dri-devel July 2012

dri-devel@lists.freedesktop.org

115 participants
688 discussions

general protection fault on ttm_init()
by Fengguang Wu 14 Jul '12

14 Jul '12

Hi Thomas, I got this error while running kvm boot tests on linux-next. Attached is the kconfig. [ 7.064962] [drm:drm_core_init] *ERROR* Cannot create /proc/dri [ 7.067443] general protection fault: 0000 [#1] PREEMPT [ 7.068713] CPU 0 [ 7.068713] Pid: 1, comm: swapper Not tainted 3.5.0-rc5-07204-g646e30b #8 Bochs Bochs [ 7.068713] RIP: 0010:[<ffffffff8105e55b>] [<ffffffff8105e55b>] __lock_acquire.isra.19+0x64/0x7c9 [ 7.068713] RSP: 0018:ffff88000006fd10 … [View More]EFLAGS: 00010097 [ 7.068713] RAX: 0000000000000000 RBX: ffff880000070000 RCX: 0000000000000000 [ 7.068713] RDX: 0000000000000000 RSI: 0000000000000000 RDI: a56b6b6b6b6b6e73 [ 7.068713] RBP: ffff88000006fd80 R08: 0000000000000000 R09: 0000000000000000 [ 7.068713] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000246 [ 7.068713] R13: 0000000000000000 R14: 0000000000000000 R15: a56b6b6b6b6b6e73 [ 7.068713] FS: 0000000000000000(0000) GS:ffffffff81a24000(0000) knlGS:0000000000000000 [ 7.068713] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 7.068713] CR2: 0000000000000000 CR3: 0000000001a12000 CR4: 00000000000006f0 [ 7.068713] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 7.068713] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 7.068713] Process swapper (pid: 1, threadinfo ffff88000006e000, task ffff880000070000) [ 7.068713] Stack: [ 7.068713] 0000000000000000 0000000000000046 ffff88000006fd50 0000000081c40020 [ 7.068713] 0000000000000000 ffffffff8105d409 ffff88000006fd90 0000000000000000 [ 7.068713] ffff88000006fde0 ffff880000070000 0000000000000246 ffff8800001c0ab8 [ 7.068713] Call Trace: [ 7.068713] [<ffffffff8105d409>] ? mutex_remove_waiter+0x6a/0x7e [ 7.068713] [<ffffffff8105f074>] lock_acquire+0x5b/0x78 [ 7.068713] [<ffffffff81352a9a>] ? get_device_parent+0x98/0x1b0 [ 7.068713] [<ffffffff81560b72>] _raw_spin_lock+0x3c/0x6f [ 7.068713] [<ffffffff81352a9a>] ? get_device_parent+0x98/0x1b0 [ 7.068713] [<ffffffff81352a9a>] get_device_parent+0x98/0x1b0 [ 7.068713] [<ffffffff8135312e>] device_add+0xf4/0x589 [ 7.068713] [<ffffffff81cff357>] ? drm_core_init+0xca/0xca [ 7.068713] [<ffffffff8135ad4d>] ? pm_runtime_init+0xfa/0xfe [ 7.068713] [<ffffffff81cff357>] ? drm_core_init+0xca/0xca [ 7.068713] [<ffffffff813535d9>] device_register+0x16/0x1a [ 7.068713] [<ffffffff81224acd>] drm_class_device_register+0x17/0x19 [ 7.068713] [<ffffffff81cff38e>] ttm_init+0x37/0x62 [ 7.068713] [<ffffffff81cdbbca>] do_one_initcall+0x78/0x131 [ 7.068713] [<ffffffff81cdbd77>] kernel_init+0xf4/0x178 [ 7.068713] [<ffffffff81cdb45d>] ? loglevel+0x31/0x31 [ 7.068713] [<ffffffff81562e94>] kernel_thread_helper+0x4/0x10 [ 7.068713] [<ffffffff81cdbc83>] ? do_one_initcall+0x131/0x131 [ 7.068713] [<ffffffff81562e90>] ? gs_change+0x13/0x13 [ 7.068713] Code: 49 89 ff 41 89 f5 0f 85 fe 06 00 00 9c 58 66 66 90 66 90 f6 c4 02 0f 84 ee 06 00 00 e8 2b 00 0e 00 31 c0 e9 59 07 00 00 44 89 e8 <4d> 8b 64 c7 08 4d 85 e4 0f 85 04 03 00 00 83 3d 20 7c 18 01 00 [ 7.068713] RIP [<ffffffff8105e55b>] __lock_acquire.isra.19+0x64/0x7c9 [ 7.068713] RSP <ffff88000006fd10> [ 7.068713] ---[ end trace 6e1803df9f758306 ]--- [ 7.068713] Kernel panic - not syncing: Fatal exception Thanks, Fengguang [View Less]

2 8

Re: [RFC] dma-fence: dma-buf synchronization (v2)
by Rob Clark 13 Jul '12

13 Jul '12

On Fri, Jul 13, 2012 at 12:35 PM, Tom Cooksey <tom.cooksey(a)arm.com> wrote: > My other thought is around atomicity. Could this be extended to > (safely) allow for hardware devices which might want to access > multiple buffers simultaneously? I think it probably can with > some tweaks to the interface? An atomic function which does > something like "give me all the fences for all these buffers > and add this fence to each instead/as-well-as"? fwiw, what I'm leaning … [View More]

