Chris Wilson [Fri, 21 Feb 2014 13:43:09 +0000 (13:43 +0000)]
sna: Do not register colormaps without any CRTCs
The xserver may crash if we try to setup colormap handling without any
CRTCs, so don't.
Suggested-by: Dave Airlie <airlied@gmail.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 20 Feb 2014 19:18:46 +0000 (19:18 +0000)]
backlight: Make search routine for device specific backlight common
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 20 Feb 2014 07:51:40 +0000 (07:51 +0000)]
sna: Query cursor size from the kernel
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Mark Kettenis [Tue, 18 Feb 2014 19:02:14 +0000 (19:02 +0000)]
backlight: Build fixes for OpenBSD
Chris would like to humbly apologise for making these changes to the
OpenBSD code blind, and to thank Mark and Jonathan for their support
and understanding.
Chris Wilson [Tue, 18 Feb 2014 07:56:07 +0000 (07:56 +0000)]
intel-virtual-output: Discard unwanted events from the mouse recorder
X always sends MappingNotify events (there is no way for the client
to ignore them). In particular, MappingNotify would be sent after a VT
switch, and this would knock out our ability to track the cursor..
Reported-by: Raul Dias <raul@dias.com.br>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=75115
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Hans de Goede [Mon, 17 Feb 2014 12:16:54 +0000 (13:16 +0100)]
backlight-helper: Simplify reading the level from stdin
Since the helper is a standalone app, the usual xserver rules of not using
stdio because of signal handling don't apply.
And since the helper does run with elevated rights, it is important to keep
the code KISS so that it can be audited easily.
This commit replaces the hard to read "raw" read loop with a much simpler
loop using fgets, improving readability of the code.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Mon, 17 Feb 2014 12:16:53 +0000 (13:16 +0100)]
backlight: Drop rights before executing pkexec
Event though we've failed to open the backlight normally, we may still be
running under a suid-root xserver, so drop any elevated rights before
executing what we hope will be pkxec.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Mon, 17 Feb 2014 12:16:52 +0000 (13:16 +0100)]
backlight: Use System instead of system when checking for pkexec
Even though we've failed to open the backlight normally, we may still be
running under a suid-root xserver, so use the servers build in System instead
of system so as to properly drop root rights.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Mon, 17 Feb 2014 12:16:51 +0000 (13:16 +0100)]
backlight: Explain better why we support both pkexec and suid root for the helper
Update the comment about trying suid-root first with some explanations of
why pkexec may be preferable in some cases.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Chris Wilson [Sat, 15 Feb 2014 21:03:49 +0000 (21:03 +0000)]
backlight: tidy use of BACKLIGHT_CLASS
Use string concantenation to simply the sprintf slightly.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 15 Feb 2014 21:00:57 +0000 (21:00 +0000)]
backlight: Reject user overrides on OpenBSD
We only handle the singular wscons OpenBSD interface for backlight, so
explicitly reject any user requests for something else.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 15 Feb 2014 19:54:30 +0000 (19:54 +0000)]
backlight: Amalgamate hardcoded priority tables
Our fallback path for finding the backlight interface uses a hardcoded
table of known backlight controllers in proirity order. Rather than
maintain this table twice in the midst of the KMS logic, push it to the
new set of common backlight routines.
This incorporates bugfixes from SNA to handle unknown backlights, but
usable, gracefully.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Jan Alexander Steffens (heftig) [Sat, 15 Feb 2014 16:53:16 +0000 (17:53 +0100)]
Backlight helper build fixes
Don't hardcode $(prefix)/libexec
Fix install hook when DESTDIR is set
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=75028
Chris Wilson [Sat, 15 Feb 2014 14:55:09 +0000 (14:55 +0000)]
intel: Add a helper for setting backlight without root rights
Once the xserver stops running as root on kms capabable systems, we will need
some other way to access the backlight.
The approach taken in this patch moves most of the heavy lifting to a
helper that runs with root privileges and pipes our requested brightness
into the sysfs backlight interface. Where required, we use pkexec to
launch the helper with the elevated privilege.
v2: Amalgamate much more of the duplicate code.
