aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2016-08-18drm/i915: Tidy up flush cpu/gtt write domainsChris Wilson1-11/+4
Since we know the write domain, we can drop the local variable and make the code look a tiny bit simpler. Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Joonas Lahtinen <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/i915: Pin the pages first in shmem prepare read/writeChris Wilson1-20/+28
There is an improbable, but not impossible, case that if we leave the pages unpin as we operate on the object, then somebody via the shrinker may steal the lock (which lock? right now, it is struct_mutex, THE lock) and change the cache domains after we have already inspected them. (Whilst here, avail ourselves of the opportunity to take a couple of steps to make the two functions look more similar.) Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Joonas Lahtinen <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/i915: Wait for writes through the GTT to land before reading backChris Wilson1-1/+11
If we quickly switch from writing through the GTT to a read of the physical page directly with the CPU (e.g. performing relocations through the GTT and then running the command parser), we can observe that the writes are not visible to the CPU. It is not a coherency problem, as extensive investigations with clflush have demonstrated, but a mere timing issue - we have to wait for the GTT to complete it's write before we start our read from the CPU. The issue can be illustrated in userspace with: gtt = gem_mmap__gtt(fd, handle, 0, OBJECT_SIZE, PROT_READ | PROT_WRITE); cpu = gem_mmap__cpu(fd, handle, 0, OBJECT_SIZE, PROT_READ | PROT_WRITE); gem_set_domain(fd, handle, I915_GEM_DOMAIN_GTT, I915_GEM_DOMAIN_GTT); for (i = 0; i < OBJECT_SIZE / 64; i++) { int x = 16*i + (i%16); gtt[x] = i; clflush(&cpu[x], sizeof(cpu[x])); assert(cpu[x] == i); } Experimenting with that shows that this behaviour is indeed limited to recent Atom-class hardware. Testcase: igt/gem_exec_flush/basic-batch-default-cmd #byt Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Joonas Lahtinen <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/i915: Before accessing an object via the cpu, flush GTT writesChris Wilson1-0/+4
If we want to read the pages directly via the CPU, we have to be sure that we have to flush the writes via the GTT (as the CPU can not see the address aliasing). Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Joonas Lahtinen <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/i915: Extract i915_gem_obj_prepare_shmem_write()Chris Wilson3-65/+102
This is a companion to i915_gem_obj_prepare_shmem_read() that prepares the backing storage for direct writes. It first serialises with the GPU, pins the backing storage and then indicates what clfushes are required in order for the writes to be coherent. Whilst here, fix support for ancient CPUs without clflush for which we cannot do the GTT+clflush tricks. v2: Add i915_gem_obj_finish_shmem_access() for symmetry Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Joonas Lahtinen <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/i915: Cache kmap between relocationsChris Wilson1-54/+111
When doing relocations, we have to obtain a mapping to the page containing the target address. This is either a kmap or iomap depending on GPU and its cache coherency. Neighbouring relocation entries are typically within the same page and so we can cache our kmapping between them and avoid those pesky TLB flushes. Note that there is some sleight-of-hand in how the slow relocate works as the reloc_entry_cache implies pagefaults disabled (as we are inside a kmap_atomic section). However, the slow relocate code is meant to be the fallback from the atomic fast path failing. Fortunately it works as we already have performed the copy_from_user for the relocation array (no more pagefaults there) and the kmap_atomic cache is enabled after we have waited upon an active buffer (so no more sleeping in atomic). Magic! Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Joonas Lahtinen <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/i915: Fallback to single page pwrite/pread if unable to release fenceChris Wilson2-13/+19
If we cannot release the fence (for example if someone is inexplicably trying to write into a tiled framebuffer that is currently pinned to the display! *cough* kms_frontbuffer_tracking *cough*) fallback to using the page-by-page pwrite/pread interface, rather than fail the syscall entirely. Since this is triggerable by the user (along pwrite) we have to remove the WARN_ON(fence->pin_count). Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Joonas Lahtinen <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/i915: Mark up the GTT flush following WC writes as ORIGIN_CPUChris Wilson1-2/+2