2 2

[pull] drm-intel-next
by Daniel Vetter 13 Jul '12

13 Jul '12

Hi Dave, New pull for -next. Highlights: - rc6/turbo support for hsw (Eugeni) - improve corner-case of the reset handling code - gpu reset handling should be rock-solid now - support for fb offset > 4096 pixels on gen4+ (yeah, you need some fairly big screens to hit that) - the "Flush Me Harder" patch to fix the gen6+ fallout from disabling the flushing_list - no more /dev/agpgart on gen6+! - HAS_PCH_xxx improvements from Paulo - a few minor bits&pieces all over, most of it in … [View More]thew hsw code QA reported 2 regression, one due a bad cable (fixed by a walk to the next radioshack) and one due to the HPD v2 patch - I owe you one for refusing to take v2 for -fixes after v1 blew up on Linus' machine I guess ;-) The later has a confirmed fix already queued up in my tree. Regressions from the last pull are all fixed and some really good news: We've finally fixed the last DP regression from 3.2. Although I'm vary of that blowing up elseplaces, hence I prefer that we soak it in 3.6 a bit before submitting it to stable. Otherwise Chris is hunting down an obscure bug that got recently introduced due to a funny interaction between two seemingly unrelated patches, one improving our gpu death handling, the other preparing the removal of the flushing_list. But he has patches already, although I'm still complaining a bit about the commit messages ... Wrt further pulls for 3.6 I'll merge feature-y stuff only at the end of the current drm-intel-next cycle so that if this will miss 3.6 I can just send you a pull for the bugfixes that are currently merged (or in the case of Chris' patches, hopefully merged soon). Yours, Daniel PS: This pull will make the already existing conflict with Linus' tree a bit more fun, but I think it should be still doable (the important thing is to keep the revert from -fixes, but don't kill any other changes from -next). The following changes since commit 7b0cfee1a24efdfe0235bac62e53f686fe8a8e24: Merge tag 'v3.5-rc4' into drm-intel-next-queued (2012-06-25 19:10:36 +0200) are available in the git repository at: git://people.freedesktop.org/~danvet/drm-intel tags/drm-intel-next-2012-07-06 for you to fetch changes up to 4acf518626cdad5bbf7aac9869bd4accbbfb4ad3: drm/i915: program FDI_RX TP and FDI delays (2012-07-05 15:09:03 +0200) ---------------------------------------------------------------- Ben Widawsky (1): drm/i915: linuxify create_hw_context() Chris Wilson (2): drm/i915: Group the GT routines together in both code and vtable drm/i915: Implement w/a for sporadic read failures on waking from rc6 Daniel Vetter (15): drm/i915: wrap up gt powersave enabling functions drm/i915: make enable/disable_gt_powersave locking consistent drm/i915: don't use dev->agp drm/i915: disable drm agp support for !gen3 with kms enabled agp/intel-agp: remove snb+ host bridge pciids drm/i915: "Flush Me Harder" required on gen6+ drm/i915: fix up ilk rc6 disabling confusion drm/i915: don't trylock in the gpu reset code drm/i915: non-interruptible sleeps can't handle -EAGAIN drm/i915: don't hang userspace when the gpu reset is stuck drm/i915: properly SIGBUS on I/O errors drm/i915: don't return a spurious -EIO from intel_ring_begin drm/i915: introduce crtc->dspaddr_offset drm/i915: adjust framebuffer base address on gen4+ drm/i915: introduce for_each_encoder_on_crtc Eugeni Dodonov (11): drm/i915: support Haswell force waking drm/i915: add RPS configuration for Haswell drm/i915: slightly improve gt enable/disable routines drm/i915: enable RC6 by default on Haswell drm/i915: disable RC6 when disabling rps drm/i915: introduce haswell_init_clock_gating drm/i915: enable RC6 workaround on Haswell drm/i915: move force wake support into intel_pm drm/i915: re-initialize DDI buffer translations after resume drm/i915: prevent bogus intel_update_fbc notifications drm/i915: program FDI_RX TP and FDI delays Jesper Juhl (1): drm/i915/sprite: Fix mem leak in intel_plane_init() Jesse Barnes (3): drm/i915: mask tiled bit when updating IVB sprites drm/i915: correct IVB default sprite format drm/i915: prefer wide & slow to fast & narrow in DP configs Paulo Zanoni (5): drm/i915: fix PIPE_WM_LINETIME definition drm/i915: add PCH_NONE to enum intel_pch drm/i915: get rid of dev_priv->info->has_pch_split drm/i915: don't ironlake_init_pch_refclk() on LPT drm/i915: fix PIPE_DDI_PORT_MASK Ville Syrjälä (2): drm/i915: Zero initialize mode_cmd drm/i915: Reject page flips with changed format/offset/pitch drivers/char/agp/intel-agp.c | 11 - drivers/gpu/drm/i915/i915_dma.c | 9 +- drivers/gpu/drm/i915/i915_drv.c | 172 ++------------ drivers/gpu/drm/i915/i915_drv.h | 28 ++- drivers/gpu/drm/i915/i915_gem.c | 44 +++- drivers/gpu/drm/i915/i915_gem_context.c | 50 ++-- drivers/gpu/drm/i915/i915_reg.h | 21 +- drivers/gpu/drm/i915/i915_suspend.c | 5 +- drivers/gpu/drm/i915/intel_ddi.c | 9 + drivers/gpu/drm/i915/intel_display.c | 162 ++++++------- drivers/gpu/drm/i915/intel_dp.c | 26 +- drivers/gpu/drm/i915/intel_drv.h | 16 +- drivers/gpu/drm/i915/intel_fb.c | 2 +- drivers/gpu/drm/i915/intel_lvds.c | 6 +- drivers/gpu/drm/i915/intel_pm.c | 396 +++++++++++++++++++++++++------ drivers/gpu/drm/i915/intel_ringbuffer.c | 34 +-- drivers/gpu/drm/i915/intel_sprite.c | 5 +- drivers/gpu/drm/i915/intel_tv.c | 10 +- 18 files changed, 579 insertions(+), 427 deletions(-) -- Daniel Vetter Mail: daniel(a)ffwll.ch Mobile: +41 (0)79 365 57 48 [View Less]

1 0

libdrm: Fix some warnings reported by clang's scan-build tool
by Johannes Obermayr 13 Jul '12

13 Jul '12

Patches 1 to 4 were sent to mesa-dev.

