Chris Wilson [Sun, 15 Dec 2013 10:41:22 +0000 (10:41 +0000)]
sna: Remove stale assertion
Now that we do not remove the move-to-gpu for a plain read from the
TearFree buffer, it is viable to use that buffer as source for cloning.
So we need to kill the stale assertion.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 15 Dec 2013 09:45:24 +0000 (09:45 +0000)]
intel-virtual-output: Disable build if timerfd is not present
Otherwise the build breaks on *BSD.
Reported-by: Yuta SATOH <nigoro.gentoo@0x100.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=72707
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 15 Dec 2013 09:39:54 +0000 (09:39 +0000)]
sna: Avoid recursive out-of-order operations in the middle of TearFree
During TearFree, if we have to wait for the shadow flip to complete, we
run the event handler for the device. This can then cause us to evaluate
pending completed vblank events, which may in turn then attempt to use
the shadow bo and recurse into the TearFree handler. Try to prevent this
and requeue the vblank event.
Reported-by: Pascal de Bruijn <pmjdebruijn@pcode.nl>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=72690
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 12 Dec 2013 16:49:02 +0000 (16:49 +0000)]
sna/gen4+: Drop new assertion that we only finish a vbo after use in this batch
A full vbo might not be discarded and so upon first use in the following
batch we decided to finish the old vbo and allocate a new. This can
happen before we even emit any relocations and so the assert is bogus.
Reported-by: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 11 Dec 2013 21:30:06 +0000 (21:30 +0000)]
sna: Remove unused variable
sna_display.c: In function 'sna_crtc_redisplay__fallback':
sna_display.c:4104:19: warning: unused variable 'sna_crtc' [-Wunused-variable]
struct sna_crtc *sna_crtc = to_sna_crtc(crtc);
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 22 Jul 2013 17:56:03 +0000 (18:56 +0100)]
intel: Recognise gen8
Assign gen=8 to the Broadwell PCI IDs, no marketing names are known at
this point in time.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 23 Jul 2013 07:17:58 +0000 (08:17 +0100)]
sna: Enable scanline waits for Broadwell
Broadwell uses the same mechanism as Haswell for vsync.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 6 Dec 2012 22:11:32 +0000 (22:11 +0000)]
sna/gen8: Initial backend for Broadwell
Should match the functionality of the earlier generations, but untuned.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 13 Nov 2013 11:50:52 +0000 (11:50 +0000)]
Update i915_pciids.h
Pulling in changes from:
kernel commit
4d4dead67a4ab1d5de393f15ed5e4e2aa63d3bcf
Author: Ben Widawsky <benjamin.widawsky@intel.com>
Date: Sun Nov 3 16:47:33 2013 -0800
drm/i915/bdw: Add device IDs
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 10 Dec 2013 10:25:56 +0000 (10:25 +0000)]
sna/gen2,3: Replace stale bo replacement with simple asserts
The surface creation routines should ensure that the minimum pitch
alignments are met. However, when rendering to a CPU bo, the pitch is
not under our control so we need to be more careful.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 10 Dec 2013 12:37:44 +0000 (12:37 +0000)]
sna: Set minimum pitch alignment to 8 bytes
This prevents render corruption on gen3 and should not adversely affect
BLT pitch requirements or other GPUs.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 9 Dec 2013 10:49:56 +0000 (10:49 +0000)]
sna: Only discard READ hint for TearFree copy iff it subsumes the pixmap
Remembering that the target window may be smaller than its backing
pixmap.
