Chris Wilson [Fri, 8 Nov 2013 12:58:06 +0000 (12:58 +0000)]
intel-virtual-output: Add DBG for the latest XSync
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 8 Nov 2013 12:54:01 +0000 (12:54 +0000)]
intel-virtual-output: Perform an explicit sync
If we fail to track rendering using ShmCompletionEvents and begin to
drop frames, insert an explicit XSync.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 8 Nov 2013 12:46:08 +0000 (12:46 +0000)]
intel-virtual-output: Check for pending events before skipping draw
Make sure we check if the destination display has caught up before
skipping a frame.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 8 Nov 2013 12:30:21 +0000 (12:30 +0000)]
intel-virtual-output: Add some DBG for bad_visual()
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 8 Nov 2013 08:53:55 +0000 (08:53 +0000)]
sna: Always schedule upload buffers for retirement after use
Even if they are multiply referenced due to cached references.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 7 Nov 2013 14:54:39 +0000 (14:54 +0000)]
sna: Only assert that the bo is small enough to be mmapped
Not that is currently in a position to be mapped without eviction.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 7 Nov 2013 13:09:25 +0000 (13:09 +0000)]
sna/gen7: Request secure batches for Haswell vsync
Since commit
8ff8eb2b38dc705f5c86f524c1cd74a811a7b04c
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon Sep 9 16:23:04 2013 +0100
sna/hsw: Scanline waits require both DERRMR and forcewake
we have been emitting LRI to enable vsync on the render ring. This
requires a privileged batch buffer, and whilst we were checking for
kernel support, we forgot to actually tell the kernel to submit the
batch with the right privileges.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71328
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 6 Nov 2013 16:24:28 +0000 (16:24 +0000)]
sna: Always copy from the tile source
The first step when tiling with an intermediatory is to copy from the
source bo to the temporary. The alu should only be applied when copying
from the temporary to the destination.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 6 Nov 2013 15:42:35 +0000 (15:42 +0000)]
sna: Remove erroneous assertion from sna_tiling_blt_composite
We have to rely on the caller only calling us for a PictOpSrc equivalent
operation as they don't all set op->op.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 6 Nov 2013 15:04:24 +0000 (15:04 +0000)]
sna: Add a DBG breadcrumb to kgem_bo_convert_to_gpu()
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 6 Nov 2013 14:51:42 +0000 (14:51 +0000)]
sna: Use tiling BLT fallback for BLT composite operations
This avoid a circuituous route through the render pathways and multiple
levels of tiling fallbacks to accomplish the same copy.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 6 Nov 2013 13:42:27 +0000 (13:42 +0000)]
sna: Trim create flags if tiled sizes are too large
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 6 Nov 2013 13:41:39 +0000 (13:41 +0000)]
sna: Fences are power-of-two sizes
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 6 Nov 2013 11:38:53 +0000 (11:38 +0000)]
sna: Fixup unwind checks for future BLT commands
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 6 Nov 2013 09:05:05 +0000 (09:05 +0000)]
sna: Clear the execlist if we completely unwind a batch after using 8x8 tiling
Another location, but the right function this time!
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 6 Nov 2013 09:40:56 +0000 (09:40 +0000)]
sna: Move initialisation of loop counter to common code
It was only being defined in x86-64 and left unitialized for x86-32.
Fixes regression from
commit
587c4866652e40e1e228b333028114766a6d3b08
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon Nov 4 15:10:40 2013 +0000
sna: Promote uint16_t to a full int to avoid overflow in computing w*h in memcpy_xor
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71286
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 6 Nov 2013 09:05:05 +0000 (09:05 +0000)]
sna: Clear the execlist if we completely unwind a batch after using 8x8 tiling
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 6 Nov 2013 08:56:01 +0000 (08:56 +0000)]
sna: Be more pessimistic for tiling sizes on older gen
On the older generation, we have severe alignment penalties for fenced
regions which dramatically reduce the amount of space we can effectively
use in a batch. To accommodate this, reduce the tiling step size.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 5 Nov 2013 22:41:06 +0000 (22:41 +0000)]
sna: Only operate inplace if no existing CPU damage for a read
With a large object, we try harder to operate inplace (to avoid creating
a second large CPU bo). This introduced an issue where we tried to read
from the GPU bo when there was already existing damage in the CPU -
triggering an assertion.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 5 Nov 2013 21:59:37 +0000 (21:59 +0000)]
sna: Trim the overestimate of required aperture space for fence alignment
We can optimistically only require that we waste the largest fence
region in a batch, as all other fences will then be naturally aligned as
well. So long as the kernel succeeds in defragmenting the aperture...
