(push, mode->clock * 1000);
+		evo_mthd(push, 0x0404 + (nv_crtc->index * 0x300), 1);
+		evo_data(push, syncs);
+		evo_kick(push, crtc->dev, 0);
+	}
+
+	nv_connector = nouveau_crtc_connector_get(nv_crtc);
+	nvd0_crtc_set_dither(nv_crtc, nv_connector->use_dithering, false);
+	nvd0_crtc_set_scale(nv_crtc, nv_connector->scaling_mode, false);
+	nvd0_crtc_set_image(nv_crtc, crtc->fb, x, y, false);
+	return 0;
+}
+
+static int
+nvd0_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
+			struct drm_framebuffer *old_fb)
+{
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+	int ret;
+
+	if (!crtc->fb) {
+		NV_DEBUG_KMS(crtc->dev, "No FB bound\n");
+		return 0;
+	}
+
+	ret = nvd0_crtc_swap_fbs(crtc, old_fb);
+	if (ret)
+		return ret;
+
+	nvd0_crtc_set_image(nv_crtc, crtc->fb, x, y, true);
+	return 0;
+}
+
+static int
+nvd0_crtc_mode_set_base_atomic(struct drm_crtc *crtc,
+			       struct drm_framebuffer *fb, int x, int y,
+			       enum mode_set_atomic state)
+{
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+	nvd0_crtc_set_image(nv_crtc, fb, x, y, true);
+	return 0;
+}
+
+static void
+nvd0_crtc_lut_load(struct drm_crtc *crtc)
+{
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+	void __iomem *lut = nvbo_kmap_obj_iovirtual(nv_crtc->lut.nvbo);
+	int i;
+
+	for (i = 0; i < 256; i++) {
+		writew(0x6000 + (nv_crtc->lut.r[i] >> 2), lut + (i * 0x20) + 0);
+		writew(0x6000 + (nv_crtc->lut.g[i] >> 2), lut + (i * 0x20) + 2);
+		writew(0x6000 + (nv_crtc->lut.b[i] >> 2), lut + (i * 0x20) + 4);
+	}
+}
+
+static int
+nvd0_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv,
+		     uint32_t handle, uint32_t width, uint32_t height)
+{
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+	struct drm_device *dev = crtc->dev;
+	struct drm_gem_object *gem;
+	struct nouveau_bo *nvbo;
+	bool visible = (handle != 0);
+	int i, ret = 0;
+
+	if (visible) {
+		if (width != 64 || height != 64)
+			return -EINVAL;
+
+		gem = drm_gem_object_lookup(dev, file_priv, handle);
+		if (unlikely(!gem))
+			return -ENOENT;
+		nvbo = nouveau_gem_object(gem);
+
+		ret = nouveau_bo_map(nvbo);
+		if (ret == 0) {
+			for (i = 0; i < 64 * 64; i++) {
+				u32 v = nouveau_bo_rd32(nvbo, i);
+				nouveau_bo_wr32(nv_crtc->cursor.nvbo, i, v);
+			}
+			nouveau_bo_unmap(nvbo);
+		}
+
+		drm_gem_object_unreference_unlocked(gem);
+	}
+
+	if (visible != nv_crtc->cursor.visible) {
+		nvd0_crtc_cursor_show(nv_crtc, visible, true);
+		nv_crtc->cursor.visible = visible;
+	}
+
+	return ret;
+}
+
+static int
+nvd0_crtc_cursor_move(struct drm_crtc *crtc, int x, int y)
+{
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+	const u32 data = (y << 16) | x;
+
+	nv_wr32(crtc->dev, 0x64d084 + (nv_crtc->index * 0x1000), data);
+	nv_wr32(crtc->dev, 0x64d080 + (nv_crtc->index * 0x1000), 0x00000000);
+	return 0;
+}
+
+static void
+nvd0_crtc_gamma_set(struct drm_crtc *crtc, u16 *r, u16 *g, u16 *b,
+		    uint32_t start, uint32_t size)
+{
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+	u32 end = max(start + size, (u32)256);
+	u32 i;
+
+	for (i = start; i < end; i++) {
+		nv_crtc->lut.r[i] = r[i];
+		nv_crtc->lut.g[i] = g[i];
+		nv_crtc->lut.b[i] = b[i];
+	}
+
+	nvd0_crtc_lut_load(crtc);
+}
+
+static void
+nvd0_crtc_destroy(struct drm_crtc *crtc)
+{
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+	nouveau_bo_unmap(nv_crtc->cursor.nvbo);
+	nouveau_bo_ref(NULL, &nv_crtc->cursor.nvbo);
+	nouveau_bo_unmap(nv_crtc->lut.nvbo);
+	nouveau_bo_ref(NULL, &nv_crtc->lut.nvbo);
+	drm_crtc_cleanup(crtc);
+	kfree(crtc);
+}
+
+static const struct drm_crtc_helper_funcs nvd0_crtc_hfunc = {
+	.dpms = nvd0_crtc_dpms,
+	.prepare = nvd0_crtc_prepare,
+	.commit = nvd0_crtc_commit,
+	.mode_fixup = nvd0_crtc_mode_fixup,
+	.mode_set = nvd0_crtc_mode_set,
+	.mode_set_base = nvd0_crtc_mode_set_base,
+	.mode_set_base_atomic = nvd0_crtc_mode_set_base_atomic,
+	.load_lut = nvd0_crtc_lut_load,
+};
+
+static const struct drm_crtc_funcs nvd0_crtc_func = {
+	.cursor_set = nvd0_crtc_cursor_set,
+	.cursor_move = nvd0_crtc_cursor_move,
+	.gamma_set = nvd0_crtc_gamma_set,
+	.set_config = drm_crtc_helper_set_config,
+	.destroy = nvd0_crtc_destroy,
+};
+
+static void
+nvd0_cursor_set_pos(struct nouveau_crtc *nv_crtc, int x, int y)
+{
+}
+
+static void
+nvd0_cursor_set_offset(struct nouveau_crtc *nv_crtc, uint32_t offset)
+{
+}
+
+static int
+nvd0_crtc_create(struct drm_device *dev, int index)
+{
+	struct nouveau_crtc *nv_crtc;
+	struct drm_crtc *crtc;
+	int ret, i;
+
+	nv_crtc = kzalloc(sizeof(*nv_crtc), GFP_KERNEL);
+	if (!nv_crtc)
+		return -ENOMEM;
+
+	nv_crtc->index = index;
+	nv_crtc->set_dither = nvd0_crtc_set_dither;
+	nv_crtc->set_scale = nvd0_crtc_set_scale;
+	nv_crtc->cursor.set_offset = nvd0_cursor_set_offset;
+	nv_crtc->cursor.set_pos = nvd0_cursor_set_pos;
+	for (i = 0; i < 256; i++) {
+		nv_crtc->lut.r[i] = i << 8;
+		nv_crtc->lut.g[i] = i << 8;
+		nv_crtc->lut.b[i] = i << 8;
+	}
+
+	crtc = &nv_crtc->base;
+	drm_crtc_init(dev, crtc, &nvd0_crtc_func);
+	drm_crtc_helper_add(crtc, &nvd0_crtc_hfunc);
+	drm_mode_crtc_set_gamma_size(crtc, 256);
+
+	ret = nouveau_bo_new(dev, 64 * 64 * 4, 0x100, TTM_PL_FLAG_VRAM,
+			     0, 0x0000, &nv_crtc->cursor.nvbo);
+	if (!ret) {
+		ret = nouveau_bo_pin(nv_crtc->cursor.nvbo, TTM_PL_FLAG_VRAM);
+		if (!ret)
+			ret = nouveau_bo_map(nv_crtc->cursor.nvbo);
+		if (ret)
+			nouveau_bo_ref(NULL, &nv_crtc->cursor.nvbo);
+	}
+
+	if (ret)
+		goto out;
+
+	ret = nouveau_bo_new(dev, 8192, 0x100, TTM_PL_FLAG_VRAM,
+			     0, 0x0000, &nv_crtc->lut.nvbo);
+	if (!ret) {
+		ret = nouveau_bo_pin(nv_crtc->lut.nvbo, TTM_PL_FLAG_VRAM);
+		if (!ret)
+			ret = nouveau_bo_map(nv_crtc->lut.nvbo);
+		if (ret)
+			nouveau_bo_ref(NULL, &nv_crtc->lut.nvbo);
+	}
+
+	if (ret)
+		goto out;
+
+	nvd0_crtc_lut_load(crtc);
+
+out:
+	if (ret)
+		nvd0_crtc_destroy(crtc);
+	return ret;
+}
+
+/******************************************************************************
+ * DAC
+ *****************************************************************************/
+static void
+nvd0_dac_dpms(struct drm_encoder *encoder, int mode)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct drm_device *dev = encoder->dev;
+	int or = nv_encoder->or;
+	u32 dpms_ctrl;
+
+	dpms_ctrl = 0x80000000;
+	if (mode == DRM_MODE_DPMS_STANDBY || mode == DRM_MODE_DPMS_OFF)
+		dpms_ctrl |= 0x00000001;
+	if (mode == DRM_MODE_DPMS_SUSPEND || mode == DRM_MODE_DPMS_OFF)
+		dpms_ctrl |= 0x00000004;
+
+	nv_wait(dev, 0x61a004 + (or * 0x0800), 0x80000000, 0x00000000);
+	nv_mask(dev, 0x61a004 + (or * 0x0800), 0xc000007f, dpms_ctrl);
+	nv_wait(dev, 0x61a004 + (or * 0x0800), 0x80000000, 0x00000000);
+}
+
+static bool
+nvd0_dac_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *mode,
+		    struct drm_display_mode *adjusted_mode)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_connector *nv_connector;
+
+	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	if (nv_connector && nv_connector->native_mode) {
+		if (nv_connector->scaling_mode != DRM_MODE_SCALE_NONE) {
+			int id = adjusted_mode->base.id;
+			*adjusted_mode = *nv_connector->native_mode;
+			adjusted_mode->base.id = id;
+		}
+	}
+
+	return true;
+}
+
+static void
+nvd0_dac_prepare(struct drm_encoder *encoder)
+{
+}
+
+static void
+nvd0_dac_commit(struct drm_encoder *encoder)
+{
+}
+
+static void
+nvd0_dac_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode,
+		  struct drm_display_mode *adjusted_mode)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	u32 *push;
+
+	nvd0_dac_dpms(encoder, DRM_MODE_DPMS_ON);
+
+	push = evo_wait(encoder->dev, 0, 4);
+	if (push) {
+		evo_mthd(push, 0x0180 + (nv_encoder->or * 0x20), 2);
+		evo_data(push, 1 << nv_crtc->index);
+		evo_data(push, 0x00ff);
+		evo_kick(push, encoder->dev, 0);
+	}
+
+	nv_encoder->crtc = encoder->crtc;
+}
+
+static void
+nvd0_dac_disconnect(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct drm_device *dev = encoder->dev;
+	u32 *push;
+
+	if (nv_encoder->crtc) {
+		nvd0_crtc_prepare(nv_encoder->crtc);
+
+		push = evo_wait(dev, 0, 4);
+		if (push) {
+			evo_mthd(push, 0x0180 + (nv_encoder->or * 0x20), 1);
+			evo_data(push, 0x00000000);
+			evo_mthd(push, 0x0080, 1);
+			evo_data(push, 0x00000000);
+			evo_kick(push, dev, 0);
+		}
+
+		nv_encoder->crtc = NULL;
+	}
+}
+
+static enum drm_connector_status
+nvd0_dac_detect(struct drm_encoder *encoder, struct drm_connector *connector)
+{
+	enum drm_connector_status status = connector_status_disconnected;
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct drm_device *dev = encoder->dev;
+	int or = nv_encoder->or;
+	u32 load;
+
+	nv_wr32(dev, 0x61a00c + (or * 0x800), 0x00100000);
+	udelay(9500);
+	nv_wr32(dev, 0x61a00c + (or * 0x800), 0x80000000);
+
+	load = nv_rd32(dev, 0x61a00c + (or * 0x800));
+	if ((load & 0x38000000) == 0x38000000)
+		status = connector_status_connected;
+
+	nv_wr32(dev, 0x61a00c + (or * 0x800), 0x00000000);
+	return status;
+}
+
+static void
+nvd0_dac_destroy(struct drm_encoder *encoder)
+{
+	drm_encoder_cleanup(encoder);
+	kfree(encoder);
+}
+
+static const struct drm_encoder_helper_funcs nvd0_dac_hfunc = {
+	.dpms = nvd0_dac_dpms,
+	.mode_fixup = nvd0_dac_mode_fixup,
+	.prepare = nvd0_dac_prepare,
+	.commit = nvd0_dac_commit,
+	.mode_set = nvd0_dac_mode_set,
+	.disable = nvd0_dac_disconnect,
+	.get_crtc = nvd0_display_crtc_get,
+	.detect = nvd0_dac_detect
+};
+
+static const struct drm_encoder_funcs nvd0_dac_func = {
+	.destroy = nvd0_dac_destroy,
+};
+
+static int
+nvd0_dac_create(struct drm_connector *connector, struct dcb_entry *dcbe)
+{
+	struct drm_device *dev = connector->dev;
+	struct nouveau_encoder *nv_encoder;
+	struct drm_encoder *encoder;
+
+	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
+	if (!nv_encoder)
+		return -ENOMEM;
+	nv_encoder->dcb = dcbe;
+	nv_encoder->or = ffs(dcbe->or) - 1;
+
+	encoder = to_drm_encoder(nv_encoder);
+	encoder->possible_crtcs = dcbe->heads;
+	encoder->possible_clones = 0;
+	drm_encoder_init(dev, encoder, &nvd0_dac_func, DRM_MODE_ENCODER_DAC);
+	drm_encoder_helper_add(encoder, &nvd0_dac_hfunc);
+
+	drm_mode_connector_attach_encoder(connector, encoder);
+	return 0;
+}
+
+/******************************************************************************
+ * SOR
+ *****************************************************************************/
+static void
+nvd0_sor_dpms(struct drm_encoder *encoder, int mode)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct drm_device *dev = encoder->dev;
+	struct drm_encoder *partner;
+	int or = nv_encoder->or;
+	u32 dpms_ctrl;
+
+	nv_encoder->last_dpms = mode;
+
+	list_for_each_entry(partner, &dev->mode_config.encoder_list, head) {
+		struct nouveau_encoder *nv_partner = nouveau_encoder(partner);
+
+		if (partner->encoder_type != DRM_MODE_ENCODER_TMDS)
+			continue;
+
+		if (nv_partner != nv_encoder &&
+		    nv_partner->dcb->or == nv_encoder->or) {
+			if (nv_partner->last_dpms == DRM_MODE_DPMS_ON)
+				return;
+			break;
+		}
+	}
+
+	dpms_ctrl  = (mode == DRM_MODE_DPMS_ON);
+	dpms_ctrl |= 0x80000000;
+
+	nv_wait(dev, 0x61c004 + (or * 0x0800), 0x80000000, 0x00000000);
+	nv_mask(dev, 0x61c004 + (or * 0x0800), 0x80000001, dpms_ctrl);
+	nv_wait(dev, 0x61c004 + (or * 0x0800), 0x80000000, 0x00000000);
+	nv_wait(dev, 0x61c030 + (or * 0x0800), 0x10000000, 0x00000000);
+}
+
+static bool
+nvd0_sor_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *mode,
+		    struct drm_display_mode *adjusted_mode)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_connector *nv_connector;
+
+	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	if (nv_connector && nv_connector->native_mode) {
+		if (nv_connector->scaling_mode != DRM_MODE_SCALE_NONE) {
+			int id = adjusted_mode->base.id;
+			*adjusted_mode = *nv_connector->native_mode;
+			adjusted_mode->base.id = id;
+		}
+	}
+
+	return true;
+}
+
+static void
+nvd0_sor_prepare(struct drm_encoder *encoder)
+{
+}
+
+static void
+nvd0_sor_commit(struct drm_encoder *encoder)
+{
+}
+
+static void
+nvd0_sor_mode_set(struct drm_encoder *encoder, struct drm_display_mode *umode,
+		  struct drm_display_mode *mode)
+{
+	struct drm_nouveau_private *dev_priv = encoder->dev->dev_private;
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct nouveau_connector *nv_connector;
+	struct nvbios *bios = &dev_priv->vbios;
+	u32 mode_ctrl = (1 << nv_crtc->index);
+	u32 *push, or_config;
+
+	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	switch (nv_encoder->dcb->type) {
+	case OUTPUT_TMDS:
+		if (nv_encoder->dcb->sorconf.link & 1) {
+			if (mode->clock < 165000)
+				mode_ctrl |= 0x00000100;
+			else
+				mode_ctrl |= 0x00000500;
+		} else {
+			mode_ctrl |= 0x00000200;
+		}
+
+		or_config = (mode_ctrl & 0x00000f00) >> 8;
+		if (mode->clock >= 165000)
+			or_config |= 0x0100;
+		break;
+	case OUTPUT_LVDS:
+		or_config = (mode_ctrl & 0x00000f00) >> 8;
+		if (bios->fp_no_ddc) {
+			if (bios->fp.dual_link)
+				or_config |= 0x0100;
+			if (bios->fp.if_is_24bit)
+				or_config |= 0x0200;
+		} else {
+			if (nv_connector->dcb->type == DCB_CONNECTOR_LVDS_SPWG) {
+				if (((u8 *)nv_connector->edid)[121] == 2)
+					or_config |= 0x0100;
+			} else
+			if (mode->clock >= bios->fp.duallink_transition_clk) {
+				or_config |= 0x0100;
+			}
+
+			if (or_config & 0x0100) {
+				if (bios->fp.strapless_is_24bit & 2)
+					or_config |= 0x0200;
+			} else {
+				if (bios->fp.strapless_is_24bit & 1)
+					or_config |= 0x0200;
+			}
+
+			if (nv_connector->base.display_info.bpc == 8)
+				or_config |= 0x0200;
+
+		}
+		break;
+	default:
+		BUG_ON(1);
+		break;
+	}
+
+	nvd0_sor_dpms(encoder, DRM_MODE_DPMS_ON);
+
+	push = evo_wait(encoder->dev, 0, 4);
+	if (push) {
+		evo_mthd(push, 0x0200 + (nv_encoder->or * 0x20), 2);
+		evo_data(push, mode_ctrl);
+		evo_data(push, or_config);
+		evo_kick(push, encoder->dev, 0);
+	}
+
+	nv_encoder->crtc = encoder->crtc;
+}
+
+static void
+nvd0_sor_disconnect(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct drm_device *dev = encoder->dev;
+	u32 *push;
+
+	if (nv_encoder->crtc) {
+		nvd0_crtc_prepare(nv_encoder->crtc);
+
+		push = evo_wait(dev, 0, 4);
+		if (push) {
+			evo_mthd(push, 0x0200 + (nv_encoder->or * 0x20), 1);
+			evo_data(push, 0x00000000);
+			evo_mthd(push, 0x0080, 1);
+			evo_data(push, 0x00000000);
+			evo_kick(push, dev, 0);
+		}
+
+		nv_encoder->crtc = NULL;
+		nv_encoder->last_dpms = DRM_MODE_DPMS_OFF;
+	}
+}
+
+static void
+nvd0_sor_destroy(struct drm_encoder *encoder)
+{
+	drm_encoder_cleanup(encoder);
+	kfree(encoder);
+}
+
+static const struct drm_encoder_helper_funcs nvd0_sor_hfunc = {
+	.dpms = nvd0_sor_dpms,
+	.mode_fixup = nvd0_sor_mode_fixup,
+	.prepare = nvd0_sor_prepare,
+	.commit = nvd0_sor_commit,
+	.mode_set = nvd0_sor_mode_set,
+	.disable = nvd0_sor_disconnect,
+	.get_crtc = nvd0_display_crtc_get,
+};
+
+static const struct drm_encoder_funcs nvd0_sor_func = {
+	.destroy = nvd0_sor_destroy,
+};
+
+static int
+nvd0_sor_create(struct drm_connector *connector, struct dcb_entry *dcbe)
+{
+	struct drm_device *dev = connector->dev;
+	struct nouveau_encoder *nv_encoder;
+	struct drm_encoder *encoder;
+
+	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
+	if (!nv_encoder)
+		return -ENOMEM;
+	nv_encoder->dcb = dcbe;
+	nv_encoder->or = ffs(dcbe->or) - 1;
+	nv_encoder->last_dpms = DRM_MODE_DPMS_OFF;
+
+	encoder = to_drm_encoder(nv_encoder);
+	encoder->possible_crtcs = dcbe->heads;
+	encoder->possible_clones = 0;
+	drm_encoder_init(dev, encoder, &nvd0_sor_func, DRM_MODE_ENCODER_TMDS);
+	drm_encoder_helper_add(encoder, &nvd0_sor_hfunc);
+
+	drm_mode_connector_attach_encoder(connector, encoder);
+	return 0;
+}
+
+/******************************************************************************
+ * IRQ
+ *****************************************************************************/
+static struct dcb_entry *
+lookup_dcb(struct drm_device *dev, int id, u32 mc)
+{
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	int type, or, i;
+
+	if (id < 4) {
+		type = OUTPUT_ANALOG;
+		or   = id;
+	} else {
+		switch (mc & 0x00000f00) {
+		case 0x00000000: type = OUTPUT_LVDS; break;
+		case 0x00000100: type = OUTPUT_TMDS; break;
+		case 0x00000200: type = OUTPUT_TMDS; break;
+		case 0x00000500: type = OUTPUT_TMDS; break;
+		default:
+			NV_ERROR(dev, "PDISP: unknown SOR mc 0x%08x\n", mc);
+			return NULL;
+		}
+
+		or = id - 4;
+	}
+
+	for (i = 0; i < dev_priv->vbios.dcb.entries; i++) {
+		struct dcb_entry *dcb = &dev_priv->vbios.dcb.entry[i];
+		if (dcb->type == type && (dcb->or & (1 << or)))
+			return dcb;
+	}
+
+	NV_ERROR(dev, "PDISP: DCB for %d/0x%08x not found\n", id, mc);
+	return NULL;
+}
+
+static void
+nvd0_display_unk1_handler(struct drm_device *dev, u32 crtc, u32 mask)
+{
+	struct dcb_entry *dcb;
+	int i;
+
+	for (i = 0; mask && i < 8; i++) {
+		u32 mcc = nv_rd32(dev, 0x640180 + (i * 0x20));
+		if (!(mcc & (1 << crtc)))
+			continue;
+
+		dcb = lookup_dcb(dev, i, mcc);
+		if (!dcb)
+			continue;
+
+		nouveau_bios_run_display_table(dev, 0x0000, -1, dcb, crtc);
+	}
+
+	nv_wr32(dev, 0x6101d4, 0x00000000);
+	nv_wr32(dev, 0x6109d4, 0x00000000);
+	nv_wr32(dev, 0x6101d0, 0x80000000);
+}
+
+static void
+nvd0_display_unk2_handler(struct drm_device *dev, u32 crtc, u32 mask)
+{
+	struct dcb_entry *dcb;
+	u32 or, tmp, pclk;
+	int i;
+
+	for (i = 0; mask && i < 8; i++) {
+		u32 mcc = nv_rd32(dev, 0x640180 + (i * 0x20));
+		if (!(mcc & (1 << crtc)))
+			continue;
+
+		dcb = lookup_dcb(dev, i, mcc);
+		if (!dcb)
+			continue;
+
+		nouveau_bios_run_display_table(dev, 0x0000, -2, dcb, crtc);
+	}
+
+	pclk = nv_rd32(dev, 0x660450 + (crtc * 0x300)) / 1000;
+	if (mask & 0x00010000) {
+		nv50_crtc_set_clock(dev, crtc, pclk);
+	}
+
+	for (i = 0; mask && i < 8; i++) {
+		u32 mcp = nv_rd32(dev, 0x660180 + (i * 0x20));
+		u32 cfg = nv_rd32(dev, 0x660184 + (i * 0x20));
+		if (!(mcp & (1 << crtc)))
+			continue;
+
+		dcb = lookup_dcb(dev, i, mcp);
+		if (!dcb)
+			continue;
+		or = ffs(dcb->or) - 1;
+
+		nouveau_bios_run_display_table(dev, cfg, pclk, dcb, crtc);
+
+		nv_wr32(dev, 0x612200 + (crtc * 0x800), 0x00000000);
+		switch (dcb->type) {
+		case OUTPUT_ANALOG:
+			nv_wr32(dev, 0x612280 + (or * 0x800), 0x00000000);
+			break;
+		case OUTPUT_TMDS:
+		case OUTPUT_LVDS:
+			if (cfg & 0x00000100)
+				tmp = 0x00000101;
+			else
+				tmp = 0x00000000;
+
+			nv_mask(dev, 0x612300 + (or * 0x800), 0x00000707, tmp);
+			break;
+		default:
+			break;
+		}
+
+		break;
+	}
+
+	nv_wr32(dev, 0x6101d4, 0x00000000);
+	nv_wr32(dev, 0x6109d4, 0x00000000);
+	nv_wr32(dev, 0x6101d0, 0x80000000);
+}
+
+static void
+nvd0_display_unk4_handler(struct drm_device *dev, u32 crtc, u32 mask)
+{
+	struct dcb_entry *dcb;
+	int pclk, i;
+
+	pclk = nv_rd32(dev, 0x660450 + (crtc * 0x300)) / 1000;
+
+	for (i = 0; mask && i < 8; i++) {
+		u32 mcp = nv_rd32(dev, 0x660180 + (i * 0x20));
+		u32 cfg = nv_rd32(dev, 0x660184 + (i * 0x20));
+		if (!(mcp & (1 << crtc)))
+			continue;
+
+		dcb = lookup_dcb(dev, i, mcp);
+		if (!dcb)
+			continue;
+
+		nouveau_bios_run_display_table(dev, cfg, -pclk, dcb, crtc);
+	}
+
+	nv_wr32(dev, 0x6101d4, 0x00000000);
+	nv_wr32(dev, 0x6109d4, 0x00000000);
+	nv_wr32(dev, 0x6101d0, 0x80000000);
+}
+
+static void
+nvd0_display_bh(unsigned long data)
+{
+	struct drm_device *dev = (struct drm_device *)data;
+	struct nvd0_display *disp = nvd0_display(dev);
+	u32 mask, crtc;
+	int i;
+
+	if (drm_debug & (DRM_UT_DRIVER | DRM_UT_KMS)) {
+		NV_INFO(dev, "PDISP: modeset req %d\n", disp->modeset);
+		NV_INFO(dev, " STAT: 0x%08x 0x%08x 0x%08x\n",
+			 nv_rd32(dev, 0x6101d0),
+			 nv_rd32(dev, 0x6101d4), nv_rd32(dev, 0x6109d4));
+		for (i = 0; i < 8; i++) {
+			NV_INFO(dev, " %s%d: 0x%08x 0x%08x\n",
+				i < 4 ? "DAC" : "SOR", i,
+				nv_rd32(dev, 0x640180 + (i * 0x20)),
+				nv_rd32(dev, 0x660180 + (i * 0x20)));
+		}
+	}
+
+	mask = nv_rd32(dev, 0x6101d4);
+	crtc = 0;
+	if (!mask) {
+		mask = nv_rd32(dev, 0x6109d4);
+		crtc = 1;
+	}
+
+	if (disp->modeset & 0x00000001)
+		nvd0_display_unk1_handler(dev, crtc, mask);
+	if (disp->modeset & 0x00000002)
+		nvd0_display_unk2_handler(dev, crtc, mask);
+	if (disp->modeset & 0x00000004)
+		nvd0_display_unk4_handler(dev, crtc, mask);
+}
+
+static void
+nvd0_display_intr(struct drm_device *dev)
+{
+	struct nvd0_display *disp = nvd0_display(dev);
+	u32 intr = nv_rd32(dev, 0x610088);
+
+	if (intr & 0x00000002) {
+		u32 stat = nv_rd32(dev, 0x61009c);
+		int chid = ffs(stat) - 1;
+		if (chid >= 0) {
+			u32 mthd = nv_rd32(dev, 0x6101f0 + (chid * 12));
+			u32 data = nv_rd32(dev, 0x6101f4 + (chid * 12));
+			u32 unkn = nv_rd32(dev, 0x6101f8 + (chid * 12));
+
+			NV_INFO(dev, "EvoCh: chid %d mthd 0x%04x data 0x%08x "
+				     "0x%08x 0x%08x\n",
+				chid, (mthd & 0x0000ffc), data, mthd, unkn);
+			nv_wr32(dev, 0x61009c, (1 << chid));
+			nv_wr32(dev, 0x6101f0 + (chid * 12), 0x90000000);
+		}
+
+		intr &= ~0x00000002;
+	}
+
+	if (intr & 0x00100000) {
+		u32 stat = nv_rd32(dev, 0x6100ac);
+
+		if (stat & 0x00000007) {
+			disp->modeset = stat;
+			tasklet_schedule(&disp->tasklet);
+
+			nv_wr32(dev, 0x6100ac, (stat & 0x00000007));
+			stat &= ~0x00000007;
+		}
+
+		if (stat) {
+			NV_INFO(dev, "PDISP: unknown intr24 0x%08x\n", stat);
+			nv_wr32(dev, 0x6100ac, stat);
+		}
+
+		intr &= ~0x00100000;
+	}
+
+	if (intr & 0x01000000) {
+		u32 stat = nv_rd32(dev, 0x6100bc);
+		nv_wr32(dev, 0x6100bc, stat);
+		intr &= ~0x01000000;
+	}
+
+	if (intr & 0x02000000) {
+		u32 stat = nv_rd32(dev, 0x6108bc);
+		nv_wr32(dev, 0x6108bc, stat);
+		intr &= ~0x02000000;
+	}
+
+	if (intr)
+		NV_INFO(dev, "PDISP: unknown intr 0x%08x\n", intr);
+}
+
+/******************************************************************************
+ * Init
+ *****************************************************************************/
+static void
+nvd0_display_fini(struct drm_device *dev)
+{
+	int i;
+
+	/* fini cursors */
+	for (i = 14; i >= 13; i--) {
+		if (!(nv_rd32(dev, 0x610490 + (i * 0x10)) & 0x00000001))
+			continue;
+
+		nv_mask(dev, 0x610490 + (i * 0x10), 0x00000001, 0x00000000);
+		nv_wait(dev, 0x610490 + (i * 0x10), 0x00010000, 0x00000000);
+		nv_mask(dev, 0x610090, 1 << i, 0x00000000);
+		nv_mask(dev, 0x6100a0, 1 << i, 0x00000000);
+	}
+
+	/* fini master */
+	if (nv_rd32(dev, 0x610490) & 0x00000010) {
+		nv_mask(dev, 0x610490, 0x00000010, 0x00000000);
+		nv_mask(dev, 0x610490, 0x00000003, 0x00000000);
+		nv_wait(dev, 0x610490, 0x80000000, 0x00000000);
+		nv_mask(dev, 0x610090, 0x00000001, 0x00000000);
+		nv_mask(dev, 0x6100a0, 0x00000001, 0x00000000);
+	}
+}
+
+int
+nvd0_display_init(struct drm_device *dev)
+{
+	struct nvd0_display *disp = nvd0_display(dev);
+	u32 *push;
+	int i;
+
+	if (nv_rd32(dev, 0x6100ac) & 0x00000100) {
+		nv_wr32(dev, 0x6100ac, 0x00000100);
+		nv_mask(dev, 0x6194e8, 0x00000001, 0x00000000);
+		if (!nv_wait(dev, 0x6194e8, 0x00000002, 0x00000000)) {
+			NV_ERROR(dev, "PDISP: 0x6194e8 0x%08x\n",
+				 nv_rd32(dev, 0x6194e8));
+			return -EBUSY;
+		}
+	}
+
+	/* nfi what these are exactly, i do know that SOR_MODE_CTRL won't
+	 * work at all unless you do the SOR part below.
+	 */
+	for (i = 0; i < 3; i++) {
+		u32 dac = nv_rd32(dev, 0x61a000 + (i * 0x800));
+		nv_wr32(dev, 0x6101c0 + (i * 0x800), dac);
+	}
+
+	for (i = 0; i < 4; i++) {
+		u32 sor = nv_rd32(dev, 0x61c000 + (i * 0x800));
+		nv_wr32(dev, 0x6301c4 + (i * 0x800), sor);
+	}
+
+	for (i = 0; i < 2; i++) {
+		u32 crtc0 = nv_rd32(dev, 0x616104 + (i * 0x800));
+		u32 crtc1 = nv_rd32(dev, 0x616108 + (i * 0x800));
+		u32 crtc2 = nv_rd32(dev, 0x61610c + (i * 0x800));
+		nv_wr32(dev, 0x6101b4 + (i * 0x800), crtc0);
+		nv_wr32(dev, 0x6101b8 + (i * 0x800), crtc1);
+		nv_wr32(dev, 0x6101bc + (i * 0x800), crtc2);
+	}
+
+	/* point at our hash table / objects, enable interrupts */
+	nv_wr32(dev, 0x610010, (disp->mem->vinst >> 8) | 9);
+	nv_mask(dev, 0x6100b0, 0x00000307, 0x00000307);
+
+	/* init master */
+	nv_wr32(dev, 0x610494, (disp->evo[0].handle >> 8) | 3);
+	nv_wr32(dev, 0x610498, 0x00010000);
+	nv_wr32(dev, 0x61049c, 0x00000001);
+	nv_mask(dev, 0x610490, 0x00000010, 0x00000010);
+	nv_wr32(dev, 0x640000, 0x00000000);
+	nv_wr32(dev, 0x610490, 0x01000013);
+	if (!nv_wait(dev, 0x610490, 0x80000000, 0x00000000)) {
+		NV_ERROR(dev, "PDISP: master 0x%08x\n",
+			 nv_rd32(dev, 0x610490));
+		return -EBUSY;
+	}
+	nv_mask(dev, 0x610090, 0x00000001, 0x00000001);
+	nv_mask(dev, 0x6100a0, 0x00000001, 0x00000001);
+
+	/* init cursors */
+	for (i = 13; i <= 14; i++) {
+		nv_wr32(dev, 0x610490 + (i * 0x10), 0x00000001);
+		if (!nv_wait(dev, 0x610490 + (i * 0x10), 0x00010000, 0x00010000)) {
+			NV_ERROR(dev, "PDISP: curs%d 0x%08x\n", i,
+				 nv_rd32(dev, 0x610490 + (i * 0x10)));
+			return -EBUSY;
+		}
+
+		nv_mask(dev, 0x610090, 1 << i, 1 << i);
+		nv_mask(dev, 0x6100a0, 1 << i, 1 << i);
+	}
+
+	push = evo_wait(dev, 0, 32);
+	if (!push)
+		return -EBUSY;
+	evo_mthd(push, 0x0088, 1);
+	evo_data(push, NvEvoSync);
+	evo_mthd(push, 0x0084, 1);
+	evo_data(push, 0x00000000);
+	evo_mthd(push, 0x0084, 1);
+	evo_data(push, 0x80000000);
+	evo_mthd(push, 0x008c, 1);
+	evo_data(push, 0x00000000);
+	evo_kick(push, dev, 0);
+
+	return 0;
+}
+
+void
+nvd0_display_destroy(struct drm_device *dev)
+{
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct nvd0_display *disp = nvd0_display(dev);
+	struct pci_dev *pdev = dev->pdev;
+
+	nvd0_display_fini(dev);
+
+	pci_free_consistent(pdev, PAGE_SIZE, disp->evo[0].ptr, disp->evo[0].handle);
+	nouveau_gpuobj_ref(NULL, &disp->mem);
+	nouveau_irq_unregister(dev, 26);
+
+	dev_priv->engine.display.priv = NULL;
+	kfree(disp);
+}
+
+int
+nvd0_display_create(struct drm_device *dev)
+{
+	struct drm_nouveau_private *dev_priv = dev->dev_private;
+	struct nouveau_instmem_engine *pinstmem = &dev_priv->engine.instmem;
+	struct dcb_table *dcb = &dev_priv->vbios.dcb;
+	struct drm_connector *connector, *tmp;
+	struct pci_dev *pdev = dev->pdev;
+	struct nvd0_display *disp;
+	struct dcb_entry *dcbe;
+	int ret, i;
+
+	disp = kzalloc(sizeof(*disp), GFP_KERNEL);
+	if (!disp)
+		return -ENOMEM;
+	dev_priv->engine.display.priv = disp;
+
+	/* create crtc objects to represent the hw heads */
+	for (i = 0; i < 2; i++) {
+		ret = nvd0_crtc_create(dev, i);
+		if (ret)
+			goto out;
+	}
+
+	/* create encoder/connector objects based on VBIOS DCB table */
+	for (i = 0, dcbe = &dcb->entry[0]; i < dcb->entries; i++, dcbe++) {
+		connector = nouveau_connector_create(dev, dcbe->connector);
+		if (IS_ERR(connector))
+			continue;
+
+		if (dcbe->location != DCB_LOC_ON_CHIP) {
+			NV_WARN(dev, "skipping off-chip encoder %d/%d\n",
+				dcbe->type, ffs(dcbe->or) - 1);
+			continue;
+		}
+
+		switch (dcbe->type) {
+		case OUTPUT_TMDS:
+		case OUTPUT_LVDS:
+			nvd0_sor_create(connector, dcbe);
+			break;
+		case OUTPUT_ANALOG:
+			nvd0_dac_create(connector, dcbe);
+			break;
+		default:
+			NV_WARN(dev, "skipping unsupported encoder %d/%d\n",
+				dcbe->type, ffs(dcbe->or) - 1);
+			continue;
+		}
+	}
+
+	/* cull any connectors we created that don't have an encoder */
+	list_for_each_entry_safe(connector, tmp, &dev->mode_config.connector_list, head) {
+		if (connector->encoder_ids[0])
+			continue;
+
+		NV_WARN(dev, "%s has no encoders, removing\n",
+			drm_get_connector_name(connector));
+		connector->funcs->destroy(connector);
+	}
+
+	/* setup interrupt handling */
+	tasklet_init(&disp->tasklet, nvd0_display_bh, (unsigned long)dev);
+	nouveau_irq_register(dev, 26, nvd0_display_intr);
+
+	/* hash table and dma objects for the memory areas we care about */
+	ret = nouveau_gpuobj_new(dev, NULL, 0x4000, 0x10000,
+				 NVOBJ_FLAG_ZERO_ALLOC, &disp->mem);
+	if (ret)
+		goto out;
+
+	nv_wo32(disp->mem, 0x1000, 0x00000049);
+	nv_wo32(disp->mem, 0x1004, (disp->mem->vinst + 0x2000) >> 8);
+	nv_wo32(disp->mem, 0x1008, (disp->mem->vinst + 0x2fff) >> 8);
+	nv_wo32(disp->mem, 0x100c, 0x00000000);
+	nv_wo32(disp->mem, 0x1010, 0x00000000);
+	nv_wo32(disp->mem, 0x1014, 0x00000000);
+	nv_wo32(disp->mem, 0x0000, NvEvoSync);
+	nv_wo32(disp->mem, 0x0004, (0x1000 << 9) | 0x00000001);
+
+	nv_wo32(disp->mem, 0x1020, 0x00000049);
+	nv_wo32(disp->mem, 0x1024, 0x00000000);
+	nv_wo32(disp->mem, 0x1028, (dev_priv->vram_size - 1) >> 8);
+	nv_wo32(disp->mem, 0x102c, 0x00000000);
+	nv_wo32(disp->mem, 0x1030, 0x00000000);
+	nv_wo32(disp->mem, 0x1034, 0x00000000);
+	nv_wo32(disp->mem, 0x0008, NvEvoVRAM);
+	nv_wo32(disp->mem, 0x000c, (0x1020 << 9) | 0x00000001);
+
+	nv_wo32(disp->mem, 0x1040, 0x00000009);
+	nv_wo32(disp->mem, 0x1044, 0x00000000);
+	nv_wo32(disp->mem, 0x1048, (dev_priv->vram_size - 1) >> 8);
+	nv_wo32(disp->mem, 0x104c, 0x00000000);
+	nv_wo32(disp->mem, 0x1050, 0x00000000);
+	nv_wo32(disp->mem, 0x1054, 0x00000000);
+	nv_wo32(disp->mem, 0x0010, NvEvoVRAM_LP);
+	nv_wo32(disp->mem, 0x0014, (0x1040 << 9) | 0x00000001);
+
+	nv_wo32(disp->mem, 0x1060, 0x0fe00009);
+	nv_wo32(disp->mem, 0x1064, 0x00000000);
+	nv_wo32(disp->mem, 0x1068, (dev_priv->vram_size - 1) >> 8);
+	nv_wo32(disp->mem, 0x106c, 0x00000000);
+	nv_wo32(disp->mem, 0x1070, 0x00000000);
+	nv_wo32(disp->mem, 0x1074, 0x00000000);
+	nv_wo32(disp->mem, 0x0018, NvEvoFB32);
+	nv_wo32(disp->mem, 0x001c, (0x1060 << 9) | 0x00000001);
+
+	pinstmem->flush(dev);
+
+	/* push buffers for evo channels */
+	disp->evo[0].ptr =
+		pci_alloc_consistent(pdev, PAGE_SIZE, &disp->evo[0].handle);
+	if (!disp->evo[0].ptr) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	ret = nvd0_display_init(dev);
+	if (ret)
+		goto out;
+
+out:
+	if (ret)
+		nvd0_display_destroy(dev);
+	return ret;
+}
diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c
index 0b517e16fd3..a72dbb3e133 100644
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@@ -41,6 +41,31 @@ static void evergreen_gpu_init(struct radeon_device *rdev);
 void evergreen_fini(struct radeon_device *rdev);
 static void evergreen_pcie_gen2_enable(struct radeon_device *rdev);
 