Keep the daemon and pipe alive for the lifetime of the backlight interface.
Provide an alternative for systems without PolicyKit.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> [v1]
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Hans de Goede [Fri, 14 Feb 2014 23:02:36 +0000 (00:02 +0100)]
intel: export fd_set_cloexec / fd_set_nonblock
Allow fd_set_cloexec / fd_set_nonblock to be used outside of intel_device.c.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Fri, 14 Feb 2014 23:02:35 +0000 (00:02 +0100)]
intel: Fix fd_set_nonblock
O_NONBLOCK is a status flag not a descriptor flag, so F_GETFL / F_SETFL should
be used to modify it.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Chris Wilson [Fri, 14 Feb 2014 19:36:47 +0000 (19:36 +0000)]
sna: Inherit the native rotation on initial output probing
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 14 Feb 2014 16:18:34 +0000 (16:18 +0000)]
sna: Reorganise native rotation ioctls to compile on old Linux
It is not just the BSDs that lack these ioctls in their userspace
headers, but everything older than about a year...
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 14 Feb 2014 15:11:21 +0000 (15:11 +0000)]
sna: Guard use of DRM_MODE_OBJECT type ids
These are not currently exported on OpenBSD and cause the build to fail.
Hide their use behind an ifdef guard for the time being.
"I noticed the following when trying to compile xf86-video-intel git on OpenBSD:
/usr/users/jsg/src/xf86-video-intel/src/sna/sna_display.c:2106: error: 'DRM_MODE_OBJECT_CRTC' undeclared (first use in this function)
/usr/users/jsg/src/xf86-video-intel/src/sna/sna_display.c:2106: error: (Each undeclared identifier is reported only once
/usr/users/jsg/src/xf86-video-intel/src/sna/sna_display.c:2106: error: for each function it appears in.)
/usr/users/jsg/src/xf86-video-intel/src/sna/sna_display.c:2107: error: 'DRM_MODE_OBJECT_PLANE' undeclared (first use in this function)"
Reported-by: Jonathan Gray
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 14 Feb 2014 14:53:14 +0000 (14:53 +0000)]
sna: Check that the backlight interface is in sysfs before access
Before we write values into the file, we want to be sure that it is our
sysfs backlight interface, and not some ordinary file we are about to
destroy.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 13 Feb 2014 22:46:22 +0000 (22:46 +0000)]
sna: Increase the acceptable screen size range
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 13 Feb 2014 22:36:36 +0000 (22:36 +0000)]
sna: Filter out requests to create a 0x0 frontbuffer
Attempting to create a 0x0 frontbuffer results in erroneous pixmap
generation, and is liable to explode at a random point in future, so
just reject it outright.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 13 Feb 2014 21:36:26 +0000 (21:36 +0000)]
sna: Fix comparison of previous BLT commands
After altering the command bits to accommodate Broadwell, we needed to
fixup the comparison of previous BLT commands to detect overwritten
commands.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 13 Feb 2014 16:18:43 +0000 (16:18 +0000)]
uxa: Check for access to backlight interface
Before attaching properties to any connector, check that we can indeed
control the backlight through the sysfs interface by doing an access
check. If the xserver is started without root privileges, we cannot
write through /sys/class/backlight and so should take care not to
advertise that we can.
Reported-by: Hans de Geode <hdegeode@redhat.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 13 Feb 2014 16:18:43 +0000 (16:18 +0000)]
sna: Check for access to backlight interface
Before attaching properties to any connector, check that we can indeed
control the backlight through the sysfs interface by doing an access
check. If the xserver is started without root privileges, we cannot
write through /sys/class/backlight and so should take care not to
advertise that we can.
Reported-by: Hans de Geode <hdegeode@redhat.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 13 Feb 2014 12:57:50 +0000 (12:57 +0000)]
sna: Clear DPMS tracking on VT switch
Assume the worst and force us to reset the property when we restore
control.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 13 Feb 2014 11:58:15 +0000 (11:58 +0000)]
sna: Extend native rotation support to sprites
The sprite plane can be independently rotated to the CRTC primary plane.
