When a client is being destroyed, the display_resource is set to NULL.
If then some destroy handler calls wl_client_post_no_memory() or
wl_resource_post_no_memory() we crash.
Signed-off-by: Marek Chalupa
---
src/wayland-server.c | 10 ++
1 file changed, 10 insertions(+)
diff --git a
when the client is being destroyed, we can use wl_client_get_object()
to get an object. Test whether it returns NULL when the object
has been destroyed
Signed-off-by: Marek Chalupa
---
tests/client-test.c | 63 +
1 file changed, 63 insertions
we return NULL insted of a pointer to invalid memory,
which could be OK API break.
Signed-off-by: Marek Chalupa
---
src/wayland-server.c | 10 ++
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/src/wayland-server.c b/src/wayland-server.c
index f745e62..c93a426 100644
--- a
while client is being destroyed, it destroys all its resources.
Destroy handlers are called for the resources and if some
destroy handler post no memory error, we get crash, because
the display resource is NULL (it is destroyed as a first resource)
Signed-off-by: Marek Chalupa
---
tests/client
we split a function while refactoring in c643781 and now
the comment makes no sense
Signed-off-by: Marek Chalupa
---
tests/display-test.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tests/display-test.c b/tests/display-test.c
index f9f8160..17956db 100644
--- a/tests
On 04/25/16 12:20, Pekka Paalanen wrote:
On Mon, 25 Apr 2016 11:33:00 +0200
Marek Chalupa wrote:
If posix_fallocate or ftruncate is interrupted by signal
while working, we return -1 as fd and the allocation process
returns BadAlloc error. That causes xwayland clients to abort
with 'Bad
the code is something like:
if (object == NULL && ...) {
object = NULL;
return;
}
first, the object is already NULL, second, the assignment has no effect
since we return from the function right away
Signed-off-by: Marek Chalupa
---
src/connection.c | 1 -
1 file ch
steps the weston-simple-egl aborts, because
it has not got shell neither ivi-shell objects
Signed-off-by: Marek Chalupa
---
clients/simple-egl.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/clients/simple-egl.c b/clients/simple-egl.c
index 78b4226..d8233c1 100644
--- a/clients
Hi
On 02/22/16 06:52, Jonas Ådahl wrote:
Add a test case that tests the servers behaviour when creating a pool
of size 0. The test suite will do the memory and fd leak check for us,
so the test case is only a triggerer.
Signed-off-by: Jonas Ådahl
---
Makefile.am | 5 ++-
tests/shm-te
Hi,
Reviewed-by: Marek Chalupa
Best,
Marek
On 02/18/16 05:46, Bryce Harrington wrote:
Signed-off-by: Bryce Harrington
---
v2: Also check strdup return
shared/config-parser.c | 22 ++
1 file changed, 22 insertions(+)
diff --git a/shared/config-parser.c b/shared
test if receiving an error on already destroyed object won't
do any harm
Signed-off-by: Marek Chalupa
---
tests/display-test.c | 50 ++
1 file changed, 50 insertions(+)
diff --git a/tests/display-test.c b/tests/display-test.c
index 1a
using wl_display_get_protocol_error() for the error analysis, so IMO its
OK break.
Reviewed-by: Marek Chalupa
Cheers,
Marek
On 02/22/16 06:37, Jonas Ådahl wrote:
If an error is received on a destroyed object, we'd get NULL passed
to display_handle_error() instead of a pointer to a valid wl_proxy.
T
Hi,
added reviewed-by to the whole series in some previous e-mail, so just
adding it here again, so that it can be seen in the patchwork
Reviewed-by: Marek Chalupa
Cheers,
Marek
On 01/14/16 21:32, Derek Foreman wrote:
We're creating resources with versions up to 4. wl_display
On 02/03/16 09:39, Jonas Ådahl wrote:
An xdg_tooltip is a new window type used to implement tooltip like
surfaces. See the interface documentation for details.
Signed-off-by: Jonas Ådahl
---
Changes since v1:
Various wording changes as suggested by Mike.
Added missing version attribute and
clients that implement pointer interface of version 5
wait for the frame event, so without it the scrolling
does not work (GTK+ clients do not scroll now for example).