2 13

[PATCH 1/2] drm: Add colouring to the range allocator
by Chris Wilson 12 Jul '12

12 Jul '12

In order to support snoopable memory on non-LLC architectures (so that we can bind vgem objects into the i915 GATT for example), we have to avoid the prefetcher on the GPU from crossing memory domains and so prevent allocation of a snoopable PTE immediately following an uncached PTE. To do that, we need to extend the range allocator with support for tracking and segregating different node colours. This will be used by i915 to segregate memory domains within the GTT. v2: Now with more drm_mm … [View More]helpers and less driver interference. Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: Dave Airlie <airlied(a)redhat.com Cc: Benjamin Herrenschmidt <benh(a)kernel.crashing.org> Cc: Ben Skeggs <bskeggs(a)redhat.com> Cc: Jerome Glisse <jglisse(a)redhat.com> Cc: Alex Deucher <alexander.deucher(a)amd.com> Cc: Daniel Vetter <daniel.vetter(a)ffwll.ch> Cc: dri-devel(a)lists.freedesktop.org --- drivers/gpu/drm/drm_gem.c | 2 +- drivers/gpu/drm/drm_mm.c | 169 ++++++++++++++++++++------------- drivers/gpu/drm/i915/i915_gem.c | 6 +- drivers/gpu/drm/i915/i915_gem_evict.c | 9 +- include/drm/drm_mm.h | 93 +++++++++++++++--- 5 files changed, 191 insertions(+), 88 deletions(-) diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c index d58e69d..fbe0842 100644 --- a/drivers/gpu/drm/drm_gem.c +++ b/drivers/gpu/drm/drm_gem.c @@ -354,7 +354,7 @@ drm_gem_create_mmap_offset(struct drm_gem_object *obj) /* Get a DRM GEM mmap offset allocated... */ list->file_offset_node = drm_mm_search_free(&mm->offset_manager, - obj->size / PAGE_SIZE, 0, 0); + obj->size / PAGE_SIZE, 0, false); if (!list->file_offset_node) { DRM_ERROR("failed to allocate offset for bo %d\n", obj->name); diff --git a/drivers/gpu/drm/drm_mm.c b/drivers/gpu/drm/drm_mm.c index 961fb54..9bb82f7 100644 --- a/drivers/gpu/drm/drm_mm.c +++ b/drivers/gpu/drm/drm_mm.c @@ -118,45 +118,53 @@ static inline unsigned long drm_mm_hole_node_end(struct drm_mm_node *hole_node) static void drm_mm_insert_helper(struct drm_mm_node *hole_node, struct drm_mm_node *node, - unsigned long size, unsigned alignment) + unsigned long size, unsigned alignment, + unsigned long color) { struct drm_mm *mm = hole_node->mm; - unsigned long tmp = 0, wasted = 0; unsigned long hole_start = drm_mm_hole_node_start(hole_node); unsigned long hole_end = drm_mm_hole_node_end(hole_node); + unsigned long adj_start = hole_start; + unsigned long adj_end = hole_end; BUG_ON(!hole_node->hole_follows || node->allocated); - if (alignment) - tmp = hole_start % alignment; + if (mm->color_adjust) + mm->color_adjust(hole_node, color, &adj_start, &adj_end); - if (!tmp) { + if (alignment) { + unsigned tmp = adj_start % alignment; + if (tmp) + adj_start += alignment - tmp; + } + + if (adj_start == hole_start) { hole_node->hole_follows = 0; - list_del_init(&hole_node->hole_stack); - } else - wasted = alignment - tmp; + list_del(&hole_node->hole_stack); + } - node->start = hole_start + wasted; + node->start = adj_start; node->size = size; node->mm = mm; + node->color = color; node->allocated = 1; INIT_LIST_HEAD(&node->hole_stack); list_add(&node->node_list, &hole_node->node_list); - BUG_ON(node->start + node->size > hole_end); + BUG_ON(node->start + node->size > adj_end); + node->hole_follows = 0; if (node->start + node->size < hole_end) { list_add(&node->hole_stack, &mm->hole_stack); node->hole_follows = 1; - } else { - node->hole_follows = 0; } } struct drm_mm_node *drm_mm_get_block_generic(struct drm_mm_node *hole_node, unsigned long size, unsigned alignment, + unsigned long color, int atomic) { struct drm_mm_node *node; @@ -165,7 +173,7 @@ struct drm_mm_node *drm_mm_get_block_generic(struct drm_mm_node *hole_node, if (unlikely(node == NULL)) return NULL; - drm_mm_insert_helper(hole_node, node, size, alignment); + drm_mm_insert_helper(hole_node, node, size, alignment, color); return node; } @@ -181,11 +189,11 @@ int drm_mm_insert_node(struct drm_mm *mm, struct drm_mm_node *node, { struct drm_mm_node *hole_node; - hole_node = drm_mm_search_free(mm, size, alignment, 0); + hole_node = drm_mm_search_free(mm, size, alignment, false); if (!hole_node) return -ENOSPC; - drm_mm_insert_helper(hole_node, node, size, alignment); + drm_mm_insert_helper(hole_node, node, size, alignment, 0); return 0; } @@ -194,50 +202,57 @@ EXPORT_SYMBOL(drm_mm_insert_node); static void drm_mm_insert_helper_range(struct drm_mm_node *hole_node, struct drm_mm_node *node, unsigned long size, unsigned alignment, + unsigned long color, unsigned long start, unsigned long end) { struct drm_mm *mm = hole_node->mm; - unsigned long tmp = 0, wasted = 0; unsigned long hole_start = drm_mm_hole_node_start(hole_node); unsigned long hole_end = drm_mm_hole_node_end(hole_node); + unsigned long adj_start = hole_start; + unsigned long adj_end = hole_end; BUG_ON(!hole_node->hole_follows || node->allocated); - if (hole_start < start) - wasted += start - hole_start; - if (alignment) - tmp = (hole_start + wasted) % alignment; + if (mm->color_adjust) + mm->color_adjust(hole_node, color, &adj_start, &adj_end); - if (tmp) - wasted += alignment - tmp; + if (adj_start < start) + adj_start = start; + + if (alignment) { + unsigned tmp = adj_start % alignment; + if (tmp) + adj_start += alignment - tmp; + } - if (!wasted) { + if (adj_start == hole_start) { hole_node->hole_follows = 0; - list_del_init(&hole_node->hole_stack); + list_del(&hole_node->hole_stack); } - node->start = hole_start + wasted; + node->start = adj_start; node->size = size; node->mm = mm; + node->color = color; node->allocated = 1; INIT_LIST_HEAD(&node->hole_stack); list_add(&node->node_list, &hole_node->node_list); - BUG_ON(node->start + node->size > hole_end); + BUG_ON(node->start + node->size > adj_end); BUG_ON(node->start + node->size > end); + node->hole_follows = 0; if (node->start + node->size < hole_end) { list_add(&node->hole_stack, &mm->hole_stack); node->hole_follows = 1; - } else { - node->hole_follows = 0; } } struct drm_mm_node *drm_mm_get_block_range_generic(struct drm_mm_node *hole_node, unsigned long size, unsigned alignment, + unsigned long color, unsigned long start, unsigned long end, int atomic) @@ -248,7 +263,7 @@ struct drm_mm_node *drm_mm_get_block_range_generic(struct drm_mm_node *hole_node if (unlikely(node == NULL)) return NULL; - drm_mm_insert_helper_range(hole_node, node, size, alignment, + drm_mm_insert_helper_range(hole_node, node, size, alignment, color, start, end); return