On Fri, 4 Mar 2011 16:47:45 -0800, James Jones wrote:
> On 3/1/11 6:56 PM, "James Jones" wrote:
>
> > On Tuesday 01 March 2011 08:02:24 Keith Packard wrote:
> > *snip*
> >
> >> Scanout pixmaps get resized automatically when the associated crtc gets
> >> a new mode. This lets a compositing manag
On Sat, Mar 5, 2011 at 2:20 AM, Alan Coopersmith
wrote:
> No calls from libXmu were used, just one instance of one simple macro
> that made the code harder to understand, so stop requiring it and
> linking directly to it. (Of course, libXaw still depends on it, so
> it will still be loaded indir
On Sat, Mar 5, 2011 at 2:04 AM, Alan Coopersmith
wrote:
> Originally fixed by Leo Binchy for Sun Bug 4163152 to resolve bus error
> on VSW5 (XTS) testing on 64-bit SPARC test run on Solaris 7.
> ( Testcase: XtCvtIntToPixmap, Purpose: 1)
>
> Since we're converting from Int to Pixmap, need to acces
No calls from libXmu were used, just one instance of one simple macro
that made the code harder to understand, so stop requiring it and
linking directly to it. (Of course, libXaw still depends on it, so
it will still be loaded indirectly.)
Signed-off-by: Alan Coopersmith
---
ScrollByL.c |
It's the only thing built in that directory, so we can use AM_CFLAGS
and drop the requirement for AM_PROG_CC_C_O in configure.ac
Signed-off-by: Alan Coopersmith
---
configure.ac |3 ---
util/Makefile.am |2 +-
2 files changed, 1 insertions(+), 4 deletions(-)
diff --git a/configure.a
Originally fixed by Leo Binchy for Sun Bug 4163152 to resolve bus error
on VSW5 (XTS) testing on 64-bit SPARC test run on Solaris 7.
( Testcase: XtCvtIntToPixmap, Purpose: 1)
Since we're converting from Int to Pixmap, need to access the argument
passed in as an int, not a Pixmap, which is consist
On 3/1/11 6:56 PM, "James Jones" wrote:
> On Tuesday 01 March 2011 08:02:24 Keith Packard wrote:
> *snip*
>
>> Scanout pixmaps get resized automatically when the associated crtc gets
>> a new mode. This lets a compositing manager deal with the scanout pixmap
>> creation while separate screen con
On 3/4/11 6:38 PM, Mark Kettenis wrote:
From: Adam Jackson
Date: Fri, 4 Mar 2011 16:15:28 -0500
These aren't hardware access, unless they are, in which case the rest of
the wrapping will catch us.
Sorry, but I don't think that is true. The xf86-video-glint driver
has a BlockHandler that can
> From: Adam Jackson
> Date: Fri, 4 Mar 2011 16:15:28 -0500
>
> These aren't hardware access, unless they are, in which case the rest of
> the wrapping will catch us.
Sorry, but I don't think that is true. The xf86-video-glint driver
has a BlockHandler that can call Permedia2vLoadCursorCallbac
On Thursday 03 March 2011 4:14:04 pm Soeren Sandmann wrote:
> James Jones writes:
> >> With a scheme like this, the Composite extension itself would become
> >> simply a way to support legacy applications since window redirection
> >> doesn't really mean anything for InputOnly windows.
> >
> > I
On Fri, 2011-03-04 at 13:37 -0800, Jeremy Huddleston wrote:
> Thanks. That was it. Very odd, but it works. I pushed the changes
> to libXt and libxcb. Did you have any pending cleanup for libXt? If
> not, I'll push a release since it's seems to need one.
>
Nope, go ahead.
signature.asc
De
On Mar 4, 2011, at 11:42, Gaetan Nadon wrote:
> On Fri, 2011-03-04 at 10:50 -0800, Jeremy Huddleston wrote:
>
>> How do I quote []s in m4? I've tried escaping with \, 's, "s, ... I'm out
>> of tricks... I'm trying to update the poll() checks in libxcb and libXt to
>> allow poll on darwin11,
These aren't hardware access, unless they are, in which case the rest of
the wrapping will catch us.
Signed-off-by: Adam Jackson
---
hw/xfree86/common/xf86VGAarbiter.c | 25 -
hw/xfree86/common/xf86VGAarbiterPriv.h |4
2 files changed, 0 insertions(+), 29 d
Remove the UNDRAWABLE_WINDOW macro (which was just needlessly cute) and
always check ->class to decide if a window is InputOnly or InputOutput.
->type is now always either window or pixmap.
Signed-off-by: Adam Jackson
---
dix/dispatch.c |3 ++-
dix/window.c |2 --
include/pixmap.h
The mask argument was almost always 0 or (equivalently) M_ANY. Drop the
mask, and fix up the scant few callers who needed anything stronger.
Signed-off-by: Adam Jackson
---
Xext/panoramiXprocs.c | 13 ++---
Xext/saver.c | 12
Xext/shm.c
No more internal users, this can be dropped now.
Signed-off-by: Adam Jackson
---
dix/Makefile.am|1 -
dix/deprecated.c | 165
include/dix.h | 31 --
include/resource.h | 29 -
4 files changed, 0 insertions(+
The class field was unused for pixmaps, and we don't have enough classes
to justify a whole uint32 anyway.
Signed-off-by: Adam Jackson
---
exa/exa.h |2 +-
exa/exa_classic.c |4 ++--
exa/exa_driver.c |6 +++---
e
Nothing too controversial, just incidental cleanups on the way to finally
getting shatter implemented (someday). Should make every pixmap four bytes
smaller on ILP32, which is pleasant; LP64 doesn't see any change though due
to the struct padding rules.