Xcb axis events are discrete, so it's fine to send
frame after every single axis event
Signed-off-by: Marek Chalupa
---
On 02/01/2016 04:20 PM, Jonas Ådahl wrote:
On Mon, Feb 01, 2016 at 04:16:17PM +0100, Marek Chalupa wrote:
Clients that implement version 5 of pointer wait for the frame
after pointer events (e. g. GTK+ clients). We send it after motion
and button, but not after the axis event in default grab
kinetic scrolling or so)
can decide when to send the frame itself
Signed-off-by: Marek Chalupa
---
src/input.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/src/input.c b/src/input.c
index 2539fa7..32ad20a 100644
--- a/src/input.c
+++ b/src/input.c
@@ -404,6 +404,7 @@ default_grab_pointer_axis
I ran into the same issue with gtk3-demo.
When the popup was being destroyed, the focus was set to 0x0 and I got
sigsegv, because popup grab tried to sent the frame.
Reviewed-by: Marek Chalupa
Regards,
Marek
On 01/29/2016 05:29 PM, Derek Foreman wrote:
Prevents a segfault when mousing
we're returning a pointer
Signed-off-by: Marek Chalupa
---
src/wayland-util.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/wayland-util.c b/src/wayland-util.c
index e782309..3f95877 100644
--- a/src/wayland-util.c
+++ b/src/wayland-util.c
@@ -127,7 +
version, interface->name, interface->version);
+ return NULL;
+ }
+
resource = malloc(sizeof *resource);
if (resource == NULL)
return NULL;
I think that checking the version is right, so
Re
Hi
On 01/13/2016 09:30 PM, Bryce Harrington wrote:
Signed-off-by: Bryce Harrington
---
shared/config-parser.c | 4
1 file changed, 4 insertions(+)
diff --git a/shared/config-parser.c b/shared/config-parser.c
index a50773b..137a9f1 100644
--- a/shared/config-parser.c
+++ b/shared/config
te)
https://bugs.freedesktop.org/show_bug.cgi?id=91356
Reported-by: Ashim
Signed-off-by: Marek Chalupa
---
src/wayland-server.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/src/wayland-server.c b/src/wayland-server.c
index 9e26b18..017a975 100644
--- a/src/wayland-server.c
+++ b/src/waylan
On 12/18/2015 02:08 AM, Bryce Harrington wrote:
This adds functionality to allow system-level control over handing out
file descriptors for sockets, to allow tighter security when running a
Wayland compositor under a Wayland session server.
Signed-off-by: Bryce Harrington
Cc: Sung-Jin Park
C
Reviewed-by: Marek Chalupa
On 12/29/2015 03:10 AM, Jonas Ådahl wrote:
e a shm_pool
___
wayland-devel mailing list
wayland-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/wayland-devel
Hi,
patches 1-2
Reviewed-by: Marek Chalupa
Cheers,
Marek
On 12/29/2015 03:10 AM, Jonas Ådahl wrote:
If an event or request have a "since" attribute that is larger than
the version of the interface it is in, fail with an explaining error
message.
Signed-off-by: Jonas Ådahl
so compiler complains about implicit declaration of calloc()
Maybe we could use zalloc even in tests/test-compositor.c?
otherwise
Reviewed-by: Marek Chalupa
Regards,
Marek
+
#endif
diff --git a/src/wayland-server.c b/src/wayland-server.c
index 55c0cf9..9e26b18 100644
--- a/src/wayland-server.c
+++
Hi,
obviously correct
Reviewed-by: Marek Chalupa
Cheers,
Marek
On 01/04/2016 11:09 AM, Victor Berger wrote:
Without this 'proxy' argument, the '%p' formatter prints a constant
garbage value.
Signed-off-by: Victor Berger
---
src/wayland-client.c | 2 +-
1 file change
Hi,
Reviewed-by: Marek Chalupa
(one small comment below)
Cheers,
Marek
On 01/05/2016 05:18 PM, Lyude wrote:
Noticed this while working on primary selection, in the event we run out
of memory when trying to create a new data source, there's a chance
we'll fail on wl_resource_cr
Hi,
On 12/07/2015 06:54 PM, Bill Spitzak wrote:
On Mon, Dec 7, 2015 at 12:42 AM, Marek Chalupa mailto:mchqwe...@gmail.com>> wrote:
diff --git a/src/wayland-server.c b/src/wayland-server.c
index 1364d5d..b372aa9 100644
--- a/src/wayland-server.c
+++ b/src/wayland-se
On 12/07/2015 01:05 PM, Pekka Paalanen wrote:
On Mon, 7 Dec 2015 11:21:15 +0100
Marek Chalupa wrote:
Hi,
I like this idea, it can be handy when debugging just a part of code (I
wrote similar patch once actually, but then I forgot about it and never
sent it to the list). However, the
Hi,
I like this idea, it can be handy when debugging just a part of code (I
wrote similar patch once actually, but then I forgot about it and never
sent it to the list). However, the implementation could be better IMHO:
I don't think we should move the variables to wayland-util.c. You don't
ash if some
resource destructor or destroy signal handler calls wl_client_post_no_memory()
while destroying client.