To rotate the sprite plane, we just set a property on the plane similar
to how we rotate the CRTC, so we can refactor them together to use the
same routines.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 13 Feb 2014 14:46:39 +0000 (14:46 +0000)]
sna: Enable coordinate transforms for native CRTC rotation
Otherwise X tells Clients that the cursor is still in the unrotated
location when it appears rotated on the display.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 13 Feb 2014 09:46:13 +0000 (09:46 +0000)]
Revert "uxa: Do not change DPMS mode on unconnected outputs"
This reverts commit
4497212307dee5e35bc6836201738a2fdb559020.
Unfortunately, this simple fix does not work for UXA as DPMS is used by
the xserver to turn off CRTCs and outputs. Since UXA does not implement
CRTC DPMS, this commit causes us to fail to turn off outputs.
The kernel has been fixed up in the meantime and that commit has been
recommended to be backported to all stable kernels:
commit
c9976dcf55c8aaa7037427b239f15e5acfc01a3a
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sun Sep 29 19:15:07 2013 +0100
drm/i915: Only apply DPMS to the encoder if enabled
so it should be safe for UXA to rely on its old behaviour.
Bugzilla: https://code.google.com/p/chromium/issues/detail?id=341135
References: https://bugs.freedesktop.org/show_bug.cgi?id=68030
Suggested-by: Dominik Behr <dbehr@google.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 13 Feb 2014 01:06:41 +0000 (01:06 +0000)]
sna: Always assign a name to the modes
In some cases, such as querying the mode from the CRTC, we may not have
a name associated with the mode. However, RandR always expects a valid
name. To satisfy this requirement, always generate the canonical mode
name if no other is specified.
References: https://bugs.freedesktop.org/show_bug.cgi?id=70132
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 16:41:51 +0000 (16:41 +0000)]
sna/gen5: Flush the render cache between operations
When we change the blend mode between operations, it appears that we
must flush the render cache or else we risk render corruption. This is
usually noticeable in rendering of single glyphs.
This was originally fixed for bug 51422, but was reintroduced by
commit
37eb7343be1aeeb90a860096756603a577df1a77
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat Dec 1 09:40:11 2012 +0000
sna/gen5: Inspired by gen4, reorder the flushing
and the desire to reduce the impact of this w/a.
Reported-by: itumaykin@gmail.com
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74882
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 16:19:25 +0000 (16:19 +0000)]
intel-virtual-output: Do not resize the screen whilst all outputs are off
If all the outputs are off, we try to resize the screen to 1x1, which is
typically illegal. So, just keep the existing screen and xfer buffer for
next time it is enabled.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 15:48:20 +0000 (15:48 +0000)]
intel-virtual-output: Force an output probe on initial connection to remote display
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 15:47:00 +0000 (15:47 +0000)]
intel-virtual-output: Fix checking for no change in output modes
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 15:13:42 +0000 (15:13 +0000)]
intel-virtual-output: Always requery modes after OutputNotify
References: https://bugs.freedesktop.org/show_bug.cgi?id=74800
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 14:48:24 +0000 (14:48 +0000)]
intel-virtul-output: Double check timestamps
Compare both res->timestamp and res->configTimestamp for changes in
state.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 11:33:45 +0000 (11:33 +0000)]
sna: Support native primary plane rotations
Use the display hardware for simple rotations, when exported through the
rotation property on the CRTC.
As the kernel support is not yet merged upstream, the feature is hidden
behind --enable-rotation.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 10:43:46 +0000 (10:43 +0000)]
Revert "sna/gen6: Serialise write to DERRMR register using STORE_REGISTER_MEM"
This reverts commit
2a5ad9c015bbb41550536dd3cf4ad18b7f21fd80.