Similarly to invalidating beforehand, if the object is mmapped via I915_MMAP_WC we cannot track writes through the I915_GEM_DOMAIN_GTT. At the conclusion of the write, i915_gem_object_flush_gtt_writes() we also need to treat the origin carefully in case it may have been untracked. See also commit aeecc9696aa0 ("drm/i915: use ORIGIN_CPU for frontbuffer invalidation on WC mmaps"). Signed-off-by: Chris Wilson <[email protected]> Cc: Daniel Vetter <[email protected]> Cc: Paulo Zanoni <[email protected]> Reviewed-by: Paulo Zanoni <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/i915: Use ORIGIN_CPU for fb invalidation from pwriteChris Wilson1-2/+2
As pwrite does not use the fence for its GTT access, and may even go through a secondary interface avoiding the main VMA, we cannot treat the write as automatically invalidated by the hardware and so we require ORIGIN_CPU frontbufer invalidate/flushes. Signed-off-by: Chris Wilson <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected] Reviewed-by: Daniel Vetter <[email protected]>
2016-08-18drm/i915: vfree() no longer ignores the low bits of the addressChris Wilson2-4/+12
Since vfree() now likes to WARN when passed a non-page-aligned pointer, we need to discard the low bits to comply with it. Fixes: d31d7cb1460c ("drm/i915: Support for creating write combined type vmaps") Signed-off-by: Chris Wilson <[email protected]> Cc: Joonas Lahtinen <[email protected]> Cc: Tvrtko Ursulin <[email protected]> Reviewed-by: Tvrtko Ursulin <[email protected]> Reviewed-by: Joonas Lahtinen <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18agp/intel: Flush chipset writes after updating a single PTEChris Wilson1-0/+2
After we update one PTE for a page, the caller expects to be able to immediately use that through a GGTT read/write. To comply with the callers expectations we therefore need to flush the chipset buffers before returning. Reported-by: Matti Hämäläinen <[email protected]> Fixes: d6473f566417 ("drm/i915: Add support for mapping an object page...") Signed-off-by: Chris Wilson <[email protected]> Cc: Ankitprasad Sharma <[email protected]> Cc: Tvrtko Ursulin <[email protected]> Tested-by: Matti Hämäläinen <[email protected]> Cc: [email protected] Reviewed-by: Mika Kuoppala <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/i915: Unconditionally flush any chipset buffers before execbufChris Wilson2-10/+4
If userspace is asynchronously streaming into the batch or other execobjects, we may not flush those writes along with a change in cache domain (as there is no change). Therefore those writes may end up in internal chipset buffers and not visible to the GPU upon execution. We must issue a flush command or otherwise we encounter incoherency in the batchbuffers and the GPU executing invalid commands (i.e. hanging) quite regularly. v2: Throw a paranoid wmb() into the general flush so that we remain consistent with before. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90841 Fixes: 1816f9236303 ("drm/i915: Support creation of unbound wc user...") Signed-off-by: Chris Wilson <[email protected]> Cc: Akash Goel <[email protected]> Cc: Daniel Vetter <[email protected]> Cc: Tvrtko Ursulin <[email protected]> Tested-by: Matti Hämäläinen <[email protected]> Cc: [email protected] Reviewed-by: Mika Kuoppala <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/rockchip: Use DRM_DEV_ERROR in vopSean Paul1-10/+14
Since we can have multiple vops, use DRM_DEV_ERROR to make logs easier to process. Acked-by: Mark Yao <[email protected]> Signed-off-by: Sean Paul <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm: Introduce DRM_DEV_* log messagesSean Paul2-64/+103
This patch consolidates all the various log functions/macros into one uber function, drm_log. It also introduces some new DRM_DEV_* variants that print the device name to delineate multiple devices of the same type. Reviewed-by: Chris Wilson <[email protected]> Reviewed-by: Eric Engestrom <[email protected]> Signed-off-by: Sean Paul <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/edid: CEA mode 64 1080p100 vsync pulse width incorrectClint Taylor1-1/+1
In the CEA-861 specification VIC 64 specifies a vsync pulse of 5 and a backporch of 36. Adjust vsync pulse width to match specification. Cc: Ville Syrjälä <[email protected]> Signed-off-by: Clint Taylor <[email protected]> Signed-off-by: Sean Paul <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-18drm/i915/gen9: Drop invalid WARN() during data rate calculationMatt Roper1-2/+0
It's possible to have a non-zero plane mask and still wind up with a total data rate of zero. There are two cases where this can happen: * planes are active (from the KMS point of view), but are all fully clipped (positioned offscreen) * the only active plane on a CRTC is the cursor (which is handled independently and not counted into the general data rate computations These are both valid display setups (although unusual), so we need to drop the WARN(). Signed-off-by: Matt Roper <[email protected]> Reviewed-by: Maarten Lankhorst <[email protected]> Testcase: kms_universal_planes.cursor-only-pipe-* Signed-off-by: Maarten Lankhorst <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected] Cc: [email protected] #v4.7+