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 5 Nov 2013 21:33:18 +0000 (21:33 +0000)]
sna: Mark partial move_area_to_gpu with MOVE_READ on promotion to move_to_gpu
When promoting a partial move_area_to_gpu to a full move_to_gpu, we have
to disable certain optimisations that we try to use if MOVE_READ==0.
Reported-and-tested-by: Matti Hamalainen <ccr@tnsp.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71198
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 5 Nov 2013 18:36:45 +0000 (18:36 +0000)]
sna: Submit execution on the bo before changing its caching status
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 5 Nov 2013 18:29:46 +0000 (18:29 +0000)]
sna: Clear snoop flag after converting from a CPU bo
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 5 Nov 2013 08:49:28 +0000 (08:49 +0000)]
sna: Use an inplace exchange for large untiled BO
On older architectures, large BO have to be untiled and so we can reuse
an existing CPU bo by adjusting its caching mode.
References: https://bugs.freedesktop.org/show_bug.cgi?id=70924
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 5 Nov 2013 08:38:22 +0000 (08:38 +0000)]
sna: Be move conservative with tiling sizes for older fenced gen
The older generations have stricter requirements for alignment of fenced
GPU surfaces, so accommodate this by reducing our estimate available
space for the temporary tile.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 4 Nov 2013 22:14:05 +0000 (22:14 +0000)]
test
Chris Wilson [Mon, 4 Nov 2013 16:45:00 +0000 (16:45 +0000)]
sna: Check for loss of state after setting up a fill BLT op
If we install a BLT fill operation early in the drawing sequence (i.e.
before calling a mi routine), we may lose our state to delayed
initialisation of sources and so need to subsequently recheck.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 4 Nov 2013 15:17:51 +0000 (15:17 +0000)]
sna: s/active/busy/ to silence compiler warning
kgem.c: In function 'kgem_check_bo':
kgem.c:4768:7: warning: declaration of 'active' shadows a global declaration [-Wshadow]
kgem.c:692:21: warning: shadowed declaration is here [-Wshadow]
kgem.c: In function 'kgem_check_many_bo_fenced':
kgem.c:4907:7: warning: declaration of 'active' shadows a global declaration [-Wshadow]
kgem.c:692:21: warning: shadowed declaration is here [-Wshadow]
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 4 Nov 2013 15:10:40 +0000 (15:10 +0000)]
sna: Promote uint16_t to a full int to avoid overflow in computing w*h in memcpy_xor
Reported-by: Conley Moorhous <conleymoorhous@gmail.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=70527
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 4 Nov 2013 12:57:01 +0000 (12:57 +0000)]
sna: Apply the BLT source offset for individual copies
Following a complex path through multiple layers of indirections and
tiling fallbacks, resulted in hitting a path where the source offset was
subsequently ignored. This leads to the operation reading from invalid
memory (or hitting the assert warning about the same).
References: https://bugs.freedesktop.org/show_bug.cgi?id=70924
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 4 Nov 2013 11:48:04 +0000 (11:48 +0000)]
sna: Only use the simple stipple upload path if wholly contained
If the stipple box is outside of the stipple pixmap, we need to
carefully upload the stipple using the modulus operation.
Buzilla: https://bugs.launchpad.net/bugs/1247785
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 4 Nov 2013 11:47:29 +0000 (11:47 +0000)]
sna: Update DBG formats for larger BO offset integers
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 4 Nov 2013 11:30:39 +0000 (11:30 +0000)]
sna: Remove the replace indirection prior to performing write_boxes
As write_boxes itself decides whether or not to stage the upload into
the destination bo, we can destroy the temporary allocation along the
write_boxes fallback path (i.e. after failing to map the destination
bo).
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 4 Nov 2013 11:12:30 +0000 (11:12 +0000)]
sna: Only destroy the old GPU after successfully replacing it
Along the sna_replace__xor path we destroyed the priv->gpu_bo twice upon
successfully replacing it.
References: https://bugs.freedesktop.org/show_bug.cgi?id=70527
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 4 Nov 2013 09:34:25 +0000 (09:34 +0000)]
sna: Rename the two variants of pwrite so very relative dangers are obvious
Since we extend the write in the cache-aligned routine, it runs the risk
of reading from beyond the end of the allocation. As such, callers
should be carefully vetted to make sure that their allocations are
already cache-aligned (typically page-aligned). To make it obvious that
this complexity exists, rename the routine.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 4 Nov 2013 09:31:05 +0000 (09:31 +0000)]
sna: Use the unoptimized pwrite for general buffers
When we call kgem_bo_write() we have less control over the allocation of
the buffer, and do not ensure it meets the alignment criteria required
for the cacheline optimisation. So use the simple pwrite routine to
avoid reading beyond the end of the allocation.