Reported-by: Andreas Klauer <Andreas.Klauer@metamorpher.de>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=72430
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 9 Dec 2013 10:32:05 +0000 (10:32 +0000)]
sna: Fallback gracefully when redisplaying after a GPU hang
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 8 Dec 2013 16:35:27 +0000 (16:35 +0000)]
sna: Simplify kgem_bo_can_map()
Remove the attempt to trick us into mapping large bo.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 5 Dec 2013 16:12:22 +0000 (16:12 +0000)]
sna: Pass MOVE_READ for partial TearFree move-to-gpu
When fixing up a pending shadow pageflip for TearFree, we must tell it
that it needs to flush the damage for a partial move-to-gpu. Even though
the current area does not read from the destination, the move-to-gpu
callback is a global pass, and we need to assume the worst for the
region outside of the target.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=72343
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 5 Dec 2013 11:09:39 +0000 (11:09 +0000)]
sna: OsRegisterSigWrapper was only introduced in 1.6.99.900
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 5 Dec 2013 11:06:59 +0000 (11:06 +0000)]
configure: Correct closing ']'
Too many closed, too few opened.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 5 Dec 2013 11:00:44 +0000 (11:00 +0000)]
sna: Compilation fixes for squeeze
Older version of glibc and valgrind require a slight massage.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 7 Nov 2013 22:27:50 +0000 (22:27 +0000)]
sna: Update Baytrail VSync logic
My current best guess at the glaring hole in the spec that is
synchronisation to vertical refresh.
Note that this leaves VSync disabled for BYT for now as it is
ineffective - but at least it now doesn't hang!
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=69869
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 4 Dec 2013 11:55:54 +0000 (11:55 +0000)]
sna: Print detected GPU first
This should make the GPU identification easier to find in future.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 4 Dec 2013 11:06:08 +0000 (11:06 +0000)]
intel: Tidy up driver identification string
Split the identification strings between the older integrated graphic
chipsets and the more recent integrated processor graphics. This helps
to emphasis the recent branding and should reduce confusion about which
processors are supported by the driver.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 3 Dec 2013 15:06:10 +0000 (15:06 +0000)]
sna/dri: Always clear the clear hint after pageflipping
commit
3fd116782bb1c05a140f4783e92a8ee6a57143cd
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon Sep 23 21:03:07 2013 +0100
sna/dri: Clear the clear hint upon applying DRI damage
missed the short-circuit path for an already all-damaged pixmap - which
is quite common for the clear pixmaps...
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=72194
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 3 Dec 2013 12:10:23 +0000 (12:10 +0000)]
sna/gen2,3: Unmap the old GPU bo when replacing due to render target constaints
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 3 Dec 2013 12:10:01 +0000 (12:10 +0000)]
sna: Mark up an ordinary pixmap for reuse
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 3 Dec 2013 11:14:53 +0000 (11:14 +0000)]
sna: More assertions around tracking mapped state
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 3 Dec 2013 11:03:39 +0000 (11:03 +0000)]
sna: Clear the clear hint after performing a userptr upload
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 3 Dec 2013 09:37:37 +0000 (09:37 +0000)]
sna: Include Pineview in fence over-allocation checks
When creating the object, we need to take account that fenced regions
in the aperture are in power-of-two sizes and so can be much larger than
the actual object. This includes Pineview, which although is more
relaxed with regards to fence alignment of GPU operations, still imposes
the same restrictions on the fence registers.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 2 Dec 2013 15:48:50 +0000 (15:48 +0000)]
sna: Fix stipple offset for drawing into child windows
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 2 Dec 2013 10:43:51 +0000 (10:43 +0000)]
tests: Exercise stippled fills
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 2 Dec 2013 11:08:05 +0000 (11:08 +0000)]
sna: Fix tile origin for pattern blits
References: https://bugs.freedesktop.org/show_bug.cgi?id=71260
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 2 Dec 2013 14:20:51 +0000 (14:20 +0000)]
sna: Drop forced alignment to 64 on pre-gen4 devices
Some linear GPU bo that we create must be naturally aligned, and the
extra alignment imposed for pure paranoia is counter productive.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 2 Dec 2013 10:43:51 +0000 (10:43 +0000)]
tests: Exercise tiled fills
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 29 Nov 2013 14:46:54 +0000 (14:46 +0000)]
sna: Dump GTT info if we fail to execute a batch due to ENOSPC
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 28 Nov 2013 22:03:39 +0000 (22:03 +0000)]
sna: Fix typo inside DBG message
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 28 Nov 2013 21:13:33 +0000 (21:13 +0000)]
sna/video: Correct handling of cropped images along packed fast path
In particular, it was offseting the read from the source image, but not
correcting the length to read - causing a read from beyond the end of
the source and a segfault.
