API for using poll/epoll instead of select

I'm finishing up replacing select with either poll or epoll (or other) and wanted to present the API I've constructed for this work. The goal is to hide the implementation details away from the server so that we can easily swap in OS-specific replacements. I've implemented poll and epoll back

[ANNOUNCE] libXfixes 5.0.2

Matt Turner (1): libXfixes 5.0.2 Michael Joost (1): Remove fallback for _XEatDataWords, require libX11 1.6 for it git tag: libXfixes-5.0.2 http://xorg.freedesktop.org/archive/individual/lib/libXfixes-5.0.2.tar.bz2 MD5: 544d73df94e638ba7b64147be416e576 libXfixes-5.0.2.tar.bz2 SHA1:

[ANNOUNCE] libXfixes 5.0.2

Matt Turner (1): libXfixes 5.0.2 Michael Joost (1): Remove fallback for _XEatDataWords, require libX11 1.6 for it git tag: libXfixes-5.0.2 http://xorg.freedesktop.org/archive/individual/lib/libXfixes-5.0.2.tar.bz2 MD5: 544d73df94e638ba7b64147be416e576 libXfixes-5.0.2.tar.bz2 SHA1:

[PATCH libpciaccess 2/2]: use cached devinfo snapshots, remove unnecessary di_init()

Use cached devinfo (DINFOCACHE) in di_init(). Remove unnecessary di_init() on nexus nodes. Signed-off-by: Henry Zhao --- src/solx_devfs.c | 25 + 1 file changed, 5 insertions(+), 20 deletions(-) diff --git a/src/solx_devfs.c

[PATCH libpciaccess 1/2]: probe should not hold pci nexus drivers open

Probe should not hold pci nexus drivers open - close after use. Signed-off-by: Henry Zhao --- src/solx_devfs.c | 29 - 1 file changed, 20 insertions(+), 9 deletions(-) diff --git a/src/solx_devfs.c b/src/solx_devfs.c index cf96467..c94daaf

Re: [PATCH xserver] glamor: Check for composite operations which are equivalent to copies

On Wed, 2016-05-25 at 09:24 -0400, Alex Deucher wrote: > On Wed, May 25, 2016 at 6:09 AM, Michel Dänzer wrote: > > From: Michel Dänzer > > > > Increases x11perf -compwinwin500 numbers by a factor of 10 for me with > > radeonsi. > > > > Conditions

Re: [PATCH xserver] glamor: Fix sampling outside of RGBx source/mask pictures

On Wed, 2016-05-25 at 09:22 -0400, Alex Deucher wrote: > On Tue, May 24, 2016 at 5:12 AM, Michel Dänzer wrote: > > From: Michel Dänzer > > > > RENDER requires that sampling outside of any source/mask picture results > > in alpha == 0.0. > > > > The

Re: [PATCH 0/2] Hide pointer invalidation details from xwayland

On Wed, 2016-05-18 at 16:03 -0500, Keith Packard wrote: > Here's a couple of patches which makes miPointerRec private to the > miPointer code; the only user of that struct was Xwayland which > used it to invalidate the current sprite. An API is provided to > perform that invalidation, and then the

Re: [PATCH xserver 2/2] xfree86: Add IndirectGLX server flag (v2)

On Mon, 2016-05-23 at 21:09 -0700, Alan Coopersmith wrote: > On 05/18/16 07:53 AM, Adam Jackson wrote: > > Not all display managers make it easy (or possible) to modify the > > command line flags passed to the server, so add a way to get to it > > from > > xorg.conf. > > > > v2: Fix the

Re: [PULL 1.18] XQuartz configure cherry-picks

On Wed, 2016-05-18 at 14:15 -0700, Jeremy Huddleston Sequoia wrote: > The following changes since commit > 9454cd51da9b38b974cff7c8b7125901f6403848: > >   xserver 1.18.3 (2016-04-04 14:39:36 -0400) > > are available in the git repository at: > >   https://github.com/XQuartz/xorg-server.git  >

[PATCH kdrive/ephyr v3 3/3] ephyr: move SetNotifyFd()/RemoveNotifyFd() calls outside ephyr mouse driver

When Xephyr is launched in multi-seat mode (i.e. with command-line option "-seat seat"), ephyr virtual keyboard/mouse driver is not loaded. Without this patch, Xephyr won't listen to any events in multi-seat mode, including XCB_EXPOSE and XCB_CONFIGURE_NOTIFY, which may be needed even for