Reported-and-tested-by: Mark Kettenis <mark.kettenis@xs4all.nl>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 3 Nov 2013 22:07:22 +0000 (22:07 +0000)]
sna: Use fast-path target placement if we are also IGNORE_CPU
If we are ignoring CPU damage, we also need only to check GPU damage
when considering placement of the target bo.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 3 Nov 2013 19:51:56 +0000 (19:51 +0000)]
sna: Wrap staging buffer access with sigtrap handler
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 2 Nov 2013 21:45:00 +0000 (21:45 +0000)]
sna: Discard bo->exec when unwinding a glyph no-op
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 2 Nov 2013 21:13:45 +0000 (21:13 +0000)]
sna: Eliminate the redundancy between is_mappable/can_map
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 2 Nov 2013 21:01:26 +0000 (21:01 +0000)]
sna: Jump to the right escape target when bypassing a self-copy
Another fix for
commit
e3f15cbf39696edae9f716bdcfbb7032ec7d7e3f [2.99.905]
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue Oct 22 15:19:15 2013 +0100
sna: Move gc back to GPU after failure to move it to CPU
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 2 Nov 2013 15:12:51 +0000 (15:12 +0000)]
sna: Add some more asserts around inactive VMA lists
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 2 Nov 2013 14:43:49 +0000 (14:43 +0000)]
sna: Tweak estimate of maximum usable aperture for strict fencing
Old generations have very strict alignment requirements even for
unfenced tiled accesses which restricts the amount of aperture space
available for use, and in the process estimate for the effect of
framebuffer fragmentation on the mappable aperture.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 30 Nov 2012 16:28:02 +0000 (16:28 +0000)]
Prepare for changes in the BLT unit on future generations
Adapt the legacy BLT commands in preparation for future changes.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 2 Nov 2013 13:11:31 +0000 (13:11 +0000)]
sna: Fallback when wedged and trying to use the BLT copy routines
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 1 Nov 2013 15:57:56 +0000 (15:57 +0000)]
sna: Guard the replace-with-xor fallback path
Before attempting to map the destination for uploading into after a
failure to use the BLT, we need to recheck that it is indeed mappable.
References: https://bugs.freedesktop.org/show_bug.cgi?id=70924
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 1 Nov 2013 14:30:14 +0000 (14:30 +0000)]
sna: Process all clip boxes when rastering segments
The macro still had buried returns which were no longer valid after the
translation to handle clipping. They needed to be breaks from the inner
most loops to the outer clip box instead.
Reported-by: Clemens Eisserer <linuxhippy@gmail.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=70802
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 1 Nov 2013 12:49:36 +0000 (12:49 +0000)]
sna/gen7: Undo overzealous flushing
commit
d580a30aaf97687f9669ea72fbc3310c2fea26f1
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu Oct 31 15:58:47 2013 +0000
sna/gen7: Flush render cache when changing CC state
ultimately doesn't prevent the issue and in the process adversely
affects perforamnce.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 1 Nov 2013 10:48:06 +0000 (10:48 +0000)]
sna: Scale uses of aperture_mappable by PAGE_SIZE
After converting aperture_mappable to count in pages, there were a few
residual users expecting a byte count.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71117
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 31 Oct 2013 15:58:47 +0000 (15:58 +0000)]
sna/gen7: Flush render cache when changing CC state
References: https://bugs.freedesktop.org/show_bug.cgi?id=68410
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 31 Oct 2013 13:35:59 +0000 (13:35 +0000)]
sna/gen6: Tweak flush around CC state changes
In order to fix some font corruption, it appears that we need an extra
flush in the Sandybridge pipeline when we change the CC stage and the
render cache is dirty. We previously triggered a full pipeline stall
for this case.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 31 Oct 2013 10:32:46 +0000 (10:32 +0000)]
sna: Tweak deletion of used buffers
Make sure we never unwind a used buffer.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 30 Oct 2013 18:48:10 +0000 (18:48 +0000)]
sna: Allow limited recursion within sigtrapped routines
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 30 Oct 2013 18:42:06 +0000 (18:42 +0000)]
Revert "sna: Detect and handle mi recursion"
This reverts commit
c6b0e3fe0c299488932ba0392847f1faf298d079.
