On Tue, Mar 16, 2010 at 04:59:47PM +0100, Simon Thum wrote:
Hi,
the hack that is pointer acceleration's init parts bit me. Currently,
this is done during driver calls into xf86InitValuatorDefaults(..., 0).
But since I increased exposure to driver details by fiddling with the
feedback class
On Mon, Mar 22, 2010 at 7:19 PM, Peter Hutterer
peter.hutte...@who-t.net wrote:
A driver that is assigned by an input class is only present as idev-driver.
The driver itself has no access to this information once PreInit is called.
For devices that rely on chain-hotplugging (wacom), this means
2010/3/23 Ian Romanick i...@freedesktop.org:
Kristian Høgsberg wrote:
2010/3/22 Ian Romanick i...@freedesktop.org:
Kristian Høgsberg wrote:
2010/2/8 Francisco Jerez curroje...@riseup.net:
The spec says (regarding glXCreateWindow): If there is already a
GLXFBConfig associated with win (as a
From: Yaakov Selkowitz yselkow...@users.sourceforge.net
In the case where xserver is built in a different prefix from xkbcomp
and xkeyboard-config, e.g. when building from git with autogen.sh
without arguments (which defaults to /usr/local) against
distro-installed components (usually in /usr),
On Mar 22, 10 17:50:38 -0700, Jeremy Huddleston wrote:
Actually, it was a valid error. The assignment was doing |= rather than =,
and the current value was garbage.
? |= looks correct.
Alan's comment is perfectly correct. 1L 31 is undefined on
architectures that define long to have 32 bit.
Ian Romanick i...@freedesktop.org writes:
Kristian Høgsberg wrote:
2010/3/22 Ian Romanick i...@freedesktop.org:
Kristian Høgsberg wrote:
2010/2/8 Francisco Jerez curroje...@riseup.net:
The spec says (regarding glXCreateWindow): If there is already a
GLXFBConfig associated with win (as a
On Tue, Mar 23, 2010 at 01:49:35AM +0100, ext Jeremy Huddleston wrote:
I was thinking smaller would be more acceptable ... but I too would
prefer something like OsRandom() in os/utils.c ... is that something
that should be exported to drivers or just internal to the server?
Not sure if
On Mon, 2010-03-22 at 15:20 -0400, Peter Harris wrote:
This is how the crash can be triggered with only two clients on the system:
Client A: (already running)
Client B: Connect
Client B: CreateGlyphSet(depthN)
Client A: Disconnect
Server: free globalGlyphs(depthN)
Client B:
On Tue, Mar 23, 2010 at 3:22 AM, Yaakov (Cygwin/X)
yselkow...@users.sourceforge.net wrote:
From: Yaakov Selkowitz yselkow...@users.sourceforge.net
In the case where xserver is built in a different prefix from xkbcomp
and xkeyboard-config, e.g. when building from git with autogen.sh
without
Matthias Hopf wrote:
On Mar 22, 10 17:50:38 -0700, Jeremy Huddleston wrote:
Actually, it was a valid error. The assignment was doing |= rather than =,
and the current value was garbage.
? |= looks correct.
Jeremy's right though - the struct is allocated on the stack, uninitialized,
2
From: Yann Droneaud ydrone...@mandriva.com
Date: Tue, 23 Mar 2010 13:05:22 +0100
Le lundi 22 mars 2010 à 17:49 -0700, Jeremy Huddleston a écrit :
I was thinking smaller would be more acceptable ... but I too would
prefer something like OsRandom() in os/utils.c ...
I would prefer
On Mon, 2010-03-22 at 15:20 -0400, Peter Harris wrote:
This crash was introduced with the FindGlyphsByHash function
in 516b96387b0e57b524a37a96da22dbeeeb041712. Before that revision,
ResizeGlyphSet was always called before FindGlyphRef, which would
re-create globalGlyphs(depthN) if necessary.
2010/3/22 Michel Dänzer mic...@daenzer.net:
From: Michel Dänzer daen...@vmware.com
Seems to work as well...
