From 0206e353a0416ad63ce07f53c807c2c725633b87 Mon Sep 17 00:00:00 2001 From: Akshay Joshi Date: Tue, 16 Aug 2011 15:34:10 -0400 Subject: Drivers: i915: Fix all space related issues. Various issues involved with the space character were generating warnings in the checkpatch.pl file. This patch removes most of those warnings. Signed-off-by: Akshay Joshi Signed-off-by: Keith Packard --- drivers/gpu/drm/i915/intel_ringbuffer.c | 76 ++++++++++++++++----------------- 1 file changed, 38 insertions(+), 38 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_ringbuffer.c') diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index c30626ea9f9..a40f3456448 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -1026,7 +1026,7 @@ static const struct intel_ring_buffer render_ring = { .irq_get = render_ring_get_irq, .irq_put = render_ring_put_irq, .dispatch_execbuffer = render_ring_dispatch_execbuffer, - .cleanup = render_ring_cleanup, + .cleanup = render_ring_cleanup, }; /* ring buffer for bit-stream decoder */ @@ -1050,23 +1050,23 @@ static const struct intel_ring_buffer bsd_ring = { static void gen6_bsd_ring_write_tail(struct intel_ring_buffer *ring, u32 value) { - drm_i915_private_t *dev_priv = ring->dev->dev_private; + drm_i915_private_t *dev_priv = ring->dev->dev_private; /* Every tail move must follow the sequence below */ - I915_WRITE(GEN6_BSD_SLEEP_PSMI_CONTROL, - GEN6_BSD_SLEEP_PSMI_CONTROL_RC_ILDL_MESSAGE_MODIFY_MASK | - GEN6_BSD_SLEEP_PSMI_CONTROL_RC_ILDL_MESSAGE_DISABLE); - I915_WRITE(GEN6_BSD_RNCID, 0x0); - - if (wait_for((I915_READ(GEN6_BSD_SLEEP_PSMI_CONTROL) & - GEN6_BSD_SLEEP_PSMI_CONTROL_IDLE_INDICATOR) == 0, - 50)) - DRM_ERROR("timed out waiting for IDLE Indicator\n"); - - I915_WRITE_TAIL(ring, value); - I915_WRITE(GEN6_BSD_SLEEP_PSMI_CONTROL, - GEN6_BSD_SLEEP_PSMI_CONTROL_RC_ILDL_MESSAGE_MODIFY_MASK | - GEN6_BSD_SLEEP_PSMI_CONTROL_RC_ILDL_MESSAGE_ENABLE); + I915_WRITE(GEN6_BSD_SLEEP_PSMI_CONTROL, + GEN6_BSD_SLEEP_PSMI_CONTROL_RC_ILDL_MESSAGE_MODIFY_MASK | + GEN6_BSD_SLEEP_PSMI_CONTROL_RC_ILDL_MESSAGE_DISABLE); + I915_WRITE(GEN6_BSD_RNCID, 0x0); + + if (wait_for((I915_READ(GEN6_BSD_SLEEP_PSMI_CONTROL) & + GEN6_BSD_SLEEP_PSMI_CONTROL_IDLE_INDICATOR) == 0, + 50)) + DRM_ERROR("timed out waiting for IDLE Indicator\n"); + + I915_WRITE_TAIL(ring, value); + I915_WRITE(GEN6_BSD_SLEEP_PSMI_CONTROL, + GEN6_BSD_SLEEP_PSMI_CONTROL_RC_ILDL_MESSAGE_MODIFY_MASK | + GEN6_BSD_SLEEP_PSMI_CONTROL_RC_ILDL_MESSAGE_ENABLE); } static int gen6_ring_flush(struct intel_ring_buffer *ring, @@ -1094,18 +1094,18 @@ static int gen6_ring_dispatch_execbuffer(struct intel_ring_buffer *ring, u32 offset, u32 len) { - int ret; + int ret; - ret = intel_ring_begin(ring, 2); - if (ret) - return ret; + ret = intel_ring_begin(ring, 2); + if (ret) + return ret; - intel_ring_emit(ring, MI_BATCH_BUFFER_START | MI_BATCH_NON_SECURE_I965); - /* bit0-7 is the length on GEN6+ */ - intel_ring_emit(ring, offset); - intel_ring_advance(ring); + intel_ring_emit(ring, MI_BATCH_BUFFER_START | MI_BATCH_NON_SECURE_I965); + /* bit0-7 is the length on GEN6+ */ + intel_ring_emit(ring, offset); + intel_ring_advance(ring); - return 0; + return 0; } static bool @@ -1272,19 +1272,19 @@ static void blt_ring_cleanup(struct intel_ring_buffer *ring) } static const struct intel_ring_buffer gen6_blt_ring = { - .name = "blt ring", - .id = RING_BLT, - .mmio_base = BLT_RING_BASE, - .size = 32 * PAGE_SIZE, - .init = blt_ring_init, - .write_tail = ring_write_tail, - .flush = blt_ring_flush, - .add_request = gen6_add_request, - .get_seqno = ring_get_seqno, - .irq_get = blt_ring_get_irq, - .irq_put = blt_ring_put_irq, - .dispatch_execbuffer = gen6_ring_dispatch_execbuffer, - .cleanup = blt_ring_cleanup, + .name = "blt ring", + .id = RING_BLT, + .mmio_base = BLT_RING_BASE, + .size = 32 * PAGE_SIZE, + .init = blt_ring_init, + .write_tail = ring_write_tail, + .flush = blt_ring_flush, + .add_request = gen6_add_request, + .get_seqno = ring_get_seqno, + .irq_get = blt_ring_get_irq, + .irq_put = blt_ring_put_irq, + .dispatch_execbuffer = gen6_ring_dispatch_execbuffer, + .cleanup = blt_ring_cleanup, }; int intel_init_render_ring_buffer(struct drm_device *dev) -- cgit v1.2.3-70-g09d2 From c8c99b0f0dea1ced5d0e10cdb9143356cc16b484 Mon Sep 17 00:00:00 2001 From: Ben Widawsky Date: Wed, 14 Sep 2011 20:32:47 -0700 Subject: drm/i915: Dumb down the semaphore logic While I think the previous code is correct, it was hard to follow and hard to debug. Since we already have a ring abstraction, might as well use it to handle the semaphore updates and compares. I don't expect this code to make semaphores better or worse, but you never know... v2: Remove magic per Keith's suggestions. Ran Daniel's gem_ring_sync_loop test on this. v3: Ignored one of Keith's suggestions. v4: Removed some bloat per Daniel's recommendation. Cc: Daniel Vetter Cc: Keith Packard Signed-off-by: Ben Widawsky Signed-off-by: Keith Packard --- drivers/gpu/drm/i915/i915_gem_execbuffer.c | 3 +- drivers/gpu/drm/i915/i915_reg.h | 13 +++ drivers/gpu/drm/i915/intel_ringbuffer.c | 149 ++++++++++++++++++++--------- drivers/gpu/drm/i915/intel_ringbuffer.h | 8 +- 4 files changed, 126 insertions(+), 47 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_ringbuffer.c') diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index 4934cf84c32..3693e83a97f 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -784,7 +784,8 @@ i915_gem_execbuffer_sync_rings(struct drm_i915_gem_object *obj, } from->sync_seqno[idx] = seqno; - return intel_ring_sync(to, from, seqno - 1); + + return to->sync_to(to, from, seqno - 1); } static int diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index a363fdd54ba..138eae15bec 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -194,6 +194,13 @@ #define MI_SEMAPHORE_UPDATE (1<<21) #define MI_SEMAPHORE_COMPARE (1<<20) #define MI_SEMAPHORE_REGISTER (1<<18) +#define MI_SEMAPHORE_SYNC_RV (2<<16) +#define MI_SEMAPHORE_SYNC_RB (0<<16) +#define MI_SEMAPHORE_SYNC_VR (0<<16) +#define MI_SEMAPHORE_SYNC_VB (2<<16) +#define MI_SEMAPHORE_SYNC_BR (2<<16) +#define MI_SEMAPHORE_SYNC_BV (0<<16) +#define MI_SEMAPHORE_SYNC_INVALID (1<<0) /* * 3D instructions used by the kernel */ @@ -296,6 +303,12 @@ #define RING_CTL(base) ((base)+0x3c) #define RING_SYNC_0(base) ((base)+0x40) #define RING_SYNC_1(base) ((base)+0x44) +#define GEN6_RVSYNC (RING_SYNC_0(RENDER_RING_BASE)) +#define GEN6_RBSYNC (RING_SYNC_1(RENDER_RING_BASE)) +#define GEN6_VRSYNC (RING_SYNC_1(GEN6_BSD_RING_BASE)) +#define GEN6_VBSYNC (RING_SYNC_0(GEN6_BSD_RING_BASE)) +#define GEN6_BRSYNC (RING_SYNC_0(BLT_RING_BASE)) +#define GEN6_BVSYNC (RING_SYNC_1(BLT_RING_BASE)) #define