+void evergreen_fix_pci_max_read_req_size(struct radeon_device *rdev)
+{
+	u16 ctl, v;
+	int cap, err;
+
+	cap = pci_pcie_cap(rdev->pdev);
+	if (!cap)
+		return;
+
+	err = pci_read_config_word(rdev->pdev, cap + PCI_EXP_DEVCTL, &ctl);
+	if (err)
+		return;
+
+	v = (ctl & PCI_EXP_DEVCTL_READRQ) >> 12;
+
+	/* if bios or OS sets MAX_READ_REQUEST_SIZE to an invalid value, fix it
+	 * to avoid hangs or perfomance issues
+	 */
+	if ((v == 0) || (v == 6) || (v == 7)) {
+		ctl &= ~PCI_EXP_DEVCTL_READRQ;
+		ctl |= (2 << 12);
+		pci_write_config_word(rdev->pdev, cap + PCI_EXP_DEVCTL, ctl);
+	}
+}
+
 void evergreen_pre_page_flip(struct radeon_device *rdev, int crtc)
 {
 	/* enable the pflip int */
@@ -1866,6 +1891,8 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
 
 	WREG32(GRBM_CNTL, GRBM_READ_TIMEOUT(0xff));
 
+	evergreen_fix_pci_max_read_req_size(rdev);
+
 	cc_gc_shader_pipe_config = RREG32(CC_GC_SHADER_PIPE_CONFIG) & ~2;
 
 	cc_gc_shader_pipe_config |=
diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c
index 3b8b849643d..bf4fce7c43f 100644
--- a/drivers/gpu/drm/radeon/ni.c
+++ b/drivers/gpu/drm/radeon/ni.c
@@ -39,6 +39,7 @@ extern int evergreen_mc_wait_for_idle(struct radeon_device *rdev);
 extern void evergreen_mc_program(struct radeon_device *rdev);
 extern void evergreen_irq_suspend(struct radeon_device *rdev);
 extern int evergreen_mc_init(struct radeon_device *rdev);
+extern void evergreen_fix_pci_max_read_req_size(struct radeon_device *rdev);
 
 #define EVERGREEN_PFP_UCODE_SIZE 1120
 #define EVERGREEN_PM4_UCODE_SIZE 1376
@@ -669,6 +670,8 @@ static void cayman_gpu_init(struct radeon_device *rdev)
 
 	WREG32(GRBM_CNTL, GRBM_READ_TIMEOUT(0xff));
 
+	evergreen_fix_pci_max_read_req_size(rdev);
+
 	mc_shared_chmap = RREG32(MC_SHARED_CHMAP);
 	mc_arb_ramcfg = RREG32(MC_ARB_RAMCFG);
 
diff --git a/drivers/gpu/drm/radeon/radeon_clocks.c b/drivers/gpu/drm/radeon/radeon_clocks.c
index dcd0863e31a..b6e18c8db9f 100644
--- a/drivers/gpu/drm/radeon/radeon_clocks.c
+++ b/drivers/gpu/drm/radeon/radeon_clocks.c
@@ -219,6 +219,9 @@ void radeon_get_clock_info(struct drm_device *dev)
 		} else {
 			DRM_INFO("Using generic clock info\n");
 
+			/* may need to be per card */
+			rdev->clock.max_pixel_clock = 35000;
+
 			if (rdev->flags & RADEON_IS_IGP) {
 				p1pll->reference_freq = 1432;
 				p2pll->reference_freq = 1432;
diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c
index 4f0c1ecac72..c4b8741dbf5 100644
--- a/drivers/gpu/drm/radeon/radeon_connectors.c
+++ b/drivers/gpu/drm/radeon/radeon_connectors.c
@@ -1297,12 +1297,33 @@ radeon_dp_detect(struct drm_connector *connector, bool force)
 		if (!radeon_dig_connector->edp_on)
 			atombios_set_edp_panel_power(connector,
 						     ATOM_TRANSMITTER_ACTION_POWER_OFF);
-	} else {
-		/* need to setup ddc on the bridge */
-		if (radeon_connector_encoder_is_dp_bridge(connector)) {
+	} else if (radeon_connector_encoder_is_dp_bridge(connector)) {
+		/* DP bridges are always DP */
+		radeon_dig_connector->dp_sink_type = CONNECTOR_OBJECT_ID_DISPLAYPORT;
+		/* get the DPCD from the bridge */
+		radeon_dp_getdpcd(radeon_connector);
+
+		if (radeon_hpd_sense(rdev, radeon_connector->hpd.hpd))
+			ret = connector_status_connected;
+		else {
+			/* need to setup ddc on the bridge */
 			if (encoder)
 				radeon_atom_ext_encoder_setup_ddc(encoder);
+			if (radeon_ddc_probe(radeon_connector,
+					     radeon_connector->requires_extended_probe))
+				ret = connector_status_connected;
+		}
+
+		if ((ret == connector_status_disconnected) &&
+		    radeon_connector->dac_load_detect) {
+			struct drm_encoder *encoder = radeon_best_single_encoder(connector);
+			struct drm_encoder_helper_funcs *encoder_funcs;
+			if (encoder) {
+				encoder_funcs = encoder->helper_private;
+				ret = encoder_funcs->detect(encoder, connector);
+			}
 		}
+	} else {
 		radeon_dig_connector->dp_sink_type = radeon_dp_getsinktype(radeon_connector);
 		if (radeon_hpd_sense(rdev, radeon_connector->hpd.hpd)) {
 			ret = connector_status_connected;
@@ -1318,16 +1339,6 @@ radeon_dp_detect(struct drm_connector *connector, bool force)
 					ret = connector_status_connected;
 			}
 		}
-
-		if ((ret == connector_status_disconnected) &&
-		    radeon_connector->dac_load_detect) {
-			struct drm_encoder *encoder = radeon_best_single_encoder(connector);
-			struct drm_encoder_helper_funcs *encoder_funcs;
-			if (encoder) {
-				encoder_funcs = encoder->helper_private;
-				ret = encoder_funcs->detect(encoder, connector);
-			}
-		}
 	}
 
 	radeon_connector_update_scratch_regs(connector, ret);
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c
index 1a858944e4f..6cc17fb96a5 100644
--- a/drivers/gpu/drm/radeon/radeon_display.c
+++ b/drivers/gpu/drm/radeon/radeon_display.c
@@ -707,16 +707,21 @@ int radeon_ddc_get_modes(struct radeon_connector *radeon_connector)
 		radeon_router_select_ddc_port(radeon_connector);
 
 	if ((radeon_connector->base.connector_type == DRM_MODE_CONNECTOR_DisplayPort) ||
-	    (radeon_connector->base.connector_type == DRM_MODE_CONNECTOR_eDP)) {
+	    (radeon_connector->base.connector_type == DRM_MODE_CONNECTOR_eDP) ||
+	    radeon_connector_encoder_is_dp_bridge(&radeon_connector->base)) {
 		struct radeon_connector_atom_dig *dig = radeon_connector->con_priv;
+
 		if ((dig->dp_sink_type == CONNECTOR_OBJECT_ID_DISPLAYPORT ||
 		     dig->dp_sink_type == CONNECTOR_OBJECT_ID_eDP) && dig->dp_i2c_bus)
-			radeon_connector->edid = drm_get_edid(&radeon_connector->base, &dig->dp_i2c_bus->adapter);
-	}
-	if (!radeon_connector->ddc_bus)
-		return -1;
-	if (!radeon_connector->edid) {
-		radeon_connector->edid = drm_get_edid(&radeon_connector->base, &radeon_connector->ddc_bus->adapter);
+			radeon_connector->edid = drm_get_edid(&radeon_connector->base,
+							      &dig->dp_i2c_bus->adapter);
+		else if (radeon_connector->ddc_bus && !radeon_connector->edid)
+			radeon_connector->edid = drm_get_edid(&radeon_connector->base,
+							      &radeon_connector->ddc_bus->adapter);
+	} else {
+		if (radeon_connector->ddc_bus && !radeon_connector->edid)
+			radeon_connector->edid = drm_get_edid(&radeon_connector->base,
+							      &radeon_connector->ddc_bus->adapter);
 	}
 
 	if (!radeon_connector->edid) {
diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h
index 7d27d2b0445..7484e1b6724 100644
--- a/drivers/hid/hid-ids.h
+++ b/drivers/hid/hid-ids.h
@@ -277,6 +277,7 @@
 #define USB_DEVICE_ID_PENPOWER		0x00f4
 
 #define USB_VENDOR_ID_GREENASIA		0x0e8f
+#define USB_DEVICE_ID_GREENASIA_DUAL_USB_JOYPAD	0x3013
 
 #define USB_VENDOR_ID_GRETAGMACBETH	0x0971
 #define USB_DEVICE_ID_GRETAGMACBETH_HUEY	0x2005
diff --git a/drivers/hid/hid-magicmouse.c b/drivers/hid/hid-magicmouse.c
index 0ec91c18a42..f0fbd7bd239 100644
--- a/drivers/hid/hid-magicmouse.c
+++ b/drivers/hid/hid-magicmouse.c
@@ -81,6 +81,28 @@ MODULE_PARM_DESC(report_undeciphered, "Report undeciphered multi-touch state fie
 #define NO_TOUCHES -1
 #define SINGLE_TOUCH_UP -2
 
+/* Touch surface information. Dimension is in hundredths of a mm, min and max
+ * are in units. */
+#define MOUSE_DIMENSION_X (float)9056
+#define MOUSE_MIN_X -1100
+#define MOUSE_MAX_X 1258
+#define MOUSE_RES_X ((MOUSE_MAX_X - MOUSE_MIN_X) / (MOUSE_DIMENSION_X / 100))
+#define MOUSE_DIMENSION_Y (float)5152
+#define MOUSE_MIN_Y -1589
+#define MOUSE_MAX_Y 2047
+#define MOUSE_RES_Y ((MOUSE_MAX_Y - MOUSE_MIN_Y) / (MOUSE_DIMENSION_Y / 100))
+
+#define TRACKPAD_DIMENSION_X (float)13000
+#define TRACKPAD_MIN_X -2909
+#define TRACKPAD_MAX_X 3167
+#define TRACKPAD_RES_X \
+	((TRACKPAD_MAX_X - TRACKPAD_MIN_X) / (TRACKPAD_DIMENSION_X / 100))
+#define TRACKPAD_DIMENSION_Y (float)11000
+#define TRACKPAD_MIN_Y -2456
+#define TRACKPAD_MAX_Y 2565
+#define TRACKPAD_RES_Y \
+	((TRACKPAD_MAX_Y - TRACKPAD_MIN_Y) / (TRACKPAD_DIMENSION_Y / 100))
+
 /**
  * struct magicmouse_sc - Tracks Magic Mouse-specific data.
  * @input: Input device through which we report events.
@@ -406,17 +428,31 @@ static void magicmouse_setup_input(struct input_dev *input, struct hid_device *h
 		 * inverse of the reported Y.
 		 */
 		if (input->id.product == USB_DEVICE_ID_APPLE_MAGICMOUSE) {
-			input_set_abs_params(input, ABS_MT_POSITION_X, -1100,
-				1358, 4, 0);
-			input_set_abs_params(input, ABS_MT_POSITION_Y, -1589,
-				2047, 4, 0);
+			input_set_abs_params(input, ABS_MT_POSITION_X,
+				MOUSE_MIN_X, MOUSE_MAX_X, 4, 0);
+			input_set_abs_params(input, ABS_MT_POSITION_Y,
+				MOUSE_MIN_Y, MOUSE_MAX_Y, 4, 0);
+
+			input_abs_set_res(input, ABS_MT_POSITION_X,
+				MOUSE_RES_X);
+			input_abs_set_res(input, ABS_MT_POSITION_Y,
+				MOUSE_RES_Y);
 		} else { /* USB_DEVICE_ID_APPLE_MAGICTRACKPAD */
-			input_set_abs_params(input, ABS_X, -2909, 3167, 4, 0);
-			input_set_abs_params(input, ABS_Y, -2456, 2565, 4, 0);
-			input_set_abs_params(input, ABS_MT_POSITION_X, -2909,
-				3167, 4, 0);
-			input_set_abs_params(input, ABS_MT_POSITION_Y, -2456,
-				2565, 4, 0);
+			input_set_abs_params(input, ABS_X, TRACKPAD_MIN_X,
+				TRACKPAD_MAX_X, 4, 0);
+			input_set_abs_params(input, ABS_Y, TRACKPAD_MIN_Y,
+				TRACKPAD_MAX_Y, 4, 0);
+			input_set_abs_params(input, ABS_MT_POSITION_X,
+				TRACKPAD_MIN_X, TRACKPAD_MAX_X, 4, 0);
+			input_set_abs_params(input, ABS_MT_POSITION_Y,
+				TRACKPAD_MIN_Y, TRACKPAD_MAX_Y, 4, 0);
+
+			input_abs_set_res(input, ABS_X, TRACKPAD_RES_X);
+			input_abs_set_res(input, ABS_Y, TRACKPAD_RES_Y);
+			input_abs_set_res(input, ABS_MT_POSITION_X,
+				TRACKPAD_RES_X);
+			input_abs_set_res(input, ABS_MT_POSITION_Y,
+				TRACKPAD_RES_Y);
 		}
 
 		input_set_events_per_packet(input, 60);
@@ -501,9 +537,17 @@ static int magicmouse_probe(struct hid_device *hdev,
 	}
 	report->size = 6;
 
+	/*
+	 * Some devices repond with 'invalid report id' when feature
+	 * report switching it into multitouch mode is sent to it.
+	 *
+	 * This results in -EIO from the _raw low-level transport callback,
+	 * but there seems to be no other way of switching the mode.
+	 * Thus the super-ugly hacky success check below.
+	 */
 	ret = hdev->hid_output_raw_report(hdev, feature, sizeof(feature),
 			HID_FEATURE_REPORT);
-	if (ret != sizeof(feature)) {
+	if (ret != -EIO && ret != sizeof(feature)) {
 		hid_err(hdev, "unable to request touch data (%d)\n", ret);
 		goto err_stop_hw;
 	}
diff --git a/drivers/hid/hid-wacom.c b/drivers/hid/hid-wacom.c
index 06888323828..a597039d075 100644
--- a/drivers/hid/hid-wacom.c
+++ b/drivers/hid/hid-wacom.c
@@ -353,11 +353,7 @@ static int wacom_probe(struct hid_device *hdev,
 	if (ret) {
 		hid_warn(hdev, "can't create sysfs battery attribute, err: %d\n",
 			 ret);
-		/*
-		 * battery attribute is not critical for the tablet, but if it
-		 * failed then there is no need to create ac attribute
-		 */
-		goto move_on;
+		goto err_battery;
 	}
 
 	wdata->ac.properties = wacom_ac_props;
@@ -371,14 +367,8 @@ static int wacom_probe(struct hid_device *hdev,
 	if (ret) {
 		hid_warn(hdev,
 			 "can't create ac battery attribute, err: %d\n", ret);
-		/*
-		 * ac attribute is not critical for the tablet, but if it
-		 * failed then we don't want to battery attribute to exist
-		 */
-		power_supply_unregister(&wdata->battery);
+		goto err_ac;
 	}
-
-move_on:
 #endif
 	hidinput = list_entry(hdev->inputs.next, struct hid_input, list);
 	input = hidinput->input;
@@ -416,6 +406,13 @@ move_on:
 
 	return 0;
 
+#ifdef CONFIG_HID_WACOM_POWER_SUPPLY
+err_ac:
+	power_supply_unregister(&wdata->battery);
+err_battery:
+	device_remove_file(&hdev->dev, &dev_attr_speed);
+	hid_hw_stop(hdev);
+#endif
 err_free:
 	kfree(wdata);
 	return ret;
@@ -426,6 +423,7 @@ static void wacom_remove(struct hid_device *hdev)
 #ifdef CONFIG_HID_WACOM_POWER_SUPPLY
 	struct wacom_data *wdata = hid_get_drvdata(hdev);
 #endif
+	device_remove_file(&hdev->dev, &dev_attr_speed);
 	hid_hw_stop(hdev);
 
 #ifdef CONFIG_HID_WACOM_POWER_SUPPLY
diff --git a/drivers/hid/usbhid/hid-quirks.c b/drivers/hid/usbhid/hid-quirks.c
index 4bdb5d46c52..3146fdcda27 100644
--- a/drivers/hid/usbhid/hid-quirks.c
+++ b/drivers/hid/usbhid/hid-quirks.c
@@ -47,6 +47,7 @@ static const struct hid_blacklist {
 	{ USB_VENDOR_ID_AFATECH, USB_DEVICE_ID_AFATECH_AF9016, HID_QUIRK_FULLSPEED_INTERVAL },
 
 	{ USB_VENDOR_ID_ETURBOTOUCH, USB_DEVICE_ID_ETURBOTOUCH, HID_QUIRK_MULTI_INPUT },
+	{ USB_VENDOR_ID_GREENASIA, USB_DEVICE_ID_GREENASIA_DUAL_USB_JOYPAD, HID_QUIRK_MULTI_INPUT },
 	{ USB_VENDOR_ID_PANTHERLORD, USB_DEVICE_ID_PANTHERLORD_TWIN_USB_JOYSTICK, HID_QUIRK_MULTI_INPUT | HID_QUIRK_SKIP_OUTPUT_REPORTS },
 	{ USB_VENDOR_ID_PLAYDOTCOM, USB_DEVICE_ID_PLAYDOTCOM_EMS_USBII, HID_QUIRK_MULTI_INPUT },
 	{ USB_VENDOR_ID_TOUCHPACK, USB_DEVICE_ID_TOUCHPACK_RTS, HID_QUIRK_MULTI_INPUT },
diff --git a/drivers/hwmon/max16065.c b/drivers/hwmon/max16065.c
index d94a24fdf4b..dd2d7b9620c 100644
--- a/drivers/hwmon/max16065.c
+++ b/drivers/hwmon/max16065.c
@@ -124,7 +124,7 @@ static inline int MV_TO_LIMIT(int mv, int range)
 
 static inline int ADC_TO_CURR(int adc, int gain)
 {
-	return adc * 1400000 / gain * 255;
+	return adc * 1400000 / (gain * 255);
 }
 
 /*
diff --git a/drivers/hwmon/pmbus/ucd9000.c b/drivers/hwmon/pmbus/ucd9000.c
index ace1c731973..d0ddb60155c 100644
--- a/drivers/hwmon/pmbus/ucd9000.c
+++ b/drivers/hwmon/pmbus/ucd9000.c
@@ -141,13 +141,11 @@ static int ucd9000_probe(struct i2c_client *client,
 	block_buffer[ret] = '\0';
 	dev_info(&client->dev, "Device ID %s\n", block_buffer);
 
-	mid = NULL;
-	for (i = 0; i < ARRAY_SIZE(ucd9000_id); i++) {
-		mid = &ucd9000_id[i];
+	for (mid = ucd9000_id; mid->name[0]; mid++) {
 		if (!strncasecmp(mid->name, block_buffer, strlen(mid->name)))
 			break;
 	}
-	if (!mid || !strlen(mid->name)) {
+	if (!mid->name[0]) {
 		dev_err(&client->dev, "Unsupported device\n");
 		return -ENODEV;
 	}
diff --git a/drivers/hwmon/pmbus/ucd9200.c b/drivers/hwmon/pmbus/ucd9200.c
index ffcc1cf3609..c65e9da707c 100644
--- a/drivers/hwmon/pmbus/ucd9200.c
+++ b/drivers/hwmon/pmbus/ucd9200.c
@@ -68,13 +68,11 @@ static int ucd9200_probe(struct i2c_client *client,
 	block_buffer[ret] = '\0';
 	dev_info(&client->dev, "Device ID %s\n", block_buffer);
 
-	mid = NULL;
-	for (i = 0; i < ARRAY_SIZE(ucd9200_id); i++) {
-		mid = &ucd9200_id[i];
+	for (mid = ucd9200_id; mid->name[0]; mid++) {
 		if (!strncasecmp(mid->name, block_buffer, strlen(mid->name)))
 			break;
 	}
-	if (!mid || !strlen(mid->name)) {
+	if (!mid->name[0]) {
 		dev_err(&client->dev, "Unsupported device\n");
 		return -ENODEV;
 	}
diff --git a/drivers/i2c/busses/i2c-pxa-pci.c b/drivers/i2c/busses/i2c-pxa-pci.c
index 6659d269b84..b73da6cd6f9 100644
--- a/drivers/i2c/busses/i2c-pxa-pci.c
+++ b/drivers/i2c/busses/i2c-pxa-pci.c
@@ -109,12 +109,15 @@ static int __devinit ce4100_i2c_probe(struct pci_dev *dev,
 		return -EINVAL;
 	}
 	sds = kzalloc(sizeof(*sds), GFP_KERNEL);
-	if (!sds)
+	if (!sds) {
+		ret = -ENOMEM;
 		goto err_mem;
+	}
 
 	for (i = 0; i < ARRAY_SIZE(sds->pdev); i++) {
 		sds->pdev[i] = add_i2c_device(dev, i);
 		if (IS_ERR(sds->pdev[i])) {
+			ret = PTR_ERR(sds->pdev[i]);
 			while (--i >= 0)
 				platform_device_unregister(sds->pdev[i]);
 			goto err_dev_add;
diff --git a/drivers/i2c/busses/i2c-tegra.c b/drivers/i2c/busses/i2c-tegra.c
index 2440b741197..3c94c4a81a5 100644
--- a/drivers/i2c/busses/i2c-tegra.c
+++ b/drivers/i2c/busses/i2c-tegra.c
@@ -270,14 +270,30 @@ static int tegra_i2c_fill_tx_fifo(struct tegra_i2c_dev *i2c_dev)
 