2016-08-18drm/i915/gen9: Initialize intel_state->active_crtcs during WM sanitization (v2)Matt Roper1-1/+16
intel_state->active_crtcs is usually only initialized when doing a modeset. During our first atomic commit after boot, we're effectively faking a modeset to sanitize the DDB/wm setup, so ensure that this field gets initialized before use. v2: - Don't clobber active_crtcs if our first commit really is a modeset (Maarten) - Grab connection_mutex when faking a modeset during sanitization (Maarten) Reported-by: Tvrtko Ursulin <[email protected]> Cc: Tvrtko Ursulin <[email protected]> Cc: Maarten Lankhorst <[email protected]> Signed-off-by: Matt Roper <[email protected]> Tested-by: Tvrtko Ursulin <[email protected]> Signed-off-by: Maarten Lankhorst <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected] Cc: [email protected] #v4.7+
2016-08-17drm/amd/amdgpu: Add more debugfs config dataTom St Denis1-1/+5
Adds family and external_rev_id to config data Signed-off-by: Tom St Denis <[email protected]> Acked-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-17drm/amdgpu: Hardcode virtual DCE vblank / scanout position return valuesEmily Deng1-12/+4
For virtual display feature, by hardcoding 0 for the vblank counter and -EINVAL for the scanout position return value, we signal to the core DRM code that there are no hardware counters we can use for these. Signed-off-by: Emily Deng <[email protected]> Reviewed-by: Michel Dänzer <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-17drm/i915: Add missing kerneldoc for guc_client_alloc:enginesChris Wilson1-0/+1
Brief parameter text to describe @engines. Signed-off-by: Chris Wilson <[email protected]> Cc: Dave Gordon <[email protected]> Cc: Tvrtko Ursulin <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected] Reviewed-by: Tvrtko Ursulin <[email protected]>
2016-08-17drm/i915: Remember to set vma->pages for the preallocated stolen objectChris Wilson1-0/+1
In commit 247177ddd517 ("drm/i915: Always set the vma->pages"), as it title implies, we always set vma->pages for bound objects. Even before that, we would set vma->ggtt_view.pages, for globally bound objects. This was forgotten for the fixup inside the preallocated stolen objects, which has to recreate a global GTT binding outside of the usual VMA insertion path Fixes: 247177ddd517 ("drm/i915: Always set the vma->pages") Signed-off-by: Chris Wilson <[email protected]> Cc: Joonas Lahtinen <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected] Reviewed-by: Joonas Lahtinen <[email protected]>
2016-08-17Revert "drm/hisilicon: Don't set drm_device->platformdev"Daniel Vetter3-9/+11
This reverts commit d25bcfb8c2e18b9b36f037f38be4d4792ebf8d57. I somehow missed that it only compiles on arm64 and broke the driver rather badly. Signed-off-by: Daniel Vetter <[email protected]>
2016-08-17drm/i915: Mark i915_hpd_poll_init_work as staticChris Wilson1-1/+2
Local function with forgotten static declaration. Fixes: 19625e85c6ec ("drm/i915: Enable polling when we don't have hpd") Signed-off-by: Chris Wilson <[email protected]> Cc: Ville Syrjälä <[email protected]> Cc: Daniel Vetter <[email protected]> Cc: Lyude <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected] Reviewed-by: Tvrtko Ursulin <[email protected]>
2016-08-17drm/i915: Mark the static key for movntqda as staticChris Wilson1-1/+1
Silence sparse who warns that the global variable is not declared static. Fixes: 0b1de5d58e19 ("drm/i915: Use SSE4.1 movntdqa to ...") Signed-off-by: Chris Wilson <[email protected]> Cc: Akash Goel <[email protected]> Cc: Damien Lespiau <[email protected]> Cc: Mika Kuoppala <[email protected]> Cc: Tvrtko Ursulin <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected] Reviewed-by: Tvrtko Ursulin <[email protected]>
2016-08-17drm/i915: Initialize legacy semaphores from engine hw id indexed arrayTvrtko Ursulin2-28/+34
Build the legacy semaphore initialisation array using the engine hardware ids instead of driver internal ones. This makes the static array size dependent only on the number of gen6 semaphore engines. Also makes the per-engine semaphore wait and signal tables hardware id indexed saving some more space. v2: Refactor I915_GEN6_NUM_ENGINES to GEN6_SEMAPHORE_LAST. (Chris Wilson) v3: More polish. (Chris Wilson) Signed-off-by: Tvrtko Ursulin <[email protected]> Reviewed-by: Chris Wilson <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-17drm/i915: Add enum for hardware engine identifiersTvrtko Ursulin2-9/+15
Put the engine hardware id in the common header so they are not only associated with the GuC since they are needed for the legacy semaphores implementation. Signed-off-by: Tvrtko Ursulin <[email protected]> Reviewed-by: Chris Wilson <[email protected]>