RING_MAX_IDLE(base) ((base)+0x54) #define RING_HWS_PGA(base) ((base)+0x80) #define RING_HWS_PGA_GEN6(base) ((base)+0x2080) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index a40f3456448..0e99589b54e 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -315,79 +315,127 @@ static void render_ring_cleanup(struct intel_ring_buffer *ring) } static void -update_semaphore(struct intel_ring_buffer *ring, int i, u32 seqno) +update_mboxes(struct intel_ring_buffer *ring, + u32 seqno, + u32 mmio_offset) { - struct drm_device *dev = ring->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - int id; - - /* - * cs -> 1 = vcs, 0 = bcs - * vcs -> 1 = bcs, 0 = cs, - * bcs -> 1 = cs, 0 = vcs. - */ - id = ring - dev_priv->ring; - id += 2 - i; - id %= 3; - - intel_ring_emit(ring, - MI_SEMAPHORE_MBOX | - MI_SEMAPHORE_REGISTER | - MI_SEMAPHORE_UPDATE); + intel_ring_emit(ring, MI_SEMAPHORE_MBOX | + MI_SEMAPHORE_GLOBAL_GTT | + MI_SEMAPHORE_REGISTER | + MI_SEMAPHORE_UPDATE); intel_ring_emit(ring, seqno); - intel_ring_emit(ring, - RING_SYNC_0(dev_priv->ring[id].mmio_base) + 4*i); + intel_ring_emit(ring, mmio_offset); } +/** + * gen6_add_request - Update the semaphore mailbox registers + * + * @ring - ring that is adding a request + * @seqno - return seqno stuck into the ring + * + * Update the mailbox registers in the *other* rings with the current seqno. + * This acts like a signal in the canonical semaphore. + */ static int gen6_add_request(struct intel_ring_buffer *ring, - u32 *result) + u32 *seqno) { - u32 seqno; + u32 mbox1_reg; + u32 mbox2_reg; int ret; ret = intel_ring_begin(ring, 10); if (ret) return ret; - seqno = i915_gem_get_seqno(ring->dev); - update_semaphore(ring, 0, seqno); - update_semaphore(ring, 1, seqno); + mbox1_reg = ring->signal_mbox[0]; + mbox2_reg = ring->signal_mbox[1]; + *seqno = i915_gem_get_seqno(ring->dev); + + update_mboxes(ring, *seqno, mbox1_reg); + update_mboxes(ring, *seqno, mbox2_reg); intel_ring_emit(ring, MI_STORE_DWORD_INDEX); intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT); - intel_ring_emit(ring, seqno); + intel_ring_emit(ring, *seqno); intel_ring_emit(ring, MI_USER_INTERRUPT); intel_ring_advance(ring); - *result = seqno; return 0; } -int -intel_ring_sync(struct intel_ring_buffer *ring, - struct intel_ring_buffer *to, +/** + * intel_ring_sync - sync the waiter to the signaller on seqno + * + * @waiter - ring that is waiting + * @signaller - ring which has, or will signal + * @seqno - seqno which the waiter will block on + */ +static int +intel_ring_sync(struct intel_ring_buffer *waiter, + struct intel_ring_buffer *signaller, + int ring, u32 seqno) { int ret; + u32 dw1 = MI_SEMAPHORE_MBOX | + MI_SEMAPHORE_COMPARE | + MI_SEMAPHORE_REGISTER; - ret = intel_ring_begin(ring, 4); + ret = intel_ring_begin(waiter, 4); if (ret) return ret; - intel_ring_emit(ring, - MI_SEMAPHORE_MBOX | - MI_SEMAPHORE_REGISTER | - intel_ring_sync_index(ring, to) << 17 | - MI_SEMAPHORE_COMPARE); - intel_ring_emit(ring, seqno); - intel_ring_emit(ring, 0); - intel_ring_emit(ring, MI_NOOP); - intel_ring_advance(ring); + intel_ring_emit(waiter, dw1 | signaller->semaphore_register[ring]); + intel_ring_emit(waiter, seqno); + intel_ring_emit(waiter, 0); + intel_ring_emit(waiter, MI_NOOP); + intel_ring_advance(waiter); return 0; } +/* VCS->RCS (RVSYNC) or BCS->RCS (RBSYNC) */ +int +render_ring_sync_to(struct intel_ring_buffer *waiter, + struct intel_ring_buffer *signaller, + u32 seqno) +{ + WARN_ON(signaller->semaphore_register[RCS] == MI_SEMAPHORE_SYNC_INVALID); + return intel_ring_sync(waiter, + signaller, + RCS, + seqno); +} + +/* RCS->VCS (VRSYNC) or BCS->VCS (VBSYNC) */ +int +gen6_bsd_ring_sync_to(struct intel_ring_buffer *waiter, + struct intel_ring_buffer *signaller, + u32 seqno) +{ + WARN_ON(signaller->semaphore_register[VCS] == MI_SEMAPHORE_SYNC_INVALID); + return intel_ring_sync(waiter, + signaller, + VCS, + seqno); +} + +/* RCS->BCS (BRSYNC) or VCS->BCS (BVSYNC) */ +int +gen6_blt_ring_sync_to(struct intel_ring_buffer *waiter, + struct intel_ring_buffer *signaller, + u32 seqno) +{ + WARN_ON(signaller->semaphore_register[BCS] == MI_SEMAPHORE_SYNC_INVALID); + return intel_ring_sync(waiter, + signaller, + BCS, + seqno); +} + + + #define PIPE_CONTROL_FLUSH(ring__, addr__) \ do { \ intel_ring_emit(ring__, GFX_OP_PIPE_CONTROL | PIPE_CONTROL_QW_WRITE | \ @@ -1027,6 +1075,11 @@ static const struct intel_ring_buffer render_ring = { .irq_put = render_ring_put_irq, .dispatch_execbuffer = render_ring_dispatch_execbuffer, .cleanup = render_ring_cleanup, + .sync_to = render_ring_sync_to, + .semaphore_register = {MI_SEMAPHORE_SYNC_INVALID, + MI_SEMAPHORE_SYNC_RV, + MI_SEMAPHORE_SYNC_RB}, + .signal_mbox = {GEN6_VRSYNC, GEN6_BRSYNC}, }; /* ring buffer for bit-stream decoder */ @@ -1154,6 +1207,11 @@ static const struct intel_ring_buffer gen6_bsd_ring = { .irq_get = gen6_bsd_ring_get_irq, .irq_put = gen6_bsd_ring_put_irq, .dispatch_execbuffer = gen6_ring_dispatch_execbuffer, + .sync_to = gen6_bsd_ring_sync_to, + .semaphore_register = {MI_SEMAPHORE_SYNC_VR, + MI_SEMAPHORE_SYNC_INVALID, + MI_SEMAPHORE_SYNC_VB}, + .signal_mbox = {GEN6_RVSYNC, GEN6_BVSYNC}, }; /* Blitter support (SandyBridge+) */ @@ -1281,10 +1339,15 @@ static const struct intel_ring_buffer gen6_blt_ring = { .flush = blt_ring_flush, .add_request = gen6_add_request, .get_seqno = ring_get_seqno, - .irq_get = blt_ring_get_irq, - .irq_put = blt_ring_put_irq, + .irq_get = blt_ring_get_irq, + .irq_put = blt_ring_put_irq, .dispatch_execbuffer = gen6_ring_dispatch_execbuffer, - .cleanup = blt_ring_cleanup, + .cleanup = blt_ring_cleanup, + .sync_to = gen6_blt_ring_sync_to, + .semaphore_register = {MI_SEMAPHORE_SYNC_BR, + MI_SEMAPHORE_SYNC_BV, + MI_SEMAPHORE_SYNC_INVALID}, + .signal_mbox = {GEN6_RBSYNC, GEN6_VBSYNC}, }; int intel_init_render_ring_buffer(struct drm_device *dev) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 600e67abd8d..68281c96c55 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -75,7 +75,12 @@ struct intel_ring_buffer { int (*dispatch_execbuffer)(struct intel_ring_buffer *ring, u32 offset, u32 length); void (*cleanup)(struct intel_ring_buffer *ring); + int (*sync_to)(struct intel_ring_buffer *ring, + struct intel_ring_buffer *to, + u32 seqno); + u32 semaphore_register[3]; /*our mbox written by others */ + u32 signal_mbox[2]; /* mboxes this ring signals to */ /** * List of objects currently involved in rendering from the * ringbuffer. @@ -180,9 +185,6 @@ static inline void intel_ring_emit(struct intel_ring_buffer *ring, void intel_ring_advance(struct intel_ring_buffer *ring); u32 intel_ring_get_seqno(struct