Chris Wilson [Wed, 30 Oct 2013 13:26:23 +0000 (13:26 +0000)]
sna: Assert that the batch is in the correct mode prior to inserting BLT commands
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 30 Oct 2013 11:52:05 +0000 (11:52 +0000)]
sna: Detect and handle mi recursion
Some of the mi routines recuse back into the generic accel routines and
so confuse our trapping of SIGBUS. Add extra assertions to pinpoint the
recursion and unwrap sufficiently to avoid that recursion.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 30 Oct 2013 10:02:28 +0000 (10:02 +0000)]
sna: Queue retirement when placing a BO on the deferred flushing list
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 30 Oct 2013 09:55:57 +0000 (09:55 +0000)]
sna: Check for wedged after submitting
If we submit prior to an operation, check that we didn't just wedge the
GPU.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 29 Oct 2013 21:19:02 +0000 (21:19 +0000)]
sna: Wrap access of upload buffers with sigtrap
For an extra layer of paranoia, catch any sigbus when trying to upload
a bitmap, and convert it to a no-op.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 29 Oct 2013 20:16:06 +0000 (20:16 +0000)]
sna/gen7: Refactor BYT vsync using IVB routines
The BYT vsync is closer to the IVB vsync, and using gen6 is just
erroneous. Apparently. At least that is what is in bspec today.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 29 Oct 2013 18:27:41 +0000 (18:27 +0000)]
sna: Fix assertion checks for fake flushing requests
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 29 Oct 2013 18:00:49 +0000 (18:00 +0000)]
sna: Remove an overzealous assert
Testing with a slow gen2 triggers this harmless assertion, so drop it.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 29 Oct 2013 16:59:57 +0000 (16:59 +0000)]
sna: Handle deferred retiring of active-buffers
Yikes, the new assertions found cases where we would retain an active
buffer even though it was still owned by the next batch. Fortunately,
this should only be a bookkeeping issue rather than lead to corruption.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 29 Oct 2013 16:20:14 +0000 (16:20 +0000)]
sna: Always run the retire_buffers during ring_idle
As we know that we had a request to retire, we know that we may make
progress retiring active buffers.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 29 Oct 2013 14:53:02 +0000 (14:53 +0000)]
sna: Perform an explicit retire for old active upload buffers
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 29 Oct 2013 15:12:17 +0000 (15:12 +0000)]
sna: Check for retired upload buffers after checking for an idle ring
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 29 Oct 2013 13:50:51 +0000 (13:50 +0000)]
sna: asserts bitmap uploads are correct
Place guards around the pointer accesses to verify that they are within
the bitmap.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 29 Oct 2013 12:46:09 +0000 (12:46 +0000)]
sna: Don't attempt to move the GC back to the GPU before it is moved away
Fixes regression from
commit
e3f15cbf39696edae9f716bdcfbb7032ec7d7e3f [2.99.905]
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue Oct 22 15:19:15 2013 +0100
sna: Move gc back to GPU after failure to move it to CPU
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 29 Oct 2013 09:56:10 +0000 (09:56 +0000)]
sna: Try harder to complete writes
Expunge our caches if we fail to write into a bo (presuming that
allocation failure is the likely fixable cause).
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 28 Oct 2013 20:48:17 +0000 (20:48 +0000)]
sna: Quieten a couple of valgrind warnings about unknown ioctls
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 28 Oct 2013 13:06:16 +0000 (13:06 +0000)]
sna: Tidy RegionNil checks
After computing the clip intersection, we immediately check for the
empty result, so refactor the check into the common routine.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 28 Oct 2013 11:29:13 +0000 (11:29 +0000)]
sna: And be pessimistic when checking aperture limits
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 28 Oct 2013 11:28:01 +0000 (11:28 +0000)]
sna: Account for extra guard pages around snooped BO in aperture checks
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 28 Oct 2013 11:16:29 +0000 (11:16 +0000)]
sna: Defer opportunistic flush if all bo are current on the GPU
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 28 Oct 2013 10:33:57 +0000 (10:33 +0000)]
sna: Leave extra room in the mappable aperture for fence alignment
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 28 Oct 2013 10:45:52 +0000 (10:45 +0000)]
sna: Use page-count for mappable aperture size
For consistency with the other aperture metrics and correctness when
passing around required number of pages.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 27 Oct 2013 16:45:53 +0000 (16:45 +0000)]
sna: Fallthrough to opportunistic flushing after aperture checks
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 27 Oct 2013 15:33:13 +0000 (15:33 +0000)]
sna: Handle transient TearFree flip failures
If we get a pageflip fail when trying to do a TearFree update, just
fallback to a copy (before turning off the display for complete