[PATCH kdrive/ephyr v3 2/3] ephyr: fix RandR version check in hostx_get_output_geometry()

Current hostx_get_output_geometry() has a little problem when checking RandR version, treating e.g. version 2.0 as lower than 1.2. Signed-off-by: Laércio de Sousa --- hw/kdrive/ephyr/hostx.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff

[PATCH kdrive/ephyr v3 1/3] ephyr: drop a lot of deprecated code

After commit 623ff251, a lot of functions and data structures in ephyr/hostx.{c,h} became uneeded. We are removing them now. Signed-off-by: Laércio de Sousa --- hw/kdrive/ephyr/hostx.c | 227

Re: [PATCH xserver] glamor: Check for composite operations which are equivalent to copies

On Wed, May 25, 2016 at 6:09 AM, Michel Dänzer wrote: > From: Michel Dänzer > > Increases x11perf -compwinwin500 numbers by a factor of 10 for me with > radeonsi. > > Conditions copied from exaComposite(). > > Signed-off-by: Michel Dänzer

Re: [PATCH xserver] glamor: Fix sampling outside of RGBx source/mask pictures

On Tue, May 24, 2016 at 5:12 AM, Michel Dänzer wrote: > From: Michel Dänzer > > RENDER requires that sampling outside of any source/mask picture results > in alpha == 0.0. > > The OpenGL border colour cannot set alpha = 0.0 if the texture format >

[Bug 95475] DRI3 screen corruption

https://bugs.freedesktop.org/show_bug.cgi?id=95475 --- Comment #15 from Ivan Majhen --- Created attachment 124082 --> https://bugs.freedesktop.org/attachment.cgi?id=124082=edit old xorg log driver with dri3 enabled -- You are receiving this mail because: You are the

[Bug 95475] DRI3 screen corruption

https://bugs.freedesktop.org/show_bug.cgi?id=95475 --- Comment #14 from Ivan Majhen --- Created attachment 124081 --> https://bugs.freedesktop.org/attachment.cgi?id=124081=edit Old dri3 -- You are receiving this mail because: You are the assignee for the

[Bug 95475] DRI3 screen corruption

https://bugs.freedesktop.org/show_bug.cgi?id=95475 --- Comment #13 from Ivan Majhen --- I always had glitches with dri3. I tested it about 30 times in the past year. Commit 64e1e4db didn't compile. First commit that compiles with current setup is

[PATCH xserver] glamor: Check for composite operations which are equivalent to copies

From: Michel Dänzer Increases x11perf -compwinwin500 numbers by a factor of 10 for me with radeonsi. Conditions copied from exaComposite(). Signed-off-by: Michel Dänzer --- glamor/glamor_render.c | 23 +++ 1 file changed, 23

[Bug 95475] DRI3 screen corruption

https://bugs.freedesktop.org/show_bug.cgi?id=95475 --- Comment #12 from Michel Dänzer --- Was this problem already present with xf86-video-ati Git commit 64e1e4db ("Add DRI3 support v2"), which was when DRI3 support was initially added? Note that this commit will only

[Bug 95475] DRI3 screen corruption

https://bugs.freedesktop.org/show_bug.cgi?id=95475 Michel Dänzer changed: What|Removed |Added Attachment #123932|text/x-log |text/plain

Re: Novice trying to understand Xrandr and multi monitor setup

On Tue, May 24, 2016 at 08:38:53PM -0700, TJ Olaes wrote: >Good evening. >I have a dual monitor setup and am trying to figure out why ScreenCount >returns 1. My code is as follows: Good day. The ScreenCount is not a RANDR call. This is from CORE multihead support attempt. That

Re: [xserver-xorg][PATCH V2 1/1] xwayland-shm: block signals during fallocate

On Wed, 25 May 2016 10:41:53 +0300 Ian Ray wrote: > posix_fallocate() does an explicit rollback if it gets EINTR, and > this is a problem on slow systems because when the allocation size > is sufficiently large posix_fallocate() will always be interrupted > by the smart

[xserver-xorg][PATCH V2 1/1] xwayland-shm: block signals during fallocate

posix_fallocate() does an explicit rollback if it gets EINTR, and this is a problem on slow systems because when the allocation size is sufficiently large posix_fallocate() will always be interrupted by the smart scheduler's SIGALRM. Changes since v1 - big comment in the code to explain what is