intel_ring_buffer *ring); -int intel_ring_sync(struct intel_ring_buffer *ring, - struct intel_ring_buffer *to, - u32 seqno); int intel_init_render_ring_buffer(struct drm_device *dev); int intel_init_bsd_ring_buffer(struct drm_device *dev); -- cgit v1.2.3-70-g09d2 From fcbc34e4dc9ed43f5b147e95beec56c987c683c5 Mon Sep 17 00:00:00 2001 From: Kenneth Graunke Date: Tue, 11 Oct 2011 23:41:08 +0200 Subject: drm/i915: Remove implied length of 2 from GFX_OP_PIPE_CONTROL #define. Not all PIPE_CONTROLs have a length of 2, so remove it from the #define and make each invocation specify the desired length. Signed-off-by: Kenneth Graunke [danvet: implement style suggestion from Ben Widawsdy] Signed-Off-by: Daniel Vetter Signed-off-by: Keith Packard --- drivers/gpu/drm/i915/i915_reg.h | 2 +- drivers/gpu/drm/i915/intel_ringbuffer.c | 8 ++++---- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_ringbuffer.c') diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 28a313a0492..1b07547e26c 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -242,7 +242,7 @@ #define ASYNC_FLIP (1<<22) #define DISPLAY_PLANE_A (0<<20) #define DISPLAY_PLANE_B (1<<20) -#define GFX_OP_PIPE_CONTROL ((0x3<<29)|(0x3<<27)|(0x2<<24)|2) +#define GFX_OP_PIPE_CONTROL(len) ((0x3<<29)|(0x3<<27)|(0x2<<24)|(len-2)) #define PIPE_CONTROL_QW_WRITE (1<<14) #define PIPE_CONTROL_DEPTH_STALL (1<<13) #define PIPE_CONTROL_WC_FLUSH (1<<12) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index 0e99589b54e..7f188c9caf0 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -438,8 +438,8 @@ gen6_blt_ring_sync_to(struct intel_ring_buffer *waiter, #define PIPE_CONTROL_FLUSH(ring__, addr__) \ do { \ - intel_ring_emit(ring__, GFX_OP_PIPE_CONTROL | PIPE_CONTROL_QW_WRITE | \ - PIPE_CONTROL_DEPTH_STALL | 2); \ + intel_ring_emit(ring__, GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE | \ + PIPE_CONTROL_DEPTH_STALL); \ intel_ring_emit(ring__, (addr__) | PIPE_CONTROL_GLOBAL_GTT); \ intel_ring_emit(ring__, 0); \ intel_ring_emit(ring__, 0); \ @@ -467,7 +467,7 @@ pc_render_add_request(struct intel_ring_buffer *ring, if (ret) return ret; - intel_ring_emit(ring, GFX_OP_PIPE_CONTROL | PIPE_CONTROL_QW_WRITE | + intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE | PIPE_CONTROL_WC_FLUSH | PIPE_CONTROL_TC_FLUSH); intel_ring_emit(ring, pc->gtt_offset | PIPE_CONTROL_GLOBAL_GTT); intel_ring_emit(ring, seqno); @@ -483,7 +483,7 @@ pc_render_add_request(struct intel_ring_buffer *ring, PIPE_CONTROL_FLUSH(ring, scratch_addr); scratch_addr += 128; PIPE_CONTROL_FLUSH(ring, scratch_addr); - intel_ring_emit(ring, GFX_OP_PIPE_CONTROL | PIPE_CONTROL_QW_WRITE | + intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE | PIPE_CONTROL_WC_FLUSH | PIPE_CONTROL_TC_FLUSH | PIPE_CONTROL_NOTIFY); intel_ring_emit(ring, pc->gtt_offset | PIPE_CONTROL_GLOBAL_GTT); -- cgit v1.2.3-70-g09d2 From 9d971b37534fb268251f74cc04a36a0a16f7da04 Mon Sep 17 00:00:00 2001 From: Kenneth Graunke Date: Tue, 11 Oct 2011 23:41:09 +0200 Subject: drm/i915: Rename PIPE_CONTROL bit defines to be less terse. "STALL_AT_SCOREBOARD" is much clearer than "STALL_EN" now that there are several different kinds of stalls. Also, "INSTRUCTION_CACHE_INVALIDATE" is a lot easier to understand at a glance than the terse "IS_FLUSH." Signed-off-by: Kenneth Graunke [danvet: use INVALIDATE