Oops, pushed from the wrong machine and merged in a incomplete bugfix
branch.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 10:34:34 +0000 (10:34 +0000)]
sna/gen5: Use the DBG option ALWAYS_FLUSH to force a full flush
When flushing between operations, we can choose between doing a full
flush to memory, or just a pipeline flush. For debugging it is better to
do the full flush to rule out cache effects.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 10:32:50 +0000 (10:32 +0000)]
Merge branch 'master' of hsw:/usr/src/xf86-video-intel
Chris Wilson [Wed, 12 Feb 2014 09:58:06 +0000 (09:58 +0000)]
sna/gen5: Add a DBG option to force flushes between operations
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 09:50:35 +0000 (09:50 +0000)]
sna: Retrieve tiling for fbcon
Yikes, there appear to be tiled fbcon framebuffers on the horizon.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 12 Feb 2014 07:56:10 +0000 (07:56 +0000)]
sna: Delete redundant branch
The MOVE_READ else branch is repeated immediately below when marking up
damage for MOVE_WRITE, so remove it for simplicity.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 11 Feb 2014 23:08:35 +0000 (23:08 +0000)]
Record Video ABI version that we configure against
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 11 Feb 2014 21:57:23 +0000 (21:57 +0000)]
Merge branch 'master' of hsw:/usr/src/xf86-video-intel
Chris Wilson [Tue, 11 Feb 2014 21:21:17 +0000 (21:21 +0000)]
sna: Defer move-to-gpu until we need to transfer GPU damage to CPU
When preparing a pixmap for CPU rendering, one of the last steps is to
copy the outstanding GPU damage back to the CPU. Currently, we flush any
outstanding pageflips (for TearFree) and in the process accidentally
destroy the shadow buffer if there is no outstanding GPU damage. Rearrange
the code to avoid tripping over that by only processing the move-to-gpu if
we need to touch the GPU bo.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 11 Feb 2014 11:23:37 +0000 (11:23 +0000)]
sna/gen6: Serialise write to DERRMR register using STORE_REGISTER_MEM
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 10 Feb 2014 16:42:06 +0000 (16:42 +0000)]
sna: Remove damaged region after converting CPU bo to GPU bo
Reported-by: Jiri Slaby <jirislaby@gmail.com>
References: https://bugs.freedesktop.org/show_bug.cgi?id=70461
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 10 Feb 2014 16:37:38 +0000 (16:37 +0000)]
sna: Check for overwriting damage before applying clear
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 10 Feb 2014 14:19:53 +0000 (14:19 +0000)]
sna: Clear GPU damage flag for 1x1 writes onto a clear buffer
An optimisation for 1x1 reads of a clear buffer lead to a bug whereby
the damage for a 1x1 write would not be tracked correctly (conflicting
full GPU damage vs partial CPU damage).
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 10 Feb 2014 09:35:11 +0000 (09:35 +0000)]
2.99.910 snapshot
Chris Wilson [Sun, 9 Feb 2014 12:28:27 +0000 (12:28 +0000)]
sna/gen6: Restore stall dropped when not flushing instead
commit
82e6d41c2f4f343bd1854d3d8ee4b624b5d68971
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu Oct 31 13:35:59 2013 +0000
sna/gen6: Tweak flush around CC state changes
Replaced the pipeline stall with a flush - but only when the target was
dirty. The missing stall however seems to be required as well.