failure). The rare tearing copy should mar the user experience far less.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 25 Oct 2013 14:09:19 +0000 (15:09 +0100)]
sna: Trim usuable fenced aperture by unfenced usage
This will overcount buffers that both fenced and unfenced (likely most
fenced buffers!) but is required to prevent ENOSPC due to alignment
issues.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 25 Oct 2013 13:27:21 +0000 (14:27 +0100)]
sna: Don't assert indirect GPU state
kgem_check_bo_fenced queries whether the GPU is idle, and so its result
is timing dependent and not suitable for assertion.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 25 Oct 2013 13:22:05 +0000 (14:22 +0100)]
sna: Remove stale mappings when replacing GPU bo
References: https://bugs.freedesktop.org/show_bug.cgi?id=70527
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 25 Oct 2013 12:20:10 +0000 (13:20 +0100)]
sna: Fix overallocation of fences for BLT commands (gen2/3)
Fixes a regression introduced in
commit
4d2840919fd00e4ae0bf1957f66009f55a01eb7b
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu Sep 26 15:30:58 2013 +0100
sna: In desperation, query the actual available aperture space
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 25 Oct 2013 10:35:39 +0000 (11:35 +0100)]
sna: Add a line of DBG for when we discard uploads into CPU bo
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 25 Oct 2013 10:23:21 +0000 (11:23 +0100)]
sna/io: Propagate failure to XOR uploads
Similar to the handling required for the normal upload paths.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 25 Oct 2013 10:03:05 +0000 (11:03 +0100)]
sna/io: Wrap the XOR upload paths with SIGBUS handling
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 24 Oct 2013 18:45:30 +0000 (19:45 +0100)]
man: Mention the new default accel
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 24 Oct 2013 18:38:53 +0000 (19:38 +0100)]
man: State the negative aspects of TearFree
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 24 Oct 2013 14:42:12 +0000 (15:42 +0100)]
sna: Do not apply drawable offsets to DamageRegionAppend
DamageRegionAppend() will itself apply the drawable offsets, so we need
to pass it the untransformed region. This also fixes an issue where we
might fallback without applying any Damage.
Reported-by: Christian Nassau <nassau@nullhomotopie.de>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=32734
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 24 Oct 2013 12:05:16 +0000 (13:05 +0100)]
sna: Fix canonical mode name to correctly use asprintf
Whoops. Only the kernel sprintf() accepts NULL as a valid target string,
not libc's.
Reported-by: Jay Little <jaylittle@jaylittle.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=70835
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 22 Oct 2013 13:22:14 +0000 (14:22 +0100)]
sna: Remove the move-to-gpu shortcircuiting for partial GPU, no CPU damage
Otherwise we may not correctly undo COW or pending move-to-gpu updates.
v2: goto move_to_gpu instead for correcting the damage handling.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=70821
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 24 Oct 2013 06:28:18 +0000 (07:28 +0100)]
Revert "sna: Remove the move-to-gpu shortcircuiting for partial GPU, no CPU damage"
This reverts commit
bccbbf8a68faf7aac274efd289d4916083b1a1e3.
Chris Wilson [Wed, 23 Oct 2013 15:15:35 +0000 (16:15 +0100)]
man: Describe the TearFree option
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 23 Oct 2013 10:15:28 +0000 (11:15 +0100)]
sna: Reset bo after allocation failure during wait-for-shadow
If we fail to allocate an alternate bo to replace the still visible
scanout bo, we stall and were supposed to continue using the old
wait->bo. However, we chose to use NULL instead which is going to lead
to an eventual crash.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 23 Oct 2013 08:50:24 +0000 (09:50 +0100)]
2.99.905 snapshot
Chris Wilson [Wed, 23 Oct 2013 07:40:43 +0000 (08:40 +0100)]
sna: Tidy opportunistic flushing
Rearrange the common code so that it should be tail-call optimised.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 22 Oct 2013 16:36:58 +0000 (17:36 +0100)]
sna: Flush the pending move-to-gpu upon Pixmap free
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 22 Oct 2013 16:35:33 +0000 (17:35 +0100)]
sna: Handle pending move-to-gpu for userptr uploads
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 22 Oct 2013 16:34:59 +0000 (17:34 +0100)]
sna: Pass the async hint when pushing GPU damage for redisplay
So that we don't accidentally clear any other hints.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 22 Oct 2013 15:59:29 +0000 (16:59 +0100)]
sna: Fix inconsistency in has_coherent_ptr assertion
We occasionally will read from an "incoherent" ptr in the belief that
the clflush hit will be worth it.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 22 Oct 2013 13:22:14 +0000 (14:22 +0100)]
sna: Remove the move-to-gpu shortcircuiting for partial GPU, no CPU damage
Otherwise we may not correctly undo COW or pending move-to-gpu updates.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>