- ajax
___
On Fri, Mar 4, 2011 at 10:50 AM, Jeremy Huddleston wrote:
> How do I quote []s in m4? I've tried escaping with \, 's, "s, ... I'm out of
> tricks... I'm trying to update the poll() checks in libxcb and libXt to
> allow poll on darwin11, but the []s are getting stripped. darwin[789]* is
> bec
On Fri, 2011-03-04 at 10:50 -0800, Jeremy Huddleston wrote:
> How do I quote []s in m4? I've tried escaping with \, 's, "s, ... I'm out of
> tricks... I'm trying to update the poll() checks in libxcb and libXt to
> allow poll on darwin11, but the []s are getting stripped. darwin[789]* is
> b
Back in October Chase Douglas [1] kicked off a flurry of patches and
discussions on this list about adding Clickpad support, and some form
of multitouch gesture processing (well, at least what is possible with
Synaptics Advanced Gesture Mode) to the xf86-input-synaptics driver (and/or
the kernel).
How do I quote []s in m4? I've tried escaping with \, 's, "s, ... I'm out of
tricks... I'm trying to update the poll() checks in libxcb and libXt to allow
poll on darwin11, but the []s are getting stripped. darwin[789]* is becoming
darwin789*, and the \([^ ]*\) in sed is becoming \(^ *\) ...
On Fri, 04 Mar 2011 13:29:17 -0500, Adam Jackson wrote:
> Reviewed-by: Adam Jackson
Peter -- these look like they should get merged through your
tree. Please let me know if you want me to pull them directly.
Adam -- feel free to Cc' an appropriate maintainer when you review stuff
so that we'll
On 3/4/11 10:55 AM, Rami Ylimäki wrote:
Hi,
Here are some memory leak fixes written and tested against X server
1.9. By looking at the code, they should be valid against latest X
server as well. The use case to expose the leaks was the same used to
expose the evdev leaks: keep connecting and dis
Hi,
Sorry about top-posting, currently on my N900 on a airplane.
Pauli, it would have been nice to include:
Reported-by: Oliver McFadden <...
but if it has already been merged, no worries, you did write the patch after
all. :-)
-- Oliver, N900.
- Original message -
> On Wed, Mar 02, 20
Hi,
Here are some memory leak fixes written and tested against X server
1.9. By looking at the code, they should be valid against latest X
server as well. The use case to expose the leaks was the same used to
expose the evdev leaks: keep connecting and disconnecting a bluetooth
keyboard repeatedly
NewInputDeviceRequest steals the contents of option list elements but
doesn't use the elements themselves for anything. Therefore the list
elements need to be released always.
Signed-off-by: Rami Ylimäki
Reviewed-by: Erkki Seppälä
---
config/hal.c |6 +++---
config/udev.c |6 +++---
2
Signed-off-by: Rami Ylimäki
Reviewed-by: Erkki Seppälä
---
dix/devices.c |2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/dix/devices.c b/dix/devices.c
index 9aaadc4..5ea6ad6 100644
--- a/dix/devices.c
+++ b/dix/devices.c
@@ -936,6 +936,8 @@ CloseDevice(DeviceIntPtr dev)
Signed-off-by: Rami Ylimäki
Reviewed-by: Erkki Seppälä
---
xkb/xkbActions.c | 11 ++-
1 files changed, 2 insertions(+), 9 deletions(-)
diff --git a/xkb/xkbActions.c b/xkb/xkbActions.c
index c020444..a29aeb7 100644
--- a/xkb/xkbActions.c
+++ b/xkb/xkbActions.c
@@ -67,20 +67,13 @@ xkbUn
Hello,
Sorry to take your time with this, but I'm a bit confused about the
status of
Bug 23705 - xserver 1.7.0rc0 uses wrong dimensions / dpi [1]
Based on the bug log (I'm not a coder) I understand this, please *do
correct* any misunderstandings on my part:
1. Since this commit[2] Xor
Currently the XKB options duplicated in EvdevAddKeyClass are never
released. For example, connecting and disconnecting a bluetooth
keyboard repeatedly causes a steadily growing memory leak.
Signed-off-by: Rami Ylimäki
Reviewed-by: Erkki Seppälä
---
src/evdev.c | 19 ++-
1 file
Signed-off-by: Rami Ylimäki
Reviewed-by: Erkki Seppälä
---
src/evdev.c |3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/src/evdev.c b/src/evdev.c
index a77be9e..165400f 100644
--- a/src/evdev.c
+++ b/src/evdev.c
@@ -1839,6 +1839,9 @@ EvdevUnInit(InputDriverPtr drv, Inpu
Hi,
These memory leaks were detected with Valgrind on X server 1.9/evdev
2.6. The original patches have been written and tested against evdev
2.6. Because the leaks seem to be a problem in evdev master also, the
patches were rebased but not tested on X server 1.10/evdev master.
Rami Ylimäki (2):
On 03/04/2011 10:32 AM, Rami Ylimäki wrote:
On 03/04/2011 01:20 AM, Karl Tomlinson wrote:
On Thu, 24 Feb 2011 12:27:28 +1300, xm...@karlt.net wrote:
Rami Ylimäki writes:
On 02/22/2011 11:26 PM, xm...@karlt.net wrote:
Although wrong, our code has been working well enough for us to get
useful
On 03/04/2011 01:20 AM, Karl Tomlinson wrote:
On Thu, 24 Feb 2011 12:27:28 +1300, xm...@karlt.net wrote:
Rami Ylimäki writes:
On 02/22/2011 11:26 PM, xm...@karlt.net wrote:
Although wrong, our code has been working well enough for us to get
useful information. If someone is able to point ou
35 matches
Mail list logo