node; @@ -267,11 +282,11 @@ int drm_mm_insert_node_in_range(struct drm_mm *mm, struct drm_mm_node *node, struct drm_mm_node *hole_node; hole_node = drm_mm_search_free_in_range(mm, size, alignment, - start, end, 0); + start, end, false); if (!hole_node) return -ENOSPC; - drm_mm_insert_helper_range(hole_node, node, size, alignment, + drm_mm_insert_helper_range(hole_node, node, size, alignment, 0, start, end); return 0; @@ -336,27 +351,23 @@ EXPORT_SYMBOL(drm_mm_put_block); static int check_free_hole(unsigned long start, unsigned long end, unsigned long size, unsigned alignment) { - unsigned wasted = 0; - if (end - start < size) return 0; if (alignment) { unsigned tmp = start % alignment; if (tmp) - wasted = alignment - tmp; - } - - if (end >= start + size + wasted) { - return 1; + start += alignment - tmp; } - return 0; + return end >= start + size; } -struct drm_mm_node *drm_mm_search_free(const struct drm_mm *mm, - unsigned long size, - unsigned alignment, int best_match) +struct drm_mm_node *drm_mm_search_free_generic(const struct drm_mm *mm, + unsigned long size, + unsigned alignment, + unsigned long color, + bool best_match) { struct drm_mm_node *entry; struct drm_mm_node *best; @@ -368,10 +379,17 @@ struct drm_mm_node *drm_mm_search_free(const struct drm_mm *mm, best_size = ~0UL; list_for_each_entry(entry, &mm->hole_stack, hole_stack) { + unsigned long adj_start = drm_mm_hole_node_start(entry); + unsigned long adj_end = drm_mm_hole_node_end(entry); + + if (mm->color_adjust) { + mm->color_adjust(entry, color, &adj_start, &adj_end); + if (adj_end <= adj_start) + continue; + } + BUG_ON(!entry->hole_follows); - if (!check_free_hole(drm_mm_hole_node_start(entry), - drm_mm_hole_node_end(entry), - size, alignment)) + if (!check_free_hole(adj_start, adj_end, size, alignment)) continue; if (!best_match) @@ -385,14 +403,15 @@ struct drm_mm_node *drm_mm_search_free(const struct drm_mm *mm, return best; } -EXPORT_SYMBOL(drm_mm_search_free); - -struct drm_mm_node *drm_mm_search_free_in_range(const struct drm_mm *mm, - unsigned long size, - unsigned alignment, - unsigned long start, - unsigned long end, - int best_match) +EXPORT_SYMBOL(drm_mm_search_free_generic); + +struct drm_mm_node *drm_mm_search_free_in_range_generic(const struct drm_mm *mm, + unsigned long size, + unsigned alignment, + unsigned long color, + unsigned long start, + unsigned long end, + bool best_match) { struct drm_mm_node *entry; struct drm_mm_node *best; @@ -410,6 +429,13 @@ struct drm_mm_node *drm_mm_search_free_in_range(const struct drm_mm *mm, end : drm_mm_hole_node_end(entry); BUG_ON(!entry->hole_follows); + + if (mm->color_adjust) { + mm->color_adjust(entry, color, &adj_start, &adj_end); + if (adj_end <= adj_start) + continue; + } + if (!check_free_hole(adj_start, adj_end, size, alignment)) continue; @@ -424,7 +450,7 @@ struct drm_mm_node *drm_mm_search_free_in_range(const struct drm_mm *mm, return best; } -EXPORT_SYMBOL(drm_mm_search_free_in_range); +EXPORT_SYMBOL(drm_mm_search_free_in_range_generic); /** * Moves an allocation. To be used with embedded struct drm_mm_node. @@ -437,6 +463,7 @@ void drm_mm_replace_node(struct drm_mm_node *old, struct drm_mm_node *new) new->mm = old->mm; new->start = old->start; new->size = old->size; + new->color = old->color; old->allocated = 0; new->allocated = 1; @@ -452,9 +479,12 @@ EXPORT_SYMBOL(drm_mm_replace_node); * Warning: As long as the scan list is non-empty, no other operations than * adding/removing nodes to/from the scan list are allowed. */ -void drm_mm_init_scan(struct drm_mm *mm, unsigned long size, - unsigned alignment) +void drm_mm_init_scan(struct drm_mm *mm, + unsigned long size, + unsigned alignment, + unsigned long color) { + mm->scan_color = color; mm->scan_alignment = alignment; mm->scan_size = size; mm->scanned_blocks = 0; @@ -474,11 +504,14 @@ EXPORT_SYMBOL(drm_mm_init_scan); * Warning: As long as the scan list is non-empty, no other operations than * adding/removing nodes to/from the scan list are allowed. */ -void drm_mm_init_scan_with_range(struct drm_mm *mm, unsigned long size, +void drm_mm_init_scan_with_range(struct drm_mm *mm, + unsigned long size, unsigned alignment, + unsigned long color, unsigned long start, unsigned long end) { + mm->scan_color = color; mm->scan_alignment = alignment; mm->scan_size = size; mm->scanned_blocks = 0; @@ -522,17 +555,21 @@ int drm_mm_scan_add_block(struct drm_mm_node *node) hole_start = drm_mm_hole_node_start(prev_node); hole_end = drm_mm_hole_node_end(prev_node); + + adj_start = hole_start; + adj_end = hole_end; + + if (mm->color_adjust) + mm->color_adjust(prev_node, mm->scan_color, &adj_start, &adj_end); + if (mm->scan_check_range) { - adj_start = hole_start < mm->scan_start ? - mm->scan_start : hole_start; - adj_end = hole_end > mm->scan_end ? - mm->scan_end : hole_end; - } else { - adj_start = hole_start; - adj_end = hole_end; + if (adj_start < mm->scan_start) + adj_start = mm->scan_start; + if (adj_end > mm->scan_end) + adj_end = mm->scan_end; } - if (check_free_hole(adj_start , adj_end, + if (check_free_hole(adj_start, adj_end, mm->scan_size, mm->scan_alignment)) { mm->scan_hit_start = hole_start; mm->scan_hit_size = hole_end; @@ -616,6 +653,8 @@ int drm_mm_init(struct drm_mm * mm, unsigned long start, unsigned long size) mm->head_node.size = start - mm->head_node.start; list_add_tail(&mm->head_node.hole_stack, &mm->hole_stack); + mm->color_adjust = NULL; + return 0; } EXPORT_SYMBOL(drm_mm_init); diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index db438f0..8a34061 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -2755,8 +2755,8 @@ i915_gem_object_bind_to_gtt(struct drm_i915_gem_object *obj, if (map_and_fenceable) free_space = drm_mm_search_free_in_range(&dev_priv->mm.gtt_space, - size, alignment, 0, - dev_priv->mm.gtt_mappable_end, + size, alignment, + 0, dev_priv->mm.gtt_mappable_end, 0); else free_space = drm_mm_search_free(&dev_priv->mm.gtt_space, @@ -2767,7 +2767,7 @@ i915_gem_object_bind_to_gtt(struct drm_i915_gem_object *obj, obj->gtt_space = drm_mm_get_block_range_generic(free_space, size, alignment, 0, - dev_priv->mm.gtt_mappable_end, + 0, dev_priv->mm.gtt_mappable_end, 0); else obj->gtt_space = diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c index ae7c24e..eba0308 100644 --- a/drivers/gpu/drm/i915/i915_gem_evict.c +++ b/drivers/gpu/drm/i915/i915_gem_evict.c @@ -78,11 +78,12 @@ i915_gem_evict_something(struct drm_device *dev, int min_size, INIT_LIST_HEAD(&unwind_list); if (mappable) - drm_mm_init_scan_with_range(&dev_priv->mm.gtt_space, min_size, - alignment, 0, - dev_priv->mm.gtt_mappable_end); + drm_mm_init_scan_with_range(&dev_priv->mm.gtt_space, + min_size, alignment, 0, + 0, dev_priv->mm.gtt_mappable_end); else - drm_mm_init_scan(&dev_priv->mm.gtt_space, min_size, alignment); + drm_mm_init_scan(&dev_priv->mm.gtt_space, + min_size, alignment, 0); /* First see if there is a large enough contiguous idle region... */ list_for_each_entry(obj, &dev_priv->mm.inactive_list, mm_list) { diff --git a/include/drm/drm_mm.h b/include/drm/drm_mm.h index 564b14a..06d7f79 100644 --- a/include/drm/drm_mm.h +++ b/include/drm/drm_mm.h @@ -50,6 +50,7 @@ struct drm_mm_node { unsigned scanned_next_free : 1; unsigned scanned_preceeds_hole : 1; unsigned allocated : 1; + unsigned long color; unsigned long start; unsigned long size; struct drm_mm *mm; @@ -66,6 +67,7 @@ struct drm_mm { spinlock_t unused_lock; unsigned int scan_check_range : 1; unsigned scan_alignment; + unsigned long scan_color; unsigned long scan_size; unsigned long scan_hit_start; unsigned scan_hit_size; @@ -73,6 +75,9 @@ struct drm_mm { unsigned long scan_start; unsigned long scan_end; struct drm_mm_node *prev_scanned_node; + + void (*color_adjust)(struct drm_mm_node *node, unsigned long color, + unsigned long *start, unsigned long *end); }; static inline bool drm_mm_node_allocated(struct drm_mm_node *node) @@ -100,11 +105,13 @@ static inline bool drm_mm_initialized(struct drm_mm *mm) extern struct drm_mm_node *drm_mm_get_block_generic(struct drm_mm_node *node, unsigned long size, unsigned alignment, + unsigned long color, int atomic); extern struct drm_mm_node *drm_mm_get_block_range_generic( struct drm_mm_node *node, unsigned long size, unsigned alignment, + unsigned long color, unsigned long start, unsigned long end, int atomic); @@ -112,13 +119,13 @@ static inline struct drm_mm_node *drm_mm_get_block(struct drm_mm_node *parent, unsigned long size, unsigned alignment) { - return drm_mm_get_block_generic(parent, size, alignment, 0); + return drm_mm_get_block_generic(parent, size, alignment, 0, 0); } static inline struct drm_mm_node *drm_mm_get_block_atomic(struct drm_mm_node *parent, unsigned long size, unsigned alignment) { - return drm_mm_get_block_generic(parent, size, alignment, 1); + return drm_mm_get_block_generic(parent, size, alignment, 0, 1); } static inline struct drm_mm_node *drm_mm_get_block_range( struct drm_mm_node *parent, @@ -127,8 +134,19 @@ static inline struct drm_mm_node *drm_mm_get_block_range( unsigned long start, unsigned long end) { - return drm_mm_get_block_range_generic(parent, size, alignment, - start, end, 0); + return drm_mm_get_block_range_generic(parent, size, alignment, 0, + start, end, 0); +} +static inline struct drm_mm_node *drm_mm_get_color_block_range( + struct drm_mm_node *parent, + unsigned long size, + unsigned alignment, + unsigned long color, + unsigned long start, + unsigned long end) +{ + return drm_mm_get_block_range_generic(parent, size, alignment, color, + start, end, 0); } static inline struct drm_mm_node *drm_mm_get_block_atomic_range( struct drm_mm_node *parent, @@ -137,7 +155,7 @@ static inline struct drm_mm_node *drm_mm_get_block_atomic_range( unsigned long start, unsigned long end) { - return drm_mm_get_block_range_generic(parent, size, alignment, + return drm_mm_get_block_range_generic(parent, size, alignment, 0, start, end, 1); } extern int drm_mm_insert_node(struct drm_mm *mm, struct drm_mm_node *node, @@ -149,18 +167,59 @@ extern int drm_mm_insert_node_in_range(struct drm_mm *mm, extern void drm_mm_put_block(struct drm_mm_node *cur); extern void drm_mm_remove_node(struct drm_mm_node *node); extern void drm_mm_replace_node(struct drm_mm_node *old, struct drm_mm_node *new); -extern struct drm_mm_node *drm_mm_search_free(const struct drm_mm *mm, - unsigned long size, - unsigned alignment, - int best_match); -extern struct drm_mm_node *drm_mm_search_free_in_range( +extern struct drm_mm_node *drm_mm_search_free_generic(const struct drm_mm *mm, + unsigned long size, + unsigned alignment, + unsigned long color, + bool best_match); +extern struct drm_mm_node *drm_mm_search_free_in_range_generic( + const struct drm_mm *mm, + unsigned long size, + unsigned alignment, + unsigned long color, + unsigned long start, + unsigned long end, + bool best_match); +static inline struct drm_mm_node *drm_mm_search_free(const struct drm_mm *mm, + unsigned long size, + unsigned alignment, + bool best_match) +{ + return drm_mm_search_free_generic(mm,size, alignment, 0, best_match); +} +static inline struct drm_mm_node *drm_mm_search_free_in_range( const struct drm_mm *mm, unsigned long size, unsigned alignment, unsigned long start, unsigned long end, - int best_match); -extern int drm_mm_init(struct drm_mm *mm, unsigned long start, + bool best_match) +{ + return drm_mm_search_free_in_range_generic(mm, size, alignment, 0, + start, end, best_match); +} +static inline struct drm_mm_node *drm_mm_search_free_color(const struct drm_mm *mm, + unsigned long size, + unsigned alignment, + unsigned long color, + bool best_match) +{ + return drm_mm_search_free_generic(mm,size, alignment, color, best_match); +} +static inline struct drm_mm_node *drm_mm_search_free_in_range_color( + const struct drm_mm *mm, + unsigned long size, + unsigned alignment, + unsigned long color, + unsigned long start, + unsigned long end, + bool best_match) +{ + return drm_mm_search_free_in_range_generic(mm, size, alignment, color, + start, end, best_match); +} +extern int drm_mm_init(struct drm_mm *mm, + unsigned long start, unsigned long size); extern void drm_mm_takedown(struct drm_mm *mm); extern int drm_mm_clean(struct drm_mm *mm); @@ -171,10 +230,14 @@ static inline struct drm_mm *drm_get_mm(struct drm_mm_node *block) return block->mm; } -void drm_mm_init_scan(struct drm_mm *mm, unsigned long size, - unsigned alignment); -void drm_mm_init_scan_with_range(struct drm_mm *mm, unsigned long size, +void drm_mm_init_scan(struct drm_mm *mm, + unsigned long size, + unsigned alignment, + unsigned long color); +void drm_mm_init_scan_with_range(struct drm_mm *mm, + unsigned long size, unsigned alignment, + unsigned long color, unsigned long start, unsigned long end); int drm_mm_scan_add_block(struct drm_mm_node *node); -- 1.7.10.4 [View Less]