 	/* Rounds down to not include partial word at the end of buf */
 	words_to_transfer = buf_remaining / BYTES_PER_FIFO_WORD;
-	if (words_to_transfer > tx_fifo_avail)
-		words_to_transfer = tx_fifo_avail;
 
-	i2c_writesl(i2c_dev, buf, I2C_TX_FIFO, words_to_transfer);
-
-	buf += words_to_transfer * BYTES_PER_FIFO_WORD;
-	buf_remaining -= words_to_transfer * BYTES_PER_FIFO_WORD;
-	tx_fifo_avail -= words_to_transfer;
+	/* It's very common to have < 4 bytes, so optimize that case. */
+	if (words_to_transfer) {
+		if (words_to_transfer > tx_fifo_avail)
+			words_to_transfer = tx_fifo_avail;
+
+		/*
+		 * Update state before writing to FIFO.  If this casues us
+		 * to finish writing all bytes (AKA buf_remaining goes to 0) we
+		 * have a potential for an interrupt (PACKET_XFER_COMPLETE is
+		 * not maskable).  We need to make sure that the isr sees
+		 * buf_remaining as 0 and doesn't call us back re-entrantly.
+		 */
+		buf_remaining -= words_to_transfer * BYTES_PER_FIFO_WORD;
+		tx_fifo_avail -= words_to_transfer;
+		i2c_dev->msg_buf_remaining = buf_remaining;
+		i2c_dev->msg_buf = buf +
+			words_to_transfer * BYTES_PER_FIFO_WORD;
+		barrier();
+
+		i2c_writesl(i2c_dev, buf, I2C_TX_FIFO, words_to_transfer);
+
+		buf += words_to_transfer * BYTES_PER_FIFO_WORD;
+	}
 
 	/*
 	 * If there is a partial word at the end of buf, handle it manually to
@@ -287,14 +303,15 @@ static int tegra_i2c_fill_tx_fifo(struct tegra_i2c_dev *i2c_dev)
 	if (tx_fifo_avail > 0 && buf_remaining > 0) {
 		BUG_ON(buf_remaining > 3);
 		memcpy(&val, buf, buf_remaining);
+
+		/* Again update before writing to FIFO to make sure isr sees. */
+		i2c_dev->msg_buf_remaining = 0;
+		i2c_dev->msg_buf = NULL;
+		barrier();
+
 		i2c_writel(i2c_dev, val, I2C_TX_FIFO);
-		buf_remaining = 0;
-		tx_fifo_avail--;
 	}
 
-	BUG_ON(tx_fifo_avail > 0 && buf_remaining > 0);
-	i2c_dev->msg_buf_remaining = buf_remaining;
-	i2c_dev->msg_buf = buf;
 	return 0;
 }
 
@@ -411,9 +428,10 @@ static irqreturn_t tegra_i2c_isr(int irq, void *dev_id)
 			tegra_i2c_mask_irq(i2c_dev, I2C_INT_TX_FIFO_DATA_REQ);
 	}
 
-	if ((status & I2C_INT_PACKET_XFER_COMPLETE) &&
-			!i2c_dev->msg_buf_remaining)
+	if (status & I2C_INT_PACKET_XFER_COMPLETE) {
+		BUG_ON(i2c_dev->msg_buf_remaining);
 		complete(&i2c_dev->msg_complete);
+	}
 
 	i2c_writel(i2c_dev, status, I2C_INT_STATUS);
 	if (i2c_dev->is_dvc)
@@ -531,7 +549,7 @@ static int tegra_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[],
 
 static u32 tegra_i2c_func(struct i2c_adapter *adap)
 {
-	return I2C_FUNC_I2C;
+	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL;
 }
 
 static const struct i2c_algorithm tegra_i2c_algo = {
@@ -719,6 +737,17 @@ static int tegra_i2c_resume(struct platform_device *pdev)
 }
 #endif
 
+#if defined(CONFIG_OF)
+/* Match table for of_platform binding */
+static const struct of_device_id tegra_i2c_of_match[] __devinitconst = {
+	{ .compatible = "nvidia,tegra20-i2c", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, tegra_i2c_of_match);
+#else
+#define tegra_i2c_of_match NULL
+#endif
+
 static struct platform_driver tegra_i2c_driver = {
 	.probe   = tegra_i2c_probe,
 	.remove  = tegra_i2c_remove,
@@ -729,6 +758,7 @@ static struct platform_driver tegra_i2c_driver = {
 	.driver  = {
 		.name  = "tegra-i2c",
 		.owner = THIS_MODULE,
+		.of_match_table = tegra_i2c_of_match,
 	},
 };
 
diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c
index a14f8dc2346..0e4227f457a 100644
--- a/drivers/iommu/amd_iommu.c
+++ b/drivers/iommu/amd_iommu.c
@@ -605,7 +605,9 @@ static void build_inv_all(struct iommu_cmd *cmd)
  * Writes the command to the IOMMUs command buffer and informs the
  * hardware about the new command.
  */
-static int iommu_queue_command(struct amd_iommu *iommu, struct iommu_cmd *cmd)
+static int iommu_queue_command_sync(struct amd_iommu *iommu,
+				    struct iommu_cmd *cmd,
+				    bool sync)
 {
 	u32 left, tail, head, next_tail;
 	unsigned long flags;
@@ -639,13 +641,18 @@ again:
 	copy_cmd_to_buffer(iommu, cmd, tail);
 
 	/* We need to sync now to make sure all commands are processed */
-	iommu->need_sync = true;
+	iommu->need_sync = sync;
 
 	spin_unlock_irqrestore(&iommu->lock, flags);
 
 	return 0;
 }
 
+static int iommu_queue_command(struct amd_iommu *iommu, struct iommu_cmd *cmd)
+{
+	return iommu_queue_command_sync(iommu, cmd, true);
+}
+
 /*
  * This function queues a completion wait command into the command
  * buffer of an IOMMU
@@ -661,7 +668,7 @@ static int iommu_completion_wait(struct amd_iommu *iommu)
 
 	build_completion_wait(&cmd, (u64)&sem);
 
-	ret = iommu_queue_command(iommu, &cmd);
+	ret = iommu_queue_command_sync(iommu, &cmd, false);
 	if (ret)
 		return ret;
 
@@ -840,14 +847,9 @@ static void domain_flush_complete(struct protection_domain *domain)
 static void domain_flush_devices(struct protection_domain *domain)
 {
 	struct iommu_dev_data *dev_data;
-	unsigned long flags;
-
-	spin_lock_irqsave(&domain->lock, flags);
 
 	list_for_each_entry(dev_data, &domain->dev_list, list)
 		device_flush_dte(dev_data);
-
-	spin_unlock_irqrestore(&domain->lock, flags);
 }
 
 /****************************************************************************
diff --git a/drivers/md/linear.h b/drivers/md/linear.h
index 0ce29b61605..2f2da05b2ce 100644
--- a/drivers/md/linear.h
+++ b/drivers/md/linear.h
@@ -10,9 +10,9 @@ typedef struct dev_info dev_info_t;
 
 struct linear_private_data
 {
+	struct rcu_head		rcu;
 	sector_t		array_sectors;
 	dev_info_t		disks[0];
-	struct rcu_head		rcu;
 };
 
 
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 8e221a20f5d..5404b229582 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -848,7 +848,7 @@ void md_super_write(mddev_t *mddev, mdk_rdev_t *rdev,
 	bio->bi_end_io = super_written;
 
 	atomic_inc(&mddev->pending_writes);
-	submit_bio(REQ_WRITE | REQ_SYNC | REQ_FLUSH | REQ_FUA, bio);
+	submit_bio(WRITE_FLUSH_FUA, bio);
 }
 
 void md_super_wait(mddev_t *mddev)
@@ -1138,8 +1138,11 @@ static int super_90_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version
 			ret = 0;
 	}
 	rdev->sectors = rdev->sb_start;
+	/* Limit to 4TB as metadata cannot record more than that */
+	if (rdev->sectors >= (2ULL << 32))
+		rdev->sectors = (2ULL << 32) - 2;
 
-	if (rdev->sectors < sb->size * 2 && sb->level > 1)
+	if (rdev->sectors < ((sector_t)sb->size) * 2 && sb->level >= 1)
 		/* "this cannot possibly happen" ... */
 		ret = -EINVAL;
 
@@ -1173,7 +1176,7 @@ static int super_90_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 		mddev->clevel[0] = 0;
 		mddev->layout = sb->layout;
 		mddev->raid_disks = sb->raid_disks;
-		mddev->dev_sectors = sb->size * 2;
+		mddev->dev_sectors = ((sector_t)sb->size) * 2;
 		mddev->events = ev1;
 		mddev->bitmap_info.offset = 0;
 		mddev->bitmap_info.default_offset = MD_SB_BYTES >> 9;
@@ -1415,6 +1418,11 @@ super_90_rdev_size_change(mdk_rdev_t *rdev, sector_t num_sectors)
 	rdev->sb_start = calc_dev_sboffset(rdev);
 	if (!num_sectors || num_sectors > rdev->sb_start)
 		num_sectors = rdev->sb_start;
+	/* Limit to 4TB as metadata cannot record more than that.
+	 * 4TB == 2^32 KB, or 2*2^32 sectors.
+	 */
+	if (num_sectors >= (2ULL << 32))
+		num_sectors = (2ULL << 32) - 2;
 	md_super_write(rdev->mddev, rdev, rdev->sb_start, rdev->sb_size,
 		       rdev->sb_page);
 	md_super_wait(rdev->mddev);
@@ -1738,6 +1746,11 @@ static void super_1_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 	sb->level = cpu_to_le32(mddev->level);
 	sb->layout = cpu_to_le32(mddev->layout);
 
+	if (test_bit(WriteMostly, &rdev->flags))
+		sb->devflags |= WriteMostly1;
+	else
+		sb->devflags &= ~WriteMostly1;
+
 	if (mddev->bitmap && mddev->bitmap_info.file == NULL) {
 		sb->bitmap_offset = cpu_to_le32((__u32)mddev->bitmap_info.offset);
 		sb->feature_map = cpu_to_le32(MD_FEATURE_BITMAP_OFFSET);
@@ -2561,7 +2574,10 @@ state_store(mdk_rdev_t *rdev, const char *buf, size_t len)
 	int err = -EINVAL;
 	if (cmd_match(buf, "faulty") && rdev->mddev->pers) {
 		md_error(rdev->mddev, rdev);
-		err = 0;
+		if (test_bit(Faulty, &rdev->flags))
+			err = 0;
+		else
+			err = -EBUSY;
 	} else if (cmd_match(buf, "remove")) {
 		if (rdev->raid_disk >= 0)
 			err = -EBUSY;
@@ -2584,7 +2600,7 @@ state_store(mdk_rdev_t *rdev, const char *buf, size_t len)
 		err = 0;
 	} else if (cmd_match(buf, "-blocked")) {
 		if (!test_bit(Faulty, &rdev->flags) &&
-		    test_bit(BlockedBadBlocks, &rdev->flags)) {
+		    rdev->badblocks.unacked_exist) {
 			/* metadata handler doesn't understand badblocks,
 			 * so we need to fail the device
 			 */
@@ -5983,6 +5999,8 @@ static int set_disk_faulty(mddev_t *mddev, dev_t dev)
 		return -ENODEV;
 
 	md_error(mddev, rdev);
+	if (!test_bit(Faulty, &rdev->flags))
+		return -EBUSY;
 	return 0;
 }
 
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 32323f0afd8..f4622dd8fc5 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -1099,12 +1099,11 @@ read_again:
 		bio_list_add(&conf->pending_bio_list, mbio);
 		spin_unlock_irqrestore(&conf->device_lock, flags);
 	}
-	r1_bio_write_done(r1_bio);
-
-	/* In case raid1d snuck in to freeze_array */
-	wake_up(&conf->wait_barrier);
-
+	/* Mustn't call r1_bio_write_done before this next test,
+	 * as it could result in the bio being freed.
+	 */
 	if (sectors_handled < (bio->bi_size >> 9)) {
+		r1_bio_write_done(r1_bio);
 		/* We need another r1_bio.  It has already been counted
 		 * in bio->bi_phys_segments
 		 */
@@ -1117,6 +1116,11 @@ read_again:
 		goto retry_write;
 	}
 
+	r1_bio_write_done(r1_bio);
+
+	/* In case raid1d snuck in to freeze_array */
+	wake_up(&conf->wait_barrier);
+
 	if (do_sync || !bitmap || !plugged)
 		md_wakeup_thread(mddev->thread);
 
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index 8b29cd4f01c..d7a8468ddea 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -337,6 +337,21 @@ static void close_write(r10bio_t *r10_bio)
 	md_write_end(r10_bio->mddev);
 }
 
+static void one_write_done(r10bio_t *r10_bio)
+{
+	if (atomic_dec_and_test(&r10_bio->remaining)) {
+		if (test_bit(R10BIO_WriteError, &r10_bio->state))
+			reschedule_retry(r10_bio);
+		else {
+			close_write(r10_bio);
+			if (test_bit(R10BIO_MadeGood, &r10_bio->state))
+				reschedule_retry(r10_bio);
+			else
+				raid_end_bio_io(r10_bio);
+		}
+	}
+}
+
 static void raid10_end_write_request(struct bio *bio, int error)
 {
 	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
@@ -387,17 +402,7 @@ static void raid10_end_write_request(struct bio *bio, int error)
 	 * Let's see if all mirrored write operations have finished
 	 * already.
 	 */
-	if (atomic_dec_and_test(&r10_bio->remaining)) {
-		if (test_bit(R10BIO_WriteError, &r10_bio->state))
-			reschedule_retry(r10_bio);
-		else {
-			close_write(r10_bio);
-			if (test_bit(R10BIO_MadeGood, &r10_bio->state))
-				reschedule_retry(r10_bio);
-			else
-				raid_end_bio_io(r10_bio);
-		}
-	}
+	one_write_done(r10_bio);
 	if (dec_rdev)
 		rdev_dec_pending(conf->mirrors[dev].rdev, conf->mddev);
 }
@@ -1127,20 +1132,12 @@ retry_write:
 		spin_unlock_irqrestore(&conf->device_lock, flags);
 	}
 
-	if (atomic_dec_and_test(&r10_bio->remaining)) {
-		/* This matches the end of raid10_end_write_request() */
-		bitmap_endwrite(r10_bio->mddev->bitmap, r10_bio->sector,
-				r10_bio->sectors,
-				!test_bit(R10BIO_Degraded, &r10_bio->state),
-				0);
-		md_write_end(mddev);
-		raid_end_bio_io(r10_bio);
-	}
-
-	/* In case raid10d snuck in to freeze_array */
-	wake_up(&conf->wait_barrier);
+	/* Don't remove the bias on 'remaining' (one_write_done) until
+	 * after checking if we need to go around again.
+	 */
 
 	if (sectors_handled < (bio->bi_size >> 9)) {
+		one_write_done(r10_bio);
 		/* We need another r10_bio.  It has already been counted
 		 * in bio->bi_phys_segments.
 		 */
@@ -1154,6 +1151,10 @@ retry_write:
 		r10_bio->state = 0;
 		goto retry_write;
 	}
+	one_write_done(r10_bio);
+
+	/* In case raid10d snuck in to freeze_array */
+	wake_up(&conf->wait_barrier);
 
 	if (do_sync || !mddev->bitmap || !plugged)
 		md_wakeup_thread(mddev->thread);
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index dbae459fb02..43709fa6b6d 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -3336,7 +3336,7 @@ static void handle_stripe(struct stripe_head *sh)
 
 finish:
 	/* wait for this device to become unblocked */
-	if (unlikely(s.blocked_rdev))
+	if (conf->mddev->external && unlikely(s.blocked_rdev))
 		md_wait_for_blocked_rdev(s.blocked_rdev, conf->mddev);
 
 	if (s.handle_bad_blocks)
diff --git a/drivers/media/dvb/dvb-usb/vp7045.c b/drivers/media/dvb/dvb-usb/vp7045.c
index 3db89e3cb0b..536c16c943b 100644
--- a/drivers/media/dvb/dvb-usb/vp7045.c
+++ b/drivers/media/dvb/dvb-usb/vp7045.c
@@ -224,26 +224,8 @@ static struct dvb_usb_device_properties vp7045_properties;
 static int vp7045_usb_probe(struct usb_interface *intf,
 		const struct usb_device_id *id)
 {
-	struct dvb_usb_device *d;
-	int ret = dvb_usb_device_init(intf, &vp7045_properties,
-				   THIS_MODULE, &d, adapter_nr);
-	if (ret)
-		return ret;
-
-	d->priv = kmalloc(20, GFP_KERNEL);
-	if (!d->priv) {
-		dvb_usb_device_exit(intf);
-		return -ENOMEM;
-	}
-
-	return ret;
-}
-
-static void vp7045_usb_disconnect(struct usb_interface *intf)
-{
-	struct dvb_usb_device *d = usb_get_intfdata(intf);
-	kfree(d->priv);
-	dvb_usb_device_exit(intf);
+	return dvb_usb_device_init(intf, &vp7045_properties,
+				   THIS_MODULE, NULL, adapter_nr);
 }
 
 static struct usb_device_id vp7045_usb_table [] = {
@@ -258,7 +240,7 @@ MODULE_DEVICE_TABLE(usb, vp7045_usb_table);
 static struct dvb_usb_device_properties vp7045_properties = {
 	.usb_ctrl = CYPRESS_FX2,
 	.firmware = "dvb-usb-vp7045-01.fw",
-	.size_of_priv = sizeof(u8 *),
+	.size_of_priv = 20,
 
 	.num_adapters = 1,
 	.adapter = {
@@ -305,7 +287,7 @@ static struct dvb_usb_device_properties vp7045_properties = {
 static struct usb_driver vp7045_usb_driver = {
 	.name		= "dvb_usb_vp7045",
 	.probe		= vp7045_usb_probe,
-	.disconnect	= vp7045_usb_disconnect,
+	.disconnect	= dvb_usb_device_exit,
 	.id_table	= vp7045_usb_table,
 };
 
diff --git a/drivers/media/rc/nuvoton-cir.c b/drivers/media/rc/nuvoton-cir.c
index eae05b50047..144f3f55d76 100644
--- a/drivers/media/rc/nuvoton-cir.c
+++ b/drivers/media/rc/nuvoton-cir.c
@@ -618,7 +618,6 @@ static void nvt_dump_rx_buf(struct nvt_dev *nvt)
 static void nvt_process_rx_ir_data(struct nvt_dev *nvt)
 {
 	DEFINE_IR_RAW_EVENT(rawir);
-	unsigned int count;
 	u32 carrier;
 	u8 sample;
 	int i;
@@ -631,65 +630,38 @@ static void nvt_process_rx_ir_data(struct nvt_dev *nvt)
 	if (nvt->carrier_detect_enabled)
 		carrier = nvt_rx_carrier_detect(nvt);
 
-	count = nvt->pkts;
-	nvt_dbg_verbose("Processing buffer of len %d", count);
+	nvt_dbg_verbose("Processing buffer of len %d", nvt->pkts);
 
 	init_ir_raw_event(&rawir);
 
-	for (i = 0; i < count; i++) {
-		nvt->pkts--;
+	for (i = 0; i < nvt->pkts; i++) {
 		sample = nvt->buf[i];
 
 		rawir.pulse = ((sample & BUF_PULSE_BIT) != 0);
 		rawir.duration = US_TO_NS((sample & BUF_LEN_MASK)
 					  * SAMPLE_PERIOD);
 
-		if ((sample & BUF_LEN_MASK) == BUF_LEN_MASK) {
-			if (nvt->rawir.pulse == rawir.pulse)
-				nvt->rawir.duration += rawir.duration;
-			else {
-				nvt->rawir.duration = rawir.duration;
-				nvt->rawir.pulse = rawir.pulse;
-			}
-			continue;
-		}
-
-		rawir.duration += nvt->rawir.duration;
+		nvt_dbg("Storing %s with duration %d",
+			rawir.pulse ? "pulse" : "space", rawir.duration);
 
-		init_ir_raw_event(&nvt->rawir);
-		nvt->rawir.duration = 0;
-		nvt->rawir.pulse = rawir.pulse;
-
-		if (sample == BUF_PULSE_BIT)
-			rawir.pulse = false;
-
-		if (rawir.duration) {
-			nvt_dbg("Storing %s with duration %d",
-				rawir.pulse ? "pulse" : "space",
-				rawir.duration);
-
-			ir_raw_event_store_with_filter(nvt->rdev, &rawir);
-		}
+		ir_raw_event_store_with_filter(nvt->rdev, &rawir);
 
 		/*
 		 * BUF_PULSE_BIT indicates end of IR data, BUF_REPEAT_BYTE
 		 * indicates end of IR signal, but new data incoming. In both
 		 * cases, it means we're ready to call ir_raw_event_handle
 		 */
-		if ((sample == BUF_PULSE_BIT) && nvt->pkts) {
+		if ((sample == BUF_PULSE_BIT) && (i + 1 < nvt->pkts)) {
 			nvt_dbg("Calling ir_raw_event_handle (signal end)\n");
 			ir_raw_event_handle(nvt->rdev);
 		}
 	}
 
+	nvt->pkts = 0;
+
 	nvt_dbg("Calling ir_raw_event_handle (buffer empty)\n");
 	ir_raw_event_handle(nvt->rdev);
 
-	if (nvt->pkts) {
-		nvt_dbg("Odd, pkts should be 0 now... (its %u)", nvt->pkts);
-		nvt->pkts = 0;
-	}
-
 	nvt_dbg_verbose("%s done", __func__);
 }
 
@@ -1048,7 +1020,6 @@ static int nvt_probe(struct pnp_dev *pdev, const struct pnp_device_id *dev_id)
 
 	spin_lock_init(&nvt->nvt_lock);
 	spin_lock_init(&nvt->tx.lock);
-	init_ir_raw_event(&nvt->rawir);
 
 	ret = -EBUSY;
 	/* now claim resources */
diff --git a/drivers/media/rc/nuvoton-cir.h b/drivers/media/rc/nuvoton-cir.h
index 1241fc89a36..0d5e0872a2e 100644
--- a/drivers/media/rc/nuvoton-cir.h
+++ b/drivers/media/rc/nuvoton-cir.h
@@ -67,7 +67,6 @@ static int debug;
 struct nvt_dev {
 	struct pnp_dev *pdev;
 	struct rc_dev *rdev;
-	struct ir_raw_event rawir;
 
 	spinlock_t nvt_lock;
 
diff --git a/drivers/media/video/gspca/ov519.c b/drivers/media/video/gspca/ov519.c
index 0800433b209..18305c89083 100644
--- a/drivers/media/video/gspca/ov519.c
+++ b/drivers/media/video/gspca/ov519.c
@@ -2858,7 +2858,6 @@ static void ov7xx0_configure(struct sd *sd)
 			case 0x60:
 				PDEBUG(D_PROBE, "Sensor is a OV7660");
 				sd->sensor = SEN_OV7660;
-				sd->invert_led = 0;
 				break;
 			default:
 				PDEBUG(D_PROBE, "Unknown sensor: 0x76%x", low);
@@ -3337,7 +3336,6 @@ static int sd_config(struct gspca_dev *gspca_dev,
 	case BRIDGE_OV519:
 		cam->cam_mode = ov519_vga_mode;
 		cam->nmodes = ARRAY_SIZE(ov519_vga_mode);
-		sd->invert_led = !sd->invert_led;
 		break;
 	case BRIDGE_OVFX2:
 		cam->cam_mode = ov519_vga_mode;
@@ -5005,24 +5003,24 @@ static const struct sd_desc sd_desc = {
 /* -- module initialisation -- */
 static const struct usb_device_id device_table[] = {
 	{USB_DEVICE(0x041e, 0x4003), .driver_info = BRIDGE_W9968CF },
-	{USB_DEVICE(0x041e, 0x4052), .driver_info = BRIDGE_OV519 },
-	{USB_DEVICE(0x041e, 0x405f),
+	{USB_DEVICE(0x041e, 0x4052),
 		.driver_info = BRIDGE_OV519 | BRIDGE_INVERT_LED },
+	{USB_DEVICE(0x041e, 0x405f), .driver_info = BRIDGE_OV519 },
 	{USB_DEVICE(0x041e, 0x4060), .driver_info = BRIDGE_OV519 },
 	{USB_DEVICE(0x041e, 0x4061), .driver_info = BRIDGE_OV519 },
-	{USB_DEVICE(0x041e, 0x4064),
-		.driver_info = BRIDGE_OV519 | BRIDGE_INVERT_LED },
+	{USB_DEVICE(0x041e, 0x4064), .driver_info = BRIDGE_OV519 },
 	{USB_DEVICE(0x041e, 0x4067), .driver_info = BRIDGE_OV519 },
-	{USB_DEVICE(0x041e, 0x4068),
+	{USB_DEVICE(0x041e, 0x4068), .driver_info = BRIDGE_OV519 },
+	{USB_DEVICE(0x045e, 0x028c),
 		.driver_info = BRIDGE_OV519 | BRIDGE_INVERT_LED },
-	{USB_DEVICE(0x045e, 0x028c), .driver_info = BRIDGE_OV519 },
 	{USB_DEVICE(0x054c, 0x0154), .driver_info = BRIDGE_OV519 },
-	{USB_DEVICE(0x054c, 0x0155),
-		.driver_info = BRIDGE_OV519 | BRIDGE_INVERT_LED },
+	{USB_DEVICE(0x054c, 0x0155), .driver_info = BRIDGE_OV519 },
 	{USB_DEVICE(0x05a9, 0x0511), .driver_info = BRIDGE_OV511 },
 	{USB_DEVICE(0x05a9, 0x0518), .driver_info = BRIDGE_OV518 },
-	{USB_DEVICE(0x05a9, 0x0519), .driver_info = BRIDGE_OV519 },
-	{USB_DEVICE(0x05a9, 0x0530), .driver_info = BRIDGE_OV519 },
+	{USB_DEVICE(0x05a9, 0x0519),
+		.driver_info = BRIDGE_OV519 | BRIDGE_INVERT_LED },
+	{USB_DEVICE(0x05a9, 0x0530),
+		.driver_info = BRIDGE_OV519 | BRIDGE_INVERT_LED },
 	{USB_DEVICE(0x05a9, 0x2800), .driver_info = BRIDGE_OVFX2 },
 	{USB_DEVICE(0x05a9, 0x4519), .driver_info = BRIDGE_OV519 },
 	{USB_DEVICE(0x05a9, 0x8519), .driver_info = BRIDGE_OV519 },
diff --git a/drivers/media/video/gspca/sonixj.c b/drivers/media/video/gspca/sonixj.c
index 81b8a600783..c477ad11f10 100644
--- a/drivers/media/video/gspca/sonixj.c
+++ b/drivers/media/video/gspca/sonixj.c
@@ -2386,7 +2386,7 @@ static int sd_start(struct gspca_dev *gspca_dev)
 		reg_w1(gspca_dev, 0x01, 0x22);
 		msleep(100);
 		reg01 = SCL_SEL_OD | S_PDN_INV;
-		reg17 &= MCK_SIZE_MASK;
+		reg17 &= ~MCK_SIZE_MASK;
 		reg17 |= 0x04;		/* clock / 4 */
 		break;
 	}
@@ -2532,6 +2532,10 @@ static int sd_start(struct gspca_dev *gspca_dev)
 		if (!mode) {			/* if 640x480 */
 			reg17 &= ~MCK_SIZE_MASK;
 			reg17 |= 0x04;		/* clock / 4 */
+		} else {
+			reg01 &= ~SYS_SEL_48M;	/* clk 24Mz */
+			reg17 &= ~MCK_SIZE_MASK;
+			reg17 |= 0x02;		/* clock / 2 */
 		}
 		break;
 	case SENSOR_OV7630:
diff --git a/drivers/media/video/pwc/pwc-v4l.c b/drivers/media/video/pwc/pwc-v4l.c
index e9a0e94b999..8c70e64444e 100644
--- a/drivers/media/video/pwc/pwc-v4l.c
+++ b/drivers/media/video/pwc/pwc-v4l.c
@@ -338,7 +338,7 @@ int pwc_init_controls(struct pwc_device *pdev)
 	if (pdev->restore_factory)
 		pdev->restore_factory->flags = V4L2_CTRL_FLAG_UPDATE;
 
-	if (!pdev->features & FEATURE_MOTOR_PANTILT)
+	if (!(pdev->features & FEATURE_MOTOR_PANTILT))
 		return hdl->error;
 
 	/* Motor pan / tilt / reset */
diff --git a/drivers/media/video/via-camera.c b/drivers/media/video/via-camera.c
index 85d3048c1d6..bb7f17f2a33 100644
--- a/drivers/media/video/via-camera.c
+++ b/drivers/media/video/via-camera.c
@@ -1332,6 +1332,8 @@ static __devinit bool viacam_serial_is_enabled(void)
 	struct pci_bus *pbus = pci_find_bus(0, 0);
 	u8 cbyte;
 
+	if (!pbus)
+		return false;
 	pci_bus_read_config_byte(pbus, VIACAM_SERIAL_DEVFN,
 			VIACAM_SERIAL_CREG, &cbyte);
 	if ((cbyte & VIACAM_SERIAL_BIT) == 0)
diff --git a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c
index 91a0a7460eb..b27b94078c2 100644
--- a/drivers/mmc/core/core.c
+++ b/drivers/mmc/core/core.c
@@ -133,7 +133,7 @@ void mmc_request_done(struct mmc_host *host, struct mmc_request *mrq)
 		if (mrq->done)
 			mrq->done(mrq);
 
-		mmc_host_clk_gate(host);
+		mmc_host_clk_release(host);
 	}
 }
 
@@ -192,7 +192,7 @@ mmc_start_request(struct mmc_host *host, struct mmc_request *mrq)
 			mrq->stop->mrq = mrq;
 		}
 	}
-	mmc_host_clk_ungate(host);
+	mmc_host_clk_hold(host);
 	led_trigger_event(host->led, LED_FULL);
 	host->ops->request(host, mrq);
 }
@@ -728,15 +728,17 @@ static inline void mmc_set_ios(struct mmc_host *host)
  */
 void mmc_set_chip_select(struct mmc_host *host, int mode)
 {
+	mmc_host_clk_hold(host);
 	host->ios.chip_select = mode;
 	mmc_set_ios(host);
+	mmc_host_clk_release(host);
 }
 