https://bugs.freedesktop.org/show_bug.cgi?id=91356
Reported-by: Ashim
Signed-off-by: Marek Chalupa
---
src/wayland-server.c | 11 ++-
1 file changed, 10 insertions(+), 1 del
we don't use ffi in wayland-server.c
Signed-off-by: Marek Chalupa
---
src/wayland-server.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/src/wayland-server.c b/src/wayland-server.c
index bac98bf..bee9a70 100644
--- a/src/wayland-server.c
+++ b/src/wayland-server.c
@@ -42,7
Hi
On 11/19/2015 08:23 PM, Bill Spitzak wrote:
On Thu, Nov 19, 2015 at 1:36 AM, Bryce Harrington mailto:br...@osg.samsung.com>> wrote:
From: Bryce Harrington mailto:br...@bryceharrington.org>>
Signed-off-by: Bryce Harrington mailto:br...@osg.samsung.com>>
---
tests/socket-t
Hi,
I think the indentation already follows project code style. From
doc/Contributing:
"- indent with tabs, and a tab is always 8 characters wide"
current code is:
^Ichar *val = getenv("XDG_RUNTIME_DIR");$
^Iif (!val)$
^I^Iassert(0 && "set $XDG_RUNTIME_DIR to run this test");$
with your pat
we don't use assert() anywhere in this file, so remove #include
Signed-off-by: Marek Chalupa
---
src/event-loop.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/src/event-loop.c b/src/event-loop.c
index 130c7be..ea27b69 100644
--- a/src/event-loop.c
+++ b/src/event-loop.c
@@ -36,7
On many places in the code we use wl_log + abort or wl_log + assert(0).
Replace these with one call to wl_abort, so that we don't mix abort(),
assert(0) and we'll save few lines
Signed-off-by: Marek Chalupa
---
src/connection.c | 22 +++---
src/wayland-clie
commit f814c5dc9 changed get_output_work_area behaviour
which broke the code for positioning maximized window.
The x position was set to 2*output->x instead of to output->x
fix https://bugs.freedesktop.org/show_bug.cgi?id=92357
Signed-off-by: Marek Chalupa
---
desktop-shell/shell
xwayland source is checked, so it dispatches twice on any event.
If the other turn has no events to dispatch, we flush the connection
redundantly
v2. do not flood logs with 'unhandled event' messages
Signed-off-by: Marek Chalupa
Reviewed-by: Derek Foreman
---
xwayland/window-man
Hi,
On 08/06/2015 01:49 PM, Davide Bettio wrote:
Hello,
On 2015-08-06 10:37, Marek Chalupa wrote:
the patch does not apply in the first place:
Applying: Extend WAYLAND_DISPLAY and name parameter semantics to
support absolute paths.
fatal: corrupt patch at line 15
Repository lacks necessary
Hey,
Just to clarify my review of this series. I like WAYLAND_DISPLAY being
able to handle absolute paths. I'm in for adding
WAYLAND_SERVER_SOCKET_DIR, because currently it is the only way how to
customize wl_display_add_socket_auto (In the future we could add
something like wl_display_add_so
Hi,
On 07/29/2015 11:23 AM, Davide Bettio wrote:
Introduce WAYLAND_SERVER_SOCKET_DIR to change socket
directory without changing XDG_RUNTIME_DIR. This might be useful to change
socket directory in case wl_display_add_socket_auto is used. For example
this
will change the socket path for weston wi
Hi,
IMO it would be better to implement this functionality in the compositor
(command line argument or so) instead of adding new env var. Or is there
a reason it must be env var?
However, it is just my point of view, I don't want to block this patch.
