- Original message -
> On Wed, Jul 14, 2010 at 07:49:39AM +0300, Oliver McFadden wrote:
> > This is very similar to the RunFromSmartParent (implicit) option,
> except
> > we do not send the signal to our parent process, but our own process
> > instead, and that signal is SIGSTOP, not SIGUS
On Wed, Jul 14, 2010 at 07:49:39AM +0300, Oliver McFadden wrote:
> This is very similar to the RunFromSmartParent (implicit) option, except
> we do not send the signal to our parent process, but our own process
> instead, and that signal is SIGSTOP, not SIGUSR1.
>
> Upstart will detect this, reali
This is very similar to the RunFromSmartParent (implicit) option, except
we do not send the signal to our parent process, but our own process
instead, and that signal is SIGSTOP, not SIGUSR1.
Upstart will detect this, realize that we are ready to accept clients
now, send us SIGCONT and move our jo
On Wed, Jul 14, 2010 at 12:26 AM, Jamey Sharp wrote:
> On Tue, Jul 13, 2010 at 7:12 PM, Alex Deucher wrote:
>> On Tue, Jul 13, 2010 at 6:35 PM, Jamey Sharp wrote:
>>> @@ -1387,8 +1384,12 @@ doPolyText(ClientPtr client, PTclosurePtr c)
>>> if (c->pDraw)
>>> {
>>>
On Tue, Jul 13, 2010 at 7:12 PM, Alex Deucher wrote:
> On Tue, Jul 13, 2010 at 6:35 PM, Jamey Sharp wrote:
>> @@ -1387,8 +1384,12 @@ doPolyText(ClientPtr client, PTclosurePtr c)
>> if (c->pDraw)
>> {
>> c->xorg += *((INT8 *)(c->pElt + 1)); /* must be signed
On Tue, Jul 13, 2010 at 6:35 PM, Jamey Sharp wrote:
> When a GC is allocated, it is zeroed, including all storage requested
> with dixRegisterPrivateKey. So CreateGC hooks don't need to initialize
> anything to zero.
>
> Signed-off-by: Jamey Sharp
Reviewed-by: Alex Deucher
> ---
> fb/fbgc.c
On Tue, Jul 13, 2010 at 6:35 PM, Jamey Sharp wrote:
> Signed-off-by: Jamey Sharp
> ---
> dix/dixfonts.c | 54 ---
> include/closestr.h | 24 ---
> 2 files changed, 21 insertions(+), 57 deletions(-)
>
> diff --git a/dix/d
On Tue, Jul 13, 2010 at 6:20 PM, Jamey Sharp wrote:
> I have an "It's probably fine then" from Alex, but could I get a
> reviewed-by for this?
Looks good to me.
Reviewed-by: Alex Deucher
>
> I have several other patches from the last couple of months also
> awaiting review:
> http://patchwork.
Jonathan,
Miss this mail but find in the http://lists.x.org/archives/xorg-devel
I have some understanding about your explanation. Is that to say if the
source pixel is out of the source image, the modulus operation will be used to
get the source pixel to to rendered, not for the
On Wed, May 19, 2010 at 11:31 AM, Adam Jackson wrote:
> ...This also marks a flags field for future expansion without needing to
> break ABI.
I was just reviewing this again because I wanted to add a flag to
ScreenRec, and I noticed:
> + unsigned backingStoreSupport:1;
> + unsign
RANDR_12_INTERFACE is defined in xserver/randr/randstr.h since version 1.2.
This driver only configures against xserver 1.2 or higher.
Reported-by: Aaron Plattner
Signed-off-by: Gaetan Nadon
---
src/g80_display.c |4
src/g80_sor.c |6 --
2 files changed, 0 insertions(+), 10
Looks sane to me...
Reviewed-by: Jeremy Huddleston
On Jul 13, 2010, at 15:35, Jamey Sharp wrote:
> The driOps field of this structure is never set to anything but
> &driGCOps, and this structure is not visible outside this source file.
> Just use the constant in the one place the field was used
Nothing ever set clientClip to anything but a region or NULL, so delete
the clientClipType field and change clientClip from (pointer) to
(RegionPtr).
Signed-off-by: Jamey Sharp
---
dix/gc.c | 51 +--
exa/exa_accel.c| 18 ++
Signed-off-by: Jamey Sharp
---
Xext/panoramiX.c |2 +-
exa/exa.c |4 ++--
exa/exa_priv.h |6 +++---
fb/fbgc.c |4 ++--
hw/dmx/dmxgc.h |4 ++--
hw/xfr
Use the mi implementation of these functions instead, moving them to
dix/gc.c.
Signed-off-by: Jamey Sharp
---
Xext/panoramiX.c | 32 --
dix/gc.c | 87 +--
doc/xml/Xserver-spec.xml | 74 +-
exa/exa
This changes only declarations as the globally allocated tables were
never supposed to be modified anyway.