 /*
  * Sets the host clock to the highest possible frequency that
  * is below "hz".
  */
-void mmc_set_clock(struct mmc_host *host, unsigned int hz)
+static void __mmc_set_clock(struct mmc_host *host, unsigned int hz)
 {
 	WARN_ON(hz < host->f_min);
 
@@ -747,6 +749,13 @@ void mmc_set_clock(struct mmc_host *host, unsigned int hz)
 	mmc_set_ios(host);
 }
 
+void mmc_set_clock(struct mmc_host *host, unsigned int hz)
+{
+	mmc_host_clk_hold(host);
+	__mmc_set_clock(host, hz);
+	mmc_host_clk_release(host);
+}
+
 #ifdef CONFIG_MMC_CLKGATE
 /*
  * This gates the clock by setting it to 0 Hz.
@@ -779,7 +788,7 @@ void mmc_ungate_clock(struct mmc_host *host)
 	if (host->clk_old) {
 		BUG_ON(host->ios.clock);
 		/* This call will also set host->clk_gated to false */
-		mmc_set_clock(host, host->clk_old);
+		__mmc_set_clock(host, host->clk_old);
 	}
 }
 
@@ -807,8 +816,10 @@ void mmc_set_ungated(struct mmc_host *host)
  */
 void mmc_set_bus_mode(struct mmc_host *host, unsigned int mode)
 {
+	mmc_host_clk_hold(host);
 	host->ios.bus_mode = mode;
 	mmc_set_ios(host);
+	mmc_host_clk_release(host);
 }
 
 /*
@@ -816,8 +827,10 @@ void mmc_set_bus_mode(struct mmc_host *host, unsigned int mode)
  */
 void mmc_set_bus_width(struct mmc_host *host, unsigned int width)
 {
+	mmc_host_clk_hold(host);
 	host->ios.bus_width = width;
 	mmc_set_ios(host);
+	mmc_host_clk_release(host);
 }
 
 /**
@@ -1015,8 +1028,10 @@ u32 mmc_select_voltage(struct mmc_host *host, u32 ocr)
 
 		ocr &= 3 << bit;
 
+		mmc_host_clk_hold(host);
 		host->ios.vdd = bit;
 		mmc_set_ios(host);
+		mmc_host_clk_release(host);
 	} else {
 		pr_warning("%s: host doesn't support card's voltages\n",
 				mmc_hostname(host));
@@ -1063,8 +1078,10 @@ int mmc_set_signal_voltage(struct mmc_host *host, int signal_voltage, bool cmd11
  */
 void mmc_set_timing(struct mmc_host *host, unsigned int timing)
 {
+	mmc_host_clk_hold(host);
 	host->ios.timing = timing;
 	mmc_set_ios(host);
+	mmc_host_clk_release(host);
 }
 
 /*
@@ -1072,8 +1089,10 @@ void mmc_set_timing(struct mmc_host *host, unsigned int timing)
  */
 void mmc_set_driver_type(struct mmc_host *host, unsigned int drv_type)
 {
+	mmc_host_clk_hold(host);
 	host->ios.drv_type = drv_type;
 	mmc_set_ios(host);
+	mmc_host_clk_release(host);
 }
 
 /*
@@ -1091,6 +1110,8 @@ static void mmc_power_up(struct mmc_host *host)
 {
 	int bit;
 
+	mmc_host_clk_hold(host);
+
 	/* If ocr is set, we use it */
 	if (host->ocr)
 		bit = ffs(host->ocr) - 1;
@@ -1126,10 +1147,14 @@ static void mmc_power_up(struct mmc_host *host)
 	 * time required to reach a stable voltage.
 	 */
 	mmc_delay(10);
+
+	mmc_host_clk_release(host);
 }
 
 static void mmc_power_off(struct mmc_host *host)
 {
+	mmc_host_clk_hold(host);
+
 	host->ios.clock = 0;
 	host->ios.vdd = 0;
 
@@ -1147,6 +1172,8 @@ static void mmc_power_off(struct mmc_host *host)
 	host->ios.bus_width = MMC_BUS_WIDTH_1;
 	host->ios.timing = MMC_TIMING_LEGACY;
 	mmc_set_ios(host);
+
+	mmc_host_clk_release(host);
 }
 
 /*
diff --git a/drivers/mmc/core/host.c b/drivers/mmc/core/host.c
index b29d3e8fd3a..793d0a0dad8 100644
--- a/drivers/mmc/core/host.c
+++ b/drivers/mmc/core/host.c
@@ -119,14 +119,14 @@ static void mmc_host_clk_gate_work(struct work_struct *work)
 }
 
 /**
- *	mmc_host_clk_ungate - ungate hardware MCI clocks
+ *	mmc_host_clk_hold - ungate hardware MCI clocks
  *	@host: host to ungate.
  *
  *	Makes sure the host ios.clock is restored to a non-zero value
  *	past this call.	Increase clock reference count and ungate clock
  *	if we're the first user.
  */
-void mmc_host_clk_ungate(struct mmc_host *host)
+void mmc_host_clk_hold(struct mmc_host *host)
 {
 	unsigned long flags;
 
@@ -164,14 +164,14 @@ static bool mmc_host_may_gate_card(struct mmc_card *card)
 }
 
 /**
- *	mmc_host_clk_gate - gate off hardware MCI clocks
+ *	mmc_host_clk_release - gate off hardware MCI clocks
  *	@host: host to gate.
  *
  *	Calls the host driver with ios.clock set to zero as often as possible
  *	in order to gate off hardware MCI clocks. Decrease clock reference
  *	count and schedule disabling of clock.
  */
-void mmc_host_clk_gate(struct mmc_host *host)
+void mmc_host_clk_release(struct mmc_host *host)
 {
 	unsigned long flags;
 
@@ -179,7 +179,7 @@ void mmc_host_clk_gate(struct mmc_host *host)
 	host->clk_requests--;
 	if (mmc_host_may_gate_card(host->card) &&
 	    !host->clk_requests)
-		schedule_work(&host->clk_gate_work);
+		queue_work(system_nrt_wq, &host->clk_gate_work);
 	spin_unlock_irqrestore(&host->clk_lock, flags);
 }
 
@@ -231,7 +231,7 @@ static inline void mmc_host_clk_exit(struct mmc_host *host)
 	if (cancel_work_sync(&host->clk_gate_work))
 		mmc_host_clk_gate_delayed(host);
 	if (host->clk_gated)
-		mmc_host_clk_ungate(host);
+		mmc_host_clk_hold(host);
 	/* There should be only one user now */
 	WARN_ON(host->clk_requests > 1);
 }
diff --git a/drivers/mmc/core/host.h b/drivers/mmc/core/host.h
index de199f91192..fb8a5cd2e4a 100644
--- a/drivers/mmc/core/host.h
+++ b/drivers/mmc/core/host.h
@@ -16,16 +16,16 @@ int mmc_register_host_class(void);
 void mmc_unregister_host_class(void);
 
 #ifdef CONFIG_MMC_CLKGATE
-void mmc_host_clk_ungate(struct mmc_host *host);
-void mmc_host_clk_gate(struct mmc_host *host);
+void mmc_host_clk_hold(struct mmc_host *host);
+void mmc_host_clk_release(struct mmc_host *host);
 unsigned int mmc_host_clk_rate(struct mmc_host *host);
 
 #else
-static inline void mmc_host_clk_ungate(struct mmc_host *host)
+static inline void mmc_host_clk_hold(struct mmc_host *host)
 {
 }
 
-static inline void mmc_host_clk_gate(struct mmc_host *host)
+static inline void mmc_host_clk_release(struct mmc_host *host)
 {
 }
 
diff --git a/drivers/mmc/core/sd.c b/drivers/mmc/core/sd.c
index 633975ff2bb..0370e03e314 100644
--- a/drivers/mmc/core/sd.c
+++ b/drivers/mmc/core/sd.c
@@ -469,56 +469,75 @@ static int sd_select_driver_type(struct mmc_card *card, u8 *status)
 	return 0;
 }
 
-static int sd_set_bus_speed_mode(struct mmc_card *card, u8 *status)
+static void sd_update_bus_speed_mode(struct mmc_card *card)
 {
-	unsigned int bus_speed = 0, timing = 0;
-	int err;
-
 	/*
 	 * If the host doesn't support any of the UHS-I modes, fallback on
 	 * default speed.
 	 */
 	if (!(card->host->caps & (MMC_CAP_UHS_SDR12 | MMC_CAP_UHS_SDR25 |
-	    MMC_CAP_UHS_SDR50 | MMC_CAP_UHS_SDR104 | MMC_CAP_UHS_DDR50)))
-		return 0;
+	    MMC_CAP_UHS_SDR50 | MMC_CAP_UHS_SDR104 | MMC_CAP_UHS_DDR50))) {
+		card->sd_bus_speed = 0;
+		return;
+	}
 
 	if ((card->host->caps & MMC_CAP_UHS_SDR104) &&
 	    (card->sw_caps.sd3_bus_mode & SD_MODE_UHS_SDR104)) {
-			bus_speed = UHS_SDR104_BUS_SPEED;
-			timing = MMC_TIMING_UHS_SDR104;
-			card->sw_caps.uhs_max_dtr = UHS_SDR104_MAX_DTR;
+			card->sd_bus_speed = UHS_SDR104_BUS_SPEED;
 	} else if ((card->host->caps & MMC_CAP_UHS_DDR50) &&
 		   (card->sw_caps.sd3_bus_mode & SD_MODE_UHS_DDR50)) {
-			bus_speed = UHS_DDR50_BUS_SPEED;
-			timing = MMC_TIMING_UHS_DDR50;
-			card->sw_caps.uhs_max_dtr = UHS_DDR50_MAX_DTR;
+			card->sd_bus_speed = UHS_DDR50_BUS_SPEED;
 	} else if ((card->host->caps & (MMC_CAP_UHS_SDR104 |
 		    MMC_CAP_UHS_SDR50)) && (card->sw_caps.sd3_bus_mode &
 		    SD_MODE_UHS_SDR50)) {
-			bus_speed = UHS_SDR50_BUS_SPEED;
-			timing = MMC_TIMING_UHS_SDR50;
-			card->sw_caps.uhs_max_dtr = UHS_SDR50_MAX_DTR;
+			card->sd_bus_speed = UHS_SDR50_BUS_SPEED;
 	} else if ((card->host->caps & (MMC_CAP_UHS_SDR104 |
 		    MMC_CAP_UHS_SDR50 | MMC_CAP_UHS_SDR25)) &&
 		   (card->sw_caps.sd3_bus_mode & SD_MODE_UHS_SDR25)) {
-			bus_speed = UHS_SDR25_BUS_SPEED;
-			timing = MMC_TIMING_UHS_SDR25;
-			card->sw_caps.uhs_max_dtr = UHS_SDR25_MAX_DTR;
+			card->sd_bus_speed = UHS_SDR25_BUS_SPEED;
 	} else if ((card->host->caps & (MMC_CAP_UHS_SDR104 |
 		    MMC_CAP_UHS_SDR50 | MMC_CAP_UHS_SDR25 |
 		    MMC_CAP_UHS_SDR12)) && (card->sw_caps.sd3_bus_mode &
 		    SD_MODE_UHS_SDR12)) {
-			bus_speed = UHS_SDR12_BUS_SPEED;
-			timing = MMC_TIMING_UHS_SDR12;
-			card->sw_caps.uhs_max_dtr = UHS_SDR12_MAX_DTR;
+			card->sd_bus_speed = UHS_SDR12_BUS_SPEED;
+	}
+}
+
+static int sd_set_bus_speed_mode(struct mmc_card *card, u8 *status)
+{
+	int err;
+	unsigned int timing = 0;
+
+	switch (card->sd_bus_speed) {
+	case UHS_SDR104_BUS_SPEED:
+		timing = MMC_TIMING_UHS_SDR104;
+		card->sw_caps.uhs_max_dtr = UHS_SDR104_MAX_DTR;
+		break;
+	case UHS_DDR50_BUS_SPEED:
+		timing = MMC_TIMING_UHS_DDR50;
+		card->sw_caps.uhs_max_dtr = UHS_DDR50_MAX_DTR;
+		break;
+	case UHS_SDR50_BUS_SPEED:
+		timing = MMC_TIMING_UHS_SDR50;
+		card->sw_caps.uhs_max_dtr = UHS_SDR50_MAX_DTR;
+		break;
+	case UHS_SDR25_BUS_SPEED:
+		timing = MMC_TIMING_UHS_SDR25;
+		card->sw_caps.uhs_max_dtr = UHS_SDR25_MAX_DTR;
+		break;
+	case UHS_SDR12_BUS_SPEED:
+		timing = MMC_TIMING_UHS_SDR12;
+		card->sw_caps.uhs_max_dtr = UHS_SDR12_MAX_DTR;
+		break;
+	default:
+		return 0;
 	}
 
-	card->sd_bus_speed = bus_speed;
-	err = mmc_sd_switch(card, 1, 0, bus_speed, status);
+	err = mmc_sd_switch(card, 1, 0, card->sd_bus_speed, status);
 	if (err)
 		return err;
 
-	if ((status[16] & 0xF) != bus_speed)
+	if ((status[16] & 0xF) != card->sd_bus_speed)
 		printk(KERN_WARNING "%s: Problem setting bus speed mode!\n",
 			mmc_hostname(card->host));
 	else {
@@ -618,18 +637,24 @@ static int mmc_sd_init_uhs_card(struct mmc_card *card)
 		mmc_set_bus_width(card->host, MMC_BUS_WIDTH_4);
 	}
 
+	/*
+	 * Select the bus speed mode depending on host
+	 * and card capability.
+	 */
+	sd_update_bus_speed_mode(card);
+
 	/* Set the driver strength for the card */
 	err = sd_select_driver_type(card, status);
 	if (err)
 		goto out;
 
-	/* Set bus speed mode of the card */
-	err = sd_set_bus_speed_mode(card, status);
+	/* Set current limit for the card */
+	err = sd_set_current_limit(card, status);
 	if (err)
 		goto out;
 
-	/* Set current limit for the card */
-	err = sd_set_current_limit(card, status);
+	/* Set bus speed mode of the card */
+	err = sd_set_bus_speed_mode(card, status);
 	if (err)
 		goto out;
 
diff --git a/drivers/mmc/host/sdhci-esdhc-imx.c b/drivers/mmc/host/sdhci-esdhc-imx.c
index 0e9780f5a4a..4dc0028086a 100644
--- a/drivers/mmc/host/sdhci-esdhc-imx.c
+++ b/drivers/mmc/host/sdhci-esdhc-imx.c
@@ -16,6 +16,7 @@
 #include <linux/err.h>
 #include <linux/clk.h>
 #include <linux/gpio.h>
+#include <linux/module.h>
 #include <linux/slab.h>
 #include <linux/mmc/host.h>
 #include <linux/mmc/mmc.h>
diff --git a/drivers/mmc/host/sdhci-s3c.c b/drivers/mmc/host/sdhci-s3c.c
index 2bd7bf4fece..fe886d6c474 100644
--- a/drivers/mmc/host/sdhci-s3c.c
+++ b/drivers/mmc/host/sdhci-s3c.c
@@ -302,6 +302,8 @@ static int sdhci_s3c_platform_8bit_width(struct sdhci_host *host, int width)
 		ctrl &= ~SDHCI_CTRL_8BITBUS;
 		break;
 	default:
+		ctrl &= ~SDHCI_CTRL_4BITBUS;
+		ctrl &= ~SDHCI_CTRL_8BITBUS;
 		break;
 	}
 
diff --git a/drivers/mmc/host/sh_mobile_sdhi.c b/drivers/mmc/host/sh_mobile_sdhi.c
index 774f6439d7c..0c4a672f5db 100644
--- a/drivers/mmc/host/sh_mobile_sdhi.c
+++ b/drivers/mmc/host/sh_mobile_sdhi.c
@@ -120,11 +120,11 @@ static int __devinit sh_mobile_sdhi_probe(struct platform_device *pdev)
 	mmc_data->hclk = clk_get_rate(priv->clk);
 	mmc_data->set_pwr = sh_mobile_sdhi_set_pwr;
 	mmc_data->get_cd = sh_mobile_sdhi_get_cd;
-	if (mmc_data->flags & TMIO_MMC_HAS_IDLE_WAIT)
-		mmc_data->write16_hook = sh_mobile_sdhi_write16_hook;
 	mmc_data->capabilities = MMC_CAP_MMC_HIGHSPEED;
 	if (p) {
 		mmc_data->flags = p->tmio_flags;
+		if (mmc_data->flags & TMIO_MMC_HAS_IDLE_WAIT)
+			mmc_data->write16_hook = sh_mobile_sdhi_write16_hook;
 		mmc_data->ocr_mask = p->tmio_ocr_mask;
 		mmc_data->capabilities |= p->tmio_caps;
 
diff --git a/drivers/mtd/ubi/debug.h b/drivers/mtd/ubi/debug.h
index 65b5b76cc37..64fbb002182 100644
--- a/drivers/mtd/ubi/debug.h
+++ b/drivers/mtd/ubi/debug.h
@@ -181,7 +181,7 @@ static inline int ubi_dbg_is_erase_failure(const struct ubi_device *ubi)
 
 #define ubi_dbg_msg(fmt, ...) do {                                           \
 	if (0)                                                               \
-		pr_debug(fmt "\n", ##__VA_ARGS__);                           \
+		printk(KERN_DEBUG fmt "\n", ##__VA_ARGS__);                  \
 } while (0)
 
 #define dbg_msg(fmt, ...)  ubi_dbg_msg(fmt, ##__VA_ARGS__)
diff --git a/drivers/net/arm/am79c961a.c b/drivers/net/arm/am79c961a.c
index 52fe21e1e2c..3b1416e3d21 100644
--- a/drivers/net/arm/am79c961a.c
+++ b/drivers/net/arm/am79c961a.c
@@ -308,8 +308,11 @@ static void am79c961_timer(unsigned long data)
 	struct net_device *dev = (struct net_device *)data;
 	struct dev_priv *priv = netdev_priv(dev);
 	unsigned int lnkstat, carrier;
+	unsigned long flags;
 
+	spin_lock_irqsave(&priv->chip_lock, flags);
 	lnkstat = read_ireg(dev->base_addr, ISALED0) & ISALED0_LNKST;
+	spin_unlock_irqrestore(&priv->chip_lock, flags);
 	carrier = netif_carrier_ok(dev);
 
 	if (lnkstat && !carrier) {
diff --git a/drivers/net/can/ti_hecc.c b/drivers/net/can/ti_hecc.c
index f7bbde9eb2c..a81249246ec 100644
--- a/drivers/net/can/ti_hecc.c
+++ b/drivers/net/can/ti_hecc.c
@@ -503,9 +503,9 @@ static netdev_tx_t ti_hecc_xmit(struct sk_buff *skb, struct net_device *ndev)
 	spin_unlock_irqrestore(&priv->mbx_lock, flags);
 
 	/* Prepare mailbox for transmission */
+	data = cf->can_dlc | (get_tx_head_prio(priv) << 8);
 	if (cf->can_id & CAN_RTR_FLAG) /* Remote transmission request */
 		data |= HECC_CANMCF_RTR;
-	data |= get_tx_head_prio(priv) << 8;
 	hecc_write_mbx(priv, mbxno, HECC_CANMCF, data);
 
 	if (cf->can_id & CAN_EFF_FLAG) /* Extended frame format */
@@ -923,6 +923,7 @@ static int ti_hecc_probe(struct platform_device *pdev)
 	priv->can.do_get_state = ti_hecc_get_state;
 	priv->can.ctrlmode_supported = CAN_CTRLMODE_3_SAMPLES;
 
+	spin_lock_init(&priv->mbx_lock);
 	ndev->irq = irq->start;
 	ndev->flags |= IFF_ECHO;
 	platform_set_drvdata(pdev, ndev);
diff --git a/drivers/net/cassini.c b/drivers/net/cassini.c
index 646c86bcc54..fdb7a175640 100644
--- a/drivers/net/cassini.c
+++ b/drivers/net/cassini.c
@@ -2452,14 +2452,13 @@ static irqreturn_t cas_interruptN(int irq, void *dev_id)
 	struct net_device *dev = dev_id;
 	struct cas *cp = netdev_priv(dev);
 	unsigned long flags;
-	int ring;
+	int ring = (irq == cp->pci_irq_INTC) ? 2 : 3;
 	u32 status = readl(cp->regs + REG_PLUS_INTRN_STATUS(ring));
 
 	/* check for shared irq */
 	if (status == 0)
 		return IRQ_NONE;
 
-	ring = (irq == cp->pci_irq_INTC) ? 2 : 3;
 	spin_lock_irqsave(&cp->lock, flags);
 	if (status & INTR_RX_DONE_ALT) { /* handle rx separately */
 #ifdef USE_NAPI
diff --git a/drivers/net/ibmveth.c b/drivers/net/ibmveth.c
index ba99af05bf6..3e667926940 100644
--- a/drivers/net/ibmveth.c
+++ b/drivers/net/ibmveth.c
@@ -395,7 +395,7 @@ static inline struct sk_buff *ibmveth_rxq_get_buffer(struct ibmveth_adapter *ada
 }
 
 /* recycle the current buffer on the rx queue */
-static void ibmveth_rxq_recycle_buffer(struct ibmveth_adapter *adapter)
+static int ibmveth_rxq_recycle_buffer(struct ibmveth_adapter *adapter)
 {
 	u32 q_index = adapter->rx_queue.index;
 	u64 correlator = adapter->rx_queue.queue_addr[q_index].correlator;
@@ -403,6 +403,7 @@ static void ibmveth_rxq_recycle_buffer(struct ibmveth_adapter *adapter)
 	unsigned int index = correlator & 0xffffffffUL;
 	union ibmveth_buf_desc desc;
 	unsigned long lpar_rc;
+	int ret = 1;
 
 	BUG_ON(pool >= IBMVETH_NUM_BUFF_POOLS);
 	BUG_ON(index >= adapter->rx_buff_pool[pool].size);
@@ -410,7 +411,7 @@ static void ibmveth_rxq_recycle_buffer(struct ibmveth_adapter *adapter)
 	if (!adapter->rx_buff_pool[pool].active) {
 		ibmveth_rxq_harvest_buffer(adapter);
 		ibmveth_free_buffer_pool(adapter, &adapter->rx_buff_pool[pool]);
-		return;
+		goto out;
 	}
 
 	desc.fields.flags_len = IBMVETH_BUF_VALID |
@@ -423,12 +424,16 @@ static void ibmveth_rxq_recycle_buffer(struct ibmveth_adapter *adapter)
 		netdev_dbg(adapter->netdev, "h_add_logical_lan_buffer failed "
 			   "during recycle rc=%ld", lpar_rc);
 		ibmveth_remove_buffer_from_pool(adapter, adapter->rx_queue.queue_addr[adapter->rx_queue.index].correlator);
+		ret = 0;
 	}
 
 	if (++adapter->rx_queue.index == adapter->rx_queue.num_slots) {
 		adapter->rx_queue.index = 0;
 		adapter->rx_queue.toggle = !adapter->rx_queue.toggle;
 	}
+
+out:
+	return ret;
 }
 
 static void ibmveth_rxq_harvest_buffer(struct ibmveth_adapter *adapter)
@@ -1084,8 +1089,9 @@ restart_poll:
 				if (rx_flush)
 					ibmveth_flush_buffer(skb->data,
 						length + offset);
+				if (!ibmveth_rxq_recycle_buffer(adapter))
+					kfree_skb(skb);
 				skb = new_skb;
-				ibmveth_rxq_recycle_buffer(adapter);
 			} else {
 				ibmveth_rxq_harvest_buffer(adapter);
 				skb_reserve(skb, offset);
diff --git a/drivers/net/phy/national.c b/drivers/net/phy/national.c
index 0620ba96350..04bb8fcc0cb 100644
--- a/drivers/net/phy/national.c
+++ b/drivers/net/phy/national.c
@@ -25,8 +25,9 @@
 /* DP83865 phy identifier values */
 #define DP83865_PHY_ID	0x20005c7a
 
-#define DP83865_INT_MASK_REG 0x15
-#define DP83865_INT_MASK_STATUS 0x14
+#define DP83865_INT_STATUS	0x14
+#define DP83865_INT_MASK	0x15
+#define DP83865_INT_CLEAR	0x17
 
 #define DP83865_INT_REMOTE_FAULT 0x0008
 #define DP83865_INT_ANE_COMPLETED 0x0010
@@ -68,21 +69,25 @@ static int ns_config_intr(struct phy_device *phydev)
 	int err;
 
 	if (phydev->interrupts == PHY_INTERRUPT_ENABLED)
-		err = phy_write(phydev, DP83865_INT_MASK_REG,
+		err = phy_write(phydev, DP83865_INT_MASK,
 				DP83865_INT_MASK_DEFAULT);
 	else
-		err = phy_write(phydev, DP83865_INT_MASK_REG, 0);
+		err = phy_write(phydev, DP83865_INT_MASK, 0);
 
 	return err;
 }
 
 static int ns_ack_interrupt(struct phy_device *phydev)
 {
-	int ret = phy_read(phydev, DP83865_INT_MASK_STATUS);
+	int ret = phy_read(phydev, DP83865_INT_STATUS);
 	if (ret < 0)
 		return ret;
 
-	return 0;
+	/* Clear the interrupt status bit by writing a “1”
+	 * to the corresponding bit in INT_CLEAR (2:0 are reserved) */
+	ret = phy_write(phydev, DP83865_INT_CLEAR, ret & ~0x7);
+
+	return ret;
 }
 
 static void ns_giga_speed_fallback(struct phy_device *phydev, int mode)
diff --git a/drivers/net/sh_eth.c b/drivers/net/sh_eth.c
index 190f619e421..1c1666e9910 100644
--- a/drivers/net/sh_eth.c
+++ b/drivers/net/sh_eth.c
@@ -31,6 +31,7 @@
 #include <linux/phy.h>
 #include <linux/cache.h>
 #include <linux/io.h>
+#include <linux/interrupt.h>
 #include <linux/pm_runtime.h>
 #include <linux/slab.h>
 #include <linux/ethtool.h>
diff --git a/drivers/net/wireless/ath/ath9k/ar9003_eeprom.c b/drivers/net/wireless/ath/ath9k/ar9003_eeprom.c
index c34bef1bf2b..1b9400371ea 100644
--- a/drivers/net/wireless/ath/ath9k/ar9003_eeprom.c
+++ b/drivers/net/wireless/ath/ath9k/ar9003_eeprom.c
@@ -69,7 +69,7 @@ static int ar9003_hw_power_interpolate(int32_t x,
 static const struct ar9300_eeprom ar9300_default = {
 	.eepromVersion = 2,
 	.templateVersion = 2,
-	.macAddr = {1, 2, 3, 4, 5, 6},
+	.macAddr = {0, 2, 3, 4, 5, 6},
 	.custData = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
 		     0, 0, 0, 0, 0, 0, 0, 0, 0, 0},
 	.baseEepHeader = {
diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c
index 9098aaad97a..6530694a59a 100644
--- a/drivers/net/wireless/ath/ath9k/main.c
+++ b/drivers/net/wireless/ath/ath9k/main.c
@@ -2283,7 +2283,11 @@ static void ath9k_set_coverage_class(struct ieee80211_hw *hw, u8 coverage_class)
 
 	mutex_lock(&sc->mutex);
 	ah->coverage_class = coverage_class;
+
+	ath9k_ps_wakeup(sc);
 	ath9k_hw_init_global_settings(ah);
+	ath9k_ps_restore(sc);
+
 	mutex_unlock(&sc->mutex);
 }
 
diff --git a/drivers/net/wireless/ath/carl9170/main.c b/drivers/net/wireless/ath/carl9170/main.c
index 0122930b14c..0474e6638d2 100644
--- a/drivers/net/wireless/ath/carl9170/main.c
+++ b/drivers/net/wireless/ath/carl9170/main.c
@@ -1066,8 +1066,10 @@ static int carl9170_op_set_key(struct ieee80211_hw *hw, enum set_key_cmd cmd,
 	 * the high througput speed in 802.11n networks.
 	 */
 
-	if (!is_main_vif(ar, vif))
+	if (!is_main_vif(ar, vif)) {
+		mutex_lock(&ar->mutex);
 		goto err_softw;
+	}
 
 	/*
 	 * While the hardware supports *catch-all* key, for offloading
diff --git a/drivers/net/wireless/rt2x00/rt2x00usb.c b/drivers/net/wireless/rt2x00/rt2x00usb.c
index 7fbb55c9da8..1e31050dafc 100644
--- a/drivers/net/wireless/rt2x00/rt2x00usb.c
+++ b/drivers/net/wireless/rt2x00/rt2x00usb.c
@@ -871,18 +871,8 @@ int rt2x00usb_suspend(struct usb_interface *usb_intf, pm_message_t state)
 {
 	struct ieee80211_hw *hw = usb_get_intfdata(usb_intf);
 	struct rt2x00_dev *rt2x00dev = hw->priv;
-	int retval;
-
-	retval = rt2x00lib_suspend(rt2x00dev, state);
-	if (retval)
-		return retval;
 
-	/*
-	 * Decrease usbdev refcount.
-	 */
-	usb_put_dev(interface_to_usbdev(usb_intf));
-
-	return 0;
+	return rt2x00lib_suspend(rt2x00dev, state);
 }
 EXPORT_SYMBOL_GPL(rt2x00usb_suspend);
 
@@ -891,8 +881,6 @@ int rt2x00usb_resume(struct usb_interface *usb_intf)
 	struct ieee80211_hw *hw = usb_get_intfdata(usb_intf);
 	struct rt2x00_dev *rt2x00dev = hw->priv;
 
-	usb_get_dev(interface_to_usbdev(usb_intf));
-
 	return rt2x00lib_resume(rt2x00dev);
 }
 EXPORT_SYMBOL_GPL(rt2x00usb_resume);
diff --git a/drivers/net/wireless/wl12xx/main.c b/drivers/net/wireless/wl12xx/main.c
index e58c22d21e3..b70ae40ad66 100644
--- a/drivers/net/wireless/wl12xx/main.c
+++ b/drivers/net/wireless/wl12xx/main.c
@@ -4283,6 +4283,7 @@ int wl1271_init_ieee80211(struct wl1271 *wl)
 	wl->hw->wiphy->interface_modes = BIT(NL80211_IFTYPE_STATION) |
 		BIT(NL80211_IFTYPE_ADHOC) | BIT(NL80211_IFTYPE_AP);
 	wl->hw->wiphy->max_scan_ssids = 1;
+	wl->hw->wiphy->max_sched_scan_ssids = 1;
 	/*
 	 * Maximum length of elements in scanning probe request templates
 	 * should be the maximum length possible for a template, without
diff --git a/drivers/net/wireless/wl12xx/sdio.c b/drivers/net/wireless/wl12xx/sdio.c
index 5cf18c2c23f..fb1fd5af75e 100644
--- a/drivers/net/wireless/wl12xx/sdio.c
+++ b/drivers/net/wireless/wl12xx/sdio.c
@@ -164,7 +164,7 @@ static int wl1271_sdio_power_on(struct wl1271 *wl)
 	/* If enabled, tell runtime PM not to power off the card */
 	if (pm_runtime_enabled(&func->dev)) {
 		ret = pm_runtime_get_sync(&func->dev);
-		if (ret)
+		if (ret < 0)
 			goto out;
 	} else {
 		/* Runtime PM is disabled: power up the card manually */
diff --git a/drivers/net/wireless/wl12xx/testmode.c b/drivers/net/wireless/wl12xx/testmode.c
index 88add68bd9a..4ae8effaee2 100644
--- a/drivers/net/wireless/wl12xx/testmode.c
+++ b/drivers/net/wireless/wl12xx/testmode.c
@@ -36,7 +36,6 @@ enum wl1271_tm_commands {
 	WL1271_TM_CMD_TEST,
 	WL1271_TM_CMD_INTERROGATE,
 	WL1271_TM_CMD_CONFIGURE,
-	WL1271_TM_CMD_NVS_PUSH,
 	WL1271_TM_CMD_SET_PLT_MODE,
 	WL1271_TM_CMD_RECOVER,
 
