diff options
| author | Dave Airlie <airlied@redhat.com> | 2012-06-28 13:38:39 +0100 |
|---|---|---|
| committer | Dave Airlie <airlied@redhat.com> | 2012-06-28 13:38:39 +0100 |
| commit | 4391b2cf4b09954b6e8ff31bf63826115fc149f0 (patch) | |
| tree | 0241ce29f4e588336b457338bf2c85c7818e9cce /drivers/gpu/drm/i915 | |
| parent | 0317c6cecd3d6cfd8f920544239a0ba4d3c3f458 (diff) | |
| parent | 7b0cfee1a24efdfe0235bac62e53f686fe8a8e24 (diff) | |
Merge tag 'drm-intel-next-2012-06-21-merged' of git://people.freedesktop.org/~danvet/drm-intel into drm-core-next
Daniel writes:
New -next pull request. Highlights:
- Remaining vlv patches from Jesse et al.
- Some hw workarounds from Jesse
- hw context support from Ben
- full uncore sharing on ivb
- prep work to move the gtt code from intel-gtt.c to drm/i915 for gen6+
- some backlight code improvements
- leftovers for the timeout ioctl (we've forgotten the getparam)
- ibx transcoder workarounds
- some smaller fixlets and improvements
- the new version of the "dont rely on HPD exclusively for VGA" patch
Wrt regressions QA reported quite a few this time around.
- The piglit/kernel-test fallout all has patches that are just awaiting
review and merging into the next -next cycle.
- Which just leaves a bunch of bugs about new modelines that don't work.
It looks like these are all due to the new 16:9/16:10 modes in 3.5
(yeah, only in this manual testing cycle did the git branch used by QA
contain a backmerge of mainline with these patches). Although I haven't
yet confirmed this by letting our QA test the revert of that series.
- Wrt bugs in general I'm trying to fight down some of our long-standing
backlight issues (not regressions), but this seems to be a game of
"you move, you lose" ... :("
Dropped merge bits since this had an -rc4 merge in it to fix some ugly
conflicts.
Diffstat (limited to 'drivers/gpu/drm/i915')
22 files changed, 1475 insertions, 216 deletions
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index 2e9268da58d..b0bacdba6d7 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -7,6 +7,7 @@ i915-y := i915_drv.o i915_dma.o i915_irq.o \ i915_debugfs.o \ i915_suspend.o \ i915_gem.o \ + i915_gem_context.o \ i915_gem_debug.o \ i915_gem_evict.o \ i915_gem_execbuffer.o \ diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index 4fa00fcfbc9..2909b123baf 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -713,6 +713,7 @@ static int i915_error_state(struct seq_file *m, void *unused) seq_printf(m, "EIR: 0x%08x\n", error->eir); seq_printf(m, "IER: 0x%08x\n", error->ier); seq_printf(m, "PGTBL_ER: 0x%08x\n", error->pgtbl_er); + seq_printf(m, "CCID: 0x%08x\n", error->ccid); for (i = 0; i < dev_priv->num_fence_regs; i++) seq_printf(m, " fence[%d] = %08llx\n", i, error->fence[i]); diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c index 262a74d1f85..9563ab8390e 100644 --- a/drivers/gpu/drm/i915/i915_dma.c +++ b/drivers/gpu/drm/i915/i915_dma.c @@ -1006,6 +1006,9 @@ static int i915_getparam(struct drm_device *dev, void *data, case I915_PARAM_HAS_ALIASING_PPGTT: value = dev_priv->mm.aliasing_ppgtt ? 1 : 0; break; + case I915_PARAM_HAS_WAIT_TIMEOUT: + value = 1; + break; default: DRM_DEBUG_DRIVER("Unknown parameter %d\n", param->param); @@ -1082,8 +1085,8 @@ static int i915_set_status_page(struct drm_device *dev, void *data, ring->status_page.gfx_addr = hws->addr & (0x1ffff<<12); - dev_priv->dri1.gfx_hws_cpu_addr = ioremap_wc(dev->agp->base + hws->addr, - 4096); + dev_priv->dri1.gfx_hws_cpu_addr = + ioremap_wc(dev_priv->mm.gtt_base_addr + hws->addr, 4096); if (dev_priv->dri1.gfx_hws_cpu_addr == NULL) { i915_dma_cleanup(dev); ring->status_page.gfx_addr = 0; @@ -1401,6 +1404,27 @@ i915_mtrr_setup(struct drm_i915_private *dev_priv, unsigned long base, } } +static void i915_kick_out_firmware_fb(struct drm_i915_private *dev_priv) +{ + struct apertures_struct *ap; + struct pci_dev *pdev = dev_priv->dev->pdev; + bool primary; + + ap = alloc_apertures(1); + if (!ap) + return; + + ap->ranges[0].base = dev_priv->dev->agp->base; + ap->ranges[0].size = + dev_priv->mm.gtt->gtt_mappable_entries << PAGE_SHIFT; + primary = + pdev->resource[PCI_ROM_RESOURCE].flags & IORESOURCE_ROM_SHADOW; + + remove_conflicting_framebuffers(ap, "inteldrmfb", primary); + + kfree(ap); +} + /** * i915_driver_load - setup chip and create an initial config * @dev: DRM device @@ -1446,6 +1470,22 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags) goto free_priv; } + ret = intel_gmch_probe(dev_priv->bridge_dev, dev->pdev, NULL); + if (!ret) { + DRM_ERROR("failed to set up gmch\n"); + ret = -EIO; + goto put_bridge; + } + + dev_priv->mm.gtt = intel_gtt_get(); + if (!dev_priv->mm.gtt) { + DRM_ERROR("Failed to initialize GTT\n"); + ret = -ENODEV; + goto put_gmch; + } + + i915_kick_out_firmware_fb(dev_priv); + pci_set_master(dev->pdev); /* overlay on gen2 is broken and can't address above 1G */ @@ -1468,26 +1508,22 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags) if (!dev_priv->regs) { DRM_ERROR("failed to map registers\n"); ret = -EIO; - goto put_bridge; - } - - dev_priv->mm.gtt = intel_gtt_get(); - if (!dev_priv->mm.gtt) { - DRM_ERROR("Failed to initialize GTT\n"); - ret = -ENODEV; - goto out_rmmap; + goto put_gmch; } aperture_size = dev_priv->mm.gtt->gtt_mappable_entries << PAGE_SHIFT; + dev_priv->mm.gtt_base_addr = dev_priv->mm.gtt->gma_bus_addr; dev_priv->mm.gtt_mapping = - io_mapping_create_wc(dev->agp->base, aperture_size); + io_mapping_create_wc(dev_priv->mm.gtt_base_addr, + aperture_size); if (dev_priv->mm.gtt_mapping == NULL) { ret = -EIO; goto out_rmmap; } - i915_mtrr_setup(dev_priv, dev->agp->base, aperture_size); + i915_mtrr_setup(dev_priv, dev_priv->mm.gtt_base_addr, + aperture_size); /* The i915 workqueue is primarily used for batched retirement of * requests (and thus managing bo) once the task has been completed @@ -1599,13 +1635,16 @@ out_gem_unload: destroy_workqueue(dev_priv->wq); out_mtrrfree: if (dev_priv->mm.gtt_mtrr >= 0) { - mtrr_del(dev_priv->mm.gtt_mtrr, dev->agp->base, - dev->agp->agp_info.aper_size * 1024 * 1024); + mtrr_del(dev_priv->mm.gtt_mtrr, + dev_priv->mm.gtt_base_addr, + aperture_size); dev_priv->mm.gtt_mtrr = -1; } io_mapping_free(dev_priv->mm.gtt_mapping); out_rmmap: pci_iounmap(dev->pdev, dev_priv->regs); +put_gmch: + intel_gmch_remove(); put_bridge: pci_dev_put(dev_priv->bridge_dev); free_priv: @@ -1637,8 +1676,9 @@ int i915_driver_unload(struct drm_device *dev) io_mapping_free(dev_priv->mm.gtt_mapping); if (dev_priv->mm.gtt_mtrr >= 0) { - mtrr_del(dev_priv->mm.gtt_mtrr, dev->agp->base, - dev->agp->agp_info.aper_size * 1024 * 1024); + mtrr_del(dev_priv->mm.gtt_mtrr, + dev_priv->mm.gtt_base_addr, + dev_priv->mm.gtt->gtt_mappable_entries * PAGE_SIZE); dev_priv->mm.gtt_mtrr = -1; } @@ -1679,6 +1719,7 @@ int i915_driver_unload(struct drm_device *dev) mutex_lock(&dev->struct_mutex); i915_gem_free_all_phys_object(dev); i915_gem_cleanup_ringbuffer(dev); + i915_gem_context_fini(dev); mutex_unlock(&dev->struct_mutex); i915_gem_cleanup_aliasing_ppgtt(dev); i915_gem_cleanup_stolen(dev); @@ -1718,6 +1759,8 @@ int i915_driver_open(struct drm_device *dev, struct drm_file *file) spin_lock_init(&file_priv->mm.lock); INIT_LIST_HEAD(&file_priv->mm.request_list); + idr_init(&file_priv->context_idr); + return 0; } @@ -1750,6 +1793,7 @@ void i915_driver_lastclose(struct drm_device * dev) void i915_driver_preclose(struct drm_device * dev, struct drm_file *file_priv) { + i915_gem_context_close(dev, file_priv); i915_gem_release(dev, file_priv); } @@ -1804,6 +1848,8 @@ struct drm_ioctl_desc i915_ioctls[] = { DRM_IOCTL_DEF_DRV(I915_SET_SPRITE_COLORKEY, intel_sprite_set_colorkey, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED), DRM_IOCTL_DEF_DRV(I915_GET_SPRITE_COLORKEY, intel_sprite_get_colorkey, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED), DRM_IOCTL_DEF_DRV(I915_GEM_WAIT, i915_gem_wait_ioctl, DRM_AUTH|DRM_UNLOCKED), + DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_CREATE, i915_gem_context_create_ioctl, DRM_UNLOCKED), + DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_DESTROY, i915_gem_context_destroy_ioctl, DRM_UNLOCKED), }; int i915_max_ioctl = DRM_ARRAY_SIZE(i915_ioctls); diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 9fe9ebe52a7..a378c080030 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -358,6 +358,9 @@ static const struct pci_device_id pciidlist[] = { /* aka */ INTEL_VGA_DEVICE(0x0406, &intel_haswell_m_info), /* GT1 mobile */ INTEL_VGA_DEVICE(0x0416, &intel_haswell_m_info), /* GT2 mobile */ INTEL_VGA_DEVICE(0x0c16, &intel_haswell_d_info), /* SDV */ + INTEL_VGA_DEVICE(0x0f30, &intel_valleyview_m_info), + INTEL_VGA_DEVICE(0x0157, &intel_valleyview_m_info), + INTEL_VGA_DEVICE(0x0155, &intel_valleyview_d_info), {0, 0, 0} }; @@ -805,7 +808,7 @@ static int gen6_do_reset(struct drm_device *dev) return ret; } -static int intel_gpu_reset(struct drm_device *dev) +int intel_gpu_reset(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; int ret = -ENODEV; @@ -866,8 +869,6 @@ int i915_reset(struct drm_device *dev) if (!mutex_trylock(&dev->struct_mutex)) return -EBUSY; - dev_priv->stop_rings = 0; - i915_gem_reset(dev); ret = -ENODEV; @@ -909,12 +910,16 @@ int i915_reset(struct drm_device *dev) for_each_ring(ring, dev_priv, i) ring->init(ring); + i915_gem_context_init(dev); i915_gem_init_ppgtt(dev); - mutex_unlock(&dev->struct_mutex); + /* + * It would make sense to re-init all the other hw state, at + * least the rps/rc6/emon init done within modeset_init_hw. For + * some unknown reason, this blows up my ilk, so don't. + */ - if (drm_core_check_feature(dev, DRIVER_MODESET)) - intel_modeset_init_hw(dev); + mutex_unlock(&dev->struct_mutex); drm_irq_uninstall(dev); drm_irq_install(dev); @@ -1149,6 +1154,84 @@ MODULE_LICENSE("GPL and additional rights"); ((reg) < 0x40000) && \ ((reg) != FORCEWAKE)) +static bool IS_DISPLAYREG(u32 reg) +{ + /* + * This should make it easier to transition modules over to the + * new register block scheme, since we can do it incrementally. + */ + if (reg >= 0x180000) + return false; + + if (reg >= RENDER_RING_BASE && + reg < RENDER_RING_BASE + 0xff) + return false; + if (reg >= GEN6_BSD_RING_BASE && + reg < GEN6_BSD_RING_BASE + 0xff) + return false; + if (reg >= BLT_RING_BASE && + reg < BLT_RING_BASE + 0xff) + return false; + + if (reg == PGTBL_ER) + return false; + + if (reg >= IPEIR_I965 && + reg < HWSTAM) + return false; + + if (reg == MI_MODE) + return false; + + if (reg == GFX_MODE_GEN7) + return false; + + if (reg == RENDER_HWS_PGA_GEN7 || + reg == BSD_HWS_PGA_GEN7 || + reg == BLT_HWS_PGA_GEN7) + return false; + + if (reg == GEN6_BSD_SLEEP_PSMI_CONTROL || + reg == GEN6_BSD_RNCID) + return false; + + if (reg == GEN6_BLITTER_ECOSKPD) + return false; + + if (reg >= 0x4000c && + reg <= 0x4002c) + return false; + + if (reg >= 0x4f000 && + reg <= 0x4f08f) + return false; + + if (reg >= 0x4f100 && + reg <= 0x4f11f) + return false; + + if (reg >= VLV_MASTER_IER && + reg <= GEN6_PMIER) + return false; + + if (reg >= FENCE_REG_SANDYBRIDGE_0 && + reg < (FENCE_REG_SANDYBRIDGE_0 + (16*8))) + return false; + + if (reg >= VLV_IIR_RW && + reg <= VLV_ISR) + return false; + + if (reg == FORCEWAKE_VLV || + reg == FORCEWAKE_ACK_VLV) + return false; + + if (reg == GEN6_GDRST) + return false; + + return true; +} + #define __i915_read(x, y) \ u##x i915_read##x(struct drm_i915_private *dev_priv, u32 reg) { \ u##x val = 0; \ @@ -1161,6 +1244,8 @@ u##x i915_read##x(struct drm_i915_private *dev_priv, u32 reg) { \ if (dev_priv->forcewake_count == 0) \ dev_priv->display.force_wake_put(dev_priv); \ spin_unlock_irqrestore(&dev_priv->gt_lock, irqflags); \ + } else if (IS_VALLEYVIEW(dev_priv->dev) && IS_DISPLAYREG(reg)) { \ + val = read##y(dev_priv->regs + reg + 0x180000); \ } else { \ val = read##y(dev_priv->regs + reg); \ } \ @@ -1181,7 +1266,11 @@ void i915_write##x(struct drm_i915_private *dev_priv, u32 reg, u##x val) { \ if (NEEDS_FORCE_WAKE((dev_priv), (reg))) { \ __fifo_ret = __gen6_gt_wait_for_fifo(dev_priv); \ } \ - write##y(val, dev_priv->regs + reg); \ + if (IS_VALLEYVIEW(dev_priv->dev) && IS_DISPLAYREG(reg)) { \ + write##y(val, dev_priv->regs + reg + 0x180000); \ + } else { \ + write##y(val, dev_priv->regs + reg); \ + } \ if (unlikely(__fifo_ret)) { \ gen6_gt_check_fifodbg(dev_priv); \ } \ diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 7cc36dbffbd..a0c15abbdce 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -176,6 +176,7 @@ struct drm_i915_error_state { u32 eir; u32 pgtbl_er; u32 ier; + u32 ccid; bool waiting[I915_NUM_RINGS]; u32 pipestat[I915_MAX_PIPES]; u32 tail[I915_NUM_RINGS]; @@ -309,6 +310,17 @@ struct i915_hw_ppgtt { dma_addr_t scratch_page_dma_addr; }; + +/* This must match up with the value previously used for execbuf2.rsvd1. */ +#define DEFAULT_CONTEXT_ID 0 +struct i915_hw_context { + int id; + bool is_initialized; + struct drm_i915_file_private *file_priv; + struct intel_ring_buffer *ring; + struct drm_i915_gem_object *obj; +}; + enum no_fbc_reason { FBC_NO_OUTPUT, /* no outputs enabled to compress */ FBC_STOLEN_TOO_SMALL, /* not enough space to hold compressed buffers */ @@ -652,6 +664,7 @@ typedef struct drm_i915_private { unsigned long gtt_end; struct io_mapping *gtt_mapping; + phys_addr_t gtt_base_addr; int gtt_mtrr; /** PPGTT used for aliasing the PPGTT with the GTT */ @@ -821,6 +834,8 @@ typedef struct drm_i915_private { struct drm_property *force_audio_property; struct work_struct parity_error_work; + bool hw_contexts_disabled; + uint32_t hw_context_size; } drm_i915_private_t; /* Iterate over initialised rings */ @@ -1030,6 +1045,7 @@ struct drm_i915_file_private { struct spinlock lock; struct list_head request_list; } mm; + struct idr context_idr; }; #define INTEL_INFO(dev) (((struct drm_i915_private *) (dev)->dev_private)->info) @@ -1075,7 +1091,8 @@ struct drm_i915_file_private { #define HAS_LLC(dev) (INTEL_INFO(dev)->has_llc) #define I915_NEED_GFX_HWS(dev) (INTEL_INFO(dev)->need_gfx_hws) -#define HAS_ALIASING_PPGTT(dev) (INTEL_INFO(dev)->gen >=6) +#define HAS_HW_CONTEXTS(dev) (INTEL_INFO(dev)->gen >= 6) +#define HAS_ALIASING_PPGTT(dev) (INTEL_INFO(dev)->gen >=6 && !IS_VALLEYVIEW(dev)) #define HAS_OVERLAY(dev) (INTEL_INFO(dev)->has_overlay) #define OVERLAY_NEEDS_PHYSICAL(dev) (INTEL_INFO(dev)->overlay_needs_physical) @@ -1170,6 +1187,7 @@ extern long i915_compat_ioctl(struct file *filp, unsigned int cmd, extern int i915_emit_box(struct drm_device *dev, struct drm_clip_rect *box, int DR1, int DR4); +extern int intel_gpu_reset(struct drm_device *dev); extern int i915_reset(struct drm_device *dev); extern unsigned long i915_chipset_val(struct drm_i915_private *dev_priv); extern unsigned long i915_mch_val(struct drm_i915_private *dev_priv); @@ -1365,6 +1383,16 @@ struct drm_gem_object *i915_gem_prime_import(struct drm_device *dev, struct dma_buf *i915_gem_prime_export(struct drm_device *dev, struct drm_gem_object *gem_obj, int flags); +/* i915_gem_context.c */ +void i915_gem_context_init(struct drm_device *dev); +void i915_gem_context_fini(struct drm_device *dev); +void i915_gem_context_close(struct drm_device *dev, struct drm_file *file); +int i915_switch_context(struct intel_ring_buffer *ring, + struct drm_file *file, int to_id); +int i915_gem_context_create_ioctl(struct drm_device *dev, void *data, + struct drm_file *file); +int i915_gem_context_destroy_ioctl(struct drm_device *dev, void *data, + struct drm_file *file); /* i915_gem_gtt.c */ int __must_check i915_gem_init_aliasing_ppgtt(struct drm_device *dev); diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index af67803e635..6a98c065932 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -1122,7 +1122,7 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf) obj->fault_mappable = true; - pfn = ((dev->agp->base + obj->gtt_offset) >> PAGE_SHIFT) + + pfn = ((dev_priv->mm.gtt_base_addr + obj->gtt_offset) >> PAGE_SHIFT) + page_offset; /* Finally, remap it using the new GTT offset */ @@ -1568,6 +1568,21 @@ i915_add_request(struct intel_ring_buffer *ring, int was_empty; int ret; + /* + * Emit any outstanding flushes - execbuf can fail to emit the flush + * after having emitted the batchbuffer command. Hence we need to fix + * things up similar to emitting the lazy request. The difference here + * is that the flush _must_ happen before the next request, no matter + * what. + */ + if (ring->gpu_caches_dirty) { + ret = i915_gem_flush_ring(ring, 0, I915_GEM_GPU_DOMAINS); + if (ret) + return ret; + + ring->gpu_caches_dirty = false; + } + BUG_ON(request == NULL); seqno = i915_gem_next_request_seqno(ring); @@ -1613,6 +1628,9 @@ i915_add_request(struct intel_ring_buffer *ring, queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work, HZ); } + + WARN_ON(!list_empty(&ring->gpu_write_list)); + return 0; } @@ -1827,14 +1845,11 @@ i915_gem_retire_work_handler(struct work_struct *work) */ idle = true; for_each_ring(ring, dev_priv, i) { - if (!list_empty(&ring->gpu_write_list)) { + if (ring->gpu_caches_dirty) { struct drm_i915_gem_request *request; - int ret; - ret = i915_gem_flush_ring(ring, - 0, I915_GEM_GPU_DOMAINS); request = kzalloc(sizeof(*request), GFP_KERNEL); - if (ret || request == NULL || + if (request == NULL || i915_add_request(ring, NULL, request)) kfree(request); } @@ -2082,11 +2097,14 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file) struct drm_i915_gem_wait *args = data; struct drm_i915_gem_object *obj; struct intel_ring_buffer *ring = NULL; - struct timespec timeout; + struct timespec timeout_stack, *timeout = NULL; u32 seqno = 0; int ret = 0; - timeout = ns_to_timespec(args->timeout_ns); + if (args->timeout_ns >= 0) { + timeout_stack = ns_to_timespec(args->timeout_ns); + timeout = &timeout_stack; + } ret = i915_mutex_lock_interruptible(dev); if (ret) @@ -2122,9 +2140,11 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file) drm_gem_object_unreference(&obj->base); mutex_unlock(&dev->struct_mutex); - ret = __wait_seqno(ring, seqno, true, &timeout); - WARN_ON(!timespec_valid(&timeout)); - args->timeout_ns = timespec_to_ns(&timeout); + ret = __wait_seqno(ring, seqno, true, timeout); + if (timeout) { + WARN_ON(!timespec_valid(timeout)); + args->timeout_ns = timespec_to_ns(timeout); + } return ret; out: @@ -2327,6 +2347,10 @@ int i915_gpu_idle(struct drm_device *dev) /* Is the device fubar? */ if (WARN_ON(!list_empty(&ring->gpu_write_list))) return -EBUSY; + + ret = i915_switch_context(ring, NULL, DEFAULT_CONTEXT_ID); + if (ret) + return ret; } return 0; @@ -3684,6 +3708,9 @@ i915_gem_init_hw(struct drm_device *dev) drm_i915_private_t *dev_priv = dev->dev_private; int ret; + if (!intel_enable_gtt()) + return -EIO; + i915_gem_l3_remap(dev); i915_gem_init_swizzling(dev); @@ -3706,6 +3733,11 @@ i915_gem_init_hw(struct drm_device *dev) dev_priv->next_seqno = 1; + /* + * XXX: There was some w/a described somewhere suggesting loading + * contexts before PPGTT. + */ + i915_gem_context_init(dev); i915_gem_init_ppgtt(dev); return 0; diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c new file mode 100644 index 00000000000..e58e8366f47 --- /dev/null +++ b/drivers/gpu/drm/i915/i915_gem_context.c @@ -0,0 +1,535 @@ +/* + * Copyright © 2011-2012 Intel Corporation + * + * 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 (including the next + * paragraph) 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 AUTHORS OR COPYRIGHT HOLDERS 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 Widawsky <ben@bwidawsk.net> + * + */ + +/* + * This file implements HW context support. On gen5+ a HW context consists of an + * opaque GPU object which is referenced at times of context saves and restores. + * With RC6 enabled, the context is also referenced as the GPU enters and exists + * from RC6 (GPU has it's own internal power context, except on gen5). Though + * something like a context does exist for the media ring, the code only + * supports contexts for the render ring. + * + * In software, there is a distinction between contexts created by the user, + * and the default HW context. The default HW context is used by GPU clients + * that do not request setup of their own hardware context. The default + * context's state is never restored to help prevent programming errors. This + * would happen if a client ran and piggy-backed off another clients GPU state. + * The default context only exists to give the GPU some offset to load as the + * current to invoke a save of the context we actually care about. In fact, the + * code could likely be constructed, albeit in a more complicated fashion, to + * never use the default context, though that limits the driver's ability to + * swap out, and/or destroy other contexts. + * + * All other contexts are created as a request by the GPU client. These contexts + * store GPU state, and thus allow GPU clients to not re-emit state (and + * potentially query certain state) at any time. The kernel driver makes + * certain that the appropriate commands are inserted. + * + * The context life cycle is semi-complicated in that context BOs may live + * longer than the context itself because of the way the hardware, and object + * tracking works. Below is a very crude representation of the state machine + * describing the context life. + * refcount pincount active + * S0: initial state 0 0 0 + * S1: context created 1 0 0 + * S2: context is currently running 2 1 X + * S3: GPU referenced, but not current 2 0 1 + * S4: context is current, but destroyed 1 1 0 + * S5: like S3, but destroyed 1 0 1 + * + * The most common (but not all) transitions: + * S0->S1: client creates a context + * S1->S2: client submits execbuf with context + * S2->S3: other clients submits execbuf with context + * S3->S1: context object was retired + * S3->S2: clients submits another execbuf + * S2->S4: context destroy called with current context + * S3->S5->S0: destroy path + * S4->S5->S0: destroy path on current context + * + * There are two confusing