v2: Actually emit the stall for all CC state changes [Ilia Mirkin]
Reported-by: Ilia Mirkin <imirkin@alum.mit.edu>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=72375
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 6 Feb 2014 10:13:41 +0000 (10:13 +0000)]
sna: And clear drawable->id on reused pixmaps
References: https://bugs.freedesktop.org/show_bug.cgi?id=74550
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 6 Feb 2014 09:24:17 +0000 (09:24 +0000)]
uxa: Do not expose TextureVideo Xv adapters when acceleration is disabled
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 6 Feb 2014 09:04:30 +0000 (09:04 +0000)]
sna: Add some DBG to clarify the source pixmap (for tiles)
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 6 Feb 2014 08:29:20 +0000 (08:29 +0000)]
sna: Reset composite offsets when reusing freed pixmaps
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74550
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 5 Feb 2014 15:30:30 +0000 (15:30 +0000)]
sna: Add some DBG around tiled blts
References: https://bugs.freedesktop.org/show_bug.cgi?id=74550
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 5 Feb 2014 11:41:42 +0000 (11:41 +0000)]
sna: Pass read hints from move-to-cpu to wait_for_shadow
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 4 Feb 2014 20:57:24 +0000 (20:57 +0000)]
sna: Tweak assert_bo_retired() to be callable on cached bo
References: https://bugs.freedesktop.org/show_bug.cgi?id=73406
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 4 Feb 2014 17:51:51 +0000 (17:51 +0000)]
sna: Rearrange assertion to ease use of substitute cached bo
Since we call kgem_bo_submit() along one path when synchronising a
cached bo (which is known to be inactive) but still want to keep the
assertion on the refcnt, simply rearrange the code to only assert on the
active path.
References: https://bugs.freedesktop.org/show_bug.cgi?id=73406
Reported-by: Matti Hamalainen <ccr@tnsp.org>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 4 Feb 2014 10:36:21 +0000 (10:36 +0000)]
sna/gen4: Disable use of pipecontrol invalidates again
One day, just not today, we may make gen4 work correctly, efficiently and
fast. Today, we can barely pick one.
References: https://bugs.freedesktop.org/show_bug.cgi?id=55500
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 4 Feb 2014 08:51:17 +0000 (08:51 +0000)]
sna/glyphs: Reset composite state between switching glyph formats
One path uses the mask channel, the other does not. We cannot rely on
overwriting all reused state in this case, and so we must clear the
composite state prior to use each time.
Reported-by: Arkadiusz Miskiewicz <arekm@maven.pl>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74494
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Tested-by: Arkadiusz Miskiewicz <arekm@maven.pl>
Edward Sheldrake [Mon, 3 Feb 2014 09:34:33 +0000 (09:34 +0000)]
sna/gen4,5: Fix setting pipe control cache flush bits
Cache flush bits are on dword 0, not 1, on gen4 and gen5. Also texture
cache invalidate is only available from Cantiga onwards.
Chris Wilson [Sun, 2 Feb 2014 03:52:11 +0000 (03:52 +0000)]
sna: Reconstruct damage for the partially replaced discarded CPU bo
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 1 Feb 2014 21:55:09 +0000 (21:55 +0000)]
sna: If IGNORE_CPU is not set we must mark the move as MOVE_READ
Logic reversal in discarding CPU damage. An old bug revealed by the more
aggressive attempts to discard CPU damage.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 1 Feb 2014 21:54:43 +0000 (21:54 +0000)]
sna: Add some more DBG for choosing render targets
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 1 Feb 2014 21:38:47 +0000 (21:38 +0000)]
sna: Skip discarding CPU bo when using as a render target
The issue with unwanted serialisation does not affect the async
move-to-cpu of a render target.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 1 Feb 2014 21:37:41 +0000 (21:37 +0000)]
sna: Allow more inplace promotions of CPU to GPU bo
If the CPU bo is wholly damaged, then it makes an ideal candidate for
simply converting into the GPU bo.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 1 Feb 2014 20:13:53 +0000 (20:13 +0000)]
sna: Undo region translation before returning
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 1 Feb 2014 17:59:29 +0000 (17:59 +0000)]
2.99.909 snapshot
Chris Wilson [Sat, 1 Feb 2014 18:57:06 +0000 (18:57 +0000)]
sna: Apply drawable offsets to partial CPU damage being replaced
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 1 Feb 2014 18:13:04 +0000 (18:13 +0000)]
sna: Don't discard the GPU prior to using it for stashing CPU damage
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 1 Feb 2014 17:56:58 +0000 (17:56 +0000)]
sna: Remark the region as damaged after uploading the rest to the GPU
If we upload outside of the replaced region to the GPU, we need to
remark the region of operation as dirty.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 1 Feb 2014 17:37:42 +0000 (17:37 +0000)]
sna: Only discard CPU damage for an replacing region
When considering move-region-to-cpu, we need to take into account that
the region may not replace the whole drawable, in which case we cannot
simply dispose of an active CPU bo.