Regards,
Marek
On 07/28/2015 05:51 PM, Da
Hi,
the patch does not apply in the first place:
Applying: Extend WAYLAND_DISPLAY and name parameter semantics to support
absolute paths.
fatal: corrupt patch at line 15
Repository lacks necessary blobs to fall back on 3-way merge.
Cannot fall back to three-way merge.
Most of my comment regar
scanner does not complain if we put into version attribute
things like -1 1x 1:3 etc.
Signed-off-by: Marek Chalupa
---
src/scanner.c | 39 ++-
1 file changed, 30 insertions(+), 9 deletions(-)
diff --git a/src/scanner.c b/src/scanner.c
index b0e9ef7..9b41ae4
Hi,
On 07/30/2015 04:33 AM, Jonas Ådahl wrote:
One of the principles of the Wayland protocol is that object creation
always succeeds. If the server for some reason cannot create a valid
object as a response to some request, this would typically be caused by
a protocol violation and in such cases
On 07/30/2015 04:33 AM, Jonas Ådahl wrote:
One of the principles of Wayland is that object creation never fails. If
the object cannot be created, it should be reported as a protocol error
and an offending client should be terminated.
A current violation to this principle is the input device crea
Free all the memory we have allocated during running.
v2.: split creating objects and getting rid of leaks
into two patches
move check for NULL description into free_description
v3.: rebase after previous patch fixes
Signed-off-by: Marek Chalupa
---
src/scanner.c | 106
Use xzalloc instead of xmalloc. This allows us to get rid
of manual initializing the memory to 0s and the code is
shorter and cleaner
Suggested by Bryce Harrington
Signed-off-by: Marek Chalupa
---
src/scanner.c | 83 ---
1 file changed
wrap creating and initializing objects (structures)
into functions and use them in the code.
v2. make create_.* functions consistent
(no func will return NULL)
Signed-off-by: Marek Chalupa
---
src/scanner.c | 164 +++---
1 file changed
>keyboard_focus) {
> +new_state->locked_mods = mods_locked & XkbAllModifiersMask;
> +XkbLatchModifiers(dev, XkbAllModifiersMask,
> + mods_latched & XkbAllModifiersMask);
> +}
I believe h
On 07/23/2015 08:50 PM, Bryce Harrington wrote:
On Thu, Jul 23, 2015 at 07:39:31AM +0200, Marek Chalupa wrote:
Free all the memory we have allocated during running.
v2.: split creating objects and getting rid of leaks
into two patches
move check for NULL description into
Hi,
thanks for review.
On 07/23/2015 08:41 PM, Bryce Harrington wrote:
On Thu, Jul 23, 2015 at 07:39:30AM +0200, Marek Chalupa wrote:
wrap creating and initializing objects (structures)
into functions and use them in the code.
Signed-off-by: Marek Chalupa
---
src/scanner.c | 158
wldbg is a tool for debugging wayland applications
Signed-off-by: Marek Chalupa
---
extras.html | 6 ++
1 file changed, 6 insertions(+)
diff --git a/extras.html b/extras.html
index 55ff35c..eba7133 100644
--- a/extras.html
+++ b/extras.html
@@ -120,6 +120,12 @@ act as a middle-man Wayland
On 07/23/2015 11:22 AM, Pekka Paalanen wrote:
On Wed, 22 Jul 2015 16:27:46 +0200
Marek Chalupa wrote:
Hi folks,
It's been some time since I started developing a tool called wldbg in my
free time. I'm using it while debugging applications based on wayland
and I think it is in e
On 07/16/2015 09:37 AM, Marek Chalupa wrote:
On 07/16/2015 09:27 AM, Marek Chalupa wrote:
Nice catch (some static analysis tool? :).
I'm not sure if this is the right way to fix it, though. If somebody
calls wl_resource_post_error with NULL resource, it is his fault and
program s
Free all the memory we have allocated during running.
v2.: split creating objects and getting rid of leaks
into two patches
move check for NULL description into free_description
Signed-off-by: Marek Chalupa
---
src/scanner.c | 110
wrap creating and initializing objects (structures)
into functions and use them in the code.