@@ -190,48 +189,6 @@ static int wl1271_tm_cmd_configure(struct wl1271 *wl, struct nlattr *tb[])
 	return 0;
 }
 
-static int wl1271_tm_cmd_nvs_push(struct wl1271 *wl, struct nlattr *tb[])
-{
-	int ret = 0;
-	size_t len;
-	void *buf;
-
-	wl1271_debug(DEBUG_TESTMODE, "testmode cmd nvs push");
-
-	if (!tb[WL1271_TM_ATTR_DATA])
-		return -EINVAL;
-
-	buf = nla_data(tb[WL1271_TM_ATTR_DATA]);
-	len = nla_len(tb[WL1271_TM_ATTR_DATA]);
-
-	mutex_lock(&wl->mutex);
-
-	kfree(wl->nvs);
-
-	if ((wl->chip.id == CHIP_ID_1283_PG20) &&
-	    (len != sizeof(struct wl128x_nvs_file)))
-		return -EINVAL;
-	else if (len != sizeof(struct wl1271_nvs_file))
-		return -EINVAL;
-
-	wl->nvs = kzalloc(len, GFP_KERNEL);
-	if (!wl->nvs) {
-		wl1271_error("could not allocate memory for the nvs file");
-		ret = -ENOMEM;
-		goto out;
-	}
-
-	memcpy(wl->nvs, buf, len);
-	wl->nvs_len = len;
-
-	wl1271_debug(DEBUG_TESTMODE, "testmode pushed nvs");
-
-out:
-	mutex_unlock(&wl->mutex);
-
-	return ret;
-}
-
 static int wl1271_tm_cmd_set_plt_mode(struct wl1271 *wl, struct nlattr *tb[])
 {
 	u32 val;
@@ -288,8 +245,6 @@ int wl1271_tm_cmd(struct ieee80211_hw *hw, void *data, int len)
 		return wl1271_tm_cmd_interrogate(wl, tb);
 	case WL1271_TM_CMD_CONFIGURE:
 		return wl1271_tm_cmd_configure(wl, tb);
-	case WL1271_TM_CMD_NVS_PUSH:
-		return wl1271_tm_cmd_nvs_push(wl, tb);
 	case WL1271_TM_CMD_SET_PLT_MODE:
 		return wl1271_tm_cmd_set_plt_mode(wl, tb);
 	case WL1271_TM_CMD_RECOVER:
diff --git a/drivers/pci/hotplug/pcihp_slot.c b/drivers/pci/hotplug/pcihp_slot.c
index 753b21aaea6..3ffd9c1acc0 100644
--- a/drivers/pci/hotplug/pcihp_slot.c
+++ b/drivers/pci/hotplug/pcihp_slot.c
@@ -169,7 +169,9 @@ void pci_configure_slot(struct pci_dev *dev)
 			(dev->class >> 8) == PCI_CLASS_BRIDGE_PCI)))
 		return;
 
-	pcie_bus_configure_settings(dev->bus, dev->bus->self->pcie_mpss);
+	if (dev->bus && dev->bus->self)
+		pcie_bus_configure_settings(dev->bus,
+					    dev->bus->self->pcie_mpss);
 
 	memset(&hpp, 0, sizeof(hpp));
 	ret = pci_get_hp_params(dev, &hpp);
diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index 0ce67423a0a..4e84fd4a431 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -77,7 +77,7 @@ unsigned long pci_cardbus_mem_size = DEFAULT_CARDBUS_MEM_SIZE;
 unsigned long pci_hotplug_io_size  = DEFAULT_HOTPLUG_IO_SIZE;
 unsigned long pci_hotplug_mem_size = DEFAULT_HOTPLUG_MEM_SIZE;
 
-enum pcie_bus_config_types pcie_bus_config = PCIE_BUS_PERFORMANCE;
+enum pcie_bus_config_types pcie_bus_config = PCIE_BUS_SAFE;
 
 /*
  * The default CLS is used if arch didn't set CLS explicitly and not
diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
index 8473727b29f..b1187ff31d8 100644
--- a/drivers/pci/probe.c
+++ b/drivers/pci/probe.c
@@ -1396,34 +1396,37 @@ static void pcie_write_mps(struct pci_dev *dev, int mps)
 
 static void pcie_write_mrrs(struct pci_dev *dev, int mps)
 {
-	int rc, mrrs;
+	int rc, mrrs, dev_mpss;
 
-	if (pcie_bus_config == PCIE_BUS_PERFORMANCE) {
-		int dev_mpss = 128 << dev->pcie_mpss;
+	/* In the "safe" case, do not configure the MRRS.  There appear to be
+	 * issues with setting MRRS to 0 on a number of devices.
+	 */
 
-		/* For Max performance, the MRRS must be set to the largest
-		 * supported value.  However, it cannot be configured larger
-		 * than the MPS the device or the bus can support.  This assumes
-		 * that the largest MRRS available on the device cannot be
-		 * smaller than the device MPSS.
-		 */
-		mrrs = mps < dev_mpss ? mps : dev_mpss;
-	} else
-		/* In the "safe" case, configure the MRRS for fairness on the
-		 * bus by making all devices have the same size
-		 */
-		mrrs = mps;
+	if (pcie_bus_config != PCIE_BUS_PERFORMANCE)
+		return;
 
+	dev_mpss = 128 << dev->pcie_mpss;
+
+	/* For Max performance, the MRRS must be set to the largest supported
+	 * value.  However, it cannot be configured larger than the MPS the
+	 * device or the bus can support.  This assumes that the largest MRRS
+	 * available on the device cannot be smaller than the device MPSS.
+	 */
+	mrrs = min(mps, dev_mpss);
 
 	/* MRRS is a R/W register.  Invalid values can be written, but a
-	 * subsiquent read will verify if the value is acceptable or not.
+	 * subsequent read will verify if the value is acceptable or not.
 	 * If the MRRS value provided is not acceptable (e.g., too large),
 	 * shrink the value until it is acceptable to the HW.
  	 */
 	while (mrrs != pcie_get_readrq(dev) && mrrs >= 128) {
+		dev_warn(&dev->dev, "Attempting to modify the PCI-E MRRS value"
+			 " to %d.  If any issues are encountered, please try "
+			 "running with pci=pcie_bus_safe\n", mrrs);
 		rc = pcie_set_readrq(dev, mrrs);
 		if (rc)
-			dev_err(&dev->dev, "Failed attempting to set the MRRS\n");
+			dev_err(&dev->dev,
+				"Failed attempting to set the MRRS\n");
 
 		mrrs /= 2;
 	}
@@ -1436,13 +1439,13 @@ static int pcie_bus_configure_set(struct pci_dev *dev, void *data)
 	if (!pci_is_pcie(dev))
 		return 0;
 
-	dev_info(&dev->dev, "Dev MPS %d MPSS %d MRRS %d\n",
+	dev_dbg(&dev->dev, "Dev MPS %d MPSS %d MRRS %d\n",
 		 pcie_get_mps(dev), 128<<dev->pcie_mpss, pcie_get_readrq(dev));
 
 	pcie_write_mps(dev, mps);
 	pcie_write_mrrs(dev, mps);
 
-	dev_info(&dev->dev, "Dev MPS %d MPSS %d MRRS %d\n",
+	dev_dbg(&dev->dev, "Dev MPS %d MPSS %d MRRS %d\n",
 		 pcie_get_mps(dev), 128<<dev->pcie_mpss, pcie_get_readrq(dev));
 
 	return 0;
@@ -1456,9 +1459,6 @@ void pcie_bus_configure_settings(struct pci_bus *bus, u8 mpss)
 {
 	u8 smpss = mpss;
 
-	if (!bus->self)
-		return;
-
 	if (!pci_is_pcie(bus->self))
 		return;
 
diff --git a/drivers/rtc/rtc-ep93xx.c b/drivers/rtc/rtc-ep93xx.c
index 335551d333b..14a42a1edc6 100644
--- a/drivers/rtc/rtc-ep93xx.c
+++ b/drivers/rtc/rtc-ep93xx.c
@@ -36,6 +36,7 @@
  */
 struct ep93xx_rtc {
 	void __iomem	*mmio_base;
+	struct rtc_device *rtc;
 };
 
 static int ep93xx_rtc_get_swcomp(struct device *dev, unsigned short *preload,
@@ -130,7 +131,6 @@ static int __init ep93xx_rtc_probe(struct platform_device *pdev)
 {
 	struct ep93xx_rtc *ep93xx_rtc;
 	struct resource *res;
-	struct rtc_device *rtc;
 	int err;
 
 	ep93xx_rtc = devm_kzalloc(&pdev->dev, sizeof(*ep93xx_rtc), GFP_KERNEL);
@@ -151,12 +151,12 @@ static int __init ep93xx_rtc_probe(struct platform_device *pdev)
 		return -ENXIO;
 
 	pdev->dev.platform_data = ep93xx_rtc;
-	platform_set_drvdata(pdev, rtc);
+	platform_set_drvdata(pdev, ep93xx_rtc);
 
-	rtc = rtc_device_register(pdev->name,
+	ep93xx_rtc->rtc = rtc_device_register(pdev->name,
 				&pdev->dev, &ep93xx_rtc_ops, THIS_MODULE);
-	if (IS_ERR(rtc)) {
-		err = PTR_ERR(rtc);
+	if (IS_ERR(ep93xx_rtc->rtc)) {
+		err = PTR_ERR(ep93xx_rtc->rtc);
 		goto exit;
 	}
 
@@ -167,7 +167,7 @@ static int __init ep93xx_rtc_probe(struct platform_device *pdev)
 	return 0;
 
 fail:
-	rtc_device_unregister(rtc);
+	rtc_device_unregister(ep93xx_rtc->rtc);
 exit:
 	platform_set_drvdata(pdev, NULL);
 	pdev->dev.platform_data = NULL;
@@ -176,11 +176,11 @@ exit:
 
 static int __exit ep93xx_rtc_remove(struct platform_device *pdev)
 {
-	struct rtc_device *rtc = platform_get_drvdata(pdev);
+	struct ep93xx_rtc *ep93xx_rtc = platform_get_drvdata(pdev);
 
 	sysfs_remove_group(&pdev->dev.kobj, &ep93xx_rtc_sysfs_files);
 	platform_set_drvdata(pdev, NULL);
-	rtc_device_unregister(rtc);
+	rtc_device_unregister(ep93xx_rtc->rtc);
 	pdev->dev.platform_data = NULL;
 
 	return 0;
diff --git a/drivers/rtc/rtc-lib.c b/drivers/rtc/rtc-lib.c
index 075f1708dea..c4cf0573111 100644
--- a/drivers/rtc/rtc-lib.c
+++ b/drivers/rtc/rtc-lib.c
@@ -85,6 +85,8 @@ void rtc_time_to_tm(unsigned long time, struct rtc_time *tm)
 	time -= tm->tm_hour * 3600;
 	tm->tm_min = time / 60;
 	tm->tm_sec = time - tm->tm_min * 60;
+
+	tm->tm_isdst = 0;
 }
 EXPORT_SYMBOL(rtc_time_to_tm);
 
diff --git a/drivers/rtc/rtc-twl.c b/drivers/rtc/rtc-twl.c
index 9a81f778d6b..20687d55e7a 100644
--- a/drivers/rtc/rtc-twl.c
+++ b/drivers/rtc/rtc-twl.c
@@ -362,14 +362,6 @@ static irqreturn_t twl_rtc_interrupt(int irq, void *rtc)
 	int res;
 	u8 rd_reg;
 
-#ifdef CONFIG_LOCKDEP
-	/* WORKAROUND for lockdep forcing IRQF_DISABLED on us, which
-	 * we don't want and can't tolerate.  Although it might be
-	 * friendlier not to borrow this thread context...
-	 */
-	local_irq_enable();
-#endif
-
 	res = twl_rtc_read_u8(&rd_reg, REG_RTC_STATUS_REG);
 	if (res)
 		goto out;
@@ -428,24 +420,12 @@ static struct rtc_class_ops twl_rtc_ops = {
 static int __devinit twl_rtc_probe(struct platform_device *pdev)
 {
 	struct rtc_device *rtc;
-	int ret = 0;
+	int ret = -EINVAL;
 	int irq = platform_get_irq(pdev, 0);
 	u8 rd_reg;
 
 	if (irq <= 0)
-		return -EINVAL;
-
-	rtc = rtc_device_register(pdev->name,
-				  &pdev->dev, &twl_rtc_ops, THIS_MODULE);
-	if (IS_ERR(rtc)) {
-		ret = PTR_ERR(rtc);
-		dev_err(&pdev->dev, "can't register RTC device, err %ld\n",
-			PTR_ERR(rtc));
-		goto out0;
-
-	}
-
-	platform_set_drvdata(pdev, rtc);
+		goto out1;
 
 	ret = twl_rtc_read_u8(&rd_reg, REG_RTC_STATUS_REG);
 	if (ret < 0)
@@ -462,14 +442,6 @@ static int __devinit twl_rtc_probe(struct platform_device *pdev)
 	if (ret < 0)
 		goto out1;
 
-	ret = request_irq(irq, twl_rtc_interrupt,
-				IRQF_TRIGGER_RISING,
-				dev_name(&rtc->dev), rtc);
-	if (ret < 0) {
-		dev_err(&pdev->dev, "IRQ is not free.\n");
-		goto out1;
-	}
-
 	if (twl_class_is_6030()) {
 		twl6030_interrupt_unmask(TWL6030_RTC_INT_MASK,
 			REG_INT_MSK_LINE_A);
@@ -480,28 +452,44 @@ static int __devinit twl_rtc_probe(struct platform_device *pdev)
 	/* Check RTC module status, Enable if it is off */
 	ret = twl_rtc_read_u8(&rd_reg, REG_RTC_CTRL_REG);
 	if (ret < 0)
-		goto out2;
+		goto out1;
 
 	if (!(rd_reg & BIT_RTC_CTRL_REG_STOP_RTC_M)) {
 		dev_info(&pdev->dev, "Enabling TWL-RTC.\n");
 		rd_reg = BIT_RTC_CTRL_REG_STOP_RTC_M;
 		ret = twl_rtc_write_u8(rd_reg, REG_RTC_CTRL_REG);
 		if (ret < 0)
-			goto out2;
+			goto out1;
 	}
 
 	/* init cached IRQ enable bits */
 	ret = twl_rtc_read_u8(&rtc_irq_bits, REG_RTC_INTERRUPTS_REG);
 	if (ret < 0)
+		goto out1;
+
+	rtc = rtc_device_register(pdev->name,
+				  &pdev->dev, &twl_rtc_ops, THIS_MODULE);
+	if (IS_ERR(rtc)) {
+		ret = PTR_ERR(rtc);
+		dev_err(&pdev->dev, "can't register RTC device, err %ld\n",
+			PTR_ERR(rtc));
+		goto out1;
+	}
+
+	ret = request_threaded_irq(irq, NULL, twl_rtc_interrupt,
+				   IRQF_TRIGGER_RISING,
+				   dev_name(&rtc->dev), rtc);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "IRQ is not free.\n");
 		goto out2;
+	}
 
-	return ret;
+	platform_set_drvdata(pdev, rtc);
+	return 0;
 
 out2:
-	free_irq(irq, rtc);
-out1:
 	rtc_device_unregister(rtc);
-out0:
+out1:
 	return ret;
 }
 
diff --git a/drivers/scsi/qla4xxx/Kconfig b/drivers/scsi/qla4xxx/Kconfig
index 2c33ce6eac1..0f5599e0abf 100644
--- a/drivers/scsi/qla4xxx/Kconfig
+++ b/drivers/scsi/qla4xxx/Kconfig
@@ -1,6 +1,6 @@
 config SCSI_QLA_ISCSI
 	tristate "QLogic ISP4XXX and ISP82XX host adapter family support"
-	depends on PCI && SCSI
+	depends on PCI && SCSI && NET
 	select SCSI_ISCSI_ATTRS
 	---help---
 	This driver supports the QLogic 40xx (ISP4XXX) and 8022 (ISP82XX)
diff --git a/drivers/sh/intc/chip.c b/drivers/sh/intc/chip.c
index f33e2dd9793..33b2ed451e0 100644
--- a/drivers/sh/intc/chip.c
+++ b/drivers/sh/intc/chip.c
@@ -186,6 +186,9 @@ static unsigned char intc_irq_sense_table[IRQ_TYPE_SENSE_MASK + 1] = {
     !defined(CONFIG_CPU_SUBTYPE_SH7709)
 	[IRQ_TYPE_LEVEL_HIGH] = VALID(3),
 #endif
+#if defined(CONFIG_ARCH_SH7372)
+	[IRQ_TYPE_EDGE_BOTH] = VALID(4),
+#endif
 };
 
 static int intc_set_type(struct irq_data *data, unsigned int type)
diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c
index a9414facda4..5ea6ec3442e 100644
--- a/drivers/tty/serial/sh-sci.c
+++ b/drivers/tty/serial/sh-sci.c
@@ -47,6 +47,7 @@
 #include <linux/ctype.h>
 #include <linux/err.h>
 #include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
 #include <linux/scatterlist.h>
 #include <linux/slab.h>
 
@@ -95,6 +96,12 @@ struct sci_port {
 #endif
 
 	struct notifier_block		freq_transition;
+
+#ifdef CONFIG_SERIAL_SH_SCI_CONSOLE
+	unsigned short saved_smr;
+	unsigned short saved_fcr;
+	unsigned char saved_brr;
+#endif
 };
 
 /* Function prototypes */
@@ -1076,7 +1083,7 @@ static unsigned int sci_get_mctrl(struct uart_port *port)
 	/* This routine is used for getting signals of: DTR, DCD, DSR, RI,
 	   and CTS/RTS */
 
-	return TIOCM_DTR | TIOCM_RTS | TIOCM_DSR;
+	return TIOCM_DTR | TIOCM_RTS | TIOCM_CTS | TIOCM_DSR;
 }
 
 #ifdef CONFIG_SERIAL_SH_SCI_DMA
@@ -1633,11 +1640,25 @@ static unsigned int sci_scbrr_calc(unsigned int algo_id, unsigned int bps,
 	return ((freq + 16 * bps) / (32 * bps) - 1);
 }
 
+static void sci_reset(struct uart_port *port)
+{
+	unsigned int status;
+
+	do {
+		status = sci_in(port, SCxSR);
+	} while (!(status & SCxSR_TEND(port)));
+
+	sci_out(port, SCSCR, 0x00);	/* TE=0, RE=0, CKE1=0 */
+
+	if (port->type != PORT_SCI)
+		sci_out(port, SCFCR, SCFCR_RFRST | SCFCR_TFRST);
+}
+
 static void sci_set_termios(struct uart_port *port, struct ktermios *termios,
 			    struct ktermios *old)
 {
 	struct sci_port *s = to_sci_port(port);
-	unsigned int status, baud, smr_val, max_baud;
+	unsigned int baud, smr_val, max_baud;
 	int t = -1;
 	u16 scfcr = 0;
 
@@ -1657,14 +1678,7 @@ static void sci_set_termios(struct uart_port *port, struct ktermios *termios,
 
 	sci_port_enable(s);
 
-	do {
-		status = sci_in(port, SCxSR);
-	} while (!(status & SCxSR_TEND(port)));
-
-	sci_out(port, SCSCR, 0x00);	/* TE=0, RE=0, CKE1=0 */
-
-	if (port->type != PORT_SCI)
-		sci_out(port, SCFCR, scfcr | SCFCR_RFRST | SCFCR_TFRST);
+	sci_reset(port);
 
 	smr_val = sci_in(port, SCSMR) & 3;
 
@@ -2037,7 +2051,8 @@ static int __devinit serial_console_setup(struct console *co, char *options)
 	if (options)
 		uart_parse_options(options, &baud, &parity, &bits, &flow);
 
-	/* TODO: disable clock */
+	sci_port_disable(sci_port);
+
 	return uart_set_options(port, co, baud, parity, bits, flow);
 }
 
@@ -2080,6 +2095,36 @@ static int __devinit sci_probe_earlyprintk(struct platform_device *pdev)
 	return 0;
 }
 
+#define uart_console(port)	((port)->cons->index == (port)->line)
+
+static int sci_runtime_suspend(struct device *dev)
+{
+	struct sci_port *sci_port = dev_get_drvdata(dev);
+	struct uart_port *port = &sci_port->port;
+
+	if (uart_console(port)) {
+		sci_port->saved_smr = sci_in(port, SCSMR);
+		sci_port->saved_brr = sci_in(port, SCBRR);
+		sci_port->saved_fcr = sci_in(port, SCFCR);
+	}
+	return 0;
+}
+
+static int sci_runtime_resume(struct device *dev)
+{
+	struct sci_port *sci_port = dev_get_drvdata(dev);
+	struct uart_port *port = &sci_port->port;
+
+	if (uart_console(port)) {
+		sci_reset(port);
+		sci_out(port, SCSMR, sci_port->saved_smr);
+		sci_out(port, SCBRR, sci_port->saved_brr);
+		sci_out(port, SCFCR, sci_port->saved_fcr);
+		sci_out(port, SCSCR, sci_port->cfg->scscr);
+	}
+	return 0;
+}
+
 #define SCI_CONSOLE	(&serial_console)
 
 #else
@@ -2089,6 +2134,8 @@ static inline int __devinit sci_probe_earlyprintk(struct platform_device *pdev)
 }
 
 #define SCI_CONSOLE	NULL
+#define sci_runtime_suspend	NULL
+#define sci_runtime_resume	NULL
 
 #endif /* CONFIG_SERIAL_SH_SCI_CONSOLE */
 
@@ -2204,6 +2251,8 @@ static int sci_resume(struct device *dev)
 }
 
 static const struct dev_pm_ops sci_dev_pm_ops = {
+	.runtime_suspend = sci_runtime_suspend,
+	.runtime_resume = sci_runtime_resume,
 	.suspend	= sci_suspend,
 	.resume		= sci_resume,
 };
diff --git a/drivers/video/backlight/backlight.c b/drivers/video/backlight/backlight.c
index 80d292fb92d..7363c1b169e 100644
--- a/drivers/video/backlight/backlight.c
+++ b/drivers/video/backlight/backlight.c
@@ -19,7 +19,7 @@
 #include <asm/backlight.h>
 #endif
 
-static const char const *backlight_types[] = {
+static const char *const backlight_types[] = {
 	[BACKLIGHT_RAW] = "raw",
 	[BACKLIGHT_PLATFORM] = "platform",
 	[BACKLIGHT_FIRMWARE] = "firmware",
diff --git a/fs/9p/v9fs_vfs.h b/fs/9p/v9fs_vfs.h
index 46ce357ca1a..410ffd6ceb5 100644
--- a/fs/9p/v9fs_vfs.h
+++ b/fs/9p/v9fs_vfs.h
@@ -54,9 +54,9 @@ extern struct kmem_cache *v9fs_inode_cache;
 
 struct inode *v9fs_alloc_inode(struct super_block *sb);
 void v9fs_destroy_inode(struct inode *inode);
-struct inode *v9fs_get_inode(struct super_block *sb, int mode);
+struct inode *v9fs_get_inode(struct super_block *sb, int mode, dev_t);
 int v9fs_init_inode(struct v9fs_session_info *v9ses,
-		    struct inode *inode, int mode);
+		    struct inode *inode, int mode, dev_t);
 void v9fs_evict_inode(struct inode *inode);
 ino_t v9fs_qid2ino(struct p9_qid *qid);
 void v9fs_stat2inode(struct p9_wstat *, struct inode *, struct super_block *);
@@ -83,4 +83,6 @@ static inline void v9fs_invalidate_inode_attr(struct inode *inode)
 	v9inode->cache_validity |= V9FS_INO_INVALID_ATTR;
 	return;
 }
+
+int v9fs_open_to_dotl_flags(int flags);
 #endif
diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c
index 3c173fcc2c5..62857a810a7 100644
--- a/fs/9p/vfs_file.c
+++ b/fs/9p/vfs_file.c
@@ -65,7 +65,7 @@ int v9fs_file_open(struct inode *inode, struct file *file)
 	v9inode = V9FS_I(inode);
 	v9ses = v9fs_inode2v9ses(inode);
 	if (v9fs_proto_dotl(v9ses))
-		omode = file->f_flags;
+		omode = v9fs_open_to_dotl_flags(file->f_flags);
 	else
 		omode = v9fs_uflags2omode(file->f_flags,
 					v9fs_proto_dotu(v9ses));
@@ -169,7 +169,18 @@ static int v9fs_file_do_lock(struct file *filp, int cmd, struct file_lock *fl)
 
 	/* convert posix lock to p9 tlock args */
 	memset(&flock, 0, sizeof(flock));
-	flock.type = fl->fl_type;
+	/* map the lock type */
+	switch (fl->fl_type) {
+	case F_RDLCK:
+		flock.type = P9_LOCK_TYPE_RDLCK;
+		break;
+	case F_WRLCK:
+		flock.type = P9_LOCK_TYPE_WRLCK;
+		break;
+	case F_UNLCK:
+		flock.type = P9_LOCK_TYPE_UNLCK;
+		break;
+	}
 	flock.start = fl->fl_start;
 	if (fl->fl_end == OFFSET_MAX)
 		flock.length = 0;
@@ -245,7 +256,7 @@ static int v9fs_file_getlock(struct file *filp, struct file_lock *fl)
 
 	/* convert posix lock to p9 tgetlock args */
 	memset(&glock, 0, sizeof(glock));
-	glock.type = fl->fl_type;
+	glock.type  = P9_LOCK_TYPE_UNLCK;
 	glock.start = fl->fl_start;
 	if (fl->fl_end == OFFSET_MAX)
 		glock.length = 0;
@@ -257,17 +268,26 @@ static int v9fs_file_getlock(struct file *filp, struct file_lock *fl)
 	res = p9_client_getlock_dotl(fid, &glock);
 	if (res < 0)
 		return res;
-	if (glock.type != F_UNLCK) {
-		fl->fl_type = glock.type;
+	/* map 9p lock type to os lock type */
+	switch (glock.type) {
+	case P9_LOCK_TYPE_RDLCK:
+		fl->fl_type = F_RDLCK;
+		break;
+	case P9_LOCK_TYPE_WRLCK:
+		fl->fl_type = F_WRLCK;
+		break;
+	case P9_LOCK_TYPE_UNLCK:
+		fl->fl_type = F_UNLCK;
+		break;
+	}
+	if (glock.type != P9_LOCK_TYPE_UNLCK) {
 		fl->fl_start = glock.start;
 		if (glock.length == 0)
 			fl->fl_end = OFFSET_MAX;
 		else
 			fl->fl_end = glock.start + glock.length - 1;
 		fl->fl_pid = glock.proc_id;
-	} else
-		fl->fl_type = F_UNLCK;
-
+	}
 	return res;
 }
 
diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c
index 8bb5507e822..e3c03db3c78 100644
--- a/fs/9p/vfs_inode.c
+++ b/fs/9p/vfs_inode.c
@@ -95,15 +95,18 @@ static int unixmode2p9mode(struct v9fs_session_info *v9ses, int mode)
 /**
  * p9mode2unixmode- convert plan9 mode bits to unix mode bits
  * @v9ses: v9fs session information
- * @mode: mode to convert
+ * @stat: p9_wstat from which mode need to be derived
+ * @rdev: major number, minor number in case of device files.
  *
  */
-
-static int p9mode2unixmode(struct v9fs_session_info *v9ses, int mode)
+static int p9mode2unixmode(struct v9fs_session_info *v9ses,
+			   struct p9_wstat *stat, dev_t *rdev)
 {
 	int res;
+	int mode = stat->mode;
 
-	res = mode & 0777;
+	res = mode & S_IALLUGO;
+	*rdev = 0;
 
 	if ((mode & P9_DMDIR) == P9_DMDIR)
 		res |= S_IFDIR;
@@ -116,9 +119,26 @@ static int p9mode2unixmode(struct v9fs_session_info *v9ses, int mode)
 		 && (v9ses->nodev == 0))
 		res |= S_IFIFO;
 	else if ((mode & P9_DMDEVICE) && (v9fs_proto_dotu(v9ses))
-		 && (v9ses->nodev == 0))
-		res |= S_IFBLK;
-	else
+		 && (v9ses->nodev == 0)) {
+		char type = 0, ext[32];
+		int major = -1, minor = -1;
+
+		strncpy(ext, stat->extension, sizeof(ext));
+		sscanf(ext, "%c %u %u", &type, &major, &minor);
+		switch (type) {
+		case 'c':
+			res |= S_IFCHR;
+			break;
+		case 'b':
+			res |= S_IFBLK;
+			break;
+		default:
+			P9_DPRINTK(P9_DEBUG_ERROR,
+				"Unknown special type %c %s\n", type,
+				stat->extension);
+		};
+		*rdev = MKDEV(major, minor);
+	} else
 		res |= S_IFREG;
 
 	if (v9fs_proto_dotu(v9ses)) {
@@ -131,7 +151,6 @@ static int p9mode2unixmode(struct v9fs_session_info *v9ses, int mode)
 		if ((mode & P9_DMSETVTX) == P9_DMSETVTX)
 			res |= S_ISVTX;
 	}
-
 	return res;
 }
 
@@ -242,13 +261,13 @@ void v9fs_destroy_inode(struct inode *inode)
 }
 
 int v9fs_init_inode(struct v9fs_session_info *v9ses,
-		    struct inode *inode, int mode)
+		    struct inode *inode, int mode, dev_t rdev)
 {
 	int err = 0;
 
 	inode_init_owner(inode, NULL, mode);
 	inode->i_blocks = 0;
-	inode->i_rdev = 0;
+	inode->i_rdev = rdev;
 	inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME;
 	inode->i_mapping->a_ops = &v9fs_addr_operations;
 
@@ -335,7 +354,7 @@ error:
  *
  */
 
-struct inode *v9fs_get_inode(struct super_block *sb, int mode)
+struct inode *v9fs_get_inode(struct super_block *sb, int mode, dev_t rdev)
 {
 	int err;
 	struct inode *inode;
@@ -348,7 +367,7 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode)
 		P9_EPRINTK(KERN_WARNING, "Problem allocating inode\n");
 		return ERR_PTR(-ENOMEM);
 	}
-	err = v9fs_init_inode(v9ses, inode, mode);
+	err = v9fs_init_inode(v9ses, inode, mode, rdev);
 	if (err) {
 		iput(inode);
 		return ERR_PTR(err);
@@ -435,11 +454,12 @@ void v9fs_evict_inode(struct inode *inode)
 static int v9fs_test_inode(struct inode *inode, void *data)
 {
 	int umode;
+	dev_t rdev;
 	struct v9fs_inode *v9inode = V9FS_I(inode);
 	struct p9_wstat *st = (struct p9_wstat *)data;
 	struct v9fs_session_info *v9ses = v9fs_inode2v9ses(inode);
 