2 1

AMD GPU: Toshiba, brightness and ATIF
by Luca Tettamanti 12 Jul '12

12 Jul '12

Hello, I'm trying to figure out how to control the brightness of the screen of my laptop, a Toshiba L885-10W. The ACPI DSDT table contains the standard control methods, the setter (_BCM), however, only sets a value in the record returned by ATIF method and send out a notification: CreateByteField (ATIB, 0x0C, BRIL) // ATIB is returned by ATIF Store (Arg0, BRIL) Notify (VGA, 0x81) When using the ACPI sysfs backlight interface nothing happens (obviously). 0x81 seems also strange, according to … [View More]

1 0

[PATCH] dma-fence: dma-buf synchronization
by Rob Clark 12 Jul '12

12 Jul '12

From: Rob Clark <rob(a)ti.com> A dma-fence can be attached to a buffer which is being filled or consumed by hw, to allow userspace to pass the buffer without waiting to another device. For example, userspace can call page_flip ioctl to display the next frame of graphics after kicking the GPU but while the GPU is still rendering. The display device sharing the buffer with the GPU would attach a callback to get notified when the GPU's rendering-complete IRQ fires, to update the scan-out … [View More]address of the display, without having to wake up userspace. A dma-fence is transient, one-shot deal. It is allocated and attached to dma-buf's list of fences. When the one that attached it is done, with the pending operation, it can signal the fence removing it from the dma-buf's list of fences: + dma_buf_attach_fence() + dma_fence_signal() Other drivers can access the current fence on the dma-buf (if any), which increment's the fences refcnt: + dma_buf_get_fence() + dma_fence_put() The one pending on the fence can add an async callback (and optionally cancel it.. for example, to recover from GPU hangs): + dma_fence_add_callback() + dma_fence_cancel_callback() Or wait synchronously (optionally with timeout or from atomic context): + dma_fence_wait() A default software-only implementation is provided, which can be used by drivers attaching a fence to a buffer when they have no other means for hw sync. But a memory backed fence is also envisioned, because it is common that GPU's can write to, or poll on some memory location for synchronization. For example: fence = dma_buf_get_fence(dmabuf); if (fence->ops == &mem_dma_fence_ops) { dma_buf *fence_buf; mem_dma_fence_get_buf(fence, &fence_buf, &offset); ... tell the hw the memory location to wait on ... } else { /* fall-back to sw sync * / dma_fence_add_callback(fence, my_cb); } The memory location is itself backed by dma-buf, to simplify mapping to the device's address space, an idea borrowed from Maarten Lankhorst. NOTE: the memory location fence is not implemented yet, the above is just for explaining how it would work. On SoC platforms, if some other hw mechanism is provided for synchronizing between IP blocks, it could be supported as an alternate implementation with it's own fence ops in a similar way. The other non-sw implementations would wrap the add/cancel_callback and wait fence ops, so that they can keep track if a device not supporting hw sync is waiting on the fence, and in this case should arrange to call dma_fence_signal() at some point after the condition has changed, to notify other devices waiting on the fence. If there are no sw waiters, this can be skipped to avoid waking the CPU unnecessarily. The intention is to provide a userspace interface (presumably via eventfd) later, to be used in conjunction with dma-buf's mmap support for sw access to buffers (or for userspace apps that would prefer to do their own synchronization). --- drivers/base/Makefile | 2 +- drivers/base/dma-buf.c | 3 + drivers/base/dma-fence.c | 325 +++++++++++++++++++++++++++++++++++++++++++++ include/linux/dma-buf.h | 3 + include/linux/dma-fence.h | 118 ++++++++++++++++ 5 files changed, 450 insertions(+), 1 deletion(-) create mode 100644 drivers/base/dma-fence.c create mode 100644 include/linux/dma-fence.h diff --git a/drivers/base/Makefile b/drivers/base/Makefile index 5aa2d70..6e9f217 100644 --- a/drivers/base/Makefile +++ b/drivers/base/Makefile @@ -10,7 +10,7 @@ obj-$(CONFIG_CMA) += dma-contiguous.o obj-y += power/ obj-$(CONFIG_HAS_DMA) += dma-mapping.o obj-$(CONFIG_HAVE_GENERIC_DMA_COHERENT) += dma-coherent.o -obj-$(CONFIG_DMA_SHARED_BUFFER) += dma-buf.o +obj-$(CONFIG_DMA_SHARED_BUFFER) += dma-buf.o dma-fence.o obj-$(CONFIG_ISA) += isa.o obj-$(CONFIG_FW_LOADER) += firmware_class.o obj-$(CONFIG_NUMA) += node.o diff --git a/drivers/base/dma-buf.c b/drivers/base/dma-buf.c index 24e88fe..b053236 100644 --- a/drivers/base/dma-buf.c +++ b/drivers/base/dma-buf.c @@ -39,6 +39,8 @@ static int dma_buf_release(struct inode *inode, struct file *file) dmabuf = file->private_data; + WARN_ON(!list_empty(&dmabuf->fence_list)); + dmabuf->ops->release(dmabuf); kfree(dmabuf); return 0; @@ -119,6 +121,7 @@ struct dma_buf *dma_buf_export(void *priv, const struct dma_buf_ops *ops, mutex_init(&dmabuf->lock); INIT_LIST_HEAD(&dmabuf->attachments); + INIT_LIST_HEAD(&dmabuf->fence_list); return dmabuf; } diff --git a/drivers/base/dma-fence.c b/drivers/base/dma-fence.c new file mode 100644 index 0000000..a94ed01 --- /dev/null +++ b/drivers/base/dma-fence.c @@ -0,0 +1,325 @@ +/* + * Fence mechanism for dma-buf to allow for asynchronous dma access + * + * Copyright (C) 2012 Texas Instruments + * Author: Rob Clark <rob.clark(a)linaro.org> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include <linux/slab.h> +#include <linux/sched.h> +#include <linux/export.h> +#include <linux/dma-fence.h> + +static DEFINE_SPINLOCK(fence_lock); + +/** + * dma_buf_attach_fence - Attach a fence to a dma-buf. + * + * @buf: the dma-buf to attach to + * @fence: the fence to attach + * + * A fence can only be attached to a single dma-buf. The dma-buf takes + * ownership of the fence, which is unref'd when the fence is signaled. + * The fence takes a reference to the dma-buf so the buffer will not be + * freed while there is a pending fence. + */ +int dma_buf_attach_fence(struct dma_buf *buf, struct dma_fence *fence) +{ + if (WARN_ON(!buf || !fence || fence->buf)) + return -EINVAL; + + spin_lock(&fence_lock); + get_dma_buf(buf); + fence->buf = buf; + list_add(&fence->list_node, &buf->fence_list); + spin_unlock(&fence_lock); + + return 0; +} +EXPORT_SYMBOL_GPL(dma_buf_attach_fence); + +/** + * dma_fence_signal - Signal a fence. + * + * @fence: The fence to signal + * + * All registered callbacks will be called directly (synchronously) and + * all blocked waters will be awoken. + */ +int dma_fence_signal(struct dma_fence *fence) +{ + unsigned long flags; + int ret; + + if (WARN_ON(!fence || !fence->buf)) + return -EINVAL; + + spin_lock_irqsave(&fence_lock, flags); + list_del(&fence->list_node); + spin_unlock_irqrestore(&fence_lock, flags); + + dma_buf_put(fence->buf); + fence->buf = NULL; + + ret = fence->ops->signal(fence); + + dma_fence_put(fence); + + return ret; +} +EXPORT_SYMBOL_GPL(dma_fence_signal); + +/** + * dma_buf_get_fence - Get the most recent pending fence attached to the + * dma-buf. + * + * @buf: the dma-buf whose fence to get + * + * If this returns NULL, there are no pending fences. Otherwise this + * takes a reference to the returned fence, so the caller must later + * call dma_fence_put() to release the reference. + */ +struct dma_fence *dma_buf_get_fence(struct dma_buf *buf) +{ + struct dma_fence *fence = NULL; + + if (WARN_ON(!buf)) + return ERR_PTR(-EINVAL); + + spin_lock(&fence_lock); + if (!list_empty(&buf->fence_list)) { + fence = list_first_entry(&buf->fence_list, + struct dma_fence, list_node); + dma_fence_get(fence); + } + spin_unlock(&fence_lock); + + return fence; +} +EXPORT_SYMBOL_GPL(dma_buf_get_fence); + +static void release_fence(struct kref *kref) +{ + struct dma_fence *fence = + container_of(kref, struct dma_fence, refcount); + + WARN_ON(waitqueue_active(&fence->event_queue) || fence->buf); + + kfree(fence); +} + +/** + * dma_fence_put - Release a reference to the fence. + */ +void dma_fence_put(struct dma_fence *fence) +{ + WARN_ON(!fence); + kref_put(&fence->refcount, release_fence); +} +EXPORT_SYMBOL_GPL(dma_fence_put); + +/** + * dma_fence_get - Take a reference to the fence. + * + * In most cases this is used only internally by dma-fence. + */ +void dma_fence_get(struct dma_fence *fence) +{ + WARN_ON(!fence); + kref_get(&fence->refcount); +} +EXPORT_SYMBOL_GPL(dma_fence_get); + +/** + * dma_fence_add_callback - Add a callback to be called when the fence + * is signaled. + * + * @fence: The fence to wait on + * @cb: The callback to register + * + * Any number of callbacks can be registered to a fence, but a callback + * can only be registered to once fence at a time. + * + * Note that the callback can be called from an atomic context. If + * fence is already signaled, this function will return -EINVAL (and + * *not* call the callback) + */ +int dma_fence_add_callback(struct dma_fence *fence, + struct dma_fence_cb *cb) +{ + int ret = -EINVAL; + + if (WARN_ON(!fence || !cb)) + return -EINVAL; + + spin_lock(&fence_lock); + if (fence->buf) { + dma_fence_get(fence); + cb->fence = fence; + ret = fence->ops->add_callback(fence, cb); + } + spin_unlock(&fence_lock); + + return ret; +} +EXPORT_SYMBOL_GPL(dma_fence_add_callback); + +/** + * dma_fence_cancel_callback - Remove a previously registered callback. + * + * @cb: The callback to unregister + * + * The callback will not be called after this function returns, but could + * be called before this function returns. + */ +int dma_fence_cancel_callback(struct dma_fence_cb *cb) +{ + struct dma_fence *fence; + int ret = -EINVAL; + + if (WARN_ON(!cb)) + return -EINVAL; + + fence = cb->fence; + + spin_lock(&fence_lock); + if (fence) { + ret = fence->ops->cancel_callback(cb); + cb->fence = NULL; + dma_fence_put(fence); + } + spin_unlock(&fence_lock); + + return ret; +} +EXPORT_SYMBOL_GPL(dma_fence_cancel_callback); + +/** + * dma_fence_wait - Wait for a fence to be signaled. + * + * @fence: The fence to wait on + * @interruptible: if true, do an interruptible wait + * @timeout: timeout, in jiffies + */ +int dma_fence_wait(struct dma_fence *fence, bool interruptible, long timeout) +{ + if (WARN_ON(!fence)) + return -EINVAL; + + return fence->ops->wait(fence, interruptible, timeout); +} +EXPORT_SYMBOL_GPL(dma_fence_wait); + +int __dma_fence_wake_func(wait_queue_t *wait, unsigned mode, + int flags, void *key) +{ + struct dma_fence_cb *cb = + container_of(wait, struct dma_fence_cb, base); + int ret; + + ret = cb->func(cb, cb->fence); + dma_fence_put(cb->fence); + cb->fence = NULL; + + return ret; +} +EXPORT_SYMBOL_GPL(__dma_fence_wake_func); + +/* + * Helpers intended to be used by the ops of the dma_fence implementation: + * + * NOTE: helpers and fxns intended to be used by other dma-fence + * implementations are not exported.. I'm not really sure if it makes + * sense to have a dma-fence implementation that is itself a module. + */ + +void __dma_fence_init(struct dma_fence *fence, struct dma_fence_ops *ops) +{ + WARN_ON(!ops || !ops->signal || !ops->add_callback || + !ops->cancel_callback || !ops->wait); + + kref_init(&fence->refcount); + fence->ops = ops; + init_waitqueue_head(&fence->event_queue); +} + +int dma_fence_helper_signal(struct dma_fence *fence) +{ + wake_up_all(&fence->event_queue); + return 0; +} + +int dma_fence_helper_add_callback(struct dma_fence *fence, + struct dma_fence_cb *cb) +{ + add_wait_queue(&fence->event_queue, &cb->base); + return 0; +} + +int dma_fence_helper_cancel_callback(struct dma_fence_cb *cb) +{ + struct dma_fence *fence = cb->fence; + remove_wait_queue(&fence->event_queue, &cb->base); + return 0; +} + +int dma_fence_helper_wait(struct dma_fence *fence, bool interruptible, + long timeout) +{ + int ret; + + if (interruptible) { + ret = wait_event_interruptible_timeout(fence->event_queue, + !fence->buf, timeout); + } else { + ret = wait_event_timeout(fence->event_queue, + !fence->buf, timeout); + } + + return ret; +} + +/* + * Pure sw implementation for dma-fence. The CPU always gets involved. + */ + +static struct dma_fence_ops sw_fence_ops = { + .signal = dma_fence_helper_signal, + .add_callback = dma_fence_helper_add_callback, + .cancel_callback = dma_fence_helper_cancel_callback, + .wait = dma_fence_helper_wait, +}; + +/** + * dma_fence_create - Create a simple sw-only fence. + * + * This fence only supports signaling from/to CPU. Other implementations + * of dma-fence can be used to support hardware to hardware signaling, if + * supported by the hardware, and use the dma_fence_helper_* functions for + * compatibility with other devices that only support sw signaling. + */ +struct dma_fence *dma_fence_create(void) +{ + struct dma_fence *fence; + + fence = kzalloc(sizeof(struct dma_fence), GFP_KERNEL); + if (!fence) + return ERR_PTR(-ENOMEM); + + __dma_fence_init(fence, &sw_fence_ops); + + return fence; +} +EXPORT_SYMBOL_GPL(dma_fence_create); diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h index eb48f38..a0a0b64 100644 --- a/include/linux/dma-buf.h +++ b/include/linux/dma-buf.h @@ -29,6 +29,7 @@ #include <linux/scatterlist.h> #include <linux/list.h> #include <linux/dma-mapping.h> +#include <linux/dma-fence.h> #include <linux/fs.h> struct device; @@ -122,6 +123,8 @@ struct dma_buf { /* mutex to serialize list manipulation and attach/detach */ struct mutex lock; void *priv; + /* list of pending dma_fence's */ + struct list_head fence_list; }; /** diff --git a/include/linux/dma-fence.h b/include/linux/dma-fence.h new file mode 100644 index 0000000..ecda334 --- /dev/null +++ b/include/linux/dma-fence.h @@ -0,0 +1,118 @@ +/* + * Fence mechanism for dma-buf to allow for asynchronous dma access + * + * Copyright (C) 2012 Texas Instruments + * Author: Rob Clark <rob.clark(a)linaro.org> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef __DMA_FENCE_H__ +#define __DMA_FENCE_H__ + +#include <linux/err.h> +#include <linux/list.h> +#include <linux/wait.h> +#include <linux/list.h> +#include <linux/dma-buf.h> + +struct dma_fence; +struct dma_fence_ops; +struct dma_fence_cb; + +struct dma_fence { + struct kref refcount; + struct dma_fence_ops *ops; + wait_queue_head_t event_queue; + struct list_head list_node; /* for dmabuf's fence_list */ + + /* + * This is initialized when the fence is attached to a dma-buf, + * and NULL'd before it is signaled. A fence can be attached to + * at most one dma-buf at a time. + */ + struct dma_buf *buf; +}; + +typedef int (*dma_fence_func_t)(struct dma_fence_cb *cb, + struct dma_fence *fence); + +struct dma_fence_ops { + int (*signal)(struct dma_fence *fence); + int (*add_callback)(struct dma_fence *fence, struct dma_fence_cb *cb); + int (*cancel_callback)(struct dma_fence_cb *cb); + int (*wait)(struct dma_fence *fence, bool interruptible, long timeout); +}; + +struct dma_fence_cb { + wait_queue_t base; + dma_fence_func_t func; + + /* + * This is initialized when the cb is added, and NULL'd when it + * is canceled or expired, so can be used to for error checking + * if the cb is already pending. A dma_fence_cb can be pending + * on at most one fence at a time. + */ + struct dma_fence *fence; +}; + +int __dma_fence_wake_func(wait_queue_t *wait, unsigned mode, + int flags, void *key); + +#define DMA_FENCE_CB_INITIALIZER(cb_func) { \ + .base = { .func = __dma_fence_wake_func }, \ + .func = (cb_func), \ + } + +#define DECLARE_DMA_FENCE(name, cb_func) \ + struct dma_fence_cb name = DMA_FENCE_CB_INITIALIZER(cb_func) + + +/* + * TODO does it make sense to be able to enable dma-fence without dma-buf, + * or visa versa? + */ +#ifdef CONFIG_DMA_SHARED_BUFFER + +int dma_buf_attach_fence(struct dma_buf *buf, struct dma_fence *fence); +struct dma_fence *dma_buf_get_fence(struct dma_buf *buf); + +/* create a basic (pure sw) fence: */ +struct dma_fence *dma_fence_create(void); + +/* intended to be used by other dma_fence implementations: */ +void __dma_fence_init(struct dma_fence *fence, struct dma_fence_ops *ops); + +void dma_fence_get(struct dma_fence *fence); +void dma_fence_put(struct dma_fence *fence); +int dma_fence_signal(struct dma_fence *fence); + +int dma_fence_add_callback(struct dma_fence *fence, + struct dma_fence_cb *cb); +int dma_fence_cancel_callback(struct dma_fence_cb *cb); +int dma_fence_wait(struct dma_fence *fence, bool interruptible, long timeout); + +/* helpers intended to be used by the ops of the dma_fence implementation: */ +int dma_fence_helper_signal(struct dma_fence *fence); +int dma_fence_helper_add_callback(struct dma_fence *fence, + struct dma_fence_cb *cb); +int dma_fence_helper_cancel_callback(struct dma_fence_cb *cb); +int dma_fence_helper_wait(struct dma_fence *fence, bool interruptible, + long timeout); + +#else +// TODO +#endif /* CONFIG_DMA_SHARED_BUFFER */ + +#endif /* __DMA_FENCE_H__ */ -- 1.7.9.5 [View Less]