2016-08-17dma-buf: fix kernel-doc warning and typosRandy Dunlap1-3/+3
Fix dma-buf kernel-doc warning and 2 minor typos in fence_array_create(). Fixes this warning: ..//drivers/dma-buf/fence-array.c:124: warning: No description found for parameter 'signal_on_any' Signed-off-by: Randy Dunlap <[email protected]> Cc: Sumit Semwal <[email protected]> Cc: [email protected] Cc: [email protected] Cc: [email protected] Signed-off-by: Sumit Semwal <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16drm: Fix kerneldoc in drm_plane_helper.cDaniel Vetter1-1/+1
Ville ocd'ed the parameter name, but forgot to update the docs! Fixes: df86af9133b4 ("drm/plane-helper: Add drm_plane_helper_check_state()") Cc: Sean Paul <[email protected]> Cc: Ville Syrjälä <[email protected]> Reviewed-by: Sean Paul <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16vgaarbiter: rst-ifiy and polish kerneldocDaniel Vetter5-197/+205
Move the documentation into Documentation/gpu, link it up and pull in the kernel doc. No actual text changes except that I did polish the kerneldoc a bit, especially for vga_client_register(). v2: Remove some rst from vga-switcheroo.rst that I don't understand, but which seems to be the reason why the new vgaarbiter.rst sometimes drops out of the sidebar index. v3: Drop one level of headings and clarify the vgaarb one a bit. v4: Fix some typos (Sean). Cc: Jonathan Corbet <[email protected]> Cc: [email protected] Cc: Sean Paul <[email protected]> Reviewed-by: Sean Paul <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16drm: document drm_display_infoDaniel Vetter5-61/+97
We seem to have a bit a mess in how to describe the bus formats, with a multitude of competing ways. Might be best to consolidate it all and use MEDIA_BUS_FMT_ also for the hdmi color formats and high color modes. Also move all the display_info related functions into drm_connector.c (there's only one) to group it all together. I did decided against also moving the edid related display info functions, they seem to fit better in drm_edid.c. Instead sprinkle a few cross references around. While at that reduce the kerneldoc for static functions, there's not point in documenting internals with that much detail really. v2: Fix typo and move misplaced hunk (Sean). Cc: Sean Paul <[email protected]> Reviewed-by: Sean Paul <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16drm: Remove display_info->min/max_(h|v)maxDaniel Vetter5-36/+0
No one looks at it, only i915/gma500 lvds even bother to fill it out. I guess a very old plan was to use this for filtering modes, but that's already done within the edid parser. v2: Move misplaced hunk to this patch. Reviewed-by: Sean Paul <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16drm: Update connector documentationDaniel Vetter3-166/+94
- Shuffle docs from drm-kms.rst into the structure docs where it makes sense. - Put the remaining bits into a new overview section. One thing I've changed is around probing: Old docs says that you _must_ use the probe helpers, which isn't correct. Helpers are always optional. v2: Review from Sean. Cc: Sean Paul <[email protected]> Reviewed-by: Sean Paul <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16drm: Don't export dp-aux devnode functionsDaniel Vetter5-39/+23
They're only used internally within the dp helpers. Also nuke the kerneldoc (we only document the driver interface in the drm shared functions). And move the header file from the public include/ directory to the source files into drm_crtc_helper_internal.h, similar to how we already have drm_crtc_internal.h. While at it also move drm_fb_helper_modinit since that belongs in there, too. I noticed this all since I spotted kerneldoc which wasn't pulled into the rst templates. v2: Update Copyright date. Cc: Sean Paul <[email protected]> Cc: Rafael Antognolli <[email protected]> Reviewed-by: Sean Paul <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16drm/doc: Include new drm_blend.cDaniel Vetter5-9/+15
There's not much point in kerneldoc if it's not included: - It won't show up in the pretty html pages. - The comments itself won't get parsed, which means 0day won't pick up changes, resulting in stale docs fast. Also, uapi really should be core, not helpers, so move drm_blend.c to that. That also means that the zpos normilize function loses it's helper status (and we might as well call it always). For that, EXPORT_SYMBOL. Just spotted while integrating docs and noticing that one was missing. With sphinx there's really no excuse any more to not build the docs and make sure it's all nice! $ make DOCBOOKS="" htmldocs Cc: Marek Szyprowski <[email protected]> Cc: Benjamin Gaignard <[email protected]> Cc: Laurent Pinchart <[email protected]> Cc: Ville Syrjälä <[email protected]> Reviewed-by: Sean Paul <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16drm: Extract drm_connector.[hc]Daniel Vetter9-1729/+1773