Reported-by: Jan Engelhardt <jengelh@inai.de>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 28 Nov 2013 16:25:52 +0000 (16:25 +0000)]
sna: Fallback from partial to full move-to-cpu
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 28 Nov 2013 14:12:56 +0000 (14:12 +0000)]
sna: Tidy up an assertion on an uninitialised variable
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 28 Nov 2013 14:02:18 +0000 (14:02 +0000)]
sna: Harden GetImage for use with very large buffers
That risk causing SIGBUS due to oom.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 28 Nov 2013 13:14:27 +0000 (13:14 +0000)]
sna: Wrap glyph composition with sigtrap handling
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 28 Nov 2013 13:03:54 +0000 (13:03 +0000)]
sna: Prefer using userptr for PutImage into large GPU bo
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 28 Nov 2013 13:03:29 +0000 (13:03 +0000)]
sna: Tidy a split conditional in an picture upload
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 28 Nov 2013 13:03:06 +0000 (13:03 +0000)]
sna: Wrap image upload with sigtrap
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 27 Nov 2013 22:06:27 +0000 (22:06 +0000)]
sna: Add a few more refcnt asserts
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 27 Nov 2013 12:59:25 +0000 (12:59 +0000)]
intel-virtual-output: Trim remote framebuffer to fit
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 27 Nov 2013 12:57:09 +0000 (12:57 +0000)]
intel-virtual-output: Workaround mutter always using the same timestamp
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 26 Nov 2013 16:13:23 +0000 (16:13 +0000)]
sna: Update connector-type-to-string mapping
Add in a couple of newcomers to the connector type and make it easier to
track updates in future.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Kirill Müller [Tue, 26 Nov 2013 10:36:54 +0000 (11:36 +0100)]
intel-virtual-output: mute spurious warnings
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=72027
Chris Wilson [Tue, 26 Nov 2013 09:41:42 +0000 (09:41 +0000)]
intel-virtual-output: Clone remote EDID for virtual outputs
References: https://bugs.freedesktop.org/show_bug.cgi?id=72020
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 26 Nov 2013 08:52:39 +0000 (08:52 +0000)]
intel-virtual-output: Correct clip region of rotated outputs in source framebuffer
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71846
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 25 Nov 2013 18:56:49 +0000 (18:56 +0000)]
sna: Correct plurization of axis for modesetting log message
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 25 Nov 2013 16:55:47 +0000 (16:55 +0000)]
intel-virtual-output: Include transforms when computing required fb size
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71846
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 25 Nov 2013 16:49:58 +0000 (16:49 +0000)]
sna: Include reflections when logging mode changes
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 25 Nov 2013 16:13:45 +0000 (16:13 +0000)]
intel-virtual-output: Tweak DBG messages for copying rotation
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 25 Nov 2013 15:33:13 +0000 (15:33 +0000)]
sna: Log mode switches on virtual outputs
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 25 Nov 2013 11:14:15 +0000 (11:14 +0000)]
sna: Keep @NOWARNFLAGS@ last
As the last option overrides the earlier options, make sure these
particular overrides always take effect by adding them last.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 25 Nov 2013 10:01:00 +0000 (10:01 +0000)]
sna/trapezoids: trapezoids_bounds() reports not-empty, so use it
Reported-by: Zdenek Kabelac <zkabelac@redhat.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Zdenek Kabelac [Mon, 25 Nov 2013 09:57:19 +0000 (09:57 +0000)]
sna: Handle failure to disable old plane when switching CRTCs
If we fail to disable the old plane, abort the PutImage and report the
error.