-	umode = p9mode2unixmode(v9ses, st->mode);
+	umode = p9mode2unixmode(v9ses, st, &rdev);
 	/* don't match inode of different type */
 	if ((inode->i_mode & S_IFMT) != (umode & S_IFMT))
 		return 0;
@@ -473,6 +493,7 @@ static struct inode *v9fs_qid_iget(struct super_block *sb,
 				   struct p9_wstat *st,
 				   int new)
 {
+	dev_t rdev;
 	int retval, umode;
 	unsigned long i_ino;
 	struct inode *inode;
@@ -496,8 +517,8 @@ static struct inode *v9fs_qid_iget(struct super_block *sb,
 	 * later.
 	 */
 	inode->i_ino = i_ino;
-	umode = p9mode2unixmode(v9ses, st->mode);
-	retval = v9fs_init_inode(v9ses, inode, umode);
+	umode = p9mode2unixmode(v9ses, st, &rdev);
+	retval = v9fs_init_inode(v9ses, inode, umode, rdev);
 	if (retval)
 		goto error;
 
@@ -532,6 +553,19 @@ v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid,
 }
 
 /**
+ * v9fs_at_to_dotl_flags- convert Linux specific AT flags to
+ * plan 9 AT flag.
+ * @flags: flags to convert
+ */
+static int v9fs_at_to_dotl_flags(int flags)
+{
+	int rflags = 0;
+	if (flags & AT_REMOVEDIR)
+		rflags |= P9_DOTL_AT_REMOVEDIR;
+	return rflags;
+}
+
+/**
  * v9fs_remove - helper function to remove files and directories
  * @dir: directory inode that is being deleted
  * @dentry:  dentry that is being deleted
@@ -558,7 +592,8 @@ static int v9fs_remove(struct inode *dir, struct dentry *dentry, int flags)
 		return retval;
 	}
 	if (v9fs_proto_dotl(v9ses))
-		retval = p9_client_unlinkat(dfid, dentry->d_name.name, flags);
+		retval = p9_client_unlinkat(dfid, dentry->d_name.name,
+					    v9fs_at_to_dotl_flags(flags));
 	if (retval == -EOPNOTSUPP) {
 		/* Try the one based on path */
 		v9fid = v9fs_fid_clone(dentry);
@@ -645,13 +680,11 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir,
 		P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", err);
 		goto error;
 	}
-	d_instantiate(dentry, inode);
 	err = v9fs_fid_add(dentry, fid);
 	if (err < 0)
 		goto error;
-
+	d_instantiate(dentry, inode);
 	return ofid;
-
 error:
 	if (ofid)
 		p9_client_clunk(ofid);
@@ -792,6 +825,7 @@ static int v9fs_vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry,
 				      struct nameidata *nameidata)
 {
+	struct dentry *res;
 	struct super_block *sb;
 	struct v9fs_session_info *v9ses;
 	struct p9_fid *dfid, *fid;
@@ -823,22 +857,35 @@ struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry,
 
 		return ERR_PTR(result);
 	}
-
-	inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb);
+	/*
+	 * Make sure we don't use a wrong inode due to parallel
+	 * unlink. For cached mode create calls request for new
+	 * inode. But with cache disabled, lookup should do this.
+	 */
+	if (v9ses->cache)
+		inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb);
+	else
+		inode = v9fs_get_new_inode_from_fid(v9ses, fid, dir->i_sb);
 	if (IS_ERR(inode)) {
 		result = PTR_ERR(inode);
 		inode = NULL;
 		goto error;
 	}
-
 	result = v9fs_fid_add(dentry, fid);
 	if (result < 0)
 		goto error_iput;
-
 inst_out:
-	d_add(dentry, inode);
-	return NULL;
-
+	/*
+	 * If we had a rename on the server and a parallel lookup
+	 * for the new name, then make sure we instantiate with
+	 * the new name. ie look up for a/b, while on server somebody
+	 * moved b under k and client parallely did a lookup for
+	 * k/b.
+	 */
+	res = d_materialise_unique(dentry, inode);
+	if (!IS_ERR(res))
+		return res;
+	result = PTR_ERR(res);
 error_iput:
 	iput(inode);
 error:
@@ -1002,7 +1049,7 @@ v9fs_vfs_getattr(struct vfsmount *mnt, struct dentry *dentry,
 		return PTR_ERR(st);
 
 	v9fs_stat2inode(st, dentry->d_inode, dentry->d_inode->i_sb);
-		generic_fillattr(dentry->d_inode, stat);
+	generic_fillattr(dentry->d_inode, stat);
 
 	p9stat_free(st);
 	kfree(st);
@@ -1086,6 +1133,7 @@ void
 v9fs_stat2inode(struct p9_wstat *stat, struct inode *inode,
 	struct super_block *sb)
 {
+	mode_t mode;
 	char ext[32];
 	char tag_name[14];
 	unsigned int i_nlink;
@@ -1121,31 +1169,9 @@ v9fs_stat2inode(struct p9_wstat *stat, struct inode *inode,
 				inode->i_nlink = i_nlink;
 		}
 	}
-	inode->i_mode = p9mode2unixmode(v9ses, stat->mode);
-	if ((S_ISBLK(inode->i_mode)) || (S_ISCHR(inode->i_mode))) {
-		char type = 0;
-		int major = -1;
-		int minor = -1;
-
-		strncpy(ext, stat->extension, sizeof(ext));
-		sscanf(ext, "%c %u %u", &type, &major, &minor);
-		switch (type) {
-		case 'c':
-			inode->i_mode &= ~S_IFBLK;
-			inode->i_mode |= S_IFCHR;
-			break;
-		case 'b':
-			break;
-		default:
-			P9_DPRINTK(P9_DEBUG_ERROR,
-				"Unknown special type %c %s\n", type,
-				stat->extension);
-		};
-		inode->i_rdev = MKDEV(major, minor);
-		init_special_inode(inode, inode->i_mode, inode->i_rdev);
-	} else
-		inode->i_rdev = 0;
-
+	mode = stat->mode & S_IALLUGO;
+	mode |= inode->i_mode & ~S_IALLUGO;
+	inode->i_mode = mode;
 	i_size_write(inode, stat->length);
 
 	/* not real number of blocks, but 512 byte ones ... */
@@ -1411,6 +1437,8 @@ v9fs_vfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t rdev)
 
 int v9fs_refresh_inode(struct p9_fid *fid, struct inode *inode)
 {
+	int umode;
+	dev_t rdev;
 	loff_t i_size;
 	struct p9_wstat *st;
 	struct v9fs_session_info *v9ses;
@@ -1419,6 +1447,12 @@ int v9fs_refresh_inode(struct p9_fid *fid, struct inode *inode)
 	st = p9_client_stat(fid);
 	if (IS_ERR(st))
 		return PTR_ERR(st);
+	/*
+	 * Don't update inode if the file type is different
+	 */
+	umode = p9mode2unixmode(v9ses, st, &rdev);
+	if ((inode->i_mode & S_IFMT) != (umode & S_IFMT))
+		goto out;
 
 	spin_lock(&inode->i_lock);
 	/*
@@ -1430,6 +1464,7 @@ int v9fs_refresh_inode(struct p9_fid *fid, struct inode *inode)
 	if (v9ses->cache)
 		inode->i_size = i_size;
 	spin_unlock(&inode->i_lock);
+out:
 	p9stat_free(st);
 	kfree(st);
 	return 0;
diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c
index b6c8ed20519..aded79fcd5c 100644
--- a/fs/9p/vfs_inode_dotl.c
+++ b/fs/9p/vfs_inode_dotl.c
@@ -153,7 +153,8 @@ static struct inode *v9fs_qid_iget_dotl(struct super_block *sb,
 	 * later.
 	 */
 	inode->i_ino = i_ino;
-	retval = v9fs_init_inode(v9ses, inode, st->st_mode);
+	retval = v9fs_init_inode(v9ses, inode,
+				 st->st_mode, new_decode_dev(st->st_rdev));
 	if (retval)
 		goto error;
 
@@ -190,6 +191,58 @@ v9fs_inode_from_fid_dotl(struct v9fs_session_info *v9ses, struct p9_fid *fid,
 	return inode;
 }
 
+struct dotl_openflag_map {
+	int open_flag;
+	int dotl_flag;
+};
+
+static int v9fs_mapped_dotl_flags(int flags)
+{
+	int i;
+	int rflags = 0;
+	struct dotl_openflag_map dotl_oflag_map[] = {
+		{ O_CREAT,	P9_DOTL_CREATE },
+		{ O_EXCL,	P9_DOTL_EXCL },
+		{ O_NOCTTY,	P9_DOTL_NOCTTY },
+		{ O_TRUNC,	P9_DOTL_TRUNC },
+		{ O_APPEND,	P9_DOTL_APPEND },
+		{ O_NONBLOCK,	P9_DOTL_NONBLOCK },
+		{ O_DSYNC,	P9_DOTL_DSYNC },
+		{ FASYNC,	P9_DOTL_FASYNC },
+		{ O_DIRECT,	P9_DOTL_DIRECT },
+		{ O_LARGEFILE,	P9_DOTL_LARGEFILE },
+		{ O_DIRECTORY,	P9_DOTL_DIRECTORY },
+		{ O_NOFOLLOW,	P9_DOTL_NOFOLLOW },
+		{ O_NOATIME,	P9_DOTL_NOATIME },
+		{ O_CLOEXEC,	P9_DOTL_CLOEXEC },
+		{ O_SYNC,	P9_DOTL_SYNC},
+	};
+	for (i = 0; i < ARRAY_SIZE(dotl_oflag_map); i++) {
+		if (flags & dotl_oflag_map[i].open_flag)
+			rflags |= dotl_oflag_map[i].dotl_flag;
+	}
+	return rflags;
+}
+
+/**
+ * v9fs_open_to_dotl_flags- convert Linux specific open flags to
+ * plan 9 open flag.
+ * @flags: flags to convert
+ */
+int v9fs_open_to_dotl_flags(int flags)
+{
+	int rflags = 0;
+
+	/*
+	 * We have same bits for P9_DOTL_READONLY, P9_DOTL_WRONLY
+	 * and P9_DOTL_NOACCESS
+	 */
+	rflags |= flags & O_ACCMODE;
+	rflags |= v9fs_mapped_dotl_flags(flags);
+
+	return rflags;
+}
+
 /**
  * v9fs_vfs_create_dotl - VFS hook to create files for 9P2000.L protocol.
  * @dir: directory inode that is being created
@@ -258,7 +311,8 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode,
 			   "Failed to get acl values in creat %d\n", err);
 		goto error;
 	}
-	err = p9_client_create_dotl(ofid, name, flags, mode, gid, &qid);
+	err = p9_client_create_dotl(ofid, name, v9fs_open_to_dotl_flags(flags),
+				    mode, gid, &qid);
 	if (err < 0) {
 		P9_DPRINTK(P9_DEBUG_VFS,
 				"p9_client_open_dotl failed in creat %d\n",
@@ -281,10 +335,10 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode,
 		P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", err);
 		goto error;
 	}
-	d_instantiate(dentry, inode);
 	err = v9fs_fid_add(dentry, fid);
 	if (err < 0)
 		goto error;
+	d_instantiate(dentry, inode);
 
 	/* Now set the ACL based on the default value */
 	v9fs_set_create_acl(dentry, &dacl, &pacl);
@@ -403,10 +457,10 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir,
 				err);
 			goto error;
 		}
-		d_instantiate(dentry, inode);
 		err = v9fs_fid_add(dentry, fid);
 		if (err < 0)
 			goto error;
+		d_instantiate(dentry, inode);
 		fid = NULL;
 	} else {
 		/*
@@ -414,7 +468,7 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir,
 		 * inode with stat. We need to get an inode
 		 * so that we can set the acl with dentry
 		 */
-		inode = v9fs_get_inode(dir->i_sb, mode);
+		inode = v9fs_get_inode(dir->i_sb, mode, 0);
 		if (IS_ERR(inode)) {
 			err = PTR_ERR(inode);
 			goto error;
@@ -540,6 +594,7 @@ int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr)
 void
 v9fs_stat2inode_dotl(struct p9_stat_dotl *stat, struct inode *inode)
 {
+	mode_t mode;
 	struct v9fs_inode *v9inode = V9FS_I(inode);
 
 	if ((stat->st_result_mask & P9_STATS_BASIC) == P9_STATS_BASIC) {
@@ -552,11 +607,10 @@ v9fs_stat2inode_dotl(struct p9_stat_dotl *stat, struct inode *inode)
 		inode->i_uid = stat->st_uid;
 		inode->i_gid = stat->st_gid;
 		inode->i_nlink = stat->st_nlink;
-		inode->i_mode = stat->st_mode;
-		inode->i_rdev = new_decode_dev(stat->st_rdev);
 
-		if ((S_ISBLK(inode->i_mode)) || (S_ISCHR(inode->i_mode)))
-			init_special_inode(inode, inode->i_mode, inode->i_rdev);
+		mode = stat->st_mode & S_IALLUGO;
+		mode |= inode->i_mode & ~S_IALLUGO;
+		inode->i_mode = mode;
 
 		i_size_write(inode, stat->st_size);
 		inode->i_blocks = stat->st_blocks;
@@ -657,14 +711,14 @@ v9fs_vfs_symlink_dotl(struct inode *dir, struct dentry *dentry,
 					err);
 			goto error;
 		}
-		d_instantiate(dentry, inode);
 		err = v9fs_fid_add(dentry, fid);
 		if (err < 0)
 			goto error;
+		d_instantiate(dentry, inode);
 		fid = NULL;
 	} else {
 		/* Not in cached mode. No need to populate inode with stat */
-		inode = v9fs_get_inode(dir->i_sb, S_IFLNK);
+		inode = v9fs_get_inode(dir->i_sb, S_IFLNK, 0);
 		if (IS_ERR(inode)) {
 			err = PTR_ERR(inode);
 			goto error;
@@ -810,17 +864,17 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, int omode,
 				err);
 			goto error;
 		}
-		d_instantiate(dentry, inode);
 		err = v9fs_fid_add(dentry, fid);
 		if (err < 0)
 			goto error;
+		d_instantiate(dentry, inode);
 		fid = NULL;
 	} else {
 		/*
 		 * Not in cached mode. No need to populate inode with stat.
 		 * socket syscall returns a fd, so we need instantiate
 		 */
-		inode = v9fs_get_inode(dir->i_sb, mode);
+		inode = v9fs_get_inode(dir->i_sb, mode, rdev);
 		if (IS_ERR(inode)) {
 			err = PTR_ERR(inode);
 			goto error;
@@ -886,6 +940,11 @@ int v9fs_refresh_inode_dotl(struct p9_fid *fid, struct inode *inode)
 	st = p9_client_getattr_dotl(fid, P9_STATS_ALL);
 	if (IS_ERR(st))
 		return PTR_ERR(st);
+	/*
+	 * Don't update inode if the file type is different
+	 */
+	if ((inode->i_mode & S_IFMT) != (st->st_mode & S_IFMT))
+		goto out;
 
 	spin_lock(&inode->i_lock);
 	/*
@@ -897,6 +956,7 @@ int v9fs_refresh_inode_dotl(struct p9_fid *fid, struct inode *inode)
 	if (v9ses->cache)
 		inode->i_size = i_size;
 	spin_unlock(&inode->i_lock);
+out:
 	kfree(st);
 	return 0;
 }
diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c
index feef6cdc1fd..c70251d47ed 100644
--- a/fs/9p/vfs_super.c
+++ b/fs/9p/vfs_super.c
@@ -149,7 +149,7 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags,
 	else
 		sb->s_d_op = &v9fs_dentry_operations;
 
-	inode = v9fs_get_inode(sb, S_IFDIR | mode);
+	inode = v9fs_get_inode(sb, S_IFDIR | mode, 0);
 	if (IS_ERR(inode)) {
 		retval = PTR_ERR(inode);
 		goto release_sb;
diff --git a/fs/block_dev.c b/fs/block_dev.c
index ff77262e887..95f786ec7f0 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -1429,6 +1429,11 @@ static int __blkdev_put(struct block_device *bdev, fmode_t mode, int for_part)
 		WARN_ON_ONCE(bdev->bd_holders);
 		sync_blockdev(bdev);
 		kill_bdev(bdev);
+		/* ->release can cause the old bdi to disappear,
+		 * so must switch it out first
+		 */
+		bdev_inode_switch_bdi(bdev->bd_inode,
+					&default_backing_dev_info);
 	}
 	if (bdev->bd_contains == bdev) {
 		if (disk->fops->release)
@@ -1442,8 +1447,6 @@ static int __blkdev_put(struct block_device *bdev, fmode_t mode, int for_part)
 		disk_put_part(bdev->bd_part);
 		bdev->bd_part = NULL;
 		bdev->bd_disk = NULL;
-		bdev_inode_switch_bdi(bdev->bd_inode,
-					&default_backing_dev_info);
 		if (bdev != bdev->bd_contains)
 			victim = bdev->bd_contains;
 		bdev->bd_contains = NULL;
diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h
index 502b9e98867..d9f99a16edd 100644
--- a/fs/btrfs/btrfs_inode.h
+++ b/fs/btrfs/btrfs_inode.h
@@ -176,7 +176,11 @@ static inline u64 btrfs_ino(struct inode *inode)
 {
 	u64 ino = BTRFS_I(inode)->location.objectid;
 
-	if (ino <= BTRFS_FIRST_FREE_OBJECTID)
+	/*
+	 * !ino: btree_inode
+	 * type == BTRFS_ROOT_ITEM_KEY: subvol dir
+	 */
+	if (!ino || BTRFS_I(inode)->location.type == BTRFS_ROOT_ITEM_KEY)
 		ino = inode->i_ino;
 	return ino;
 }
diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c
index b910694f61e..a1cb7821bec 100644
--- a/fs/btrfs/file-item.c
+++ b/fs/btrfs/file-item.c
@@ -183,8 +183,10 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 	 * read from the commit root and sidestep a nasty deadlock
 	 * between reading the free space cache and updating the csum tree.
 	 */
-	if (btrfs_is_free_space_inode(root, inode))
+	if (btrfs_is_free_space_inode(root, inode)) {
 		path->search_commit_root = 1;
+		path->skip_locking = 1;
+	}
 
 	disk_bytenr = (u64)bio->bi_sector << 9;
 	if (dio)
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index e7872e485f1..3c3abff731a 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1075,12 +1075,6 @@ static noinline int prepare_pages(struct btrfs_root *root, struct file *file,
 	start_pos = pos & ~((u64)root->sectorsize - 1);
 	last_pos = ((u64)index + num_pages) << PAGE_CACHE_SHIFT;
 
-	if (start_pos > inode->i_size) {
-		err = btrfs_cont_expand(inode, i_size_read(inode), start_pos);
-		if (err)
-			return err;
-	}
-
 again:
 	for (i = 0; i < num_pages; i++) {
 		pages[i] = find_or_create_page(inode->i_mapping, index + i,
@@ -1338,6 +1332,7 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb,
 	struct inode *inode = fdentry(file)->d_inode;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	loff_t *ppos = &iocb->ki_pos;
+	u64 start_pos;
 	ssize_t num_written = 0;
 	ssize_t err = 0;
 	size_t count, ocount;
@@ -1386,6 +1381,15 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb,
 	file_update_time(file);
 	BTRFS_I(inode)->sequence++;
 
+	start_pos = round_down(pos, root->sectorsize);
+	if (start_pos > i_size_read(inode)) {
+		err = btrfs_cont_expand(inode, i_size_read(inode), start_pos);
+		if (err) {
+			mutex_unlock(&inode->i_mutex);
+			goto out;
+		}
+	}
+
 	if (unlikely(file->f_flags & O_DIRECT)) {
 		num_written = __btrfs_direct_write(iocb, iov, nr_segs,
 						   pos, ppos, count, ocount);
diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 6a265b9f85f..41ac927401d 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -190,9 +190,11 @@ int btrfs_truncate_free_space_cache(struct btrfs_root *root,
 				    struct btrfs_path *path,
 				    struct inode *inode)
 {
+	struct btrfs_block_rsv *rsv;
 	loff_t oldsize;
 	int ret = 0;
 
+	rsv = trans->block_rsv;
 	trans->block_rsv = root->orphan_block_rsv;
 	ret = btrfs_block_rsv_check(trans, root,
 				    root->orphan_block_rsv,
@@ -210,6 +212,8 @@ int btrfs_truncate_free_space_cache(struct btrfs_root *root,
 	 */
 	ret = btrfs_truncate_inode_items(trans, root, inode,
 					 0, BTRFS_EXTENT_DATA_KEY);
+
+	trans->block_rsv = rsv;
 	if (ret) {
 		WARN_ON(1);
 		return ret;
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 0ccc7438ad3..4d14de6d121 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -1786,7 +1786,7 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
 			  &ordered_extent->list);
 
 	ret = btrfs_ordered_update_i_size(inode, 0, ordered_extent);
-	if (!ret) {
+	if (!ret || !test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags)) {
 		ret = btrfs_update_inode(trans, root, inode);
 		BUG_ON(ret);
 	}
@@ -3510,15 +3510,19 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 			err = btrfs_drop_extents(trans, inode, cur_offset,
 						 cur_offset + hole_size,
 						 &hint_byte, 1);
-			if (err)
+			if (err) {
+				btrfs_end_transaction(trans, root);
 				break;
+			}
 
 			err = btrfs_insert_file_extent(trans, root,
 					btrfs_ino(inode), cur_offset, 0,
 					0, hole_size, 0, hole_size,
 					0, 0, 0);
-			if (err)
+			if (err) {
+				btrfs_end_transaction(trans, root);
 				break;
+			}
 
 			btrfs_drop_extent_cache(inode, hole_start,
 					last_byte - 1, 0);
@@ -3952,7 +3956,6 @@ struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location,
 			 struct btrfs_root *root, int *new)
 {
 	struct inode *inode;
-	int bad_inode = 0;
 
 	inode = btrfs_iget_locked(s, location->objectid, root);
 	if (!inode)
@@ -3968,15 +3971,12 @@ struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location,
 			if (new)
 				*new = 1;
 		} else {
-			bad_inode = 1;
+			unlock_new_inode(inode);
+			iput(inode);
+			inode = ERR_PTR(-ESTALE);
 		}
 	}
 
-	if (bad_inode) {
-		iput(inode);
-		inode = ERR_PTR(-ESTALE);
-	}
-
 	return inode;
 }
 
@@ -5823,7 +5823,7 @@ again:
 
 	add_pending_csums(trans, inode, ordered->file_offset, &ordered->list);
 	ret = btrfs_ordered_update_i_size(inode, 0, ordered);
-	if (!ret)
+	if (!ret || !test_bit(BTRFS_ORDERED_PREALLOC, &ordered->flags))
 		btrfs_update_inode(trans, root, inode);
 	ret = 0;
 out_unlock:
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 970977aab22..3351b1b2457 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -2220,6 +2220,12 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
 	    !IS_ALIGNED(destoff, bs))
 		goto out_unlock;
 
+	if (destoff > inode->i_size) {
+		ret = btrfs_cont_expand(inode, inode->i_size, destoff);
+		if (ret)
+			goto out_unlock;
+	}
+
 	/* do any pending delalloc/csum calc on src, one way or
 	   another, and lock file content */
 	while (1) {
@@ -2325,14 +2331,21 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
 
 			if (type == BTRFS_FILE_EXTENT_REG ||
 			    type == BTRFS_FILE_EXTENT_PREALLOC) {
+				/*
+				 *    a  | --- range to clone ---|  b
+				 * | ------------- extent ------------- |
+				 */
+
+				/* substract range b */
+				if (key.offset + datal > off + len)
+					datal = off + len - key.offset;
+
+				/* substract range a */
 				if (off > key.offset) {
 					datao += off - key.offset;
 					datal -= off - key.offset;
 				}
 
-				if (key.offset + datal > off + len)
-					datal = off + len - key.offset;
-
 				ret = btrfs_drop_extents(trans, inode,
 							 new_key.offset,
 							 new_key.offset + datal,
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 7dc36fab4af..e24b7964a15 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -884,6 +884,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	struct btrfs_root *tree_root = fs_info->tree_root;
 	struct btrfs_root *root = pending->root;
 	struct btrfs_root *parent_root;
+	struct btrfs_block_rsv *rsv;
 	struct inode *parent_inode;
 	struct dentry *parent;
 	struct dentry *dentry;
@@ -895,6 +896,8 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	u64 objectid;
 	u64 root_flags;
 
+	rsv = trans->block_rsv;
+
 	new_root_item = kmalloc(sizeof(*new_root_item), GFP_NOFS);
 	if (!new_root_item) {
 		pending->error = -ENOMEM;
@@ -1002,6 +1005,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	btrfs_orphan_post_snapshot(trans, pending);
 fail:
 	kfree(new_root_item);
+	trans->block_rsv = rsv;
 	btrfs_block_rsv_release(root, &pending->block_rsv, (u64)-1);
 	return 0;
 }
diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c
index d733b9cfea3..69565e5fc6a 100644
--- a/fs/btrfs/xattr.c
+++ b/fs/btrfs/xattr.c
@@ -116,6 +116,12 @@ static int do_setxattr(struct btrfs_trans_handle *trans,
 		if (ret)
 			goto out;
 		btrfs_release_path(path);
+
+		/*
+		 * remove the attribute
+		 */
+		if (!value)
+			goto out;
 	}
 
 again:
@@ -158,6 +164,9 @@ out:
 	return ret;
 }
 
+/*
+ * @value: "" makes the attribute to empty, NULL removes it
+ */
 int __btrfs_setxattr(struct btrfs_trans_handle *trans,
 		     struct inode *inode, const char *name,
 		     const void *value, size_t size, int flags)
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index fee028b5332..86c59e16ba7 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -1595,7 +1595,7 @@ static int set_request_path_attr(struct inode *rinode, struct dentry *rdentry,
 		r = build_dentry_path(rdentry, ppath, pathlen, ino, freepath);
 		dout(" dentry %p %llx/%.*s\n", rdentry, *ino, *pathlen,
 		     *ppath);
-	} else if (rpath) {
+	} else if (rpath || rino) {
 		*ino = rino;
 		*ppath = rpath;
 		*pathlen = strlen(rpath);
diff --git a/fs/ceph/super.c b/fs/ceph/super.c
index d47c5ec7fb1..88bacaf385d 100644
--- a/fs/ceph/super.c
+++ b/fs/ceph/super.c
@@ -813,8 +813,8 @@ static struct dentry *ceph_mount(struct file_system_type *fs_type,
 	fsc = create_fs_client(fsopt, opt);
 	if (IS_ERR(fsc)) {
 		res = ERR_CAST(fsc);
-		kfree(fsopt);
-		kfree(opt);
+		destroy_mount_options(fsopt);
+		ceph_destroy_options(opt);
 		goto out_final;
 	}
 
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index e717dfd2f2b..b7d7bd0f066 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -175,6 +175,7 @@ struct mpage_da_data {
  */
 #define	EXT4_IO_END_UNWRITTEN	0x0001
 #define EXT4_IO_END_ERROR	0x0002
+#define EXT4_IO_END_QUEUED	0x0004
 
 struct ext4_io_page {
 	struct page	*p_page;
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index c4da98a959a..18d2558b762 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -121,9 +121,6 @@ void ext4_evict_inode(struct inode *inode)
 
 	trace_ext4_evict_inode(inode);
 
-	mutex_lock(&inode->i_mutex);
-	ext4_flush_completed_IO(inode);
-	mutex_unlock(&inode->i_mutex);
 	ext4_ioend_wait(inode);
 
 	if (inode->i_nlink) {
diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c
index 78839af7ce2..92f38ee13f8 100644
--- a/fs/ext4/page-io.c
+++ b/fs/ext4/page-io.c
@@ -142,7 +142,23 @@ static void ext4_end_io_work(struct work_struct *work)
 	unsigned long		flags;
 	int			ret;
 
-	mutex_lock(&inode->i_mutex);
+	if (!mutex_trylock(&inode->i_mutex)) {
+		/*
+		 * Requeue the work instead of waiting so that the work
+		 * items queued after this can be processed.
+		 */
+		queue_work(EXT4_SB(inode->i_sb)->dio_unwritten_wq, &io->work);
+		/*
+		 * To prevent the ext4-dio-unwritten thread from keeping
+		 * requeueing end_io requests and occupying cpu for too long,
+		 * yield the cpu if it sees an end_io request that has already
+		 * been requeued.
+		 */
+		if (io->flag & EXT4_IO_END_QUEUED)
+			yield();
+		io->flag |= EXT4_IO_END_QUEUED;
+		return;
+	}
 	ret = ext4_end_io_nolock(io);
 	if (ret < 0) {
 		mutex_unlock(&inode->i_mutex);
diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c
index 168a80f7f12..5cb8614508c 100644
--- a/fs/fuse/dev.c
+++ b/fs/fuse/dev.c
@@ -258,10 +258,14 @@ void fuse_queue_forget(struct fuse_conn *fc, struct fuse_forget_link *forget,
 	forget->forget_one.nlookup = nlookup;
 
 	spin_lock(&fc->lock);
-	fc->forget_list_tail->next = forget;
-	fc->forget_list_tail = forget;
-	wake_up(&fc->waitq);
-	kill_fasync(&fc->fasync, SIGIO, POLL_IN);
+	if (fc->connected) {
+		fc->forget_list_tail->next = forget;
+		fc->forget_list_tail = forget;
+		wake_up(&fc->waitq);
+		kill_fasync(&fc->fasync, SIGIO, POLL_IN);
+	} else {
+		kfree(forget);
+	}
 	spin_unlock(&fc->lock);
 }
 
diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c
index 12b502929da..add96f6ffda 100644
--- a/fs/fuse/inode.c
+++ b/fs/fuse/inode.c
@@ -812,6 +812,9 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req)
 			if (arg->minor >= 17) {
 				if (!(arg->flags & FUSE_FLOCK_LOCKS))
 					fc->no_flock = 1;
+			} else {
+				if (!(arg->flags & FUSE_POSIX_LOCKS))
+					fc->no_flock = 1;
 			}
 			if (arg->flags & FUSE_ATOMIC_O_TRUNC)
 				fc->atomic_o_trunc = 1;
diff --git a/fs/namei.c b/fs/namei.c
index 2826db35dc2..b52bc685465 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -727,25 +727,22 @@ static int follow_automount(struct path *path, unsigned flags,
 	if ((flags & LOOKUP_NO_AUTOMOUNT) && !(flags & LOOKUP_PARENT))
 		return -EISDIR; /* we actually want to stop here */
 