for ro cache flags for more consistency] Signed-Off-by: Daniel Vetter Signed-off-by: Keith Packard --- drivers/gpu/drm/i915/i915_reg.h | 16 ++++++++-------- drivers/gpu/drm/i915/intel_ringbuffer.c | 6 ++++-- 2 files changed, 12 insertions(+), 10 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_ringbuffer.c') diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 1b07547e26c..d675d79efd4 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -243,15 +243,15 @@ #define DISPLAY_PLANE_A (0<<20) #define DISPLAY_PLANE_B (1<<20) #define GFX_OP_PIPE_CONTROL(len) ((0x3<<29)|(0x3<<27)|(0x2<<24)|(len-2)) -#define PIPE_CONTROL_QW_WRITE (1<<14) -#define PIPE_CONTROL_DEPTH_STALL (1<<13) -#define PIPE_CONTROL_WC_FLUSH (1<<12) -#define PIPE_CONTROL_IS_FLUSH (1<<11) /* MBZ on Ironlake */ -#define PIPE_CONTROL_TC_FLUSH (1<<10) /* GM45+ only */ -#define PIPE_CONTROL_ISP_DIS (1<<9) -#define PIPE_CONTROL_NOTIFY (1<<8) +#define PIPE_CONTROL_QW_WRITE (1<<14) +#define PIPE_CONTROL_DEPTH_STALL (1<<13) +#define PIPE_CONTROL_WRITE_FLUSH (1<<12) +#define PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE (1<<11) /* MBZ on Ironlake */ +#define PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE (1<<10) /* GM45+ only */ +#define PIPE_CONTROL_INDIRECT_STATE_DISABLE (1<<9) +#define PIPE_CONTROL_NOTIFY (1<<8) +#define PIPE_CONTROL_STALL_AT_SCOREBOARD (1<<1) #define PIPE_CONTROL_GLOBAL_GTT (1<<2) /* in addr dword */ -#define PIPE_CONTROL_STALL_EN (1<<1) /* in addr word, Ironlake+ only */ /* diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index 7f188c9caf0..ca8363531a6 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -468,7 +468,8 @@ pc_render_add_request(struct intel_ring_buffer *ring, return ret; intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE | - PIPE_CONTROL_WC_FLUSH | PIPE_CONTROL_TC_FLUSH); + PIPE_CONTROL_WRITE_FLUSH | + PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE); intel_ring_emit(ring, pc->gtt_offset | PIPE_CONTROL_GLOBAL_GTT); intel_ring_emit(ring, seqno); intel_ring_emit(ring, 0); @@ -484,7 +485,8 @@ pc_render_add_request(struct intel_ring_buffer *ring, scratch_addr += 128; PIPE_CONTROL_FLUSH(ring, scratch_addr); intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE | - PIPE_CONTROL_WC_FLUSH | PIPE_CONTROL_TC_FLUSH | + PIPE_CONTROL_WRITE_FLUSH | + PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE | PIPE_CONTROL_NOTIFY); intel_ring_emit(ring, pc->gtt_offset | PIPE_CONTROL_GLOBAL_GTT); intel_ring_emit(ring, seqno); -- cgit v1.2.3-70-g09d2 From 8d31528703ceda6f631e39953130abe9b3ca52b2 Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Sun, 16 Oct 2011 10:23:31 +0200 Subject: drm/i915: Use PIPE_CONTROL for flushing on gen6+. v2 by danvet: Use a new flag to flush the render target cache on gen6+ (hw reuses the old write flush bit), as suggested by Ben Widawsdy. Signed-off-by: Jesse Barnes Signed-off-by: Kenneth Graunke [danvet: this seems to fix cairo-perf-trace hangs on my snb] Signed-Off-by: Daniel Vetter Signed-off-by: Keith Packard --- drivers/gpu/drm/i915/i915_reg.h | 6 ++ drivers/gpu/drm/i915/intel_ringbuffer.c | 136 +++++++++++++++++++++++++++++--- 2 files changed, 130 insertions(+), 12 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_ringbuffer.c') diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index