Signed-off-by: Michel Dänzer daen...@vmware.com
---
configure.ac | 2 +-
glx/glxdri2.c | 9 +
2 files changed, 6 insertions(+), 5 deletions(-)
diff --git
Am 23.03.2010 07:10, schrieb Peter Hutterer:
On Tue, Mar 16, 2010 at 04:59:47PM +0100, Simon Thum wrote:
Since the fix is unintrusive, I propose it as a tempoary solution. And I
pledge to fix things later that year :)
Reviewed-by: Peter Hutterer peter.hutte...@who-t.net
can I have your
I think Keith usually looks for an explicit CC to imply that patches
are ready to be applied.
On Tue, Mar 23, 2010 at 9:08 AM, Peter Harris phar...@opentext.com wrote:
This is how the crash can be triggered with only two clients on the system:
Client A: (already running)
Client B: Connect
On 2010-03-23 12:26, Dan Nicholson wrote:
I think Keith usually looks for an explicit CC to imply that patches
are ready to be applied.
I did CC: Keith.
I also added the note as per http://www.x.org/wiki/XServer If a patch
should be merged to master directly instead of going through a personal
On Mar 23, 2010, at 06:48, Mark Kettenis wrote:
Guys, if you ask me, introducing all this additional complecity just
to placate a static analysis tool is starting to get a bit silly.
How about just putting a comment in the code that the usage of rand()
is not security related at all and
XkbEnableDisableControls set extra garbage bits on the xkbControlsNotify
changedControls mask because it was uninitialized on the stack.
Found by clang
Signed-off-by: Jeremy Huddleston jerem...@apple.com
---
xkb/xkbUtils.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git
The parent xdm process registers RemovePid with atexit(), which means
that any child exit would trigger the (wrong) removal of the pidfile.
So in RemovePid, don't do anything if we're not the parent xdm process.
Signed-off-by: Julien Cristau jcris...@debian.org
---
dm.c |3 +++
1 files
This allows ipv4 connections mapped to ipv6, in case the system default
is backwards.
Signed-off-by: Julien Cristau jcris...@debian.org
---
socket.c | 11 +++
1 files changed, 11 insertions(+), 0 deletions(-)
diff --git a/socket.c b/socket.c
index 7e678ad..8cd9d69 100644
---
xdm uses gethostbyaddr and inet_ntop to get the display name. This
fails for e.g. link-local ipv6 addresses. Use getnameinfo instead,
which gives us what we want.
Signed-off-by: Julien Cristau jcris...@debian.org
---
I'm not sure this is correct, I don't quite understand where the
Reviewed-by: Jeremy Huddleston jerem...@apple.com
On Mar 23, 2010, at 11:40, Julien Cristau wrote:
xdm uses gethostbyaddr and inet_ntop to get the display name. This
fails for e.g. link-local ipv6 addresses. Use getnameinfo instead,
which gives us what we want.
Signed-off-by: Julien Cristau
On Tue, Mar 23, 2010 at 10:37:36AM -0700, Jeremy Huddleston wrote:
XkbEnableDisableControls set extra garbage bits on the xkbControlsNotify
changedControls mask because it was uninitialized on the stack.
Found by clang
Signed-off-by: Jeremy Huddleston jerem...@apple.com
---
[CC'ing Pter, see below]
Am 23.03.2010 18:42, schrieb Florian Echtler:
Just for my understanding: when talking about a special client, you think of
something like a (compositing) window manager?
Yes, 'special' since it registers itself for rights (and duties) only
one client shall possess.
Due to the checks in configure, this means it gets priority over HAL if
libudev is found.
Signed-off-by: Peter Hutterer peter.hutte...@who-t.net
---
Is it time yet to ring the bells on HAL? Fedora's been building with udev
enabled for quite a while now and IIRC Debian for even longer.
I think
Peter Hutterer wrote:
Due to the checks in configure, this means it gets priority over HAL if
libudev is found.
Signed-off-by: Peter Hutterer peter.hutte...@who-t.net
---
Is it time yet to ring the bells on HAL? Fedora's been building with udev
enabled for quite a while now and IIRC Debian for
On Mon, Mar 22, 2010 at 07:59:18AM -0700, Dan Nicholson wrote:
2. Make the handling of missing sections from an existing
configuration behave more like the full autoconfig. In other words, if
there's a missing Screen section, generate multiple Screen/Device
pairs with fallbacks. I think this
27 matches
Mail list logo