terms used above: + * The "current context" means the context which is currently running on the + * GPU. The GPU has loaded it's state already and has stored away the gtt + * offset of the BO. The GPU is not actively referencing the data at this + * offset, but it will on the next context switch. The only way to avoid this + * is to do a GPU reset. + * + * An "active context' is one which was previously the "current context" and is + * on the active list waiting for the next context switch to occur. Until this + * happens, the object must remain at the same gtt offset. It is therefore + * possible to destroy a context, but it is still active. + * + */ + +#include "drmP.h" +#include "i915_drm.h" +#include "i915_drv.h" + +/* This is a HW constraint. The value below is the largest known requirement + * I've seen in a spec to date, and that was a workaround for a non-shipping + * part. It should be safe to decrease this, but it's more future proof as is. + */ +#define CONTEXT_ALIGN (64<<10) + +static struct i915_hw_context * +i915_gem_context_get(struct drm_i915_file_private *file_priv, u32 id); +static int do_switch(struct drm_i915_gem_object *from_obj, + struct i915_hw_context *to, u32 seqno); + +static int get_context_size(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + int ret; + u32 reg; + + switch (INTEL_INFO(dev)->gen) { + case 6: + reg = I915_READ(CXT_SIZE); + ret = GEN6_CXT_TOTAL_SIZE(reg) * 64; + break; + case 7: + reg = I915_READ(GEN7_CTX_SIZE); + ret = GEN7_CTX_TOTAL_SIZE(reg) * 64; + break; + default: + BUG(); + } + + return ret; +} + +static void do_destroy(struct i915_hw_context *ctx) +{ + struct drm_device *dev = ctx->obj->base.dev; + struct drm_i915_private *dev_priv = dev->dev_private; + + if (ctx->file_priv) + idr_remove(&ctx->file_priv->context_idr, ctx->id); + else + BUG_ON(ctx != dev_priv->ring[RCS].default_context); + + drm_gem_object_unreference(&ctx->obj->base); + kfree(ctx); +} + +static int +create_hw_context(struct drm_device *dev, + struct drm_i915_file_private *file_priv, + struct i915_hw_context **ctx_out) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + int ret, id; + + *ctx_out = kzalloc(sizeof(struct drm_i915_file_private), GFP_KERNEL); + if (*ctx_out == NULL) + return -ENOMEM; + + (*ctx_out)->obj = i915_gem_alloc_object(dev, + dev_priv->hw_context_size); + if ((*ctx_out)->obj == NULL) { + kfree(*ctx_out); + DRM_DEBUG_DRIVER("Context object allocated failed\n"); + return -ENOMEM; + } + + /* The ring associated with the context object is handled by the normal + * object tracking code. We give an initial ring value simple to pass an + * assertion in the context switch code. + */ + (*ctx_out)->ring = &dev_priv->ring[RCS]; + + /* Default context will never have a file_priv */ + if (file_priv == NULL) + return 0; + + (*ctx_out)->file_priv = file_priv; + +again: + if (idr_pre_get(&file_priv->context_idr, GFP_KERNEL) == 0) { + ret = -ENOMEM; + DRM_DEBUG_DRIVER("idr allocation failed\n"); + goto err_out; + } + + ret = idr_get_new_above(&file_priv->context_idr, *ctx_out, + DEFAULT_CONTEXT_ID + 1, &id); + if (ret == 0) + (*ctx_out)->id = id; + + if (ret == -EAGAIN) + goto again; + else if (ret) + goto err_out; + + return 0; + +err_out: + do_destroy(*ctx_out); + return ret; +} + +static inline bool is_default_context(struct i915_hw_context *ctx) +{ + return (ctx == ctx->ring->default_context); +} + +/** + * The default context needs to exist per ring that uses contexts. It stores the + * context state of the GPU for applications that don't utilize HW contexts, as + * well as an idle case. + */ +static int create_default_context(struct drm_i915_private *dev_priv) +{ + struct i915_hw_context *ctx; + int ret; + + BUG_ON(!mutex_is_locked(&dev_priv->dev->struct_mutex)); + + ret = create_hw_context(dev_priv->dev, NULL, + &dev_priv->ring[RCS].default_context); + if (ret) + return ret; + + /* We may need to do things with the shrinker which require us to + * immediately switch back to the default context. This can cause a + * problem as pinning the default context also requires GTT space which + * may not be available. To avoid this we always pin the + * default context. + */ + ctx = dev_priv->ring[RCS].default_context; + ret = i915_gem_object_pin(ctx->obj, CONTEXT_ALIGN, false); + if (ret) { + do_destroy(ctx); + return ret; + } + + ret = do_switch(NULL, ctx, 0); + if (ret) { + i915_gem_object_unpin(ctx->obj); + do_destroy(ctx); + } else { + DRM_DEBUG_DRIVER("Default HW context loaded\n"); + } + + return ret; +} + +void i915_gem_context_init(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + uint32_t ctx_size; + + if (!HAS_HW_CONTEXTS(dev)) { + dev_priv->hw_contexts_disabled = true; + return; + } + + /* If called from reset, or thaw... we've been here already */ + if (dev_priv->hw_contexts_disabled || + dev_priv->ring[RCS].default_context) + return; + + ctx_size = get_context_size(dev); + dev_priv->hw_context_size = get_context_size(dev); + dev_priv->hw_context_size = round_up(dev_priv->hw_context_size, 4096); + + if (ctx_size <= 0 || ctx_size > (1<<20)) { + dev_priv->hw_contexts_disabled = true; + return; + } + + if (create_default_context(dev_priv)) { + dev_priv->hw_contexts_disabled = true; + return; + } + + DRM_DEBUG_DRIVER("HW context support initialized\n"); +} + +void i915_gem_context_fini(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + + if (dev_priv->hw_contexts_disabled) + return; + + /* The only known way to stop the gpu from accessing the hw context is + * to reset it. Do this as the very last operation to avoid confusing + * other code, leading to spurious errors. */ + intel_gpu_reset(dev); + + i915_gem_object_unpin(dev_priv->ring[RCS].default_context->obj); + + do_destroy(dev_priv->ring[RCS].default_context); +} + +static int context_idr_cleanup(int id, void *p, void *data) +{ + struct i915_hw_context *ctx = p; + + BUG_ON(id == DEFAULT_CONTEXT_ID); + + do_destroy(ctx); + + return 0; +} + +void i915_gem_context_close(struct drm_device *dev, struct drm_file *file) +{ + struct drm_i915_file_private *file_priv = file->driver_priv; + + mutex_lock(&dev->struct_mutex); + idr_for_each(&file_priv->context_idr, context_idr_cleanup, NULL); + idr_destroy(&file_priv->context_idr); + mutex_unlock(&dev->struct_mutex); +} + +static struct i915_hw_context * +i915_gem_context_get(struct drm_i915_file_private *file_priv, u32 id) +{ + return (struct i915_hw_context *)idr_find(&file_priv->context_idr, id); +} + +static inline int +mi_set_context(struct intel_ring_buffer *ring, + struct i915_hw_context *new_context, + u32 hw_flags) +{ + int ret; + + /* w/a: If Flush TLB Invalidation Mode is enabled, driver must do a TLB + * invalidation prior to MI_SET_CONTEXT. On GEN6 we don't set the value + * explicitly, so we rely on the value at ring init, stored in + * itlb_before_ctx_switch. + */ + if (IS_GEN6(ring->dev) && ring->itlb_before_ctx_switch) { + ret = ring->flush(ring, 0, 0); + if (ret) + return ret; + } + + ret = intel_ring_begin(ring, 6); + if (ret) + return ret; + + if (IS_GEN7(ring->dev)) + intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_DISABLE); + else + intel_ring_emit(ring, MI_NOOP); + + intel_ring_emit(ring, MI_NOOP); + intel_ring_emit(ring, MI_SET_CONTEXT); + intel_ring_emit(ring, new_context->obj->gtt_offset | + MI_MM_SPACE_GTT | + MI_SAVE_EXT_STATE_EN | + MI_RESTORE_EXT_STATE_EN | + hw_flags); + /* w/a: MI_SET_CONTEXT must always be followed by MI_NOOP */ + intel_ring_emit(ring, MI_NOOP); + + if (IS_GEN7(ring->dev)) + intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_ENABLE); + else + intel_ring_emit(ring, MI_NOOP); + + intel_ring_advance(ring); + + return ret; +} + +static int do_switch(struct drm_i915_gem_object *from_obj, + struct i915_hw_context *to, + u32 seqno) +{ + struct intel_ring_buffer *ring = NULL; + u32 hw_flags = 0; + int ret; + + BUG_ON(to == NULL); + BUG_ON(from_obj != NULL && from_obj->pin_count == 0); + + ret = i915_gem_object_pin(to->obj, CONTEXT_ALIGN, false); + if (ret) + return ret; + + if (!to->obj->has_global_gtt_mapping) + i915_gem_gtt_bind_object(to->obj, to->obj->cache_level); + + if (!to->is_initialized || is_default_context(to)) + hw_flags |= MI_RESTORE_INHIBIT; + else if (WARN_ON_ONCE(from_obj == to->obj)) /* not yet expected */ + hw_flags |= MI_FORCE_RESTORE; + + ring = to->ring; + ret = mi_set_context(ring, to, hw_flags); + if (ret) { + i915_gem_object_unpin(to->obj); + return ret; + } + + /* The backing object for the context is done after switching to the + * *next* context. Therefore we cannot retire the previous context until + * the next context has already started running. In fact, the below code + * is a bit suboptimal because the retiring can occur simply after the + * MI_SET_CONTEXT instead of when the next seqno has completed. + */ + if (from_obj != NULL) { + from_obj->base.read_domains = I915_GEM_DOMAIN_INSTRUCTION; + i915_gem_object_move_to_active(from_obj, ring, seqno); + /* As long as MI_SET_CONTEXT is serializing, ie. it flushes the + * whole damn pipeline, we don't need to explicitly mark the + * object dirty. The only exception is that the context must be + * correct in case the object gets swapped out. Ideally we'd be + * able to defer doing this until we know the object would be + * swapped, but there is no way to do that yet. + */ + from_obj->dirty = 1; + BUG_ON(from_obj->ring != to->ring); + i915_gem_object_unpin(from_obj); + } + + ring->last_context_obj = to->obj; + to->is_initialized = true; + + return 0; +} + +/** + * i915_switch_context() - perform a GPU context switch. + * @ring: ring for which we'll execute the context switch + * @file_priv: file_priv associated with the context, may be NULL + * @id: context id number + * @seqno: sequence number by which the new context will be switched to + * @flags: + * + * The context life cycle is simple. The context refcount is incremented and + * decremented by 1 and create and destroy. If the context is in use by the GPU, + * it will have a refoucnt > 1. This allows us to destroy the context abstract + * object while letting the normal object tracking destroy the backing BO. + */ +int i915_switch_context(struct intel_ring_buffer *ring, + struct drm_file *file, + int to_id) +{ + struct drm_i915_private *dev_priv = ring->dev->dev_private; + struct drm_i915_file_private *file_priv = NULL; + struct i915_hw_context *to; + struct drm_i915_gem_object *from_obj = ring->last_context_obj; + int ret; + + if (dev_priv->hw_contexts_disabled) + return 0; + + if (ring != &dev_priv->ring[RCS]) + return 0; + + if (file) + file_priv = file->driver_priv; + + if (to_id == DEFAULT_CONTEXT_ID) { + to = ring->default_context; + } else { + to = i915_gem_context_get(file_priv, to_id); + if (to == NULL) + return -ENOENT; + } + + if (from_obj == to->obj) + return 0; + + ret = do_switch(from_obj, to, i915_gem_next_request_seqno(to->ring)); + if (ret) + return ret; + + /* Just to make the code a little cleaner we take the object reference + * after the switch was successful. It would be more intuitive to ref + * the 'to' object before the switch but we know the refcount must be >0 + * if context_get() succeeded, and we hold struct mutex. So it's safe to + * do this here/now + */ + drm_gem_object_reference(&to->obj->base); + if (from_obj != NULL) + drm_gem_object_unreference(&from_obj->base); + return ret; +} + +int i915_gem_context_create_ioctl(struct drm_device *dev, void *data, + struct drm_file *file) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + struct drm_i915_gem_context_create *args = data; + struct drm_i915_file_private *file_priv = file->driver_priv; + struct i915_hw_context *ctx; + int ret; + + if (!