-	/*
-	 * We don't want to mount if someone's just doing a stat and they've
-	 * set AT_SYMLINK_NOFOLLOW - unless they're stat'ing a directory and
-	 * appended a '/' to the name.
+	/* We don't want to mount if someone's just doing a stat -
+	 * unless they're stat'ing a directory and appended a '/' to
+	 * the name.
+	 *
+	 * We do, however, want to mount if someone wants to open or
+	 * create a file of any type under the mountpoint, wants to
+	 * traverse through the mountpoint or wants to open the
+	 * mounted directory.  Also, autofs may mark negative dentries
+	 * as being automount points.  These will need the attentions
+	 * of the daemon to instantiate them before they can be used.
 	 */
-	if (!(flags & LOOKUP_FOLLOW)) {
-		/* We do, however, want to mount if someone wants to open or
-		 * create a file of any type under the mountpoint, wants to
-		 * traverse through the mountpoint or wants to open the mounted
-		 * directory.
-		 * Also, autofs may mark negative dentries as being automount
-		 * points.  These will need the attentions of the daemon to
-		 * instantiate them before they can be used.
-		 */
-		if (!(flags & (LOOKUP_PARENT | LOOKUP_DIRECTORY |
-			     LOOKUP_OPEN | LOOKUP_CREATE)) &&
-		    path->dentry->d_inode)
-			return -EISDIR;
-	}
+	if (!(flags & (LOOKUP_PARENT | LOOKUP_DIRECTORY |
+		     LOOKUP_OPEN | LOOKUP_CREATE)) &&
+	    path->dentry->d_inode)
+		return -EISDIR;
+
 	current->total_link_count++;
 	if (current->total_link_count >= 40)
 		return -ELOOP;
diff --git a/fs/ubifs/debug.h b/fs/ubifs/debug.h
index 45174b53437..feb361e252a 100644
--- a/fs/ubifs/debug.h
+++ b/fs/ubifs/debug.h
@@ -335,9 +335,9 @@ void dbg_debugfs_exit_fs(struct ubifs_info *c);
 #define DBGKEY(key)  ((char *)(key))
 #define DBGKEY1(key) ((char *)(key))
 
-#define ubifs_dbg_msg(fmt, ...) do {               \
-	if (0)                                     \
-		pr_debug(fmt "\n", ##__VA_ARGS__); \
+#define ubifs_dbg_msg(fmt, ...) do {                        \
+	if (0)                                              \
+		printk(KERN_DEBUG fmt "\n", ##__VA_ARGS__); \
 } while (0)
 
 #define dbg_dump_stack()
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
index b9c172b3fbb..673704fab74 100644
--- a/fs/xfs/xfs_iops.c
+++ b/fs/xfs/xfs_iops.c
@@ -70,9 +70,8 @@ xfs_synchronize_times(
 }
 
 /*
- * If the linux inode is valid, mark it dirty.
- * Used when committing a dirty inode into a transaction so that
- * the inode will get written back by the linux code
+ * If the linux inode is valid, mark it dirty, else mark the dirty state
+ * in the XFS inode to make sure we pick it up when reclaiming the inode.
  */
 void
 xfs_mark_inode_dirty_sync(
@@ -82,6 +81,10 @@ xfs_mark_inode_dirty_sync(
 
 	if (!(inode->i_state & (I_WILL_FREE|I_FREEING)))
 		mark_inode_dirty_sync(inode);
+	else {
+		barrier();
+		ip->i_update_core = 1;
+	}
 }
 
 void
@@ -92,6 +95,11 @@ xfs_mark_inode_dirty(
 
 	if (!(inode->i_state & (I_WILL_FREE|I_FREEING)))
 		mark_inode_dirty(inode);
+	else {
+		barrier();
+		ip->i_update_core = 1;
+	}
+
 }
 
 /*
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index 9a72dda58bd..2366c54cc4f 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -356,6 +356,8 @@ xfs_parseargs(
 			mp->m_flags |= XFS_MOUNT_DELAYLOG;
 		} else if (!strcmp(this_char, MNTOPT_NODELAYLOG)) {
 			mp->m_flags &= ~XFS_MOUNT_DELAYLOG;
+			xfs_warn(mp,
+	"nodelaylog is deprecated and will be removed in Linux 3.3");
 		} else if (!strcmp(this_char, MNTOPT_DISCARD)) {
 			mp->m_flags |= XFS_MOUNT_DISCARD;
 		} else if (!strcmp(this_char, MNTOPT_NODISCARD)) {
@@ -877,33 +879,17 @@ xfs_log_inode(
 	struct xfs_trans	*tp;
 	int			error;
 
-	xfs_iunlock(ip, XFS_ILOCK_SHARED);
 	tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS);
 	error = xfs_trans_reserve(tp, 0, XFS_FSYNC_TS_LOG_RES(mp), 0, 0, 0);
-
 	if (error) {
 		xfs_trans_cancel(tp, 0);
-		/* we need to return with the lock hold shared */
-		xfs_ilock(ip, XFS_ILOCK_SHARED);
 		return error;
 	}
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
-
-	/*
-	 * Note - it's possible that we might have pushed ourselves out of the
-	 * way during trans_reserve which would flush the inode.  But there's
-	 * no guarantee that the inode buffer has actually gone out yet (it's
-	 * delwri).  Plus the buffer could be pinned anyway if it's part of
-	 * an inode in another recent transaction.  So we play it safe and
-	 * fire off the transaction anyway.
-	 */
-	xfs_trans_ijoin(tp, ip);
+	xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL);
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-	error = xfs_trans_commit(tp, 0);
-	xfs_ilock_demote(ip, XFS_ILOCK_EXCL);
-
-	return error;
+	return xfs_trans_commit(tp, 0);
 }
 
 STATIC int
@@ -918,7 +904,9 @@ xfs_fs_write_inode(
 	trace_xfs_write_inode(ip);
 
 	if (XFS_FORCED_SHUTDOWN(mp))
-		return XFS_ERROR(EIO);
+		return -XFS_ERROR(EIO);
+	if (!ip->i_update_core)
+		return 0;
 
 	if (wbc->sync_mode == WB_SYNC_ALL) {
 		/*
@@ -929,12 +917,10 @@ xfs_fs_write_inode(
 		 * of synchronous log foces dramatically.
 		 */
 		xfs_ioend_wait(ip);
-		xfs_ilock(ip, XFS_ILOCK_SHARED);
-		if (ip->i_update_core) {
-			error = xfs_log_inode(ip);
-			if (error)
-				goto out_unlock;
-		}
+		error = xfs_log_inode(ip);
+		if (error)
+			goto out;
+		return 0;
 	} else {
 		/*
 		 * We make this non-blocking if the inode is contended, return
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 245bafdafd5..c816075c01c 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -944,8 +944,10 @@ extern void perf_pmu_unregister(struct pmu *pmu);
 
 extern int perf_num_counters(void);
 extern const char *perf_pmu_name(void);
-extern void __perf_event_task_sched_in(struct task_struct *task);
-extern void __perf_event_task_sched_out(struct task_struct *task, struct task_struct *next);
+extern void __perf_event_task_sched_in(struct task_struct *prev,
+				       struct task_struct *task);
+extern void __perf_event_task_sched_out(struct task_struct *prev,
+					struct task_struct *next);
 extern int perf_event_init_task(struct task_struct *child);
 extern void perf_event_exit_task(struct task_struct *child);
 extern void perf_event_free_task(struct task_struct *task);
@@ -1059,17 +1061,20 @@ perf_sw_event(u32 event_id, u64 nr, struct pt_regs *regs, u64 addr)
 
 extern struct jump_label_key perf_sched_events;
 
-static inline void perf_event_task_sched_in(struct task_struct *task)
+static inline void perf_event_task_sched_in(struct task_struct *prev,
+					    struct task_struct *task)
 {
 	if (static_branch(&perf_sched_events))
-		__perf_event_task_sched_in(task);
+		__perf_event_task_sched_in(prev, task);
 }
 
-static inline void perf_event_task_sched_out(struct task_struct *task, struct task_struct *next)
+static inline void perf_event_task_sched_out(struct task_struct *prev,
+					     struct task_struct *next)
 {
 	perf_sw_event(PERF_COUNT_SW_CONTEXT_SWITCHES, 1, NULL, 0);
 
-	__perf_event_task_sched_out(task, next);
+	if (static_branch(&perf_sched_events))
+		__perf_event_task_sched_out(prev, next);
 }
 
 extern void perf_event_mmap(struct vm_area_struct *vma);
@@ -1139,10 +1144,11 @@ extern void perf_event_disable(struct perf_event *event);
 extern void perf_event_task_tick(void);
 #else
 static inline void
-perf_event_task_sched_in(struct task_struct *task)			{ }
+perf_event_task_sched_in(struct task_struct *prev,
+			 struct task_struct *task)			{ }
 static inline void
-perf_event_task_sched_out(struct task_struct *task,
-			    struct task_struct *next)			{ }
+perf_event_task_sched_out(struct task_struct *prev,
+			  struct task_struct *next)			{ }
 static inline int perf_event_init_task(struct task_struct *child)	{ return 0; }
 static inline void perf_event_exit_task(struct task_struct *child)	{ }
 static inline void perf_event_free_task(struct task_struct *task)	{ }
diff --git a/include/linux/regulator/consumer.h b/include/linux/regulator/consumer.h
index 26f6ea4444e..b47771aa571 100644
--- a/include/linux/regulator/consumer.h
+++ b/include/linux/regulator/consumer.h
@@ -123,7 +123,7 @@ struct regulator_bulk_data {
 	const char *supply;
 	struct regulator *consumer;
 
-	/* Internal use */
+	/* private: Internal use */
 	int ret;
 };
 
diff --git a/include/net/9p/9p.h b/include/net/9p/9p.h
index 342dcf13d03..a6326ef8ade 100644
--- a/include/net/9p/9p.h
+++ b/include/net/9p/9p.h
@@ -288,6 +288,35 @@ enum p9_perm_t {
 	P9_DMSETVTX = 0x00010000,
 };
 
+/* 9p2000.L open flags */
+#define P9_DOTL_RDONLY        00000000
+#define P9_DOTL_WRONLY        00000001
+#define P9_DOTL_RDWR          00000002
+#define P9_DOTL_NOACCESS      00000003
+#define P9_DOTL_CREATE        00000100
+#define P9_DOTL_EXCL          00000200
+#define P9_DOTL_NOCTTY        00000400
+#define P9_DOTL_TRUNC         00001000
+#define P9_DOTL_APPEND        00002000
+#define P9_DOTL_NONBLOCK      00004000
+#define P9_DOTL_DSYNC         00010000
+#define P9_DOTL_FASYNC        00020000
+#define P9_DOTL_DIRECT        00040000
+#define P9_DOTL_LARGEFILE     00100000
+#define P9_DOTL_DIRECTORY     00200000
+#define P9_DOTL_NOFOLLOW      00400000
+#define P9_DOTL_NOATIME       01000000
+#define P9_DOTL_CLOEXEC       02000000
+#define P9_DOTL_SYNC          04000000
+
+/* 9p2000.L at flags */
+#define P9_DOTL_AT_REMOVEDIR		0x200
+
+/* 9p2000.L lock type */
+#define P9_LOCK_TYPE_RDLCK 0
+#define P9_LOCK_TYPE_WRLCK 1
+#define P9_LOCK_TYPE_UNLCK 2
+
 /**
  * enum p9_qid_t - QID types
  * @P9_QTDIR: directory
diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h
index d17f47fc9e3..401d73bd151 100644
--- a/include/net/cfg80211.h
+++ b/include/net/cfg80211.h
@@ -1744,6 +1744,8 @@ struct wiphy_wowlan_support {
  *	by default for perm_addr. In this case, the mask should be set to
  *	all-zeroes. In this case it is assumed that the device can handle
  *	the same number of arbitrary MAC addresses.
+ * @registered: protects ->resume and ->suspend sysfs callbacks against
+ *	unregister hardware
  * @debugfsdir: debugfs directory used for this wiphy, will be renamed
  *	automatically on wiphy renames
  * @dev: (virtual) struct device for this wiphy
@@ -1865,6 +1867,9 @@ struct wiphy {
 	 * you need use set_wiphy_dev() (see below) */
 	struct device dev;
 
+	/* protects ->resume, ->suspend sysfs callbacks against unregister hw */
+	bool registered;
+
 	/* dir in debugfs: ieee80211/<wiphyname> */
 	struct dentry *debugfsdir;
 
diff --git a/kernel/events/core.c b/kernel/events/core.c
index b8785e26ee1..0f857782d06 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -399,14 +399,54 @@ void perf_cgroup_switch(struct task_struct *task, int mode)
 	local_irq_restore(flags);
 }
 
-static inline void perf_cgroup_sched_out(struct task_struct *task)
+static inline void perf_cgroup_sched_out(struct task_struct *task,
+					 struct task_struct *next)
 {
-	perf_cgroup_switch(task, PERF_CGROUP_SWOUT);
+	struct perf_cgroup *cgrp1;
+	struct perf_cgroup *cgrp2 = NULL;
+
+	/*
+	 * we come here when we know perf_cgroup_events > 0
+	 */
+	cgrp1 = perf_cgroup_from_task(task);
+
+	/*
+	 * next is NULL when called from perf_event_enable_on_exec()
+	 * that will systematically cause a cgroup_switch()
+	 */
+	if (next)
+		cgrp2 = perf_cgroup_from_task(next);
+
+	/*
+	 * only schedule out current cgroup events if we know
+	 * that we are switching to a different cgroup. Otherwise,
+	 * do no touch the cgroup events.
+	 */
+	if (cgrp1 != cgrp2)
+		perf_cgroup_switch(task, PERF_CGROUP_SWOUT);
 }
 
-static inline void perf_cgroup_sched_in(struct task_struct *task)
+static inline void perf_cgroup_sched_in(struct task_struct *prev,
+					struct task_struct *task)
 {
-	perf_cgroup_switch(task, PERF_CGROUP_SWIN);
+	struct perf_cgroup *cgrp1;
+	struct perf_cgroup *cgrp2 = NULL;
+
+	/*
+	 * we come here when we know perf_cgroup_events > 0
+	 */
+	cgrp1 = perf_cgroup_from_task(task);
+
+	/* prev can never be NULL */
+	cgrp2 = perf_cgroup_from_task(prev);
+
+	/*
+	 * only need to schedule in cgroup events if we are changing
+	 * cgroup during ctxsw. Cgroup events were not scheduled
+	 * out of ctxsw out if that was not the case.
+	 */
+	if (cgrp1 != cgrp2)
+		perf_cgroup_switch(task, PERF_CGROUP_SWIN);
 }
 
 static inline int perf_cgroup_connect(int fd, struct perf_event *event,
@@ -518,11 +558,13 @@ static inline void update_cgrp_time_from_cpuctx(struct perf_cpu_context *cpuctx)
 {
 }
 
-static inline void perf_cgroup_sched_out(struct task_struct *task)
+static inline void perf_cgroup_sched_out(struct task_struct *task,
+					 struct task_struct *next)
 {
 }
 
-static inline void perf_cgroup_sched_in(struct task_struct *task)
+static inline void perf_cgroup_sched_in(struct task_struct *prev,
+					struct task_struct *task)
 {
 }
 
@@ -1988,7 +2030,7 @@ void __perf_event_task_sched_out(struct task_struct *task,
 	 * cgroup event are system-wide mode only
 	 */
 	if (atomic_read(&__get_cpu_var(perf_cgroup_events)))
-		perf_cgroup_sched_out(task);
+		perf_cgroup_sched_out(task, next);
 }
 
 static void task_ctx_sched_out(struct perf_event_context *ctx)
@@ -2153,7 +2195,8 @@ static void perf_event_context_sched_in(struct perf_event_context *ctx,
  * accessing the event control register. If a NMI hits, then it will
  * keep the event running.
  */
-void __perf_event_task_sched_in(struct task_struct *task)
+void __perf_event_task_sched_in(struct task_struct *prev,
+				struct task_struct *task)
 {
 	struct perf_event_context *ctx;
 	int ctxn;
@@ -2171,7 +2214,7 @@ void __perf_event_task_sched_in(struct task_struct *task)
 	 * cgroup event are system-wide mode only
 	 */
 	if (atomic_read(&__get_cpu_var(perf_cgroup_events)))
-		perf_cgroup_sched_in(task);
+		perf_cgroup_sched_in(prev, task);
 }
 
 static u64 perf_calculate_period(struct perf_event *event, u64 nsec, u64 count)
@@ -2427,7 +2470,7 @@ static void perf_event_enable_on_exec(struct perf_event_context *ctx)
 	 * ctxswin cgroup events which are already scheduled
 	 * in.
 	 */
-	perf_cgroup_sched_out(current);
+	perf_cgroup_sched_out(current, NULL);
 
 	raw_spin_lock(&ctx->lock);
 	task_ctx_sched_out(ctx);
@@ -3353,8 +3396,8 @@ static int perf_event_index(struct perf_event *event)
 }
 
 static void calc_timer_values(struct perf_event *event,
-				u64 *running,
-				u64 *enabled)
+				u64 *enabled,
+				u64 *running)
 {
 	u64 now, ctx_time;
 
diff --git a/kernel/sched.c b/kernel/sched.c
index ccacdbdecf4..ec5f472bc5b 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -3065,7 +3065,7 @@ static void finish_task_switch(struct rq *rq, struct task_struct *prev)
 #ifdef __ARCH_WANT_INTERRUPTS_ON_CTXSW
 	local_irq_disable();
 #endif /* __ARCH_WANT_INTERRUPTS_ON_CTXSW */
-	perf_event_task_sched_in(current);
+	perf_event_task_sched_in(prev, current);
 #ifdef __ARCH_WANT_INTERRUPTS_ON_CTXSW
 	local_irq_enable();
 #endif /* __ARCH_WANT_INTERRUPTS_ON_CTXSW */
@@ -4279,9 +4279,9 @@ pick_next_task(struct rq *rq)
 }
 
 /*
- * schedule() is the main scheduler function.
+ * __schedule() is the main scheduler function.
  */
-asmlinkage void __sched schedule(void)
+static void __sched __schedule(void)
 {
 	struct task_struct *prev, *next;
 	unsigned long *switch_count;
@@ -4322,16 +4322,6 @@ need_resched:
 				if (to_wakeup)
 					try_to_wake_up_local(to_wakeup);
 			}
-
-			/*
-			 * If we are going to sleep and we have plugged IO
-			 * queued, make sure to submit it to avoid deadlocks.
-			 */
-			if (blk_needs_flush_plug(prev)) {
-				raw_spin_unlock(&rq->lock);
-				blk_schedule_flush_plug(prev);
-				raw_spin_lock(&rq->lock);
-			}
 		}
 		switch_count = &prev->nvcsw;
 	}
@@ -4369,6 +4359,26 @@ need_resched:
 	if (need_resched())
 		goto need_resched;
 }
+
+static inline void sched_submit_work(struct task_struct *tsk)
+{
+	if (!tsk->state)
+		return;
+	/*
+	 * If we are going to sleep and we have plugged IO queued,
+	 * make sure to submit it to avoid deadlocks.
+	 */
+	if (blk_needs_flush_plug(tsk))
+		blk_schedule_flush_plug(tsk);
+}
+
+asmlinkage void schedule(void)
+{
+	struct task_struct *tsk = current;
+
+	sched_submit_work(tsk);
+	__schedule();
+}
 EXPORT_SYMBOL(schedule);
 
 #ifdef CONFIG_MUTEX_SPIN_ON_OWNER
@@ -4435,7 +4445,7 @@ asmlinkage void __sched notrace preempt_schedule(void)
 
 	do {
 		add_preempt_count_notrace(PREEMPT_ACTIVE);
-		schedule();
+		__schedule();
 		sub_preempt_count_notrace(PREEMPT_ACTIVE);
 
 		/*
@@ -4463,7 +4473,7 @@ asmlinkage void __sched preempt_schedule_irq(void)
 	do {
 		add_preempt_count(PREEMPT_ACTIVE);
 		local_irq_enable();
-		schedule();
+		__schedule();
 		local_irq_disable();
 		sub_preempt_count(PREEMPT_ACTIVE);
 
@@ -5588,7 +5598,7 @@ static inline int should_resched(void)
 static void __cond_resched(void)
 {
 	add_preempt_count(PREEMPT_ACTIVE);
-	schedule();
+	__schedule();
 	sub_preempt_count(PREEMPT_ACTIVE);
 }
 
@@ -7443,6 +7453,7 @@ static void __sdt_free(const struct cpumask *cpu_map)
 			struct sched_domain *sd = *per_cpu_ptr(sdd->sd, j);
 			if (sd && (sd->flags & SD_OVERLAP))
 				free_sched_groups(sd->groups, 0);
+			kfree(*per_cpu_ptr(sdd->sd, j));
 			kfree(*per_cpu_ptr(sdd->sg, j));
 			kfree(*per_cpu_ptr(sdd->sgp, j));
 		}
diff --git a/kernel/time/alarmtimer.c b/kernel/time/alarmtimer.c
index 59f369f98a0..ea5e1a928d5 100644
--- a/kernel/time/alarmtimer.c
+++ b/kernel/time/alarmtimer.c
@@ -441,6 +441,8 @@ static int alarm_timer_create(struct k_itimer *new_timer)
 static void alarm_timer_get(struct k_itimer *timr,
 				struct itimerspec *cur_setting)
 {
+	memset(cur_setting, 0, sizeof(struct itimerspec));
+
 	cur_setting->it_interval =
 			ktime_to_timespec(timr->it.alarmtimer.period);
 	cur_setting->it_value =
@@ -479,11 +481,17 @@ static int alarm_timer_set(struct k_itimer *timr, int flags,
 	if (!rtcdev)
 		return -ENOTSUPP;
 
-	/* Save old values */
-	old_setting->it_interval =
-			ktime_to_timespec(timr->it.alarmtimer.period);
-	old_setting->it_value =
-			ktime_to_timespec(timr->it.alarmtimer.node.expires);
+	/*
+	 * XXX HACK! Currently we can DOS a system if the interval
+	 * period on alarmtimers is too small. Cap the interval here
+	 * to 100us and solve this properly in a future patch! -jstultz
+	 */
+	if ((new_setting->it_interval.tv_sec == 0) &&
+			(new_setting->it_interval.tv_nsec < 100000))
+		new_setting->it_interval.tv_nsec = 100000;
+
+	if (old_setting)
+		alarm_timer_get(timr, old_setting);
 
 	/* If the timer was already set, cancel it */
 	alarm_cancel(&timr->it.alarmtimer);
diff --git a/lib/Makefile b/lib/Makefile
index d5d175c8a6c..3f5bc6d903e 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -12,7 +12,7 @@ lib-y := ctype.o string.o vsprintf.o cmdline.o \
 	 idr.o int_sqrt.o extable.o prio_tree.o \
 	 sha1.o md5.o irq_regs.o reciprocal_div.o argv_split.o \
 	 proportions.o prio_heap.o ratelimit.o show_mem.o \
-	 is_single_threaded.o plist.o decompress.o find_next_bit.o
+	 is_single_threaded.o plist.o decompress.o
 
 lib-$(CONFIG_MMU) += ioremap.o
 lib-$(CONFIG_SMP) += cpumask.o
@@ -22,7 +22,7 @@ lib-y	+= kobject.o kref.o klist.o
 obj-y += bcd.o div64.o sort.o parser.o halfmd4.o debug_locks.o random32.o \
 	 bust_spinlocks.o hexdump.o kasprintf.o bitmap.o scatterlist.o \
 	 string_helpers.o gcd.o lcm.o list_sort.o uuid.o flex_array.o \
-	 bsearch.o find_last_bit.o
+	 bsearch.o find_last_bit.o find_next_bit.o
 obj-y += kstrtox.o
 obj-$(CONFIG_TEST_KSTRTOX) += test-kstrtox.o
 
diff --git a/net/9p/trans_virtio.c b/net/9p/trans_virtio.c
index 175b5135bdc..e317583fcc7 100644
--- a/net/9p/trans_virtio.c
+++ b/net/9p/trans_virtio.c
@@ -263,7 +263,6 @@ p9_virtio_request(struct p9_client *client, struct p9_req_t *req)
 {
 	int in, out, inp, outp;
 	struct virtio_chan *chan = client->trans;
-	char *rdata = (char *)req->rc+sizeof(struct p9_fcall);
 	unsigned long flags;
 	size_t pdata_off = 0;
 	struct trans_rpage_info *rpinfo = NULL;
@@ -346,7 +345,8 @@ req_retry_pinned:
 		 * Arrange in such a way that server places header in the
 		 * alloced memory and payload onto the user buffer.
 		 */
-		inp = pack_sg_list(chan->sg, out, VIRTQUEUE_NUM, rdata, 11);
+		inp = pack_sg_list(chan->sg, out,
+				   VIRTQUEUE_NUM, req->rc->sdata, 11);
 		/*
 		 * Running executables in the filesystem may result in
 		 * a read request with kernel buffer as opposed to user buffer.
@@ -366,8 +366,8 @@ req_retry_pinned:
 		}
 		in += inp;
 	} else {
-		in = pack_sg_list(chan->sg, out, VIRTQUEUE_NUM, rdata,
-				req->rc->capacity);
+		in = pack_sg_list(chan->sg, out, VIRTQUEUE_NUM,
+				  req->rc->sdata, req->rc->capacity);
 	}
 
 	err = virtqueue_add_buf(chan->vq, chan->sg, out, in, req->tc);
@@ -592,7 +592,14 @@ static struct p9_trans_module p9_virtio_trans = {
 	.close = p9_virtio_close,
 	.request = p9_virtio_request,
 	.cancel = p9_virtio_cancel,
-	.maxsize = PAGE_SIZE*VIRTQUEUE_NUM,
+
+	/*
+	 * We leave one entry for input and one entry for response
+	 * headers. We also skip one more entry to accomodate, address
+	 * that are not at page boundary, that can result in an extra
+	 * page in zero copy.
+	 */
+	.maxsize = PAGE_SIZE * (VIRTQUEUE_NUM - 3),
 	.pref = P9_TRANS_PREF_PAYLOAD_SEP,
 	.def = 0,
 	.owner = THIS_MODULE,
diff --git a/net/bluetooth/af_bluetooth.c b/net/bluetooth/af_bluetooth.c
index 8add9b49991..117e0d16178 100644
--- a/net/bluetooth/af_bluetooth.c
+++ b/net/bluetooth/af_bluetooth.c
@@ -494,9 +494,8 @@ int bt_sock_wait_state(struct sock *sk, int state, unsigned long timeo)
 	BT_DBG("sk %p", sk);
 
 	add_wait_queue(sk_sleep(sk), &wait);
+	set_current_state(TASK_INTERRUPTIBLE);
 	while (sk->sk_state != state) {
-		set_current_state(TASK_INTERRUPTIBLE);
-
 		if (!timeo) {
 			err = -EINPROGRESS;
 			break;
@@ -510,12 +509,13 @@ int bt_sock_wait_state(struct sock *sk, int state, unsigned long timeo)
 		release_sock(sk);
 		timeo = schedule_timeout(timeo);
 		lock_sock(sk);
+		set_current_state(TASK_INTERRUPTIBLE);
 
 		err = sock_error(sk);
 		if (err)
 			break;
 	}
-	set_current_state(TASK_RUNNING);
+	__set_current_state(TASK_RUNNING);
 	remove_wait_queue(sk_sleep(sk), &wait);
 	return err;
 }
diff --git a/net/bluetooth/bnep/bnep.h b/net/bluetooth/bnep/bnep.h
index 8e6c06158f8..e7ee5314f39 100644
--- a/net/bluetooth/bnep/bnep.h
+++ b/net/bluetooth/bnep/bnep.h
@@ -155,6 +155,7 @@ struct bnep_session {
 	unsigned int  role;
 	unsigned long state;
 	unsigned long flags;
+	atomic_t      terminate;
 	struct task_struct *task;
 
 	struct ethhdr eh;
diff --git a/net/bluetooth/bnep/core.c b/net/bluetooth/bnep/core.c
index ca39fcf010c..d9edfe8bf9d 100644
--- a/net/bluetooth/bnep/core.c
+++ b/net/bluetooth/bnep/core.c
@@ -484,9 +484,11 @@ static int bnep_session(void *arg)
 
 	init_waitqueue_entry(&wait, current);
 	add_wait_queue(sk_sleep(sk), &wait);
-	while (!kthread_should_stop()) {
+	while (1) {
 		set_current_state(TASK_INTERRUPTIBLE);
 
+		if (atomic_read(&s->terminate))
+			break;
 		/* RX */
 		while ((skb = skb_dequeue(&sk->sk_receive_queue))) {
 			skb_orphan(skb);
@@ -504,7 +506,7 @@ static int bnep_session(void *arg)
 
 		schedule();
 	}
-	set_current_state(TASK_RUNNING);
+	__set_current_state(TASK_RUNNING);
 	remove_wait_queue(sk_sleep(sk), &wait);
 
 	/* Cleanup session */
@@ -640,9 +642,10 @@ int bnep_del_connection(struct bnep_conndel_req *req)
 	down_read(&bnep_session_sem);
 
 	s = __bnep_get_session(req->dst);
-	if (s)
-		kthread_stop(s->task);
-	else
+	if (s) {
+		atomic_inc(&s->terminate);
+		wake_up_process(s->task);
+	} else
 		err = -ENOENT;
 
 	up_read(&bnep_session_sem);
diff --git a/net/bluetooth/cmtp/capi.c b/net/bluetooth/cmtp/capi.c
index 040f67b1297..50f0d135eb8 100644
--- a/net/bluetooth/cmtp/capi.c
+++ b/net/bluetooth/cmtp/capi.c
@@ -386,7 +386,8 @@ static void cmtp_reset_ctr(struct capi_ctr *ctrl)
 
 	capi_ctr_down(ctrl);
 
-	kthread_stop(session->task);
+	atomic_inc(&session->terminate);
+	wake_up_process(session->task);
 }
 
 static void cmtp_register_appl(struct capi_ctr *ctrl, __u16 appl, capi_register_params *rp)
diff --git a/net/bluetooth/cmtp/cmtp.h b/net/bluetooth/cmtp/cmtp.h
index db43b54ac9a..c32638dddbf 100644
--- a/net/bluetooth/cmtp/cmtp.h
+++ b/net/bluetooth/cmtp/cmtp.h
@@ -81,6 +81,7 @@ struct cmtp_session {
 
 	char name[BTNAMSIZ];
 
+	atomic_t terminate;
 	struct task_struct *task;
 
 	wait_queue_head_t wait;
diff --git a/net/bluetooth/cmtp/core.c b/net/bluetooth/cmtp/core.c
index c5b11af908b..521baa4fe83 100644
--- a/net/bluetooth/cmtp/core.c
+++ b/net/bluetooth/cmtp/core.c
@@ -292,9 +292,11 @@ static int cmtp_session(void *arg)
 
 	init_waitqueue_entry(&wait, current);
 	add_wait_queue(sk_sleep(sk), &wait);
-	while (!kthread_should_stop()) {
+	while (1) {
 		set_current_state(TASK_INTERRUPTIBLE);
 
+		if (atomic_read(&session->terminate))
+			break;
 		if (sk->sk_state != BT_CONNECTED)
 			break;
 