Signed-off-by: Marek Chalupa
---
src/scanner.c | 158 ++
1 file changed, 105 insertions(+), 53 deletions(-)
diff --git a/src/scanner.c b/src
On 07/22/2015 07:43 PM, Derek Foreman wrote:
On 22/07/15 02:25 AM, Marek Chalupa wrote:
Thanks for review,
On 07/17/2015 11:02 PM, Derek Foreman wrote:
On 16/07/15 06:59 AM, Marek Chalupa wrote:
Create functions for structures allocation (instead of inlining it into
the code) and free the
Hi folks,
It's been some time since I started developing a tool called wldbg in my
free time. I'm using it while debugging applications based on wayland
and I think it is in enough good shape to bring it out (although there
is still a lot of missing features I'd like to add/fix)
It can work
Thanks for review,
On 07/17/2015 11:02 PM, Derek Foreman wrote:
On 16/07/15 06:59 AM, Marek Chalupa wrote:
Create functions for structures allocation (instead of inlining it into
the code) and free the objects after we don't use them anymore.
Signed-off-by: Marek Chalupa
I think th
On Fri, Jul 17, 2015 at 2:06 PM, Pekka Paalanen wrote:
> On Wed, 1 Jul 2015 12:29:00 +0200
> Marek Chalupa wrote:
>
> > sizeof operator returns size_t, which need not to be unsigned int, but
> can
> > be long (long) unsigned int.
> > So here the correct fix shou
Create functions for structures allocation (instead of inlining it into
the code) and free the objects after we don't use them anymore.
Signed-off-by: Marek Chalupa
---
src/scanner.c | 266 +-
1 file changed, 209 insertions(+
Wrap few long lines to the length around 80 chars
Signed-off-by: Marek Chalupa
---
src/scanner.c | 15 +++
1 file changed, 11 insertions(+), 4 deletions(-)
diff --git a/src/scanner.c b/src/scanner.c
index c652612..ac511f1 100644
--- a/src/scanner.c
+++ b/src/scanner.c
@@ -597,7
On 07/16/2015 09:27 AM, Marek Chalupa wrote:
Nice catch (some static analysis tool? :).
I'm not sure if this is the right way to fix it, though. If somebody
calls wl_resource_post_error with NULL resource, it is his fault and
program should crash to reveal this error. So the right fix in
On 07/16/2015 09:22 AM, Giulio Camuffo wrote:
2015-07-16 9:27 GMT+03:00 Ashim :
Initialising 'wl_client *client = NULL' and checking 'resource' for NULL and
returning if found.
This patch will avoid dereferencing of 'resource' if NULL
I think in this case passing a NULL resource to
wl_resou
Nice catch (some static analysis tool? :).
I'm not sure if this is the right way to fix it, though. If somebody
calls wl_resource_post_error with NULL resource, it is his fault and
program should crash to reveal this error. So the right fix in this case
would be to delete wl_client_post_no_mem
Reviewed-by: Marek Chalupa
On Sat, Jul 11, 2015 at 4:51 AM, Bryce Harrington
wrote:
> Addresses this warning found by Denis Denisov:
>
> [src/logind-util.c:702]: (warning) %d in format string (no. 1)
> requires 'int' but the argument type is 'unsigned int
this does not happen. And sometimes weston crashes, but it is probably
different (weston's) bug.
Looks like caused by call to RRCrtcDestroy in xwl_output_destroy, any
idea why? I don't know much about X's internals...
On 07/15/2015 11:21 AM, Marek Chalupa wrote:
On 05/16
On 05/21/2015 03:43 PM, Marek Chalupa wrote:
output.done event can be sent even on some property change, not only
when announcing the output. Therefore we must check if we already have it
otherwise we may corrupt the list by adding it multiple times.
This fixes bug when xwayland looped
RRCrtcPtr randr_crtc;
As I wrote before, I'm not in favor of patch 1 in this series. However,
this patch needs patch 1 to work properly, so I'm in for using the
version 1 of patch 1
(http://lists.freedesktop.org/archives/wayland-devel/2015-May/021909.html)
With above men
Reviewed-by: Marek Chalupa
(http://lists.freedesktop.org/archives/wayland-devel/2015-May/021952.html)
On 05/16/2015 07:38 AM, Dima Ryazanov wrote:
snprintf does not allocate memory, so we can never get an out-of-memory error.
(Also, the error handler would free xwl_output after it was
On 07/10/2015 06:30 AM, Bryce Harrington wrote:
From the man pages it appears this routine can return NULL on certain
error conditions.