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Zdenek Kabelac [Mon, 25 Nov 2013 08:09:35 +0000 (09:09 +0100)]
uxa: fix order of inline Bool
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Zdenek Kabelac [Mon, 25 Nov 2013 08:09:34 +0000 (09:09 +0100)]
uxa,legacy: use unsigned for comparition of unsigned
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Zdenek Kabelac [Mon, 25 Nov 2013 08:09:32 +0000 (09:09 +0100)]
sna: Add missing va_end
Put va_end() into error paths
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Zdenek Kabelac [Mon, 25 Nov 2013 08:09:31 +0000 (09:09 +0100)]
uxa: Avoid relying on an assert and cleanup iterator
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Zdenek Kabelac [Mon, 25 Nov 2013 08:09:30 +0000 (09:09 +0100)]
Fix indentation for incorrectly shifted statements
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Zdenek Kabelac [Mon, 25 Nov 2013 08:09:29 +0000 (09:09 +0100)]
uxa: rename some local variable to not shadow global ones
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Zdenek Kabelac [Mon, 25 Nov 2013 08:09:28 +0000 (09:09 +0100)]
uxa: Ensure mask_picture is defined
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Zdenek Kabelac [Mon, 25 Nov 2013 08:09:27 +0000 (09:09 +0100)]
Cast to use unsigned math
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Chris Wilson [Mon, 25 Nov 2013 09:36:32 +0000 (09:36 +0000)]
Correct type of uevent_handler
xf86AddGeneralHandler returns an opaque pointer, not a function pointer.
Reported-by: Zdenek Kabelac <zkabelac@redhat.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Zdenek Kabelac [Mon, 25 Nov 2013 08:09:25 +0000 (09:09 +0100)]
Remove uneeded headers
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Zdenek Kabelac [Mon, 25 Nov 2013 08:09:24 +0000 (09:09 +0100)]
config apostrophe
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Zdenek Kabelac [Mon, 25 Nov 2013 08:09:20 +0000 (09:09 +0100)]
Makefile convert @var@ to $(var)
Avoid using @var@ since this could not be easily overwritten through
'make var=xxx' option which is normally available.
For Makefile.am users should avoid using @var@.
Signed-off-by: Zdenek Kabelac <zkabelac@redhat.com>
Chris Wilson [Fri, 22 Nov 2013 17:37:57 +0000 (17:37 +0000)]
Revert "intel-virtual-output: Only track the most recent visibility status of the cursor"
This reverts commit
abf1a16914d993cc150005879375d4bb17fdccf3.
We need to track visibility over all clones on a display, not just the
last.
Regression from
commit
abf1a16914d993cc150005879375d4bb17fdccf3 [2.99.906]
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri Nov 8 17:09:35 2013 +0000
intel-virtual-output: Only track the most recent visibility status of the curso
Reported-by: Kirill Müller <mail@kirill-mueller.de>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71838
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 22 Nov 2013 13:34:35 +0000 (13:34 +0000)]
uxa,sna: Prevent bo exchange when pinned for non-DRI2 clients
With the advent of DRI3 (and previously with Prime and Glamor) we have
external clients who rely on the pixmap<->bo mapping being invariant.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 22 Nov 2013 12:07:06 +0000 (12:07 +0000)]
sna: Tidy up fallback from render composite to blt composite
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 21 Nov 2013 21:52:46 +0000 (21:52 +0000)]
sna: XY_SETUP_* expects the dst tiling to be set on gen4+
A few paths failed to set the tiling flag when setting up the BLT
engine, leaving the engine in a potentially inconsistent state.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71260
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 21 Nov 2013 14:48:23 +0000 (14:48 +0000)]
sna: Fix typo in sna_blt_fill_begin()
Reported-by: intelgraphics7@gmx-topmail.de
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 20 Nov 2013 18:50:40 +0000 (18:50 +0000)]
sna: Set supported rotations on virtual outputs
As the virtual outputs are created later, they do not get automatically
populated with RR properties and we must do that instantiation
ourselves.
Reported-by: Kirill Müller <mail@kirill-mueller.de>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71846
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 20 Nov 2013 11:17:20 +0000 (11:17 +0000)]
sna/gen7: Try to reduce flushes between primitive continuations
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 19 Nov 2013 12:15:24 +0000 (12:15 +0000)]
intel-virtual-output: Check for failure to create a clone
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 19 Nov 2013 12:13:26 +0000 (12:13 +0000)]
intel-virtual-output: Fix trivial and unlikely fd leak from bumblebee failure
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 19 Nov 2013 12:09:35 +0000 (12:09 +0000)]
sna: Fix use before assignment
mono.sna was being used the line before it was assigned.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 19 Nov 2013 12:03:33 +0000 (12:03 +0000)]
sna: Free composite boxes in the unlikely all-invalid path
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 19 Nov 2013 12:01:33 +0000 (12:01 +0000)]
sna: Correct check for overwriting last BLT fill
Typo crept in from the reloc rework.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 19 Nov 2013 12:00:08 +0000 (12:00 +0000)]
sna/gen4+: Clear redirect bo before falling back to BLT paths
Just in case we later try to take action on the already freed
redirection.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 19 Nov 2013 09:47:15 +0000 (09:47 +0000)]
test: Fix compilation of render-trapezoid-image
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71768
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 18 Nov 2013 16:59:08 +0000 (16:59 +0000)]
sna/gen7: Rework random GPU flushing
There seems to be no clear rationale for these flushes. They have been
empirically derived to pass test/render-copyarea and rendercheck,
without impacting upon a few simple benchmarks.