@@ -307,7 +309,7 @@ static int cmtp_session(void *arg)
 
 		schedule();
 	}
-	set_current_state(TASK_RUNNING);
+	__set_current_state(TASK_RUNNING);
 	remove_wait_queue(sk_sleep(sk), &wait);
 
 	down_write(&cmtp_session_sem);
@@ -380,16 +382,17 @@ int cmtp_add_connection(struct cmtp_connadd_req *req, struct socket *sock)
 
 	if (!(session->flags & (1 << CMTP_LOOPBACK))) {
 		err = cmtp_attach_device(session);
-		if (err < 0)
-			goto detach;
+		if (err < 0) {
+			atomic_inc(&session->terminate);
+			wake_up_process(session->task);
+			up_write(&cmtp_session_sem);
+			return err;
+		}
 	}
 
 	up_write(&cmtp_session_sem);
 	return 0;
 
-detach:
-	cmtp_detach_device(session);
-
 unlink:
 	__cmtp_unlink_session(session);
 
@@ -414,7 +417,8 @@ int cmtp_del_connection(struct cmtp_conndel_req *req)
 		skb_queue_purge(&session->transmit);
 
 		/* Stop session thread */
-		kthread_stop(session->task);
+		atomic_inc(&session->terminate);
+		wake_up_process(session->task);
 	} else
 		err = -ENOENT;
 
diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c
index ec0bc3f60f2..56943add45c 100644
--- a/net/bluetooth/hci_core.c
+++ b/net/bluetooth/hci_core.c
@@ -1209,7 +1209,6 @@ static void hci_cmd_timer(unsigned long arg)
 
 	BT_ERR("%s command tx timeout", hdev->name);
 	atomic_set(&hdev->cmd_cnt, 1);
-	clear_bit(HCI_RESET, &hdev->flags);
 	tasklet_schedule(&hdev->cmd_task);
 }
 
@@ -1327,7 +1326,7 @@ int hci_blacklist_add(struct hci_dev *hdev, bdaddr_t *bdaddr)
 
 	entry = kzalloc(sizeof(struct bdaddr_list), GFP_KERNEL);
 	if (!entry) {
-		return -ENOMEM;
+		err = -ENOMEM;
 		goto err;
 	}
 
@@ -2408,7 +2407,10 @@ static void hci_cmd_task(unsigned long arg)
 		if (hdev->sent_cmd) {
 			atomic_dec(&hdev->cmd_cnt);
 			hci_send_frame(skb);
-			mod_timer(&hdev->cmd_timer,
+			if (test_bit(HCI_RESET, &hdev->flags))
+				del_timer(&hdev->cmd_timer);
+			else
+				mod_timer(&hdev->cmd_timer,
 				  jiffies + msecs_to_jiffies(HCI_CMD_TIMEOUT));
 		} else {
 			skb_queue_head(&hdev->cmd_q, skb);
diff --git a/net/bluetooth/hidp/core.c b/net/bluetooth/hidp/core.c
index 43b4c2deb7c..fb68f344c34 100644
--- a/net/bluetooth/hidp/core.c
+++ b/net/bluetooth/hidp/core.c
@@ -764,6 +764,7 @@ static int hidp_session(void *arg)
 
 	up_write(&hidp_session_sem);
 
+	kfree(session->rd_data);
 	kfree(session);
 	return 0;
 }
@@ -841,7 +842,8 @@ static int hidp_setup_input(struct hidp_session *session,
 
 	err = input_register_device(input);
 	if (err < 0) {
-		hci_conn_put_device(session->conn);
+		input_free_device(input);
+		session->input = NULL;
 		return err;
 	}
 
@@ -1044,8 +1046,12 @@ int hidp_add_connection(struct hidp_connadd_req *req, struct socket *ctrl_sock,
 	}
 
 	err = hid_add_device(session->hid);
-	if (err < 0)
-		goto err_add_device;
+	if (err < 0) {
+		atomic_inc(&session->terminate);
+		wake_up_process(session->task);
+		up_write(&hidp_session_sem);
+		return err;
+	}
 
 	if (session->input) {
 		hidp_send_ctrl_message(session,
@@ -1059,12 +1065,6 @@ int hidp_add_connection(struct hidp_connadd_req *req, struct socket *ctrl_sock,
 	up_write(&hidp_session_sem);
 	return 0;
 
-err_add_device:
-	hid_destroy_device(session->hid);
-	session->hid = NULL;
-	atomic_inc(&session->terminate);
-	wake_up_process(session->task);
-
 unlink:
 	hidp_del_timer(session);
 
@@ -1090,7 +1090,6 @@ purge:
 failed:
 	up_write(&hidp_session_sem);
 
-	input_free_device(session->input);
 	kfree(session);
 	return err;
 }
diff --git a/net/bluetooth/l2cap_core.c b/net/bluetooth/l2cap_core.c
index 3204ba8a701..b3bdb482bbe 100644
--- a/net/bluetooth/l2cap_core.c
+++ b/net/bluetooth/l2cap_core.c
@@ -1159,9 +1159,8 @@ int __l2cap_wait_ack(struct sock *sk)
 	int timeo = HZ/5;
 
 	add_wait_queue(sk_sleep(sk), &wait);
-	while ((chan->unacked_frames > 0 && chan->conn)) {
-		set_current_state(TASK_INTERRUPTIBLE);
-
+	set_current_state(TASK_INTERRUPTIBLE);
+	while (chan->unacked_frames > 0 && chan->conn) {
 		if (!timeo)
 			timeo = HZ/5;
 
@@ -1173,6 +1172,7 @@ int __l2cap_wait_ack(struct sock *sk)
 		release_sock(sk);
 		timeo = schedule_timeout(timeo);
 		lock_sock(sk);
+		set_current_state(TASK_INTERRUPTIBLE);
 
 		err = sock_error(sk);
 		if (err)
diff --git a/net/bluetooth/l2cap_sock.c b/net/bluetooth/l2cap_sock.c
index 5c36b3e8739..61f1f623091 100644
--- a/net/bluetooth/l2cap_sock.c
+++ b/net/bluetooth/l2cap_sock.c
@@ -235,30 +235,26 @@ static int l2cap_sock_accept(struct socket *sock, struct socket *newsock, int fl
 
 	lock_sock_nested(sk, SINGLE_DEPTH_NESTING);
 
-	if (sk->sk_state != BT_LISTEN) {
-		err = -EBADFD;
-		goto done;
-	}
-
 	timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK);
 
 	BT_DBG("sk %p timeo %ld", sk, timeo);
 
 	/* Wait for an incoming connection. (wake-one). */
 	add_wait_queue_exclusive(sk_sleep(sk), &wait);
-	while (!(nsk = bt_accept_dequeue(sk, newsock))) {
+	while (1) {
 		set_current_state(TASK_INTERRUPTIBLE);
-		if (!timeo) {
-			err = -EAGAIN;
+
+		if (sk->sk_state != BT_LISTEN) {
+			err = -EBADFD;
 			break;
 		}
 
-		release_sock(sk);
-		timeo = schedule_timeout(timeo);
-		lock_sock_nested(sk, SINGLE_DEPTH_NESTING);
+		nsk = bt_accept_dequeue(sk, newsock);
+		if (nsk)
+			break;
 
-		if (sk->sk_state != BT_LISTEN) {
-			err = -EBADFD;
+		if (!timeo) {
+			err = -EAGAIN;
 			break;
 		}
 
@@ -266,8 +262,12 @@ static int l2cap_sock_accept(struct socket *sock, struct socket *newsock, int fl
 			err = sock_intr_errno(timeo);
 			break;
 		}
+
+		release_sock(sk);
+		timeo = schedule_timeout(timeo);
+		lock_sock_nested(sk, SINGLE_DEPTH_NESTING);
 	}
-	set_current_state(TASK_RUNNING);
+	__set_current_state(TASK_RUNNING);
 	remove_wait_queue(sk_sleep(sk), &wait);
 
 	if (err)
@@ -993,7 +993,7 @@ static struct sock *l2cap_sock_alloc(struct net *net, struct socket *sock, int p
 	INIT_LIST_HEAD(&bt_sk(sk)->accept_q);
 
 	sk->sk_destruct = l2cap_sock_destruct;
-	sk->sk_sndtimeo = msecs_to_jiffies(L2CAP_CONN_TIMEOUT);
+	sk->sk_sndtimeo = L2CAP_CONN_TIMEOUT;
 
 	sock_reset_flag(sk, SOCK_ZAPPED);
 
diff --git a/net/bluetooth/rfcomm/core.c b/net/bluetooth/rfcomm/core.c
index 5759bb7054f..5ba3f6df665 100644
--- a/net/bluetooth/rfcomm/core.c
+++ b/net/bluetooth/rfcomm/core.c
@@ -62,7 +62,6 @@ static DEFINE_MUTEX(rfcomm_mutex);
 #define rfcomm_lock()	mutex_lock(&rfcomm_mutex)
 #define rfcomm_unlock()	mutex_unlock(&rfcomm_mutex)
 
-static unsigned long rfcomm_event;
 
 static LIST_HEAD(session_list);
 
@@ -120,7 +119,6 @@ static inline void rfcomm_schedule(void)
 {
 	if (!rfcomm_thread)
 		return;
-	set_bit(RFCOMM_SCHED_WAKEUP, &rfcomm_event);
 	wake_up_process(rfcomm_thread);
 }
 
@@ -2038,19 +2036,18 @@ static int rfcomm_run(void *unused)
 
 	rfcomm_add_listener(BDADDR_ANY);
 
-	while (!kthread_should_stop()) {
+	while (1) {
 		set_current_state(TASK_INTERRUPTIBLE);
-		if (!test_bit(RFCOMM_SCHED_WAKEUP, &rfcomm_event)) {
-			/* No pending events. Let's sleep.
-			 * Incoming connections and data will wake us up. */
-			schedule();
-		}
-		set_current_state(TASK_RUNNING);
+
+		if (kthread_should_stop())
+			break;
 
 		/* Process stuff */
-		clear_bit(RFCOMM_SCHED_WAKEUP, &rfcomm_event);
 		rfcomm_process_sessions();
+
+		schedule();
 	}
+	__set_current_state(TASK_RUNNING);
 
 	rfcomm_kill_listener();
 
diff --git a/net/bluetooth/rfcomm/sock.c b/net/bluetooth/rfcomm/sock.c
index 8f01e6b11a7..482722bbc7a 100644
--- a/net/bluetooth/rfcomm/sock.c
+++ b/net/bluetooth/rfcomm/sock.c
@@ -485,11 +485,6 @@ static int rfcomm_sock_accept(struct socket *sock, struct socket *newsock, int f
 
 	lock_sock(sk);
 
-	if (sk->sk_state != BT_LISTEN) {
-		err = -EBADFD;
-		goto done;
-	}
-
 	if (sk->sk_type != SOCK_STREAM) {
 		err = -EINVAL;
 		goto done;
@@ -501,19 +496,20 @@ static int rfcomm_sock_accept(struct socket *sock, struct socket *newsock, int f
 
 	/* Wait for an incoming connection. (wake-one). */
 	add_wait_queue_exclusive(sk_sleep(sk), &wait);
-	while (!(nsk = bt_accept_dequeue(sk, newsock))) {
+	while (1) {
 		set_current_state(TASK_INTERRUPTIBLE);
-		if (!timeo) {
-			err = -EAGAIN;
+
+		if (sk->sk_state != BT_LISTEN) {
+			err = -EBADFD;
 			break;
 		}
 
-		release_sock(sk);
-		timeo = schedule_timeout(timeo);
-		lock_sock(sk);
+		nsk = bt_accept_dequeue(sk, newsock);
+		if (nsk)
+			break;
 
-		if (sk->sk_state != BT_LISTEN) {
-			err = -EBADFD;
+		if (!timeo) {
+			err = -EAGAIN;
 			break;
 		}
 
@@ -521,8 +517,12 @@ static int rfcomm_sock_accept(struct socket *sock, struct socket *newsock, int f
 			err = sock_intr_errno(timeo);
 			break;
 		}
+
+		release_sock(sk);
+		timeo = schedule_timeout(timeo);
+		lock_sock(sk);
 	}
-	set_current_state(TASK_RUNNING);
+	__set_current_state(TASK_RUNNING);
 	remove_wait_queue(sk_sleep(sk), &wait);
 
 	if (err)
diff --git a/net/bluetooth/sco.c b/net/bluetooth/sco.c
index 4c3621b5e0a..8270f05e3f1 100644
--- a/net/bluetooth/sco.c
+++ b/net/bluetooth/sco.c
@@ -564,30 +564,26 @@ static int sco_sock_accept(struct socket *sock, struct socket *newsock, int flag
 
 	lock_sock(sk);
 
-	if (sk->sk_state != BT_LISTEN) {
-		err = -EBADFD;
-		goto done;
-	}
-
 	timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK);
 
 	BT_DBG("sk %p timeo %ld", sk, timeo);
 
 	/* Wait for an incoming connection. (wake-one). */
 	add_wait_queue_exclusive(sk_sleep(sk), &wait);
-	while (!(ch = bt_accept_dequeue(sk, newsock))) {
+	while (1) {
 		set_current_state(TASK_INTERRUPTIBLE);
-		if (!timeo) {
-			err = -EAGAIN;
+
+		if (sk->sk_state != BT_LISTEN) {
+			err = -EBADFD;
 			break;
 		}
 
-		release_sock(sk);
-		timeo = schedule_timeout(timeo);
-		lock_sock(sk);
+		ch = bt_accept_dequeue(sk, newsock);
+		if (ch)
+			break;
 
-		if (sk->sk_state != BT_LISTEN) {
-			err = -EBADFD;
+		if (!timeo) {
+			err = -EAGAIN;
 			break;
 		}
 
@@ -595,8 +591,12 @@ static int sco_sock_accept(struct socket *sock, struct socket *newsock, int flag
 			err = sock_intr_errno(timeo);
 			break;
 		}
+
+		release_sock(sk);
+		timeo = schedule_timeout(timeo);
+		lock_sock(sk);
 	}
-	set_current_state(TASK_RUNNING);
+	__set_current_state(TASK_RUNNING);
 	remove_wait_queue(sk_sleep(sk), &wait);
 
 	if (err)
diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
index 2d85ca7111d..995cbe0ac0b 100644
--- a/net/bridge/br_multicast.c
+++ b/net/bridge/br_multicast.c
@@ -1456,7 +1456,7 @@ static int br_multicast_ipv6_rcv(struct net_bridge *br,
 {
 	struct sk_buff *skb2;
 	const struct ipv6hdr *ip6h;
-	struct icmp6hdr *icmp6h;
+	u8 icmp6_type;
 	u8 nexthdr;
 	unsigned len;
 	int offset;
@@ -1502,9 +1502,9 @@ static int br_multicast_ipv6_rcv(struct net_bridge *br,
 	__skb_pull(skb2, offset);
 	skb_reset_transport_header(skb2);
 
-	icmp6h = icmp6_hdr(skb2);
+	icmp6_type = icmp6_hdr(skb2)->icmp6_type;
 
-	switch (icmp6h->icmp6_type) {
+	switch (icmp6_type) {
 	case ICMPV6_MGM_QUERY:
 	case ICMPV6_MGM_REPORT:
 	case ICMPV6_MGM_REDUCTION:
@@ -1520,16 +1520,23 @@ static int br_multicast_ipv6_rcv(struct net_bridge *br,
 		err = pskb_trim_rcsum(skb2, len);
 		if (err)
 			goto out;
+		err = -EINVAL;
 	}
 
+	ip6h = ipv6_hdr(skb2);
+
 	switch (skb2->ip_summed) {
 	case CHECKSUM_COMPLETE:
-		if (!csum_fold(skb2->csum))
+		if (!csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr, skb2->len,
+					IPPROTO_ICMPV6, skb2->csum))
 			break;
 		/*FALLTHROUGH*/
 	case CHECKSUM_NONE:
-		skb2->csum = 0;
-		if (skb_checksum_complete(skb2))
+		skb2->csum = ~csum_unfold(csum_ipv6_magic(&ip6h->saddr,
+							&ip6h->daddr,
+							skb2->len,
+							IPPROTO_ICMPV6, 0));
+		if (__skb_checksum_complete(skb2))
 			goto out;
 	}
 
@@ -1537,7 +1544,7 @@ static int br_multicast_ipv6_rcv(struct net_bridge *br,
 
 	BR_INPUT_SKB_CB(skb)->igmp = 1;
 
-	switch (icmp6h->icmp6_type) {
+	switch (icmp6_type) {
 	case ICMPV6_MGM_REPORT:
 	    {
 		struct mld_msg *mld;
diff --git a/net/ceph/msgpool.c b/net/ceph/msgpool.c
index d5f2d97ac05..1f4cb30a42c 100644
--- a/net/ceph/msgpool.c
+++ b/net/ceph/msgpool.c
@@ -7,27 +7,37 @@
 
 #include <linux/ceph/msgpool.h>
 
-static void *alloc_fn(gfp_t gfp_mask, void *arg)
+static void *msgpool_alloc(gfp_t gfp_mask, void *arg)
 {
 	struct ceph_msgpool *pool = arg;
-	void *p;
+	struct ceph_msg *msg;
 
-	p = ceph_msg_new(0, pool->front_len, gfp_mask);
-	if (!p)
-		pr_err("msgpool %s alloc failed\n", pool->name);
-	return p;
+	msg = ceph_msg_new(0, pool->front_len, gfp_mask);
+	if (!msg) {
+		dout("msgpool_alloc %s failed\n", pool->name);
+	} else {
+		dout("msgpool_alloc %s %p\n", pool->name, msg);
+		msg->pool = pool;
+	}
+	return msg;
 }
 
-static void free_fn(void *element, void *arg)
+static void msgpool_free(void *element, void *arg)
 {
-	ceph_msg_put(element);
+	struct ceph_msgpool *pool = arg;
+	struct ceph_msg *msg = element;
+
+	dout("msgpool_release %s %p\n", pool->name, msg);
+	msg->pool = NULL;
+	ceph_msg_put(msg);
 }
 
 int ceph_msgpool_init(struct ceph_msgpool *pool,
 		      int front_len, int size, bool blocking, const char *name)
 {
+	dout("msgpool %s init\n", name);
 	pool->front_len = front_len;
-	pool->pool = mempool_create(size, alloc_fn, free_fn, pool);
+	pool->pool = mempool_create(size, msgpool_alloc, msgpool_free, pool);
 	if (!pool->pool)
 		return -ENOMEM;
 	pool->name = name;
@@ -36,14 +46,17 @@ int ceph_msgpool_init(struct ceph_msgpool *pool,
 
 void ceph_msgpool_destroy(struct ceph_msgpool *pool)
 {
+	dout("msgpool %s destroy\n", pool->name);
 	mempool_destroy(pool->pool);
 }
 
 struct ceph_msg *ceph_msgpool_get(struct ceph_msgpool *pool,
 				  int front_len)
 {
+	struct ceph_msg *msg;
+
 	if (front_len > pool->front_len) {
-		pr_err("msgpool_get pool %s need front %d, pool size is %d\n",
+		dout("msgpool_get %s need front %d, pool size is %d\n",
 		       pool->name, front_len, pool->front_len);
 		WARN_ON(1);
 
@@ -51,14 +64,19 @@ struct ceph_msg *ceph_msgpool_get(struct ceph_msgpool *pool,
 		return ceph_msg_new(0, front_len, GFP_NOFS);
 	}
 
-	return mempool_alloc(pool->pool, GFP_NOFS);
+	msg = mempool_alloc(pool->pool, GFP_NOFS);
+	dout("msgpool_get %s %p\n", pool->name, msg);
+	return msg;
 }
 
 void ceph_msgpool_put(struct ceph_msgpool *pool, struct ceph_msg *msg)
 {
+	dout("msgpool_put %s %p\n", pool->name, msg);
+
 	/* reset msg front_len; user may have changed it */
 	msg->front.iov_len = pool->front_len;
 	msg->hdr.front_len = cpu_to_le32(pool->front_len);
 
 	kref_init(&msg->kref);  /* retake single ref */
+	mempool_free(msg, pool->pool);
 }
diff --git a/net/ceph/osd_client.c b/net/ceph/osd_client.c
index ce310eee708..16836a7df7a 100644
--- a/net/ceph/osd_client.c
+++ b/net/ceph/osd_client.c
@@ -685,6 +685,18 @@ static void __remove_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
 	put_osd(osd);
 }
 
+static void remove_all_osds(struct ceph_osd_client *osdc)
+{
+	dout("__remove_old_osds %p\n", osdc);
+	mutex_lock(&osdc->request_mutex);
+	while (!RB_EMPTY_ROOT(&osdc->osds)) {
+		struct ceph_osd *osd = rb_entry(rb_first(&osdc->osds),
+						struct ceph_osd, o_node);
+		__remove_osd(osdc, osd);
+	}
+	mutex_unlock(&osdc->request_mutex);
+}
+
 static void __move_osd_to_lru(struct ceph_osd_client *osdc,
 			      struct ceph_osd *osd)
 {
@@ -701,14 +713,14 @@ static void __remove_osd_from_lru(struct ceph_osd *osd)
 		list_del_init(&osd->o_osd_lru);
 }
 
-static void remove_old_osds(struct ceph_osd_client *osdc, int remove_all)
+static void remove_old_osds(struct ceph_osd_client *osdc)
 {
 	struct ceph_osd *osd, *nosd;
 
 	dout("__remove_old_osds %p\n", osdc);
 	mutex_lock(&osdc->request_mutex);
 	list_for_each_entry_safe(osd, nosd, &osdc->osd_lru, o_osd_lru) {
-		if (!remove_all && time_before(jiffies, osd->lru_ttl))
+		if (time_before(jiffies, osd->lru_ttl))
 			break;
 		__remove_osd(osdc, osd);
 	}
@@ -751,6 +763,7 @@ static void __insert_osd(struct ceph_osd_client *osdc, struct ceph_osd *new)
 	struct rb_node *parent = NULL;
 	struct ceph_osd *osd = NULL;
 
+	dout("__insert_osd %p osd%d\n", new, new->o_osd);
 	while (*p) {
 		parent = *p;
 		osd = rb_entry(parent, struct ceph_osd, o_node);
@@ -1144,7 +1157,7 @@ static void handle_osds_timeout(struct work_struct *work)
 
 	dout("osds timeout\n");
 	down_read(&osdc->map_sem);
-	remove_old_osds(osdc, 0);
+	remove_old_osds(osdc);
 	up_read(&osdc->map_sem);
 
 	schedule_delayed_work(&osdc->osds_timeout_work,
@@ -1862,8 +1875,7 @@ void ceph_osdc_stop(struct ceph_osd_client *osdc)
 		ceph_osdmap_destroy(osdc->osdmap);
 		osdc->osdmap = NULL;
 	}
-	remove_old_osds(osdc, 1);
-	WARN_ON(!RB_EMPTY_ROOT(&osdc->osds));
+	remove_all_osds(osdc);
 	mempool_destroy(osdc->req_mempool);
 	ceph_msgpool_destroy(&osdc->msgpool_op);
 	ceph_msgpool_destroy(&osdc->msgpool_op_reply);
diff --git a/net/core/neighbour.c b/net/core/neighbour.c
index 8fab9b0bb20..1334d7e56f0 100644
--- a/net/core/neighbour.c
+++ b/net/core/neighbour.c
@@ -1319,11 +1319,15 @@ static void neigh_proxy_process(unsigned long arg)
 
 		if (tdif <= 0) {
 			struct net_device *dev = skb->dev;
+
 			__skb_unlink(skb, &tbl->proxy_queue);
-			if (tbl->proxy_redo && netif_running(dev))
+			if (tbl->proxy_redo && netif_running(dev)) {
+				rcu_read_lock();
 				tbl->proxy_redo(skb);
-			else
+				rcu_read_unlock();
+			} else {
 				kfree_skb(skb);
+			}
 
 			dev_put(dev);
 		} else if (!sched_next || tdif < sched_next)
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index adf84dd8c7b..52622517e0d 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -558,13 +558,14 @@ int __netpoll_rx(struct sk_buff *skb)
 	if (skb_shared(skb))
 		goto out;
 
-	iph = (struct iphdr *)skb->data;
 	if (!pskb_may_pull(skb, sizeof(struct iphdr)))
 		goto out;
+	iph = (struct iphdr *)skb->data;
 	if (iph->ihl < 5 || iph->version != 4)
 		goto out;
 	if (!pskb_may_pull(skb, iph->ihl*4))
 		goto out;
+	iph = (struct iphdr *)skb->data;
 	if (ip_fast_csum((u8 *)iph, iph->ihl) != 0)
 		goto out;
 
@@ -579,6 +580,7 @@ int __netpoll_rx(struct sk_buff *skb)
 	if (pskb_trim_rcsum(skb, len))
 		goto out;
 
+	iph = (struct iphdr *)skb->data;
 	if (iph->protocol != IPPROTO_UDP)
 		goto out;
 
diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c
index 283c0a26e03..d577199eabd 100644
--- a/net/ipv4/igmp.c
+++ b/net/ipv4/igmp.c
@@ -767,7 +767,7 @@ static int igmp_xmarksources(struct ip_mc_list *pmc, int nsrcs, __be32 *srcs)
 			break;
 		for (i=0; i<nsrcs; i++) {
 			/* skip inactive filters */
-			if (pmc->sfcount[MCAST_INCLUDE] ||
+			if (psf->sf_count[MCAST_INCLUDE] ||
 			    pmc->sfcount[MCAST_EXCLUDE] !=
 			    psf->sf_count[MCAST_EXCLUDE])
 				continue;
diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c
index 3e6ebcdb477..ee7839f4d6e 100644
--- a/net/ipv6/mcast.c
+++ b/net/ipv6/mcast.c
@@ -1059,7 +1059,7 @@ static int mld_xmarksources(struct ifmcaddr6 *pmc, int nsrcs,
 			break;
 		for (i=0; i<nsrcs; i++) {
 			/* skip inactive filters */
-			if (pmc->mca_sfcount[MCAST_INCLUDE] ||
+			if (psf->sf_count[MCAST_INCLUDE] ||
 			    pmc->mca_sfcount[MCAST_EXCLUDE] !=
 			    psf->sf_count[MCAST_EXCLUDE])
 				continue;
diff --git a/net/mac80211/main.c b/net/mac80211/main.c
index 866f269183c..acb44230b25 100644
--- a/net/mac80211/main.c
+++ b/net/mac80211/main.c
@@ -1012,7 +1012,6 @@ void ieee80211_unregister_hw(struct ieee80211_hw *hw)
 	cancel_work_sync(&local->reconfig_filter);
 
 	ieee80211_clear_tx_pending(local);
-	sta_info_stop(local);
 	rate_control_deinitialize(local);
 
 	if (skb_queue_len(&local->skb_queue) ||
@@ -1024,6 +1023,7 @@ void ieee80211_unregister_hw(struct ieee80211_hw *hw)
 
 	destroy_workqueue(local->workqueue);
 	wiphy_unregister(local->hw.wiphy);
+	sta_info_stop(local);
 	ieee80211_wep_free(local);
 	ieee80211_led_exit(local);
 	kfree(local->int_scan_req);
diff --git a/net/socket.c b/net/socket.c
index 24a77400b65..ffe92ca32f2 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -1965,8 +1965,9 @@ static int __sys_sendmsg(struct socket *sock, struct msghdr __user *msg,
 	 * used_address->name_len is initialized to UINT_MAX so that the first
 	 * destination address never matches.
 	 */
-	if (used_address && used_address->name_len == msg_sys->msg_namelen &&
-	    !memcmp(&used_address->name, msg->msg_name,
+	if (used_address && msg_sys->msg_name &&
+	    used_address->name_len == msg_sys->msg_namelen &&
+	    !memcmp(&used_address->name, msg_sys->msg_name,
 		    used_address->name_len)) {
 		err = sock_sendmsg_nosec(sock, msg_sys, total_len);
 		goto out_freectl;
@@ -1978,8 +1979,9 @@ static int __sys_sendmsg(struct socket *sock, struct msghdr __user *msg,
 	 */
 	if (used_address && err >= 0) {
 		used_address->name_len = msg_sys->msg_namelen;
-		memcpy(&used_address->name, msg->msg_name,
-		       used_address->name_len);
+		if (msg_sys->msg_name)
+			memcpy(&used_address->name, msg_sys->msg_name,
+			       used_address->name_len);
 	}
 
 out_freectl:
diff --git a/net/wireless/core.c b/net/wireless/core.c
index 645437cfc46..c14865172da 100644
--- a/net/wireless/core.c
+++ b/net/wireless/core.c
@@ -616,6 +616,9 @@ int wiphy_register(struct wiphy *wiphy)
 	if (res)
 		goto out_rm_dev;
 
+	rtnl_lock();
+	rdev->wiphy.registered = true;
+	rtnl_unlock();
 	return 0;
 
 out_rm_dev:
@@ -647,6 +650,10 @@ void wiphy_unregister(struct wiphy *wiphy)
 {
 	struct cfg80211_registered_device *rdev = wiphy_to_dev(wiphy);
 
+	rtnl_lock();
+	rdev->wiphy.registered = false;
+	rtnl_unlock();
+
 	rfkill_unregister(rdev->rfkill);
 
 	/* protect the device list */
diff --git a/net/wireless/sysfs.c b/net/wireless/sysfs.c
index c6e4ca6a7d2..ff574597a85 100644
--- a/net/wireless/sysfs.c
+++ b/net/wireless/sysfs.c
@@ -93,7 +93,8 @@ static int wiphy_suspend(struct device *dev, pm_message_t state)
 
 	if (rdev->ops->suspend) {
 		rtnl_lock();
-		ret = rdev->ops->suspend(&rdev->wiphy, rdev->wowlan);
+		if (rdev->wiphy.registered)
+			ret = rdev->ops->suspend(&rdev->wiphy, rdev->wowlan);
 		rtnl_unlock();
 	}
 
@@ -112,7 +113,8 @@ static int wiphy_resume(struct device *dev)
 
 	if (rdev->ops->resume) {
 		rtnl_lock();
-		ret = rdev->ops->resume(&rdev->wiphy);
+		if (rdev->wiphy.registered)
+			ret = rdev->ops->resume(&rdev->wiphy);
 		rtnl_unlock();
 	}
 
diff --git a/tools/perf/arch/arm/util/dwarf-regs.c b/tools/perf/arch/arm/util/dwarf-regs.c
index fff6450c8c9..e8d5c551c69 100644
--- a/tools/perf/arch/arm/util/dwarf-regs.c
+++ b/tools/perf/arch/arm/util/dwarf-regs.c
@@ -8,7 +8,10 @@
  * published by the Free Software Foundation.
  */
 
+#include <stdlib.h>
+#ifndef __UCLIBC__
 #include <libio.h>
+#endif
 #include <dwarf-regs.h>
 
 struct pt_regs_dwarfnum {
author	Dave Airlie <airlied@redhat.com>	2011-09-20 09:35:22 +0100
committer	Dave Airlie <airlied@redhat.com>	2011-09-20 09:35:22 +0100
commit	b2d108ba333cdff80d9e7645d7697cbb6bb0fc29 (patch)
tree	f62b2464a35e4676b5e3e9ef340758a17d0b80f8
parent	fcf4de5acf09889e3f0c131ebe385c983006d71b (diff)
parent	a0d9a8feb928465f3cb525a19e5fafd06ef66ced (diff)