Suggested by Marek Chalupa
Signed-off-by: Bryce Harrington
---
xwayland/selection.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/xwayland
Fixed benchmark uses main(int argc, char *argv[])
but does not use the arguments, so we can replace them with void
Signed-off-by: Marek Chalupa
---
tests/fixed-benchmark.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/fixed-benchmark.c b/tests/fixed-benchmark.c
index
On Wed, Jul 1, 2015 at 11:25 AM, Marek Chalupa wrote:
>
>
> On Fri, Jun 26, 2015 at 6:35 PM, Derek Foreman
> wrote:
>
>> There's no situation where a shm buffer without a pool makes sense,
>> so we enforce the pool's existence a little more rigidl
sizeof operator returns size_t, which need not to be unsigned int, but can
be long (long) unsigned int.
So here the correct fix should be either to use %zu (since C99) or cast it
to unsigned long and use %lu
Regards,
Marek
On Sat, Jun 20, 2015 at 1:28 AM, Bryce Harrington
wrote:
> Addresses thi
Reviewed-by: Marek Chalupa
On Sat, Jun 20, 2015 at 1:12 AM, Bryce Harrington
wrote:
> It is redundant to check x*alloc's return value for null pointers, since
> they are guaranteed to either return non-NULL or terminate the program.
>
> In cases where we memset the malloc'
To be up to good programming practices we should close the input even on
the error paths on lines 1381, 1392 and 1400 (it can be stdin, but on exit
it doesn't matter).
Scanner is sooo leaky, that pedantically closing read-only input file on
every error path may seem a little bit useless, but it is
Reviewed-by: Marek Chalupa
On Sat, Jun 20, 2015 at 12:47 AM, Bryce Harrington
wrote:
> error(1, ...) already will exit, per man page: "If status has a nonzero
> value, then error() calls exit(3) to terminate the program using the
> given value as the exit status." So ex
On Fri, Jun 26, 2015 at 6:35 PM, Derek Foreman
wrote:
> There's no situation where a shm buffer without a pool makes sense,
> so we enforce the pool's existence a little more rigidly.
>
> Signed-off-by: Derek Foreman
> ---
> src/wayland-shm.c | 10 ++
> 1 file changed, 6 insertions(+),
evel@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/wayland-devel
>
Just wondering: does xcb_get_property_reply always return non-NULL? Because
there's no check for that before dereferencing the reply.
Reviewed-by: Marek Chalupa
___
wayland-devel mailing list
wayland-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/wayland-devel
On Wed, Jul 1, 2015 at 10:10 AM, Jonas Ådahl wrote:
> On Wed, Jul 01, 2015 at 09:51:41AM +0200, Marek Chalupa wrote:
> > On Wed, Jul 1, 2015 at 9:16 AM, Jonas Ådahl wrote:
> >
> > > On Wed, Jul 01, 2015 at 08:57:05AM +0200, Marek Chalupa wrote:
> > > > On
On Wed, Jul 1, 2015 at 9:16 AM, Jonas Ådahl wrote:
> On Wed, Jul 01, 2015 at 08:57:05AM +0200, Marek Chalupa wrote:
> > On Mon, Jun 29, 2015 at 3:37 PM, Giulio Camuffo >
> > wrote:
> >
> > > 2015-06-29 14:30 GMT+03:00 Jonas Ådahl :
> > > > Arna
On Mon, Jun 29, 2015 at 3:37 PM, Giulio Camuffo
wrote:
> 2015-06-29 14:30 GMT+03:00 Jonas Ådahl :
> > Arnaud Vrac discovered an issue in the libwayland client API causing
> > race conditions when doing round trips using the wl_display object using
> > non-default proxy queues.
> >
> > The problem
common code into function
move expecting_events into right branch
Signed-off-by: Marek Chalupa
---
hw/xwayland/xwayland-output.c | 42 ++
1 file changed, 30 insertions(+), 12 deletions(-)
diff --git a/hw/xwayland/xwayland-output.c b/hw/xwayland/xwayland
Hi,
On Thu, May 14, 2015 at 6:37 PM, Dima Ryazanov wrote:
> Oh wow, I was playing around with outputs, and never realized
> output_handle_done
> was being called after any geometry change, not just after the output was
> created.