Reported-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reported-by: Conley Moorhous <conleymoorhous@gmail.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74327
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 1 Feb 2014 16:55:32 +0000 (16:55 +0000)]
glamor: Enable Xv support
In addition to the UXA texture paths (which in theory support VSync) and
direct overlay support, create a textured video paths that utilize the
glamor acceleration code.
Requested-by: Fabio Pedretti <fabio.ped@libero.it>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74311
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 31 Jan 2014 17:59:30 +0000 (17:59 +0000)]
2.99.908 snapshot
Chris Wilson [Fri, 31 Jan 2014 20:02:44 +0000 (20:02 +0000)]
sna: remove short-circuit for move-to-CPU when damage covers region
The short-circuit path missed translating the damage from drawable space
into the pixmap (for Composite setups) which may have resulted in
corruption. The path was also failing to consider the impact of reusing
an active CPU bo when it could be discarding the unwanted damage and
reallocating.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 31 Jan 2014 18:02:51 +0000 (18:02 +0000)]
sna: Use 24bpp rendering by default for gen2 again
Too many users where unhappy that their software was broken. Too bad the
same software is also broken at high bit depths as well.
References: https://bugs.freedesktop.org/show_bug.cgi?id=73877
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 31 Jan 2014 16:42:46 +0000 (16:42 +0000)]
sna: Balance memory accounting for buffer objects
A couple bugs ended up with CPU bo gradually accumulating whilst the
overall number of bo vanished.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 31 Jan 2014 15:39:23 +0000 (15:39 +0000)]
sna: Reorder assignments vs DBG
It helps reading the debug log if what we print is correct.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 31 Jan 2014 15:03:43 +0000 (15:03 +0000)]
sna: Actually reuse pixmap headers for ordinary pixmaps
commit
3dbf17f00e200e864818b0b8ca762183cff9085f [2.99.907]
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue Dec 3 12:10:01 2013 +0000
sna: Mark up an ordinary pixmap for reuse
started to put the frequently allocated pixmaps into the object cache -
but as those frequent allocations themselves did not use the cache, the
cache kept growing.
Bugzilla: https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/1272338
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 31 Jan 2014 14:28:56 +0000 (14:28 +0000)]
sna: Tidy up pixmap header allocations
Refactor some more of the common code, and add some debug tracking.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 31 Jan 2014 14:10:36 +0000 (14:10 +0000)]
sna: Show the memory debug info again
This needs to be visible whenever we --enable-debug=memory. Make it so,
and trim it to a single line in the process.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 30 Jan 2014 12:19:04 +0000 (12:19 +0000)]
sna: Silence a compiler warning for mismatching const strings
sna_display.c: In function 'preferred_mode':
sna_display.c:3393:7: warning: assignment discards 'const' qualifier from pointer target type [enabled by default]
mode = xf86GetOptValString(output->options, OPTION_PREFERRED_MODE);
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 30 Jan 2014 12:18:05 +0000 (12:18 +0000)]
sna: Suppress warning for freeing a "constant" string
We allocated and own the string, so we can free it!
sna_display.c: In function 'sna_output_get_modes':
sna_display.c:2314:4: warning: passing argument 1 of 'free' discards 'const' qualifier from pointer target type [enabled by default]
free(current->name);
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 30 Jan 2014 12:17:02 +0000 (12:17 +0000)]
sna: Handle char *backlight_iface
The backlight_iface is a freeable string, in places we were assigning a
constant string and later freeing it.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 30 Jan 2014 12:13:16 +0000 (12:13 +0000)]
sna: Constify GCFunc to quieten the compiler.