d675d79efd4..6f6eed16963 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -243,14 +243,20 @@ #define DISPLAY_PLANE_A (0<<20) #define DISPLAY_PLANE_B (1<<20) #define GFX_OP_PIPE_CONTROL(len) ((0x3<<29)|(0x3<<27)|(0x2<<24)|(len-2)) +#define PIPE_CONTROL_CS_STALL (1<<20) #define PIPE_CONTROL_QW_WRITE (1<<14) #define PIPE_CONTROL_DEPTH_STALL (1<<13) #define PIPE_CONTROL_WRITE_FLUSH (1<<12) +#define PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH (1<<12) /* gen6+ */ #define PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE (1<<11) /* MBZ on Ironlake */ #define PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE (1<<10) /* GM45+ only */ #define PIPE_CONTROL_INDIRECT_STATE_DISABLE (1<<9) #define PIPE_CONTROL_NOTIFY (1<<8) +#define PIPE_CONTROL_VF_CACHE_INVALIDATE (1<<4) +#define PIPE_CONTROL_CONST_CACHE_INVALIDATE (1<<3) +#define PIPE_CONTROL_STATE_CACHE_INVALIDATE (1<<2) #define PIPE_CONTROL_STALL_AT_SCOREBOARD (1<<1) +#define PIPE_CONTROL_DEPTH_CACHE_FLUSH (1<<0) #define PIPE_CONTROL_GLOBAL_GTT (1<<2) /* in addr dword */ diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index ca8363531a6..ca70e2f1044 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -34,6 +34,16 @@ #include "i915_trace.h" #include "intel_drv.h" +/* + * 965+ support PIPE_CONTROL commands, which provide finer grained control + * over cache flushing. + */ +struct pipe_control { + struct drm_i915_gem_object *obj; + volatile u32 *cpu_page; + u32 gtt_offset; +}; + static inline int ring_space(struct intel_ring_buffer *ring) { int space = (ring->head & HEAD_ADDR) - (ring->tail + 8); @@ -123,6 +133,118 @@ render_ring_flush(struct intel_ring_buffer *ring, return 0; } +/** + * Emits a PIPE_CONTROL with a non-zero post-sync operation, for + * implementing two workarounds on gen6. From section 1.4.7.1 + * "PIPE_CONTROL" of the Sandy Bridge PRM volume 2 part 1: + * + * [DevSNB-C+{W/A}] Before any depth stall flush (including those + * produced by non-pipelined state commands), software needs to first + * send a PIPE_CONTROL with no bits set except Post-Sync Operation != + * 0. + * + * [Dev-SNB{W/A}]: Before a PIPE_CONTROL with Write Cache Flush Enable + * =1, a PIPE_CONTROL with any non-zero post-sync-op is required. + * + * And the workaround for these two requires this workaround first: + * + * [Dev-SNB{W/A}]: Pipe-control with CS-stall bit set must be sent + * BEFORE the pipe-control with a post-sync op and no write-cache + * flushes. + * + * And this last workaround is tricky because of the requirements on + * that bit. From section 1.4.7.2.3 "Stall" of the Sandy Bridge PRM + * volume 2 part 1: + * + * "1 of the following must also be set: + * - Render Target Cache Flush Enable ([12] of DW1) + * - Depth Cache Flush Enable ([0] of DW1) + * - Stall at Pixel Scoreboard ([1] of DW1) + * - Depth Stall ([13] of DW1) + * - Post-Sync Operation ([13] of DW1) + * - Notify Enable ([8] of DW1)" + * + * The cache flushes require the workaround flush that triggered this + * one, so we can't use it. Depth stall would trigger the same. + * Post-sync nonzero is what triggered this second workaround, so we + * can't use that one either. Notify enable is IRQs, which aren't + * really our business. That leaves only stall at scoreboard. + */ +static int +intel_emit_post_sync_nonzero_flush(struct