Technically this is an API change for XAA, though ABI is preserved, but
there better not have been anybody modifying the XAA fallbacks table.
They'd deserve whatever they get.
Signed-off-by:
Signed-off-by: Jamey Sharp
---
hw/xnest/Color.c |5 ++---
hw/xnest/Color.h |3 ++-
hw/xnest/Screen.c |2 ++
3 files changed, 6 insertions(+), 4 deletions(-)
diff --git a/hw/xnest/Color.c b/hw/xnest/Color.c
index 883b77c..0aae8e9 100644
--- a/hw/xnest/Color.c
+++ b/hw/xnest/Color.c
Signed-off-by: Jamey Sharp
---
include/colormapst.h |4 +---
1 files changed, 1 insertions(+), 3 deletions(-)
diff --git a/include/colormapst.h b/include/colormapst.h
index c4e24a6..bb79c86 100644
--- a/include/colormapst.h
+++ b/include/colormapst.h
@@ -113,9 +113,7 @@ typedef struct _Color
This reverts commit bc0c56c407117d1545e20d21f7d30eb3472d618b since we're
breaking ABI anyway.
Conflicts:
ChangeLog
dix/colormap.c
dix/pixmap.c
dix/window.c
include/pixmapstr.h
Signed-off-by: Jamey Sharp
---
dix/colormap.c |7 ---
include/co
None of XAA, EXA, or UXA do any hardware access during CreateGC, so they
don't need VGA arbitration. I haven't found any open source drivers that
hook CreateGC, so they're safe. I'd be surprised if any driver directly
hooks CreateGC and does hardware access from it and needs VGA
arbitration.
Signe
The driOps field of this structure is never set to anything but
&driGCOps, and this structure is not visible outside this source file.
Just use the constant in the one place the field was used.
Signed-off-by: Jamey Sharp
Cc: Jeremy Huddleston
---
hw/xquartz/xpr/driWrap.c |6 +-
1 files
When a GC is allocated, it is zeroed, including all storage requested
with dixRegisterPrivateKey. So CreateGC hooks don't need to initialize
anything to zero.
Signed-off-by: Jamey Sharp
---
fb/fbgc.c |8 +---
hw/xfree86/xaa/xaaGC.c |5 -
hw/xnest/GC.c |3
Signed-off-by: Jamey Sharp
---
dix/dixfonts.c | 54 ---
include/closestr.h | 24 ---
2 files changed, 21 insertions(+), 57 deletions(-)
diff --git a/dix/dixfonts.c b/dix/dixfonts.c
index 4a8f113..91092be 100644
--- a/dix
I have an "It's probably fine then" from Alex, but could I get a
reviewed-by for this?
I have several other patches from the last couple of months also
awaiting review:
http://patchwork.freedesktop.org/project/Xorg/list/?submitter=45
Jamey
On Sun, Jun 13, 2010 at 6:22 PM, Jamey Sharp wrote:
> T
vm86 has been defaulted off since 1.6, and is still a terrible idea to
actually use. Time to say goodbye.
The stub backend is only ever built by default on freebsd/ppc, and can't
be doing any good there. The right fix is --disable-int10 if you want
to not ship int10 support.
Signed-off-by: Adam
On Mon, Jun 21, 2010 at 9:09 AM, Alan Coopersmith
wrote:
> Pat Kane wrote:
>> Does anyone try to compile the current Xorg git tree on Solaris 10?
>> (not OpenSolaris)?
>> We are having GCC problems on 64 bit systems [looks like GCC bugs but
>> I've not taken
>> a very detailed look yet.]
>
> Not
On Tue, 13 Jul 2010 11:22:10 -0700, Jeremy Huddleston
wrote:
>
> On Jul 13, 2010, at 10:18, Keith Packard wrote:
>
> > On Tue, 13 Jul 2010 08:29:22 -0700, Jeremy Huddleston > > wrote:
> >
> >> Jeremy Huddleston (1):
> >> XQuartz: Avoid a crash when mistakenly free()ing in
> >> QuartzSet
On Jul 13, 2010, at 10:18, Keith Packard wrote:
On Tue, 13 Jul 2010 08:29:22 -0700, Jeremy Huddleston > wrote:
Jeremy Huddleston (1):
XQuartz: Avoid a crash when mistakenly free()ing in
QuartzSetCursor on some configs
I'd suggest just initializing 'data' to NULL instead of adding che
On 2010-07-13 12:24, Jeremy Huddleston wrote:
>
> 0001 corresponds to commit 1884db430a5680e37e94726dff46686e2218d525
>
> and doesn't apply to the clean 1.8.99.904 tarball:
It looks like Peter and Keith both independently picked up that patch.
Keith's 018c878e is part of the tag, but Peter's wa
On Tue, 13 Jul 2010 09:24:59 -0700, Jeremy Huddleston
wrote:
> Was the tag applied to the wrong commit, or am I missing something
> here?