(dev->driver->driver_features & DRIVER_GEM)) + return -ENODEV; + + if (dev_priv->hw_contexts_disabled) + return -ENODEV; + + ret = i915_mutex_lock_interruptible(dev); + if (ret) + return ret; + + ret = create_hw_context(dev, file_priv, &ctx); + mutex_unlock(&dev->struct_mutex); + + args->ctx_id = ctx->id; + DRM_DEBUG_DRIVER("HW context %d created\n", args->ctx_id); + + return ret; +} + +int i915_gem_context_destroy_ioctl(struct drm_device *dev, void *data, + struct drm_file *file) +{ + struct drm_i915_gem_context_destroy *args = data; + struct drm_i915_file_private *file_priv = file->driver_priv; + struct i915_hw_context *ctx; + int ret; + + if (!(dev->driver->driver_features & DRIVER_GEM)) + return -ENODEV; + + ret = i915_mutex_lock_interruptible(dev); + if (ret) + return ret; + + ctx = i915_gem_context_get(file_priv, args->ctx_id); + if (!ctx) { + mutex_unlock(&dev->struct_mutex); + return -ENOENT; + } + + do_destroy(ctx); + + mutex_unlock(&dev->struct_mutex); + + DRM_DEBUG_DRIVER("HW context %d destroyed\n", args->ctx_id); + return 0; +} diff --git a/drivers/gpu/drm/i915/i915_gem_debug.c b/drivers/gpu/drm/i915/i915_gem_debug.c index a4f6aaabca9..bddf7bed183 100644 --- a/drivers/gpu/drm/i915/i915_gem_debug.c +++ b/drivers/gpu/drm/i915/i915_gem_debug.c @@ -132,7 +132,8 @@ i915_gem_object_check_coherency(struct drm_i915_gem_object *obj, int handle) __func__, obj, obj->gtt_offset, handle, obj->size / 1024); - gtt_mapping = ioremap(dev->agp->base + obj->gtt_offset, obj->base.size); + gtt_mapping = ioremap(dev_priv->mm.gtt_base_addr + obj->gtt_offset, + obj->base.size); if (gtt_mapping == NULL) { DRM_ERROR("failed to map GTT space\n"); return; diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index 974a9f1068a..88e2e114189 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -810,33 +810,16 @@ err: return ret; } -static int +static void i915_gem_execbuffer_flush(struct drm_device *dev, uint32_t invalidate_domains, - uint32_t flush_domains, - uint32_t flush_rings) + uint32_t flush_domains) { - drm_i915_private_t *dev_priv = dev->dev_private; - int i, ret; - if (flush_domains & I915_GEM_DOMAIN_CPU) intel_gtt_chipset_flush(); if (flush_domains & I915_GEM_DOMAIN_GTT) wmb(); - - if ((flush_domains | invalidate_domains) & I915_GEM_GPU_DOMAINS) { - for (i = 0; i < I915_NUM_RINGS; i++) - if (flush_rings & (1 << i)) { - ret = i915_gem_flush_ring(&dev_priv->ring[i], - invalidate_domains, - flush_domains); - if (ret) - return ret; - } - } - - return 0; } static int @@ -885,12 +868,9 @@ i915_gem_execbuffer_move_to_gpu(struct intel_ring_buffer *ring, i915_gem_object_set_to_gpu_domain(obj, ring, &cd); if (cd.invalidate_domains | cd.flush_domains) { - ret = i915_gem_execbuffer_flush(ring->dev, - cd.invalidate_domains, - cd.flush_domains, - cd.flush_rings); - if (ret) - return ret; + i915_gem_execbuffer_flush(ring->dev, + cd.invalidate_domains, + cd.flush_domains); } if (cd.flips) { @@ -905,6 +885,11 @@ i915_gem_execbuffer_move_to_gpu(struct intel_ring_buffer *ring, return ret; } + /* Unconditionally invalidate gpu caches. */ + ret = i915_gem_flush_ring(ring, I915_GEM_GPU_DOMAINS, 0); + if (ret) + return ret; + return 0; } @@ -983,26 +968,13 @@ i915_gem_execbuffer_retire_commands(struct drm_device *dev, struct intel_ring_buffer *ring) { struct drm_i915_gem_request *request; - u32 invalidate; - /* - * Ensure that the commands in the batch buffer are - * finished before the interrupt fires. - * - * The sampler always gets flushed on i965 (sigh). - */ - invalidate = I915_GEM_DOMAIN_COMMAND; - if (INTEL_INFO(dev)->gen >= 4) - invalidate |= I915_GEM_DOMAIN_SAMPLER; - if (ring->flush(ring, invalidate, 0)) { - i915_gem_next_request_seqno(ring); - return; - } + /* Unconditionally force add_request to emit a full flush. */ + ring->gpu_caches_dirty = true; /* Add a breadcrumb for the completion of the batch buffer */ request = kzalloc(sizeof(*request), GFP_KERNEL); if (request == NULL || i915_add_request(ring, file, request)) { - i915_gem_next_request_seqno(ring); kfree(request); } } @@ -1044,6 +1016,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data, struct drm_i915_gem_object *batch_obj; struct drm_clip_rect *cliprects = NULL; struct intel_ring_buffer *ring; + u32 ctx_id = i915_execbuffer2_get_context_id(*args); u32 exec_start, exec_len; u32 seqno; u32 mask; @@ -1065,9 +1038,19 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data, break; case I915_EXEC_BSD: ring = &dev_priv->ring[VCS]; + if (ctx_id != 0) { + DRM_DEBUG("Ring %s doesn't support contexts\n", + ring->name); + return -EPERM; + } break; case I915_EXEC_BLT: ring = &dev_priv->ring[BCS]; + if (ctx_id != 0) { + DRM_DEBUG("Ring %s doesn't support contexts\n", + ring->name); + return -EPERM; + } break; default: DRM_DEBUG("execbuf with unknown ring: %d\n", @@ -1261,6 +1244,10 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data, goto err; } + ret = i915_switch_context(ring, file, ctx_id); + if (ret) + goto err; + trace_i915_gem_ring_dispatch(ring, seqno); exec_start = batch_obj->gtt_offset + args->batch_start_offset; @@ -1367,6 +1354,7 @@ i915_gem_execbuffer(struct drm_device *dev, void *data, exec2.num_cliprects = args->num_cliprects; exec2.cliprects_ptr = args->cliprects_ptr; exec2.flags = I915_EXEC_RENDER; + i915_execbuffer2_set_context_id(exec2, 0); ret = i915_gem_do_execbuffer(dev, data, file, &exec2, exec2_list); if (!ret) { diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 75e0b029e1f..23f2ea0f065 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -513,15 +513,10 @@ static irqreturn_t valleyview_irq_handler(DRM_IRQ_ARGS) unsigned long irqflags; int pipe; u32 pipe_stats[I915_MAX_PIPES]; - u32 vblank_status; - int vblank = 0; bool blc_event; atomic_inc(&dev_priv->irq_received); - vblank_status = PIPE_START_VBLANK_INTERRUPT_STATUS | - PIPE_VBLANK_INTERRUPT_STATUS; - while (true) { iir = I915_READ(VLV_IIR); gt_iir = I915_READ(GTIIR); @@ -551,6 +546,16 @@ static irqreturn_t valleyview_irq_handler(DRM_IRQ_ARGS) } spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); + for_each_pipe(pipe) { + if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS) + drm_handle_vblank(dev, pipe); + + if (pipe_stats[pipe] & PLANE_FLIPDONE_INT_STATUS_VLV) { + intel_prepare_page_flip(dev, pipe); + intel_finish_page_flip(dev, pipe); + } + } + /* Consume port. Then clear IIR or we'll miss events */ if (iir & I915_DISPLAY_PORT_INTERRUPT) { u32 hotplug_status = I915_READ(PORT_HOTPLUG_STAT); @@ -565,19 +570,6 @@ static irqreturn_t valleyview_irq_handler(DRM_IRQ_ARGS) I915_READ(PORT_HOTPLUG_STAT); } - - if (iir & I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT) { - drm_handle_vblank(dev, 0); - vblank++; - intel_finish_page_flip(dev, 0); - } - - if (iir & I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT) { - drm_handle_vblank(dev, 1); - vblank++; - intel_finish_page_flip(dev, 0); - } - if (pipe_stats[pipe] & PIPE_LEGACY_BLC_EVENT_STATUS) blc_event = true; @@ -1188,6 +1180,7 @@ static void i915_capture_error_state(struct drm_device *dev) kref_init(&error->ref); error->eir = I915_READ(EIR); error->pgtbl_er = I915_READ(PGTBL_ER); + error->ccid = I915_READ(CCID); if (HAS_PCH_SPLIT(dev)) error->ier = I915_READ(DEIER) | I915_READ(GTIER); @@ -1510,23 +1503,20 @@ static int valleyview_enable_vblank(struct drm_device *dev, int pipe) { drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; unsigned long irqflags; - u32 dpfl, imr; + u32 imr; if (!i915_pipe_enabled(dev, pipe)) return -EINVAL; spin_lock_irqsave(&dev_priv->irq_lock, irqflags); - dpfl = I915_READ(VLV_DPFLIPSTAT); imr = I915_READ(VLV_IMR); - if (pipe == 0) { - dpfl |= PIPEA_VBLANK_INT_EN; + if (pipe == 0) imr &= ~I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT; - } else { - dpfl |= PIPEA_VBLANK_INT_EN; + else imr &= ~I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT; - } - I915_WRITE(VLV_DPFLIPSTAT, dpfl); I915_WRITE(VLV_IMR, imr); + i915_enable_pipestat(dev_priv, pipe, + PIPE_START_VBLANK_INTERRUPT_ENABLE); spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); return 0; @@ -1576,20 +1566,17 @@ static void valleyview_disable_vblank(struct drm_device *dev, int pipe) { drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; unsigned long irqflags; - u32 dpfl, imr; + u32 imr; spin_lock_irqsave(&dev_priv->irq_lock, irqflags); - dpfl = I915_READ(VLV_DPFLIPSTAT); + i915_disable_pipestat(dev_priv, pipe, + PIPE_START_VBLANK_INTERRUPT_ENABLE); imr = I915_READ(VLV_IMR); - if (pipe == 0) { - dpfl &= ~PIPEA_VBLANK_INT_EN; + if (pipe == 0) imr |= I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT; - } else { - dpfl &= ~PIPEB_VBLANK_INT_EN; + else imr |= I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT; - } I915_WRITE(VLV_IMR, imr); - I915_WRITE(VLV_DPFLIPSTAT, dpfl); spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); } @@ -1923,16 +1910,24 @@ static int ivybridge_irq_postinstall(struct drm_device *dev) static int valleyview_irq_postinstall(struct drm_device *dev) { drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; - u32 render_irqs; u32 enable_mask; u32 hotplug_en = I915_READ(PORT_HOTPLUG_EN); + u32 pipestat_enable = PLANE_FLIP_DONE_INT_EN_VLV; u16 msid; enable_mask = I915_DISPLAY_PORT_INTERRUPT; - enable_mask |= I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT | + enable_mask |= I915_DISPLAY_PIPE_A_EVENT_INTERRUPT | + I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT | + I915_DISPLAY_PIPE_B_EVENT_INTERRUPT | I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT; - dev_priv->irq_mask = ~enable_mask; + /* + *Leave vblank interrupts masked initially. enable/disable will + * toggle them based on usage. + */ + dev_priv->irq_mask = (~enable_mask) | + I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT | + I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT; dev_priv->pipestat[0] = 0; dev_priv->pipestat[1] = 0; @@ -1951,26 +1946,27 @@ static int valleyview_irq_postinstall(struct drm_device *dev) I915_WRITE(PIPESTAT(1), 0xffff); POSTING_READ(VLV_IER); + i915_enable_pipestat(dev_priv, 0, pipestat_enable); + i915_enable_pipestat(dev_priv, 1, pipestat_enable); + I915_WRITE(VLV_IIR, 0xffffffff); I915_WRITE(VLV_IIR, 0xffffffff); - render_irqs = GT_GEN6_BLT_FLUSHDW_NOTIFY_INTERRUPT | - GT_GEN6_BLT_CS_ERROR_INTERRUPT | - GT_GEN6_BLT_USER_INTERRUPT | - GT_GEN6_BSD_USER_INTERRUPT | - GT_GEN6_BSD_CS_ERROR_INTERRUPT | - GT_GEN7_L3_PARITY_ERROR_INTERRUPT | - GT_PIPE_NOTIFY | - GT_RENDER_CS_ERROR_INTERRUPT | - GT_SYNC_STATUS | - GT_USER_INTERRUPT; - - dev_priv->gt_irq_mask = ~render_irqs; + dev_priv->gt_irq_mask = ~0; I915_WRITE(GTIIR, I915_READ(GTIIR)); I915_WRITE(GTIIR, I915_READ(GTIIR)); - I915_WRITE(GTIMR, 0); - I915_WRITE(GTIER, render_irqs); + I915_WRITE(GTIMR, dev_priv->gt_irq_mask); + I915_WRITE(GTIER, GT_GEN6_BLT_FLUSHDW_NOTIFY_INTERRUPT | + GT_GEN6_BLT_CS_ERROR_INTERRUPT | + GT_GEN6_BLT_USER_INTERRUPT | + GT_GEN6_BSD_USER_INTERRUPT | + GT_GEN6_BSD_CS_ERROR_INTERRUPT | + GT_GEN7_L3_PARITY_ERROR_INTERRUPT | + GT_PIPE_NOTIFY | + GT_RENDER_CS_ERROR_INTERRUPT | + GT_SYNC_STATUS | + GT_USER_INTERRUPT); POSTING_READ(GTIER); /* ack & enable invalid PTE error interrupts */ diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 35113d8d7b0..9dfc4c5ff31 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -217,6 +217,9 @@ #define MI_DISPLAY_FLIP_IVB_SPRITE_B (3 << 19) #define MI_DISPLAY_FLIP_IVB_PLANE_C (4 << 19) #define MI_DISPLAY_FLIP_IVB_SPRITE_C (5 << 19) +#define MI_ARB_ON_OFF MI_INSTR(0x08, 0) +#define MI_ARB_ENABLE (1<<0) +#define MI_ARB_DISABLE (0<<0) #define MI_SET_CONTEXT MI_INSTR(0x18, 0) #define MI_MM_SPACE_GTT (1<<8) @@ -299,6 +302,7 @@ #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_TLB_INVALIDATE (1<<18) #define PIPE_CONTROL_QW_WRITE (1<<14) #define PIPE_CONTROL_DEPTH_STALL (1<<13) #define PIPE_CONTROL_WRITE_FLUSH (1<<12) @@ -908,6 +912,7 @@ #define DPLL_P2_CLOCK_DIV_MASK 0x03000000 /* i915 */ #define DPLL_FPA01_P1_POST_DIV_MASK 0x00ff0000 /* i915 */ #define DPLL_FPA01_P1_POST_DIV_MASK_PINEVIEW 0x00ff8000 /* Pineview */ +#define DPLL_LOCK_VLV (1<<15) #define DPLL_INTEGRATED_CLOCK_VLV (1<<13) #define SRX_INDEX 0x3c4 @@ -1462,6 +1467,27 @@ */ #define CCID 0x2180 #define CCID_EN (1<<0) +#define CXT_SIZE 0x21a0 +#define GEN6_CXT_POWER_SIZE(cxt_reg) ((cxt_reg >> 24) & 0x3f) +#define GEN6_CXT_RING_SIZE(cxt_reg) ((cxt_reg >> 18) & 0x3f) +#define GEN6_CXT_RENDER_SIZE(cxt_reg) ((cxt_reg >> 12) & 0x3f) +#define GEN6_CXT_EXTENDED_SIZE(cxt_reg) ((cxt_reg >> 6) & 0x3f) +#define GEN6_CXT_PIPELINE_SIZE(cxt_reg) ((cxt_reg >> 0) & 0x3f) +#define GEN6_CXT_TOTAL_SIZE(cxt_reg) (GEN6_CXT_POWER_SIZE(cxt_reg) + \ + GEN6_CXT_RING_SIZE(cxt_reg) + \ + GEN6_CXT_RENDER_SIZE(cxt_reg) + \ + GEN6_CXT_EXTENDED_SIZE(cxt_reg) + \ + GEN6_CXT_PIPELINE_SIZE(cxt_reg)) +#define GEN7_CTX_SIZE 0x21a8 +#define GEN7_CTX_RENDER_SIZE(ctx_reg) ((ctx_reg >> 16) & 0x3f) +#define GEN7_CTX_EXTENDED_SIZE(ctx_reg) ((ctx_reg >> 9) & 0x7f) +#define GEN7_CTX_GT1_SIZE(ctx_reg) ((ctx_reg >> 6) & 0x7) +#define GEN7_CTX_VFSTATE_SIZE(ctx_reg) ((ctx_reg >> 0) & 0x3f) +#define GEN7_CTX_TOTAL_SIZE(ctx_reg) (GEN7_CTX_RENDER_SIZE(ctx_reg) + \ + GEN7_CTX_EXTENDED_SIZE(ctx_reg) + \ + GEN7_CTX_GT1_SIZE(ctx_reg) + \ + GEN7_CTX_VFSTATE_SIZE(ctx_reg)) + /* * Overlay regs */ @@ -1815,18 +1841,35 @@ #define PFIT_AUTO_RATIOS 0x61238 /* Backlight control */ -#define BLC_PWM_CTL 0x61254 -#define BACKLIGHT_MODULATION_FREQ_SHIFT (17) #define BLC_PWM_CTL2 0x61250 /* 965+ only */ -#define BLM_COMBINATION_MODE (1 << 30) +#define BLM_PWM_ENABLE (1 << 31) +#define BLM_COMBINATION_MODE (1 << 30) /* gen4 only */ +#define BLM_PIPE_SELECT (1 << 29) +#define BLM_PIPE_SELECT_IVB (3 << 29) +#define BLM_PIPE_A (0 << 29) +#define BLM_PIPE_B (1 << 29) +#define BLM_PIPE_C (2 << 29) /* ivb + */ +#define BLM_PIPE(pipe) ((pipe) << 29) +#define BLM_POLARITY_I965 (1 << 28) /* gen4 only */ +#define BLM_PHASE_IN_INTERUPT_STATUS (1 << 26) +#define BLM_PHASE_IN_ENABLE (1 << 25) +#define BLM_PHASE_IN_INTERUPT_ENABL (1 << 24) +#define BLM_PHASE_IN_TIME_BASE_SHIFT (16) +#define BLM_PHASE_IN_TIME_BASE_MASK (0xff << 16) +#define BLM_PHASE_IN_COUNT_SHIFT (8) +#define BLM_PHASE_IN_COUNT_MASK (0xff << 8) +#define BLM_PHASE_IN_INCR_SHIFT (0) +#define BLM_PHASE_IN_INCR_MASK (0xff << 0) +#define BLC_PWM_CTL 0x61254 /* * This is the most significant 15 bits of the number of backlight cycles in a * complete cycle of the modulated backlight control. * * The actual value is this field multiplied by two. */ -#define BACKLIGHT_MODULATION_FREQ_MASK (0x7fff << 17) -#define BLM_LEGACY_MODE (1 << 16) +#define BACKLIGHT_MODULATION_FREQ_SHIFT (17) +#define BACKLIGHT_MODULATION_FREQ_MASK (0x7fff << 17) +#define BLM_LEGACY_MODE (1 << 16) /* gen2 only */ /* * This is the number of cycles out of the backlight modulation cycle for which * the backlight is on. @@ -1836,9 +1879,24 @@ */ #define BACKLIGHT_DUTY_CYCLE_SHIFT (0) #define BACKLIGHT_DUTY_CYCLE_MASK (0xffff) +#define BACKLIGHT_DUTY_CYCLE_MASK_PNV (0xfffe) +#define BLM_POLARITY_PNV (1 << 0) /* pnv only */ #define BLC_HIST_CTL 0x61260 +/* New registers for PCH-split platforms. Safe where new bits show up, the + * register layout machtes with gen4 BLC_PWM_CTL[12]. */ +#define BLC_PWM_CPU_CTL2 0x48250 +#define BLC_PWM_CPU_CTL 0x48254 + +/* PCH CTL1 is totally different, all but the below bits are reserved. CTL2 is + * like the normal CTL from gen4 and earlier. Hooray for confusing naming. */ +#define BLC_PWM_PCH_CTL1 0xc8250 +#define BLM_PCH_PWM_ENABLE (1 << 30) +#define BLM_PCH_OVERRIDE_ENABLE (1 << 30) +#define BLM_PCH_POLARITY (1 << 29) +#define BLC_PWM_PCH_CTL2 0xc8254 + /* TV port control */ #define TV_CTL 0x68000 /** Enables the TV encoder */ @@ -2603,13 +2661,13 @@ #define PIPESTAT(pipe) _PIPE(pipe, _PIPEASTAT, _PIPEBSTAT) #define VLV_DPFLIPSTAT 0x70028 -#define PIPEB_LINE_COMPARE_STATUS (1<<29) +#define PIPEB_LINE_COMPARE_INT_EN (1<<29) #define PIPEB_HLINE_INT_EN (1<<28) #define PIPEB_VBLANK_INT_EN (1<<27) #define SPRITED_FLIPDONE_INT_EN (1<<26) #define SPRITEC_FLIPDONE_INT_EN (1<<25) #define PLANEB_FLIPDONE_INT_EN (1<<24) -#define PIPEA_LINE_COMPARE_STATUS (1<<21) +#define PIPEA_LINE_COMPARE_INT_EN (1<<21) #define PIPEA_HLINE_INT_EN (1<<20) #define PIPEA_VBLANK_INT_EN (1<<19) #define SPRITEB_FLIPDONE_INT_EN (1<<18) @@ -3844,7 +3902,6 @@ #define ADPA_CRT_HOTPLUG_FORCE_TRIGGER (1<<16) /* or SDVOB */ -#define VLV_HDMIB 0x61140 #define HDMIB 0xe1140 #define PORT_ENABLE (1 << 31) #define TRANSCODER(pipe) ((pipe) << 30) @@ -3875,20 +3932,18 @@ #define PCH_LVDS 0xe1180 #define LVDS_DETECTED (1 << 1) -#define BLC_PWM_CPU_CTL2 0x48250 -#define PWM_ENABLE (1 << 31) -#define PWM_PIPE_A (0 << 29) -#define PWM_PIPE_B (1 << 29) -#define BLC_PWM_CPU_CTL 0x48254 +/* vlv has 2 sets of panel control regs. */ +#define PIPEA_PP_STATUS 0x61200 +#define PIPEA_PP_CONTROL 0x61204 +#define PIPEA_PP_ON_DELAYS 0x61208 +#define PIPEA_PP_OFF_DELAYS 0x6120c +#define PIPEA_PP_DIVISOR 0x61210 -#define BLC_PWM_PCH_CTL1 0xc8250 -#define PWM_PCH_ENABLE (1 << 31) -#define PWM_POLARITY_ACTIVE_LOW (1 << 29) -#define PWM_POLARITY_ACTIVE_HIGH (0 << 29) -#define PWM_POLARITY_ACTIVE_LOW2 (1 << 28) -#define PWM_POLARITY_ACTIVE_HIGH2 (0 << 28) - -#define BLC_PWM_PCH_CTL2 0xc8254 +#define PIPEB_PP_STATUS 0x61300 +#define PIPEB_PP_CONTROL 0x61304 +#define PIPEB_PP_ON_DELAYS 0x61308 +#define PIPEB_PP_OFF_DELAYS 0x6130c +#define PIPEB_PP_DIVISOR 0x61310 #define PCH_PP_STATUS 0xc7200 #define PCH_PP_CONTROL 0xc7204 @@ -4032,10 +4087,15 @@ # define GEN6_CSUNIT_CLOCK_GATE_DISABLE (1 << 7) #define GEN6_UCGCTL2 0x9404 +# define GEN7_VDSUNIT_CLOCK_GATE_DISABLE (1 << 30) +# define GEN7_TDLUNIT_CLOCK_GATE_DISABLE (1 << 22) # define GEN6_RCZUNIT_CLOCK_GATE_DISABLE (1 << 13) # define GEN6_RCPBUNIT_CLOCK_GATE_DISABLE (1 << 12) # define GEN6_RCCUNIT_CLOCK_GATE_DISABLE (1 << 11) +#define GEN7_UCGCTL4 0x940c +#define GEN7_L3BANK2X_CLOCK_GATE_DISABLE (1<<25) + #define GEN6_RPNSWREQ 0xA008 #define GEN6_TURBO_DISABLE (1<<31) #define GEN6_FREQUENCY(x) ((x)<<25) diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c index 75a70c46ef1..61d55d3141c 100644 --- a/drivers/gpu/drm/i915/intel_crt.c +++ b/drivers/gpu/drm/i915/intel_crt.c @@ -88,6 +88,9 @@ static void gmch_crt_dpms(struct drm_encoder *encoder, int mode) temp &= ~(ADPA_HSYNC_CNTL_DISABLE | ADPA_VSYNC_CNTL_DISABLE); temp &= ~ADPA_DAC_ENABLE; + if (IS_VALLEYVIEW(dev) && mode != DRM_MODE_DPMS_ON) + mode = DRM_MODE_DPMS_OFF; + switch (mode) { case DRM_MODE_DPMS_ON: temp |= ADPA_DAC_ENABLE; @@ -230,6 +233,42 @@ static bool intel_ironlake_crt_detect_hotplug(struct drm_connector *connector) return ret; } +static bool valleyview_crt_detect_hotplug(struct drm_connector *connector) +{ + struct drm_device *dev = connector->dev; + struct drm_i915_private *dev_priv = dev->dev_private; + u32 adpa; + bool ret; + u32 save_adpa; + + save_adpa = adpa = I915_READ(ADPA); + DRM_DEBUG_KMS("trigger hotplug detect cycle: adpa=0x%x\n", adpa); + + adpa |= ADPA_CRT_HOTPLUG_FORCE_TRIGGER; + + I915_WRITE(ADPA, adpa); + + if (wait_for((I915_READ(ADPA) & ADPA_CRT_HOTPLUG_FORCE_TRIGGER) == 0, + 1000)) { + DRM_DEBUG_KMS("timed out waiting for FORCE_TRIGGER"); + I915_WRITE(ADPA, save_adpa); + } + + /* Check the status to see if both blue and green are on now */ + adpa = I915_READ(ADPA); + if ((adpa & ADPA_CRT_HOTPLUG_MONITOR_MASK) != 0) + ret = true; + else + ret = false; + + DRM_DEBUG_KMS("valleyview hotplug adpa=0x%x, result %d\n", adpa, ret); + + /* FIXME: debug force function and remove */ + ret = true; + + return ret; +} + /** * Uses CRT_HOTPLUG_EN and CRT_HOTPLUG_STAT to detect CRT presence. * @@ -249,6 +288,9 @@ static bool intel_crt_detect_hotplug(struct drm_connector *connector) if (HAS_PCH_SPLIT(dev)) return intel_ironlake_crt_detect_hotplug(connector); + if (IS_VALLEYVIEW(dev)) + return valleyview_crt_detect_hotplug(connector); + /* * On 4 series desktop, CRT detect sequence need to be done twice * to get a reliable result. @@ -453,18 +495,27 @@ intel_crt_detect(struct drm_connector *connector, bool force) struct intel_load_detect_pipe tmp; if (I915_HAS_HOTPLUG(dev)) { + /* We can not rely on the HPD pin always being correctly wired + * up, for example many KVM do not pass it through, and so + * only trust an assertion that the monitor is connected. + */ if (intel_crt_detect_hotplug(connector)) { DRM_DEBUG_KMS("CRT detected via hotplug\n"); return connector_status_connected; - } else { + } else DRM_DEBUG_KMS("CRT not detected via hotplug\n"); - return connector_status_disconnected; - } } if (intel_crt_detect_ddc(connector)) return connector_status_connected; + /* Load detection is broken on HPD capable machines. Whoever wants a + * broken monitor (without edid) to work behind a broken kvm (that fails + * to have the right resistors for HP detection) needs to fix this up. + * For now just bail out. */ + if (I915_HAS_HOTPLUG(dev)) + return connector_status_disconnected; + if (!force) return connector->status; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index add1a15dc8b..b3052ef70d1 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -98,6 +98,11 @@ intel_find_pll_ironlake_dp(const intel_limit_t *, struct drm_crtc *crtc, int target, int refclk, intel_clock_t *match_clock, intel_clock_t *best_clock); +static bool +intel_vlv_find_best_pll(const intel_limit_t *limit, struct drm_crtc *crtc, + int target, int refclk, intel_clock_t *match_clock, + intel_clock_t *best_clock); + static inline u32 /* units of 100MHz */ intel_fdi_link_freq(struct drm_device *dev) { @@ -359,6 +364,48 @@ static const intel_limit_t intel_limits_ironlake_display_port = { .find_pll = intel_find_pll_ironlake_dp, }; +static const intel_limit_t intel_limits_vlv_dac = { + .dot = { .min = 25000, .max = 270000 }, + .vco = { .min = 4000000, .max = 6000000 }, + .n = { .min = 1, .max = 7 }, + .m = { .min = 22, .max = 450 }, /* guess */ + .m1 = { .min = 2, .max = 3 }, + .m2 = { .min = 11, .max = 156 }, + .p = { .min = 10, .max = 30 }, + .p1 = { .min = 2, .max = 3 }, + .p2 = { .dot_limit = 270000, + .p2_slow = 2, .p2_fast = 20 }, + .find_pll = intel_vlv_find_best_pll, +}; + +static const intel_limit_t intel_limits_vlv_hdmi = { + .dot = { .min = 20000, .max = 165000 }, + .vco = { .min = 5994000, .max = 4000000 }, + .n = { .min = 1, .max = 7 }, + .m = { .min = 60, .max = 300 }, /* guess */ + .m1 = { .min = 2, .max = 3 }, + .m2 = { .min = 11, .max = 156 }, + .p = { .min = 10, .max = 30 }, + .p1 = { .min = 2, .max = 3 }, + .p2 = { .dot_limit = 270000, + .p2_slow = 2, .p2_fast = 20 }, + .find_pll = intel_vlv_find_best_pll, +}; + +static const intel_limit_t intel_limits_vlv_dp = { + .dot = { .min = 162000, .max = 270000 }, + .vco = { .min = 5994000, .max = 4000000 }, + .n = { .min = 1, .max = 7 }, + .m = { .min = 60, .max = 300 }, /* guess */ + .m1 = { .min = 2, .max = 3 }, + .m2 = { .min = 11, .max = 156 }, + .p = { .min = 10, .max = 30 }, + .p1 = { .min = 2, .max = 3 }, + .p2 = { .dot_limit = 270000, + .p2_slow = 2, .p2_fast = 20 }, + .find_pll = intel_vlv_find_best_pll, +}; + u32 intel_dpio_read(struct drm_i915_private *dev_priv, int reg) { unsigned long flags; @@ -384,6 +431,28 @@ out_unlock: return val; } +static void intel_dpio_write(struct drm_i915_private *dev_priv, int reg, + u32 val) +{ + unsigned long flags; + + spin_lock_irqsave(&dev_priv->dpio_lock, flags); + if (wait_for_atomic_us((I915_READ(DPIO_PKT) & DPIO_BUSY) == 0, 100)) { + DRM_ERROR("DPIO idle wait timed out\n"); + goto out_unlock; + } + + I915_WRITE(DPIO_DATA, val); + I915_WRITE(DPIO_REG, reg); + I915_WRITE(DPIO_PKT, DPIO_RID | DPIO_OP_WRITE | DPIO_PORTID | + DPIO_BYTE); + if (wait_for_atomic_us((I915_READ(DPIO_PKT) & DPIO_BUSY) == 0, 100)) + DRM_ERROR("DPIO write wait timed out\n"); + +out_unlock: + spin_unlock_irqrestore(&dev_priv->dpio_lock, flags); +} + static void vlv_init_dpio(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -434,7 +503,7 @@ static bool is_dual_link_lvds(struct drm_i915_private *dev_priv, * register is uninitialized. */ val = I915_READ(reg); - if (!(val & ~LVDS_DETECTED)) + if (!(val & ~(LVDS_PIPE_MASK | LVDS_DETECTED))) val = dev_priv->bios_lvds_val; dev_priv->lvds_val = val; } @@ -510,6 +579,13 @@ static const intel_limit_t *intel_limit(struct drm_crtc *crtc, int refclk) limit = &intel_limits_pineview_lvds; else limit = &intel_limits_pineview_sdvo; + } else if (IS_VALLEYVIEW(dev)) { + if (intel_pipe_has_type(crtc, INTEL_OUTPUT_ANALOG)) + limit = &intel_limits_vlv_dac; + else if (intel_pipe_has_type(crtc, INTEL_OUTPUT_HDMI)) + limit = &intel_limits_vlv_hdmi; + else + limit = &intel_limits_vlv_dp; } else if (!IS_GEN2(dev)) { if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS)) limit = &intel_limits_i9xx_lvds; @@ -783,6 +859,73 @@ intel_find_pll_g4x_dp(const intel_limit_t *limit, struct drm_crtc *crtc, memcpy(best_clock, &clock, sizeof(intel_clock_t)); return true; } +static bool +intel_vlv_find_best_pll(const intel_limit_t *limit, struct drm_crtc *crtc, + int target, int refclk, intel_clock_t *match_clock, + intel_clock_t *best_clock) +{ + u32 p1, p2, m1, m2, vco, bestn, bestm1, bestm2, bestp1, bestp2; + u32 m, n, fastclk; + u32 updrate, minupdate, fracbits, p; + unsigned long bestppm, ppm, absppm; + int dotclk, flag; + + dotclk = target * 1000; + bestppm = 1000000; + ppm = absppm = 0; + fastclk = dotclk / (2*100); + updrate = 0; + minupdate = 19200; + fracbits = 1; + n = p = p1 = p2 = m = m1 = m2 = vco = bestn = 0; + bestm1 = bestm2 = bestp1 = bestp2 = 0; + + /* based on hardware requirement, prefer smaller n to precision */ + for (n = limit->n.min; n <= ((refclk) / minupdate); n++) { + updrate = refclk / n; + for (p1 = limit->p1.max; p1 > limit->p1.min; p1--) { + for (p2 = limit->p2.p2_fast+1; p2 > 0; p2--) { + if (p2 > 10) + p2 = p2 - 1; + p = p1 * p2; + /* based on hardware requirement, prefer bigger m1,m2 values */ + for (m1 = limit->m1.min; m1 <= limit->m1.max; m1++) { + m2 = (((2*(fastclk * p * n / m1 )) + + refclk) / (2*refclk)); + m = m1 * m2; + vco = updrate * m; + if (vco >= limit->vco.min && vco < limit->vco.max) { + ppm = 1000000 * ((vco / p) - fastclk) / fastclk; + absppm = (ppm > 0) ? ppm : (-ppm); + if (absppm < 100 && ((p1 * p2) > (bestp1 * bestp2))) { + bestppm = 0; + flag = 1; + } + if (absppm < bestppm - 10) { + bestppm = absppm; + flag = 1; + } + if (flag) { + bestn = n; + bestm1 = m1; + bestm2 = m2; + bestp1 = p1; + bestp2 = p2; + flag = 0; + } + } + } + } + } + } + best_clock->n = bestn; + best_clock->m1 = bestm1; + best_clock->m2 = bestm2; + best_clock->p1 = bestp1; + best_clock->p2 = bestp2; + + return true; +} static void ironlake_wait_for_vblank(struct drm_device *dev, int pipe) { @@ -1232,6 +1375,9 @@ static void assert_pch_dp_disabled(struct drm_i915_private *dev_priv, WARN(dp_pipe_enabled(dev_priv, pipe, port_sel, val), "PCH DP (0x%08x) enabled on transcoder %c, should be disabled\n", reg, pipe_name(pipe)); + + WARN(HAS_PCH_IBX(dev_priv->dev) && (val & SDVO_PIPE_B_SELECT), + "IBX PCH dp port still using transcoder B\n"); } static void assert_pch_hdmi_disabled(struct drm_i915_private *dev_priv, @@ -1241,6 +1387,9 @@ static void assert_pch_hdmi_disabled(struct drm_i915_private *dev_priv, WARN(hdmi_pipe_enabled(dev_priv, val, pipe), "PCH HDMI (0x%08x) enabled on transcoder %c, should be disabled\n", reg, pipe_name(pipe)); + + WARN(HAS_PCH_IBX(dev_priv->dev) && (val & SDVO_PIPE_B_SELECT), + "IBX PCH hdmi port still using transcoder B\n"); } static void assert_pch_ports_disabled(struct drm_i915_private *dev_priv, @@ -1287,7 +1436,7 @@ static void intel_enable_pll(struct drm_i915_private *dev_priv, enum pipe pipe) u32 val; /* No really, not for ILK+ */ - BUG_ON(dev_priv->info->gen >= 5); + BUG_ON(!IS_VALLEYVIEW(dev_priv->dev) && dev_priv->info->gen >= 5); /* PLL is protected by panel, make sure we can write it */ if (IS_MOBILE(dev_priv->dev) && !IS_I830(dev_priv->dev)) @@ -1344,7 +1493,7 @@ intel_sbi_write(struct drm_i915_private *dev_priv, u16 reg, u32 value) unsigned long flags; spin_lock_irqsave(&dev_priv->dpio_lock, flags); - if (wait_for((I915_READ(SBI_CTL_STAT) & SBI_READY) == 0, + if (wait_for((I915_READ(SBI_CTL_STAT) & SBI_BUSY) == 0, 100)) { DRM_ERROR("timeout waiting for SBI to become ready\n"); goto out_unlock; @@ -1358,7 +1507,7 @@ intel_sbi_write(struct drm_i915_private *dev_priv, u16 reg, u32 value) SBI_BUSY | SBI_CTL_OP_CRWR); - if (wait_for((I915_READ(SBI_CTL_STAT) & (SBI_READY | SBI_RESPONSE_SUCCESS)) == 0, + if (wait_for((I915_READ(SBI_CTL_STAT) & (SBI_BUSY | SBI_RESPONSE_FAIL)) == 0, 100)) { DRM_ERROR("timeout waiting for SBI to complete write transaction\n"); goto out_unlock; @@ -1372,10 +1521,10 @@ static u32 intel_sbi_read(struct drm_i915_private *dev_priv, u16 reg) { unsigned long flags; - u32 value; + u32 value = 0; spin_lock_irqsave(&dev_priv->dpio_lock, flags); - if (wait_for((I915_READ(SBI_CTL_STAT) & SBI_READY) == 0, + if (wait_for((I915_READ(SBI_CTL_STAT) & SBI_BUSY) == 0, 100)) { DRM_ERROR("timeout waiting for SBI to become ready\n"); goto out_unlock; @@ -1387,7 +1536,7 @@ intel_sbi_read(struct drm_i915_private *dev_priv, u16 reg) SBI_BUSY | SBI_CTL_OP_CRRD); - if (wait_for((I915_READ(SBI_CTL_STAT) & (SBI_READY | SBI_RESPONSE_SUCCESS)) == 0, + if (wait_for((I915_READ(SBI_CTL_STAT) & (SBI_BUSY | SBI_RESPONSE_FAIL)) == 0, 100)) { DRM_ERROR("timeout waiting for SBI to complete read transaction\n"); goto out_unlock; @@ -3666,13 +3815,37 @@ static bool intel_choose_pipe_bpp_dither(struct drm_crtc *crtc, return display_bpc != bpc; } +static int vlv_get_refclk(struct drm_crtc *crtc) +{ + struct drm_device *dev = crtc->dev; + struct drm_i915_private *dev_priv = dev->dev_private; + int refclk = 27000; /* for DP & HDMI */ + + return 100000; /* only one validated so far */ + + if (intel_pipe_has_type(crtc, INTEL_OUTPUT_ANALOG)) { + refclk = 96000; + } else if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS)) { + if (intel_panel_use_ssc(dev_priv)) + refclk = 100000; + else + refclk = 96000; + } else if (intel_pipe_has_type(crtc, INTEL_OUTPUT_EDP)) { + refclk = 100000; + } + + return refclk; +} + static int i9xx_get_refclk(struct drm_crtc *crtc, int num_connectors) { struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; int refclk; - if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS) && + if (IS_VALLEYVIEW(dev)) { + refclk = vlv_get_refclk(crtc); + } else if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS) && intel_panel_use_ssc(dev_priv) && num_connectors < 2) { refclk = dev_priv->lvds_ssc_freq * 1000; DRM_DEBUG_KMS("using SSC reference clock of %d MHz\n", @@ -3787,6 +3960,72 @@ static void intel_update_lvds(struct drm_crtc *crtc, intel_clock_t *clock, I915_WRITE(LVDS, temp); } +static void vlv_update_pll(struct drm_crtc *crtc, + struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode, + intel_clock_t *clock, intel_clock_t *reduced_clock, + int refclk, int num_connectors) +{ + struct drm_device *dev = crtc->dev; + struct drm_i915_private *dev_priv = dev->dev_private; + struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + int pipe = intel_crtc->pipe; + u32 dpll, mdiv, pdiv; + u32 bestn, bestm1, bestm2, bestp1, bestp2; + bool is_hdmi; + + is_hdmi = intel_pipe_has_type(crtc, INTEL_OUTPUT_HDMI); + + bestn = clock->n; + bestm1 = clock->m1; + bestm2 = clock->m2; + bestp1 = clock->p1; + bestp2 = clock->p2; + + /* Enable DPIO clock input */ + dpll = DPLL_EXT_BUFFER_ENABLE_VLV | DPLL_REFA_CLK_ENABLE_VLV | + DPLL_VGA_MODE_DIS | DPLL_INTEGRATED_CLOCK_VLV; + I915_WRITE(DPLL(pipe), dpll); + POSTING_READ(DPLL(pipe)); + + mdiv = ((bestm1 << DPIO_M1DIV_SHIFT) | (bestm2 & DPIO_M2DIV_MASK)); + mdiv |= ((bestp1 << DPIO_P1_SHIFT) | (bestp2 << DPIO_P2_SHIFT)); + mdiv |= ((bestn << DPIO_N_SHIFT)); + mdiv |= (1 << DPIO_POST_DIV_SHIFT); + mdiv |= (1 << DPIO_K_SHIFT); + mdiv |= DPIO_ENABLE_CALIBRATION; + intel_dpio_write(dev_priv, DPIO_DIV(pipe), mdiv); + + intel_dpio_write(dev_priv, DPIO_CORE_CLK(pipe), 0x01000000); + + pdiv = DPIO_REFSEL_OVERRIDE | (5 << DPIO_PLL_MODESEL_SHIFT) | + (3 << DPIO_BIAS_CURRENT_CTL_SHIFT) | (1<<20) | + (8 << DPIO_DRIVER_CTL_SHIFT) | (5 << DPIO_CLK_BIAS_CTL_SHIFT); + intel_dpio_write(dev_priv, DPIO_REFSFR(pipe), pdiv); + + intel_dpio_write(dev_priv, DPIO_LFP_COEFF(pipe), 0x009f0051); + + dpll |= DPLL_VCO_ENABLE; + I915_WRITE(DPLL(pipe), dpll); + POSTING_READ(DPLL(pipe)); + if (wait_for(((I915_READ(DPLL(pipe)) & DPLL_LOCK_VLV) == DPLL_LOCK_VLV), 1)) + DRM_ERROR("DPLL %d failed to lock\n", pipe); + + if (is_hdmi) { + u32 temp = intel_mode_get_pixel_multiplier(adjusted_mode); + + if (temp > 1) + temp = (temp - 1) << DPLL_MD_UDI_MULTIPLIER_SHIFT; + else + temp = 0; + + I915_WRITE(DPLL_MD(pipe), temp); + POSTING_READ(DPLL_MD(pipe)); + } + + intel_dpio_write(dev_priv, DPIO_FASTCLK_DISABLE, 0x641); /* ??? */ +} + static void i9xx_update_pll(struct drm_crtc *crtc, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode, @@ -4044,6 +4283,9 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc, if (IS_GEN2(dev)) i8xx_update_pll(crtc, adjusted_mode, &clock, num_connectors); + else if (IS_VALLEYVIEW(dev)) + vlv_update_pll(crtc, mode,adjusted_mode, &clock, NULL, + refclk, num_connectors); else i9xx_update_pll(crtc, mode, adjusted_mode, &clock, has_reduced_clock ? &reduced_clock : NULL, @@ -6543,7 +6785,24 @@ static void intel_setup_outputs(struct drm_device *dev) if (!dpd_is_edp && (I915_READ(PCH_DP_D) & DP_DETECTED)) intel_dp_init(dev, PCH_DP_D); + } else if (IS_VALLEYVIEW(dev)) { + int found; + + if (I915_READ(SDVOB) & PORT_DETECTED) { + /* SDVOB multiplex with HDMIB */ + found = intel_sdvo_init(dev, SDVOB, true); + if (!found) + intel_hdmi_init(dev, SDVOB); + if (!found && (I915_READ(DP_B) & DP_DETECTED)) + intel_dp_init(dev, DP_B); + } + + if (I915_READ(SDVOC) & PORT_DETECTED) + intel_hdmi_init(dev, SDVOC); + /* Shares lanes with HDMI on SDVOC */ + if (I915_READ(DP_C) & DP_DETECTED) + intel_dp_init(dev, DP_C); } else if (SUPPORTS_DIGITAL_OUTPUTS(dev)) { bool found = false; @@ -6906,9 +7165,9 @@ static void ivb_pch_pwm_override(struct drm_device *dev) * IVB has CPU eDP backlight regs too, set things up to let the * PCH regs control the backlight */ - I915_WRITE(BLC_PWM_CPU_CTL2, PWM_ENABLE); + I915_WRITE(BLC_PWM_CPU_CTL2, BLM_PWM_ENABLE); I915_WRITE(BLC_PWM_CPU_CTL, 0); - I915_WRITE(BLC_PWM_PCH_CTL1, PWM_ENABLE | (1<<30)); + I915_WRITE(BLC_PWM_PCH_CTL1, BLM_PCH_PWM_ENABLE | BLM_PCH_OVERRIDE_ENABLE); } void intel_modeset_init_hw(struct drm_device *dev) @@ -6965,7 +7224,7 @@ void intel_modeset_init(struct drm_device *dev) dev->mode_config.max_width = 8192; dev->mode_config.max_height = 8192; } - dev->mode_config.fb_base = dev->agp->base; + dev->mode_config.fb_base = dev_priv->mm.gtt_base_addr; DRM_DEBUG_KMS("%d display pipe%s available.\n", dev_priv->num_pipe, dev_priv->num_pipe > 1 ? "s" : ""); diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index c35edd7ca84..5290e9df327 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -169,6 +169,7 @@ struct intel_crtc { u8 lut_r[256], lut_g[256], lut_b[256]; int dpms_mode; bool active; /* is the crtc on? independent of the dpms mode */ + bool primary_disabled; /* is the crtc obscured by a plane? */ bool busy; /* is scanout buffer being updated frequently? */ struct timer_list idle_timer; bool lowfreq_avail; @@ -191,7 +192,6 @@ struct intel_plane { struct drm_plane base; enum pipe pipe; struct drm_i915_gem_object *obj; - bool primary_disabled; int max_downscale; u32 lut_r[1024], lut_g[1024], lut_b[1024]; void (*update_plane)(struct drm_plane *plane, @@ -379,7 +379,8 @@ extern u32 intel_panel_get_max_backlight(struct drm_device *dev); extern u32 intel_panel_get_backlight(struct drm_device *dev); extern void intel_panel_set_backlight(struct drm_device *dev, u32 level); extern int intel_panel_setup_backlight(struct drm_device *dev); -extern void intel_panel_enable_backlight(struct drm_device *dev); +extern void intel_panel_enable_backlight(struct drm_device *dev, + enum pipe pipe); extern void intel_panel_disable_backlight(struct drm_device *dev); extern void intel_panel_destroy_backlight(struct drm_device *dev); extern enum drm_connector_status intel_panel_detect(struct drm_device *dev); diff --git a/drivers/gpu/drm/i915/intel_fb.c b/drivers/gpu/drm/i915/intel_fb.c index bf8690720a0..e9f8338bd80 100644 --- a/drivers/gpu/drm/i915/intel_fb.c +++ b/drivers/gpu/drm/i915/intel_fb.c @@ -140,7 +140,9 @@ static int intelfb_create(struct intel_fbdev *ifbdev, info->fix.smem_start = dev->mode_config.fb_base + obj->gtt_offset; info->fix.smem_len = size; - info->screen_base = ioremap_wc(dev->agp->base + obj->gtt_offset, size); + info->screen_base = + ioremap_wc(dev_priv->mm.gtt_base_addr + obj->gtt_offset, + size); if (!info->screen_base) { ret = -ENOSPC; goto out_unpin; diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index b507d38faa1..5b2c88ca6ed 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -37,6 +37,19 @@ #include "i915_drm.h" #include "i915_drv.h" +static void +assert_hdmi_port_disabled(struct intel_hdmi *intel_hdmi) +{ + struct drm_device *dev = intel_hdmi->base.base.dev; + struct drm_i915_private *dev_priv = dev->dev_private; + uint32_t enabled_bits; + + enabled_bits = IS_HASWELL(dev) ? DDI_BUF_CTL_ENABLE : SDVO_ENABLE; + + WARN(I915_READ(intel_hdmi->sdvox_reg) & enabled_bits, + "HDMI port enabled, expecting disabled\n"); +} + struct intel_hdmi *enc_to_intel_hdmi(struct drm_encoder *encoder) { return container_of(encoder, struct intel_hdmi, base.base); @@ -334,6 +347,8 @@ static void g4x_set_infoframes(struct drm_encoder *encoder, u32 val = I915_READ(reg); u32 port; + assert_hdmi_port_disabled(intel_hdmi); + /* If the registers were not initialized yet, they might be zeroes, * which means we're selecting the AVI DIP and we're setting its * frequency to once. This seems to really confuse the HW and make @@ -395,6 +410,8 @@ static void ibx_set_infoframes(struct drm_encoder *encoder, u32 val = I915_READ(reg); u32 port; + assert_hdmi_port_disabled(intel_hdmi); + /* See the big comment in g4x_set_infoframes() */ val |= VIDEO_DIP_SELECT_AVI | VIDEO_DIP_FREQ_VSYNC; @@ -451,6 +468,8 @@ static void cpt_set_infoframes(struct drm_encoder *encoder, u32 reg = TVIDEO_DIP_CTL(intel_crtc->pipe); u32 val = I915_READ(reg); + assert_hdmi_port_disabled(intel_hdmi); + /* See the big comment in g4x_set_infoframes() */ val |= VIDEO_DIP_SELECT_AVI | VIDEO_DIP_FREQ_VSYNC; @@ -484,6 +503,8 @@ static void vlv_set_infoframes(struct drm_encoder *encoder, u32 reg = VLV_TVIDEO_DIP_CTL(intel_crtc->pipe); u32 val = I915_READ(reg); + assert_hdmi_port_disabled(intel_hdmi); + /* See the big comment in g4x_set_infoframes() */ val |= VIDEO_DIP_SELECT_AVI | VIDEO_DIP_FREQ_VSYNC; @@ -516,6 +537,8 @@ static void hsw_set_infoframes(struct drm_encoder *encoder, u32 reg = HSW_TVIDEO_DIP_CTL(intel_crtc->pipe); u32 val = I915_READ(reg); + assert_hdmi_port_disabled(intel_hdmi); + if (!intel_hdmi->has_hdmi_sink) { I915_WRITE(reg, 0); POSTING_READ(reg); @@ -569,7 +592,7 @@ static void intel_hdmi_mode_set(struct drm_encoder *encoder, if (HAS_PCH_CPT(dev)) sdvox |= PORT_TRANS_SEL_CPT(intel_crtc->pipe); - else if (intel_crtc->pipe == 1) + else if (intel_crtc->pipe == PIPE_B) sdvox |= SDVO_PIPE_B_SELECT; I915_WRITE(intel_hdmi->sdvox_reg, sdvox); @@ -591,6 +614,36 @@ static void intel_hdmi_dpms(struct drm_encoder *encoder, int mode) temp = I915_READ(intel_hdmi->sdvox_reg); + /* HW workaround for IBX, we need to move the port to transcoder A + * before disabling it. */ + if (HAS_PCH_IBX(dev)) { + struct drm_crtc *crtc = encoder->crtc; + int pipe = crtc ? to_intel_crtc(crtc)->pipe : -1; + + if (mode != DRM_MODE_DPMS_ON) { + if (temp & SDVO_PIPE_B_SELECT) { + temp &= ~SDVO_PIPE_B_SELECT; + I915_WRITE(intel_hdmi->sdvox_reg, temp); + POSTING_READ(intel_hdmi->sdvox_reg); + + /* Again we need to write this twice. */ + I915_WRITE(intel_hdmi->sdvox_reg, temp); + POSTING_READ(intel_hdmi->sdvox_reg); + + /* Transcoder selection bits only update + * effectively on vblank. */ + if (crtc) + intel_wait_for_vblank(dev, pipe); + else + msleep(50); + } + } else { + /* Restore the transcoder select bit. */ + if (pipe == PIPE_B) + enable_bits |= SDVO_PIPE_B_SELECT; + } + } + /* HW workaround, need to toggle enable bit off and on for 12bpc, but * we do this anyway which shows more stable in testing. */ diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index 08eb04c787e..05fcadbeac6 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -71,6 +71,7 @@ static struct intel_lvds *intel_attached_lvds(struct drm_connector *connector) static void intel_lvds_enable(struct intel_lvds *intel_lvds) { struct drm_device *dev = intel_lvds->base.base.dev; + struct intel_crtc *intel_crtc = to_intel_crtc(intel_lvds->base.base.crtc); struct drm_i915_private *dev_priv = dev->dev_private; u32 ctl_reg, lvds_reg, stat_reg; @@ -107,7 +108,7 @@ static void intel_lvds_enable(struct intel_lvds *intel_lvds) if (wait_for((I915_READ(stat_reg) & PP_ON) != 0, 1000)) DRM_ERROR("timed out waiting for panel to power on\n"); - intel_panel_enable_backlight(dev); + intel_panel_enable_backlight(dev, intel_crtc->pipe); } static void intel_lvds_disable(struct intel_lvds *intel_lvds) @@ -777,6 +778,14 @@ static const struct dmi_system_id intel_no_lvds[] = { DMI_MATCH(DMI_BOARD_NAME, "MS-7469"), }, }, + { + .callback = intel_no_lvds_dmi_callback, + .ident = "ZOTAC ZBOXSD-ID12/ID13", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "ZOTAC"), + DMI_MATCH(DMI_BOARD_NAME, "ZBOXSD-ID12/ID13"), + }, + }, { } /* terminating entry */ }; @@ -967,6 +976,8 @@ bool intel_lvds_init(struct drm_device *dev) intel_encoder->clone_mask = (1 << INTEL_LVDS_CLONE_BIT); if (HAS_PCH_SPLIT(dev)) intel_encoder->crtc_mask = (1 << 0) | (1 << 1) | (1 << 2); + else if (IS_GEN4(dev)) + intel_encoder->crtc_mask = (1 << 0) | (1 << 1); else intel_encoder->crtc_mask = (1 << 1); @@ -1074,35 +1085,14 @@ bool intel_lvds_init(struct drm_device *dev) goto failed; out: + /* + * Unlock registers and just + * leave them unlocked + */ if (HAS_PCH_SPLIT(dev)) { - u32 pwm; - - pipe = (I915_READ(PCH_LVDS) & LVDS_PIPEB_SELECT) ? 