References: https://bugs.freedesktop.org/show_bug.cgi?id=68410
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 18 Nov 2013 14:16:07 +0000 (14:16 +0000)]
sna: Tell valgrind about the returned mode list
Reported-by: Zdenek Kabelac <zkabelac@redhat.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 18 Nov 2013 12:33:05 +0000 (12:33 +0000)]
sna: Process Damage relative to dst->pDrawable not its pixmap->drawable
References: https://bugs.freedesktop.org/show_bug.cgi?id=32734
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 18 Nov 2013 12:04:29 +0000 (12:04 +0000)]
tests: Avoid dependence on undefined C
Right-shifting by a negative value is undefined.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 18 Nov 2013 11:58:43 +0000 (11:58 +0000)]
sna/damage: Tidy up a tail return
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 15 Nov 2013 21:20:30 +0000 (21:20 +0000)]
sna/damage: Guard against integer overflow before malloc
Check that the multiplication to compute the allocation will not
overflow.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 15 Nov 2013 11:55:51 +0000 (11:55 +0000)]
sna: Include valgrind leakchecks for --enable-debug=memory
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 14 Nov 2013 16:14:09 +0000 (16:14 +0000)]
sna: Add the missing braces around the conditional block
Fixes regression from
commit
f3225fcb38686f3b9701725bf3a11ecf1c100c3f
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue Nov 5 08:38:22 2013 +0000
sna: Be move conservative with tiling sizes for older fenced gen
Reported-by: Jay Little <jaylittle@jaylittle.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71605
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 14 Nov 2013 10:17:49 +0000 (10:17 +0000)]
tests: Exercise render copies under a mask
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 13 Nov 2013 16:25:54 +0000 (16:25 +0000)]
sna/gen6+: Only substitute SRC for CLEAR if using the correct source
If the source has not been verified to be clear, then using the SRC
operator will paint the wrong color.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 13 Nov 2013 15:28:37 +0000 (15:28 +0000)]
sna/gen6+: Improve debug PREFER_RENDER to cover all decision points
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 13 Nov 2013 15:27:05 +0000 (15:27 +0000)]
tests: Add render-composite-solid-mask
Another reference test to exercise simple composite pathways clearly.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 13 Nov 2013 14:17:51 +0000 (14:17 +0000)]
sna: Eliminate a compiler warning for a shadowed variable
sna_driver.c: In function 'sna_load_palette':
sna_driver.c:86:12: warning: declaration of 'index' shadows a global declaration [-Wshadow]
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 13 Nov 2013 14:17:51 +0000 (14:17 +0000)]
sna: Eliminate a compiler warning for a shadowed variable
We don't strictly need that local structure, so do without and keep the
compiler quiet.
sna_driver.c: In function 'fb_supports_depth':
sna_driver.c:414:23: warning: declaration of 'close' shadows a global declaration [-Wshadow]
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 12 Nov 2013 14:43:28 +0000 (14:43 +0000)]
2.99.906 snapshot
Chris Wilson [Wed, 13 Nov 2013 09:52:14 +0000 (09:52 +0000)]
Bump experimental ioctl command numbers
In the meantime a new ioctl has snuck in.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 12 Nov 2013 18:59:54 +0000 (18:59 +0000)]
sna: Discard cached upload proxy when writing to the pixmap via the CPU
Reported-by: Zdenek Kabelac <zkabelac@redhat.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
projects / platform / upstream / xf86-video-intel.git / log