The tag looks fine to me -- it's not off on a side branch or
anything. I'm afraid my git-fu doesn't extend to debugging this
though. Anyone else have a clue?
On Tue, 13 Jul 2010 08:29:22 -0700, Jeremy Huddleston
wrote:
> Jeremy Huddleston (1):
> XQuartz: Avoid a crash when mistakenly free()ing in QuartzSetCursor on
> some configs
I'd suggest just initializing 'data' to NULL instead of adding checks.
--
keith.pack...@intel.com
pgpCMzeqsPqm
I'm having some issues with my normal way of generating patches off of master.
With a current state, I run:
$ git format-patch xorg-server-1.8.99.904...origin/master
0001-Revert-dix-use-the-event-mask-of-the-grab-for-TryCli.patch
0002-xkb-remove-now-obsolete-comment.patch
0003-dix-fix-up-erroneo
On Tue, Jul 13, 2010 at 06:39:23AM -0700, Gaetan Nadon wrote:
> This driver only configures against xserver 1.2 or higher.
> Servers 1.2 and higher always include randr 1.2 modes code.
> Therefore BUILD_XMODES will never be set to "yes".
Fair enough. If anyone needs to be able to build the driver
The following changes since commit 3209b094a3b1466b579e8020e12a4f3fa78a5f3f:
DRI2: re-allocate DRI2 drawable if pixmap serial changes (2010-07-09 16:14:14
-0700)
are available in the git repository at:
git://anongit.freedesktop.org/~jeremyhu/xserver master
Jan Hauffa (1):
XQuartz: Rem
This driver only configures against xserver 1.2 or higher.
Servers 1.2 and higher always include randr 1.2 modes code.
Therefore BUILD_XMODES will never be set to "yes".
There is no need to allow this driver to build from an arbitrary
modes and parser tree outside the installed server.
This will
From: Michel Dänzer
Previously we assumed every pixmap destroyed during a software fallback was
also created during a software fallback and had access prepared, but that's
not always true.
Fixes a server abort
Reported-by: 邓逸昕
Signed-off-by: Michel Dänzer
---
exa/exa.c | 23 +++
Acked-by: Maarten Maathuis
2010/7/13 Michel Dänzer :
> From: Michel Dänzer
>
> Previously we assumed every pixmap destroyed during a software fallback was
> also created during a software fallback and had access prepared, but that's
> not always true.
>
> Fixes a server abort
> Reported-by: 邓逸昕
On Tue, Jul 13, 2010 at 01:14:03AM +0200, ext Keith Packard wrote:
> Because some EnterVT code needs to remove it self from the
> call chain, we need to fix all of the wrappers to correctly
> unwrap/rewrap during the call chain. This is a follow-on to the fix
> for bug 27114 in commit 68a9ee8370e6f
On Mon, Jul 12, 2010 at 04:14:03PM -0700, Keith Packard wrote:
> Because some EnterVT code needs to remove it self from the
> call chain, we need to fix all of the wrappers to correctly
> unwrap/rewrap during the call chain. This is a follow-on to the fix
> for bug 27114 in commit 68a9ee8370e6f9b38
From: Michel Dänzer
Previously we assumed every pixmap destroyed during a software fallback was
also created during a software fallback and had access prepared, but that's
not always true.
Fixes a server abort
Reported-by: 邓逸昕
Signed-off-by: Michel Dänzer
---
exa/exa.c | 23 +++
On Tue, 13 Jul 2010 11:11:42 +0100, Magnus Kessler
wrote:
> Commit 5de1b74d64f807b59c730871d4cb171484db9780 in xf86-video-intel
> introduced a fix for the container_of macro. Port this fix over to
> xorg, to fix compilation failure of xf86-video-intel.
Reviewed-by: Chris Wilson
--
Chris Wilson
Commit 5de1b74d64f807b59c730871d4cb171484db9780 in xf86-video-intel
introduced a fix for the container_of macro. Port this fix over to
xorg, to fix compilation failure of xf86-video-intel.
---
include/list.h |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/include/list.h b
Jonathan,
Thanks a lot for your tutorial:).
I got a lot from your explanation. And from the debug I got the matrix
you described from pSrc->transform when rotation is happening. Our driver
justify how about the rotation (90,180,270) from matrix[0][0] and matrix[1][0]
and does th
On Mon, 12 Jul 2010, jesserayadk...@gmail.com wrote:
From: Jesse Adkins
Signed-off-by: Jesse Adkins
---
nls/vi_VN.tcvn/Compose.pre |2 --
nls/vi_VN.viscii/Compose.pre |2 --
2 files changed, 0 insertions(+), 4 deletions(-)
diff --git a/nls/vi_VN.tcvn/Compose.pre b/nls/vi_VN.tcvn/Comp
44 matches
Mail list logo