3 3

[Bug 15844] Failed to load firmware "radeon/RV770_pfp.bin"
by bugzilla-daemon＠bugzilla.kernel.org 11 Jul '12

11 Jul '12

https://bugzilla.kernel.org/show_bug.cgi?id=15844 Alan <alan(a)lxorguk.ukuu.org.uk> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED CC| |alan(a)lxorguk.ukuu.org.uk Resolution| |DOCUMENTED -- Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?… [View More]

1 0

[Bug 15748] BUG: unable to handle kernel NULL pointer dereference at 0000000000000028
by bugzilla-daemon＠bugzilla.kernel.org 11 Jul '12

11 Jul '12

https://bugzilla.kernel.org/show_bug.cgi?id=15748 Alan <alan(a)lxorguk.ukuu.org.uk> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED CC| |alan(a)lxorguk.ukuu.org.uk Resolution| |OBSOLETE --- Comment #2 from Alan <alan(a)lxorguk.ukuu.org.uk> 2012-07-… [View More]

1 0

[Bug 15738] VGA output of Radeon HD3450 displays too bright colors with KMS radeon driver
by bugzilla-daemon＠bugzilla.kernel.org 11 Jul '12

11 Jul '12

https://bugzilla.kernel.org/show_bug.cgi?id=15738 Alan <alan(a)lxorguk.ukuu.org.uk> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED CC| |alan(a)lxorguk.ukuu.org.uk Resolution| |CODE_FIX -- Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=… [View More]

1 0

← Newer
1
...
60
61
62
63
64
65
66
...
69
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel July 2012