intel_ring_buffer *ring) +{ + struct pipe_control *pc = ring->private; + u32 scratch_addr = pc->gtt_offset + 128; + int ret; + + + ret = intel_ring_begin(ring, 6); + if (ret) + return ret; + + intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(5)); + intel_ring_emit(ring, PIPE_CONTROL_CS_STALL | + PIPE_CONTROL_STALL_AT_SCOREBOARD); + intel_ring_emit(ring, scratch_addr | PIPE_CONTROL_GLOBAL_GTT); /* address */ + intel_ring_emit(ring, 0); /* low dword */ + intel_ring_emit(ring, 0); /* high dword */ + intel_ring_emit(ring, MI_NOOP); + intel_ring_advance(ring); + + ret = intel_ring_begin(ring, 6); + if (ret) + return ret; + + intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(5)); + intel_ring_emit(ring, PIPE_CONTROL_QW_WRITE); + intel_ring_emit(ring, scratch_addr | PIPE_CONTROL_GLOBAL_GTT); /* address */ + intel_ring_emit(ring, 0); + intel_ring_emit(ring, 0); + intel_ring_emit(ring, MI_NOOP); + intel_ring_advance(ring); + + return 0; +} + +static int +gen6_render_ring_flush(struct intel_ring_buffer *ring, + u32 invalidate_domains, u32 flush_domains) +{ + u32 flags = 0; + struct pipe_control *pc = ring->private; + u32 scratch_addr = pc->gtt_offset + 128; + int ret; + + /* Force SNB workarounds for PIPE_CONTROL flushes */ + intel_emit_post_sync_nonzero_flush(ring); + + /* Just flush everything. Experiments have shown that reducing the + * number of bits based on the write domains has little performance + * impact. + */ + flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH; + flags |= PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE; + flags |= PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE; + flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH; + flags |= PIPE_CONTROL_VF_CACHE_INVALIDATE; + flags |= PIPE_CONTROL_CONST_CACHE_INVALIDATE; + flags |= PIPE_CONTROL_STATE_CACHE_INVALIDATE; + + ret = intel_ring_begin(ring, 6); + if (ret) + return ret; + + intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(5)); + intel_ring_emit(ring, flags); + intel_ring_emit(ring, scratch_addr | PIPE_CONTROL_GLOBAL_GTT); + intel_ring_emit(ring, 0); /* lower dword */ + intel_ring_emit(ring, 0); /* uppwer dword */ + intel_ring_emit(ring, MI_NOOP); + intel_ring_advance(ring); + + return 0; +} + static void ring_write_tail(struct intel_ring_buffer *ring, u32 value) { @@ -206,16 +328,6 @@ static int init_ring_common(struct intel_ring_buffer *ring) return 0; } -/* - * 965+ support PIPE_CONTROL commands, which provide finer grained control - * over cache flushing. - */ -struct pipe_control { - struct drm_i915_gem_object *obj; - volatile u32 *cpu_page; - u32 gtt_offset; -}; - static int init_pipe_control(struct intel_ring_buffer *ring) { @@ -296,8 +408,7 @@ static int init_render_ring(struct intel_ring_buffer *ring) GFX_MODE_ENABLE(GFX_REPLAY_MODE)); } - if (INTEL_INFO(dev)->gen >= 6) { - } else if (IS_GEN5(dev)) { + if (INTEL_INFO(dev)->gen >= 5) { ret = init_pipe_control(ring); if (ret) return ret; @@ -1360,6 +1471,7 @@ int intel_init_render_ring_buffer(struct drm_device *dev) *ring = render_ring; if (INTEL_INFO(dev)->gen >= 6) { ring->add_request = gen6_add_request; + ring->flush = gen6_render_ring_flush; ring->irq_get = gen6_render_ring_get_irq; ring->irq_put = gen6_render_ring_put_irq; } else if (IS_GEN5(dev)) { -- cgit v1.2.3-70-g09d2