>
> On Thu, May 14, 2015 at 2:58 AM, Mare
Hi,
On Sat, May 16, 2015 at 7:38 AM, Dima Ryazanov wrote:
> Add the output to the list when it's created rather than when its
> properties
> change (as pointed out by Marek Chalupa).
> Remove the output from the list when it's destroyed.
>
> Signed-off-by: Dima Ry
output_done, or if it's just
> an oversight.
>
> On Thu, May 14, 2015 at 9:37 AM, Dima Ryazanov wrote:
>
>> Oh wow, I was playing around with outputs, and never realized
>> output_handle_done
>> was being called after any geometry change, not just after the output
And also write out a warning when we got some event that
we cannot handle.
Signed-off-by: Marek Chalupa
---
xwayland/window-manager.c | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/xwayland/window-manager.c b/xwayland/window-manager.c
index 039f0cd..b1fd12d 100644
: Marek Chalupa
---
hw/xwayland/xwayland-output.c | 25 +++--
1 file changed, 19 insertions(+), 6 deletions(-)
diff --git a/hw/xwayland/xwayland-output.c b/hw/xwayland/xwayland-output.c
index 155cbc1..0c96e87 100644
--- a/hw/xwayland/xwayland-output.c
+++ b/hw/xwayland/xwayland
On Tue, May 12, 2015 at 7:21 PM, Dima Ryazanov wrote:
> The logic is pretty much copied from weston's clients/window.c.
>
> Signed-off-by: Dima Ryazanov
> ---
> hw/xwayland/xwayland.c | 25 -
> hw/xwayland/xwayland.h | 8
> 2 files changed, 32 insertions(+), 1
Checking for negative return value of snprintf is wrong, but snprintf can
actually "fail". It can truncate the output. However, since sizeof name is
256 and the length of "XWAYLAND%d" is always shorter that this, it is safe
here.
Reviewed-by: Marek Chalupa
On Tue, May 12, 2
On Mon, Apr 27, 2015 at 4:06 PM, Giulio Camuffo
wrote:
> 2015-04-27 21:53 GMT+03:00 Bill Spitzak :
> > On 04/26/2015 11:48 PM, Marek Chalupa wrote:
> >
> >>
> >> Is the --include-core-headers option necessary? Until now the scanner
> >> included wayla
On Fri, Apr 3, 2015 at 4:16 AM, Bryce Harrington
wrote:
> config-malformed uses an invalid configuration file to verify that the
> test harness is indeed attempting to load the test's requested config
> file.
>
> Signed-off-by: Bryce Harrington
> ---
> Makefile.am | 12 +++
Hi,
yep, I can confirm R-b for the two patches.
Thanks,
Marek
On Tue, Mar 31, 2015 at 11:26 AM, Pekka Paalanen
wrote:
> On Mon, 30 Mar 2015 06:52:37 -0400
> Marek Chalupa wrote:
>
> > Hehe, race in sending patches (I will have to rebase my patches I just
> sent
> >
Test misc races when adding/releasing devices
v2.: use one roundtrip after releasing devices
add touch support
v3.: remove useless checks
add few comments
repeat tests 30 times instead of 100 times
(it took too long, 30 is enough)
Signed-off-by: Marek Chalupa
Let the client bind to wl_touch. Since we have our own seat,
we know that the compositor will have wl_touch capability.
v2: rebased due to changes in previous commit
Signed-off-by: Marek Chalupa
---
tests/weston-test-client-helper.c | 79 +++
tests/weston
then pick the one that we need and
destroy the rest. The effect is the same, but this code
is better understandable.
Signed-off-by: Marek Chalupa
---
tests/weston-test-client-helper.c | 77 +--
tests/weston-test-client-helper.h | 9 +
tests/weston-t
On Mon, Mar 30, 2015 at 8:22 AM, Pekka Paalanen wrote:
> On Mon, 30 Mar 2015 06:37:59 -0400
> Marek Chalupa wrote:
>
> > Test misc races when adding/releasing devices
> >
> > v2.: use one roundtrip after releasing devices
> > add touch support
>
Hehe, race in sending patches (I will have to rebase my patches I just sent
if this series will get merged before main :)
Anyway, for this and the previous patch:
Reviewed-by: Marek Chalupa
On Mon, Mar 30, 2015 at 5:20 AM, Pekka Paalanen wrote:
> From: Pekka Paalanen
>
> A more de
1 - 100 of 395 matches
Mail list logo