Pulls in quite a lot of connector related structures (cmdline mode, force/status enums, display info), but I think that all makes perfect sense. Also had to move a few more core kms object stuff into drm_modeset.h. And as a first cleanup remove the kerneldoc for the 2 connector IOCTL - DRM core docs are aimed at drivers, no point documenting internal in excruciating detail. v2: And also pull in all the connector property code. Reviewed-by: Sean Paul <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16drm: Export drm_property_replace_global_blobDaniel Vetter2-6/+13
It's really part of the core blob interface, and the drm_connector.c extraction needs it too. Reviewed-by: Sean Paul <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16drm: update kerneldoc for changes introduced by commits "16fadc2568e9" and ↵Lothar Waßmann1-0/+10
"9671e228fb78" Describe the new parameter 'bus_flags' to of_get_drm_display_mode() in the kerneldoc comments and add kerneldoc comments to the new function drm_bus_flags_from_videomode(). Signed-off-by: Lothar Waßmann <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16drm/doc: Update drm_framebuffer docsDaniel Vetter3-37/+130
- Move the intro section into a DOC comment, and update it slightly. - kernel-doc for struct drm_framebuffer! v2: - Copypaste fail (Sean). - Explain the linear @offsets clearer (Ville). Cc: Sean Paul <[email protected]> Cc: Ville Syrjälä <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected]
2016-08-16drm: Extract drm_framebuffer.[hc]Daniel Vetter9-958/+1075
Also start with drm_modeset.h with the core bits, since we need to untangle this mess somehow. That allows us to move the drm_modes.h include to the right spot, except for the temporary connector status enum. That will get fixed as soon as drm_connector.h exists. v2: Rebase. v3: Move drm_crtc_force_disable_all back again, that wasn't meant to be moved (Sean). v4: Rebase. Cc: Sean Paul <[email protected]> Reviewed-by: Sean Paul <[email protected]> Signed-off-by: Daniel Vetter <[email protected]>
2016-08-16drm/amd/amdgpu: Add more config data for debugfsTom St Denis1-1/+6
Adds rev_id as well as cg/pg flags to help debug runtime. Signed-off-by: Tom St Denis <[email protected]> Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-16drm/amdgpu: fix coding style in amdgpu_object.cChristian König1-37/+49
Just a few 80 chars problems. Signed-off-by: Christian König <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-16drm/amdgpu: add function pointer to the pte_update_paramsChristian König1-22/+54
Remember what function to call while planning the commands instead of figuring it our later on. Signed-off-by: Christian König <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-16drm/amdgpu: stop splitting PTE commands into smaller onesChristian König5-149/+73
It doesn't make much sense to create bigger commands first which we then need to split into smaller one again. Just make sure the commands we create aren't to big in the first place. Signed-off-by: Christian König <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-16drm/amdgpu: remove AMDGPU_VM_NO_FLUSH defineChristian König1-3/+0
Not used any more. Signed-off-by: Christian König <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-16drm/amdgpu: cleanup the write_pte implementationsChristian König5-98/+54
We don't need the gart mapping handling here any more. Signed-off-by: Christian König <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-16drm/amdgpu: remove pages_addr handling from the VM codeChristian König1-9/+1
Not needed any more. Signed-off-by: Christian König <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-16drm/amdgpu: write PTEs directly into the IB.Christian König1-5/+21
Write the PTEs at the end of the IB instead of directly into the SDMA commands. This can save quite some CPU cycles building the entries. This doesn't change the DW estimation because PTEs where embedded into the IB before as well. It just moves them to the end of the IB. Signed-off-by: Christian König <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-16drm/amdgpu: add shadow flag V2Chunming Zhou1-0/+8
Indicate if need to sync between bo and shadow, where sync to where. V2: Rename to backup_shadow Signed-off-by: Chunming Zhou <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-16drm/amdgpu: allocate shadow for pd/pt bo V2Chunming Zhou1-3/+11
The pd/pt shadow bo will be used to backup page table, when gpu reset happens, we can restore the page table by them. V2: Free shadow bo. Signed-off-by: Chunming Zhou <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2016-08-16drm/amdgpu: validate shadow as well when validating boChunming Zhou1-30/+46
Signed-off-by: Chunming Zhou <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.