We keep a pointer to the old funcs when wrapping, this now needs to be const.
sna_accel.c: In function 'sna_gc_move_to_cpu':
sna_accel.c:4008:17: warning: assignment discards 'const' qualifier from pointer target type [enabled by default]
sgc->old_funcs = gc->funcs;
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 29 Jan 2014 16:48:21 +0000 (16:48 +0000)]
sna: Fix format specifier typo ('%D') in DBG
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 29 Jan 2014 13:42:50 +0000 (13:42 +0000)]
sna: Always clear the domain even only a potential sync
It helps to keep the domain-tracking bookkeeping consistent.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 29 Jan 2014 13:39:20 +0000 (13:39 +0000)]
sna: We can only retire a bo if is not referenced by the current batch
Fixes regression from
commit
8b0ebebcab21647348f769c25ca0c1d81d169e75
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue Jan 28 16:30:47 2014 +0000
sna: Be a little more assertive in retiring after set-domain
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74176
Reported-by: Jan Alexander Steffens <jan.steffens@gmail.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 29 Jan 2014 13:15:43 +0000 (13:15 +0000)]
sna: Add assertions for set/get binding
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 29 Jan 2014 13:06:08 +0000 (13:06 +0000)]
sna/gen4+: Assert that the cached binding location is valid
We can at least check that it is in the right region (i.e. not past
where the current surface has been allocated from).
References: https://bugs.freedesktop.org/show_bug.cgi?id=74176
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 29 Jan 2014 11:15:45 +0000 (11:15 +0000)]
sna: Do not use color corrected pixel values for comparing against gc
Dashed zero-width lines devolve into two passes (first to draw the
foreground dashes, and then to draw the gaps). The code was using the
pixel value to decide which phase was being drawn -- however the color
on the gc is not always bit exact (e.g. alpha channel) with the pixel
value. As a result, we would oft miss drawing anything when rendering
into the scanout.
Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1059152
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 29 Jan 2014 05:40:00 +0000 (05:40 +0000)]
sna/gen6+: Encourage inplace uploads for Composite
First try using the blt path for a potential inplace upload for a
simple composite operation copying from a ShmPixmap.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 29 Jan 2014 05:32:25 +0000 (05:32 +0000)]
sna: Skip undamaged TearFree redisplays
If we have not had cause to flush the wait_for_shadow buffer during the
course of the rendering, then we never wrote to the backbuffer and its
contents are still identical to the current frontbuffer. So if the
wait_for_shadow is still flagged as required on the scanout, we know we
can safely discard the redisplay request.
References: https://bugs.freedesktop.org/show_bug.cgi?id=70905
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 28 Jan 2014 21:56:23 +0000 (21:56 +0000)]
Allow selection of glamor as the default acceleration method
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74162
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 28 Jan 2014 19:17:14 +0000 (19:17 +0000)]
sna: Check for a hang after a spurious error return from set-domain-ioctl
set-domain-ioctl should only ever fail for invalid user input, beyond
the normal signal handling. As such if we do find a spurious error
return from the kernel, check for a GPU hang.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 28 Jan 2014 16:30:47 +0000 (16:30 +0000)]
sna: Be a little more assertive in retiring after set-domain
After a successful set-domain for writing with the CPU, we know that the
buffer is idle so remove it from our request tracking. (External clients
complicate matters in that they may keep the bo active even after our
set-domain.) On the contrary, because of read-read optimisations a bo
may still be active after a set-domain for reading by the CPU, in which
we need to remain conservative in retiring the bo.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 28 Jan 2014 12:51:00 +0000 (12:51 +0000)]
sna: Cancel the backlight change if set-property(DPMS) fails
If the kernel reports an error from setting the DPMS property, we need
to undo the local changes and propagate back the error to the caller.
Although the latter is prevented by the RandR API, we can at least
restore the backlight.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 27 Jan 2014 21:48:33 +0000 (21:48 +0000)]
sna: Rearrange damage allocation to handle malloc failure more gracefully
If we fail to allocate new damage boxes, first try collescing the
existing boxes to free up memory.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>