1 : 0; - - /* make sure PWM is enabled and locked to the LVDS pipe */ - pwm = I915_READ(BLC_PWM_CPU_CTL2); - if (pipe == 0 && (pwm & PWM_PIPE_B)) - I915_WRITE(BLC_PWM_CPU_CTL2, pwm & ~PWM_ENABLE); - if (pipe) - pwm |= PWM_PIPE_B; - else - pwm &= ~PWM_PIPE_B; - I915_WRITE(BLC_PWM_CPU_CTL2, pwm | PWM_ENABLE); - - pwm = I915_READ(BLC_PWM_PCH_CTL1); - pwm |= PWM_PCH_ENABLE; - I915_WRITE(BLC_PWM_PCH_CTL1, pwm); - /* - * Unlock registers and just - * leave them unlocked - */ I915_WRITE(PCH_PP_CONTROL, I915_READ(PCH_PP_CONTROL) | PANEL_UNLOCK_REGS); } else { - /* - * Unlock registers and just - * leave them unlocked - */ I915_WRITE(PP_CONTROL, I915_READ(PP_CONTROL) | PANEL_UNLOCK_REGS); } diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c index 7180cc828f9..58c7ee7238b 100644 --- a/drivers/gpu/drm/i915/intel_panel.c +++ b/drivers/gpu/drm/i915/intel_panel.c @@ -287,9 +287,18 @@ void intel_panel_disable_backlight(struct drm_device *dev) dev_priv->backlight_enabled = false; intel_panel_actually_set_backlight(dev, 0); + + if (INTEL_INFO(dev)->gen >= 4) { + uint32_t reg; + + reg = HAS_PCH_SPLIT(dev) ? BLC_PWM_CPU_CTL2 : BLC_PWM_CTL2; + + I915_WRITE(reg, I915_READ(reg) & ~BLM_PWM_ENABLE); + } } -void intel_panel_enable_backlight(struct drm_device *dev) +void intel_panel_enable_backlight(struct drm_device *dev, + enum pipe pipe) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -298,6 +307,33 @@ void intel_panel_enable_backlight(struct drm_device *dev) dev_priv->backlight_enabled = true; intel_panel_actually_set_backlight(dev, dev_priv->backlight_level); + + if (INTEL_INFO(dev)->gen >= 4) { + uint32_t reg, tmp; + + reg = HAS_PCH_SPLIT(dev) ? BLC_PWM_CPU_CTL2 : BLC_PWM_CTL2; + + + tmp = I915_READ(reg); + + /* Note that this can also get called through dpms changes. And + * we don't track the backlight dpms state, hence check whether + * we have to do anything first. */ + if (tmp & BLM_PWM_ENABLE) + return; + + if (dev_priv->num_pipe == 3) + tmp &= ~BLM_PIPE_SELECT_IVB; + else + tmp &= ~BLM_PIPE_SELECT; + + tmp |= BLM_PIPE(pipe); + tmp &= ~BLM_PWM_ENABLE; + + I915_WRITE(reg, tmp); + POSTING_READ(reg); + I915_WRITE(reg, tmp | BLM_PWM_ENABLE); + } } static void intel_panel_init_backlight(struct drm_device *dev) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index d0ce2a5b1d3..7504fbce05c 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -405,7 +405,9 @@ void intel_update_fbc(struct drm_device *dev) * - going to an unsupported config (interlace, pixel multiply, etc.) */ list_for_each_entry(tmp_crtc, &dev->mode_config.crtc_list, head) { - if (tmp_crtc->enabled && tmp_crtc->fb) { + if (tmp_crtc->enabled && + !to_intel_crtc(tmp_crtc)->primary_disabled && + tmp_crtc->fb) { if (crtc) { DRM_DEBUG_KMS("more than one pipe active, disabling compression\n"); dev_priv->no_fbc_reason = FBC_MULTIPLE_PIPES; @@ -3328,8 +3330,12 @@ static void gen6_init_clock_gating(struct drm_device *dev) * * According to the spec, bit 11 (RCCUNIT) must also be set, * but we didn't debug actual testcases to find it out. + * + * Also apply WaDisableVDSUnitClockGating and + * WaDisableRCPBUnitClockGating. */ I915_WRITE(GEN6_UCGCTL2, + GEN7_VDSUNIT_CLOCK_GATE_DISABLE | GEN6_RCPBUNIT_CLOCK_GATE_DISABLE | GEN6_RCCUNIT_CLOCK_GATE_DISABLE); @@ -3357,6 +3363,9 @@ static void gen6_init_clock_gating(struct drm_device *dev) ILK_DPARB_CLK_GATE | ILK_DPFD_CLK_GATE); + I915_WRITE(GEN6_MBCTL, I915_READ(GEN6_MBCTL) | + GEN6_MBCTL_ENABLE_BOOT_FETCH); + for_each_pipe(pipe) { I915_WRITE(DSPCNTR(pipe), I915_READ(DSPCNTR(pipe)) | @@ -3382,6 +3391,7 @@ static void ivybridge_init_clock_gating(struct drm_device *dev) struct drm_i915_private *dev_priv = dev->dev_private; int pipe; uint32_t dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE; + uint32_t snpcr; I915_WRITE(PCH_DSPCLK_GATE_D, dspclk_gate); @@ -3389,11 +3399,6 @@ static void ivybridge_init_clock_gating(struct drm_device *dev) I915_WRITE(WM2_LP_ILK, 0); I915_WRITE(WM1_LP_ILK, 0); - /* According to the spec, bit 13 (RCZUNIT) must be set on IVB. - * This implements the WaDisableRCZUnitClockGating workaround. - */ - I915_WRITE(GEN6_UCGCTL2, GEN6_RCZUNIT_CLOCK_GATE_DISABLE); - I915_WRITE(ILK_DSPCLK_GATE, IVB_VRHUNIT_CLK_GATE); I915_WRITE(IVB_CHICKEN3, @@ -3410,6 +3415,23 @@ static void ivybridge_init_clock_gating(struct drm_device *dev) I915_WRITE(GEN7_L3_CHICKEN_MODE_REGISTER, GEN7_WA_L3_CHICKEN_MODE); + /* According to the BSpec vol1g, bit 12 (RCPBUNIT) clock + * gating disable must be set. Failure to set it results in + * flickering pixels due to Z write ordering failures after + * some amount of runtime in the Mesa "fire" demo, and Unigine + * Sanctuary and Tropics, and apparently anything else with + * alpha test or pixel discard. + * + * According to the spec, bit 11 (RCCUNIT) must also be set, + * but we didn't debug actual testcases to find it out. + * + * According to the spec, bit 13 (RCZUNIT) must be set on IVB. + * This implements the WaDisableRCZUnitClockGating workaround. + */ + I915_WRITE(GEN6_UCGCTL2, + GEN6_RCZUNIT_CLOCK_GATE_DISABLE | + GEN6_RCCUNIT_CLOCK_GATE_DISABLE); + /* This is required by WaCatErrorRejectionIssue */ I915_WRITE(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG, I915_READ(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG) | @@ -3422,11 +3444,19 @@ static void ivybridge_init_clock_gating(struct drm_device *dev) intel_flush_display_plane(dev_priv, pipe); } + I915_WRITE(GEN6_MBCTL, I915_READ(GEN6_MBCTL) | + GEN6_MBCTL_ENABLE_BOOT_FETCH); + gen7_setup_fixed_func_scheduler(dev_priv); /* WaDisable4x2SubspanOptimization */ I915_WRITE(CACHE_MODE_1, _MASKED_BIT_ENABLE(PIXEL_SUBSPAN_COLLECT_OPT_DISABLE)); + + snpcr = I915_READ(GEN6_MBCUNIT_SNPCR); + snpcr &= ~GEN6_MBC_SNPCR_MASK; + snpcr |= GEN6_MBC_SNPCR_MED; + I915_WRITE(GEN6_MBCUNIT_SNPCR, snpcr); } static void valleyview_init_clock_gating(struct drm_device *dev) @@ -3441,11 +3471,6 @@ static void valleyview_init_clock_gating(struct drm_device *dev) I915_WRITE(WM2_LP_ILK, 0); I915_WRITE(WM1_LP_ILK, 0); - /* According to the spec, bit 13 (RCZUNIT) must be set on IVB. - * This implements the WaDisableRCZUnitClockGating workaround. - */ - I915_WRITE(GEN6_UCGCTL2, GEN6_RCZUNIT_CLOCK_GATE_DISABLE); - I915_WRITE(ILK_DSPCLK_GATE, IVB_VRHUNIT_CLK_GATE); I915_WRITE(IVB_CHICKEN3, @@ -3465,6 +3490,35 @@ static void valleyview_init_clock_gating(struct drm_device *dev) I915_READ(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG) | GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB); + I915_WRITE(GEN6_MBCTL, I915_READ(GEN6_MBCTL) | + GEN6_MBCTL_ENABLE_BOOT_FETCH); + + + /* According to the BSpec vol1g, bit 12 (RCPBUNIT) clock + * gating disable must be set. Failure to set it results in + * flickering pixels due to Z write ordering failures after + * some amount of runtime in the Mesa "fire" demo, and Unigine + * Sanctuary and Tropics, and apparently anything else with + * alpha test or pixel discard. + * + * According to the spec, bit 11 (RCCUNIT) must also be set, + * but we didn't debug actual testcases to find it out. + * + * According to the spec, bit 13 (RCZUNIT) must be set on IVB. + * This implements the WaDisableRCZUnitClockGating workaround. + * + * Also apply WaDisableVDSUnitClockGating and + * WaDisableRCPBUnitClockGating. + */ + I915_WRITE(GEN6_UCGCTL2, + GEN7_VDSUNIT_CLOCK_GATE_DISABLE | + GEN7_TDLUNIT_CLOCK_GATE_DISABLE | + GEN6_RCZUNIT_CLOCK_GATE_DISABLE | + GEN6_RCPBUNIT_CLOCK_GATE_DISABLE | + GEN6_RCCUNIT_CLOCK_GATE_DISABLE); + + I915_WRITE(GEN7_UCGCTL4, GEN7_L3BANK2X_CLOCK_GATE_DISABLE); + for_each_pipe(pipe) { I915_WRITE(DSPCNTR(pipe), I915_READ(DSPCNTR(pipe)) | @@ -3474,6 +3528,19 @@ static void valleyview_init_clock_gating(struct drm_device *dev) I915_WRITE(CACHE_MODE_1, _MASKED_BIT_ENABLE(PIXEL_SUBSPAN_COLLECT_OPT_DISABLE)); + + /* + * On ValleyView, the GUnit needs to signal the GT + * when flip and other events complete. So enable + * all the GUnit->GT interrupts here + */ + I915_WRITE(VLV_DPFLIPSTAT, PIPEB_LINE_COMPARE_INT_EN | + PIPEB_HLINE_INT_EN | PIPEB_VBLANK_INT_EN | + SPRITED_FLIPDONE_INT_EN | SPRITEC_FLIPDONE_INT_EN | + PLANEB_FLIPDONE_INT_EN | PIPEA_LINE_COMPARE_INT_EN | + PIPEA_HLINE_INT_EN | PIPEA_VBLANK_INT_EN | + SPRITEB_FLIPDONE_INT_EN | SPRITEA_FLIPDONE_INT_EN | + PLANEA_FLIPDONE_INT_EN); } static void g4x_init_clock_gating(struct drm_device *dev) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index f30850b079e..f30a53a8917 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -226,6 +226,7 @@ gen6_render_ring_flush(struct intel_ring_buffer *ring, * impact. */ flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH; + flags |= PIPE_CONTROL_TLB_INVALIDATE; flags |= PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE; flags |= PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE; flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH; @@ -433,6 +434,13 @@ static int init_render_ring(struct intel_ring_buffer *ring) */ I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(CM0_STC_EVICT_DISABLE_LRA_SNB)); + + /* This is not explicitly set for GEN6, so read the register. + * see intel_ring_mi_set_context() for why we care. + * TODO: consider explicitly setting the bit for GEN5 + */ + ring->itlb_before_ctx_switch = + !!(I915_READ(GFX_MODE) & GFX_TLB_INVALIDATE_ALWAYS); } if (INTEL_INFO(dev)->gen >= 6) @@ -979,6 +987,7 @@ static int intel_init_ring_buffer(struct drm_device *dev, struct intel_ring_buffer *ring) { struct drm_i915_gem_object *obj; + struct drm_i915_private *dev_priv = dev->dev_private; int ret; ring->dev = dev; @@ -1012,8 +1021,9 @@ static int intel_init_ring_buffer(struct drm_device *dev, if (ret) goto err_unpin; - ring->virtual_start = ioremap_wc(dev->agp->base + obj->gtt_offset, - ring->size); + ring->virtual_start = + ioremap_wc(dev_priv->mm.gtt->gma_bus_addr + obj->gtt_offset, + ring->size); if (ring->virtual_start == NULL) { DRM_ERROR("Failed to map ringbuffer.\n"); ret = -EINVAL; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 55d3da26bae..1d3c81fdad9 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -113,9 +113,17 @@ struct intel_ring_buffer { * Do we have some not yet emitted requests outstanding? */ u32 outstanding_lazy_request; + bool gpu_caches_dirty; wait_queue_head_t irq_queue; + /** + * Do an explicit TLB flush before MI_SET_CONTEXT + */ + bool itlb_before_ctx_switch; + struct i915_hw_context *default_context; + struct drm_i915_gem_object *last_context_obj; + void *private; }; diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 2a20fb0781d..9d7777bc154 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -326,6 +326,12 @@ intel_enable_primary(struct drm_crtc *crtc) struct intel_crtc *intel_crtc = to_intel_crtc(crtc); int reg = DSPCNTR(intel_crtc->plane); + if (!intel_crtc->primary_disabled) + return; + + intel_crtc->primary_disabled = false; + intel_update_fbc(dev); + I915_WRITE(reg, I915_READ(reg) | DISPLAY_PLANE_ENABLE); } @@ -337,7 +343,13 @@ intel_disable_primary(struct drm_crtc *crtc) struct intel_crtc *intel_crtc = to_intel_crtc(crtc); int reg = DSPCNTR(intel_crtc->plane); + if (intel_crtc->primary_disabled) + return; + I915_WRITE(reg, I915_READ(reg) & ~DISPLAY_PLANE_ENABLE); + + intel_crtc->primary_disabled = true; + intel_update_fbc(dev); } static int @@ -485,18 +497,14 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, * Be sure to re-enable the primary before the sprite is no longer * covering it fully. */ - if (!disable_primary && intel_plane->primary_disabled) { + if (!disable_primary) intel_enable_primary(crtc); - intel_plane->primary_disabled = false; - } intel_plane->update_plane(plane, fb, obj, crtc_x, crtc_y, crtc_w, crtc_h, x, y, src_w, src_h); - if (disable_primary) { + if (disable_primary) intel_disable_primary(crtc); - intel_plane->primary_disabled = true; - } /* Unpin old obj after new one is active to avoid ugliness */ if (old_obj) { @@ -527,11 +535,8 @@ intel_disable_plane(struct drm_plane *plane) struct intel_plane *intel_plane = to_intel_plane(plane); int ret = 0; - if (intel_plane->primary_disabled) { + if (plane->crtc) intel_enable_primary(plane->crtc); - intel_plane->primary_disabled = false; - } - intel_plane->disable_plane(plane); if (!intel_plane->obj) |