Moved from Environment to XServer class
Signed-off-by: Peter Hutterer
---
include/xorg/gtest/xorg-gtest-xserver.h |5 +
src/environment.cpp | 34 +-
src/xserver.cpp | 35 +++
3 files c
And initialise with the default settings
Signed-off-by: Peter Hutterer
---
include/xorg/gtest/xorg-gtest-xserver.h | 12
src/environment.cpp |4 +++-
src/xserver.cpp | 26 +-
3 files changed, 40 insertions(+
This class is a Process, so it's a drop-in replacement for the current
Environment, but it provides a few methods to talk to the server being
started.
Signed-off-by: Peter Hutterer
---
configure.ac|2 +-
include/Makefile.am |1 +
include/xo
This branch is available on
git://people.freedesktop.org/~whot/xorg-gtest :devel
The main features here are adding a new XServer class that controls most of
the server-related bits. The current API is maintained, the bits just move
into different classes so existing tests should continue to work
Peter Hutterer writes:
> I left the first patch as-is because it's easy to cherry-pick to 1.12 this
> way. I also left the xf86BlockSIGIO() calls in though I don't think outside
> of the ddx actually uses them. We can probably drop them, either this cycle
> or the next.
I like the whole series,
Peter Hutterer writes:
> void
> OsReleaseSignals(void)
> {
> #ifdef SIG_BLOCK
> if (--BlockedSignalCount == 0) {
> sigprocmask(SIG_SETMASK, &PreviousSignalMask, 0);
> +OsReleaseSIGIO();
> }
> #endif
Should you use SIG_UNBLOCK instead of SIG_SETMASK? It took me sev
Set a destructor function on pixman images and call fbFinishAccess()
from there, rather than directly from free_pixman_pict(). This ensures
that fbFinishAccess() gets called even if pixman still has a reference
to the image after free_pixman_pict(), as is the case for alphamaps.
Reviewed-by: Aaron
Peter Hutterer writes:
> We can ignore the "wasset" argument now since the DIX will keep proper
> refcounting.
Ok, so I was thinking that we should change the API here, but really
what we should do is plan to change users to use
OsBlockSIGIO/OsReleaseSIGIO directly and plan on deprecating these
Peter Hutterer writes:
> Signed-off-by: Peter Hutterer
These patches all seem innocuous enough;
For the series, Reviewed-by: Keith Packard
--
keith.pack...@intel.com
pgpPT88BkAcHC.pgp
Description: PGP signature
___
xorg-devel@lists.x.org: X.Org
Signed-off-by: Peter Hutterer
---
hw/dmx/input/dmxevents.c | 24
hw/dmx/input/dmxsigio.c | 22 --
hw/dmx/input/dmxsigio.h |2 --
hw/kdrive/ephyr/ephyr.c | 24 ++--
hw/kdrive/src/kinput.c | 40 ++-
The hooks are left for this cycle, we can drop it next cycle once the
drivers that need it (e.g. wacom) have been updated.
Signed-off-by: Peter Hutterer
---
hw/xfree86/common/xf86Cursor.c | 12 +---
hw/xfree86/common/xf86Events.c | 14 +++---
hw/xfree86/common/xf8
We can ignore the "wasset" argument now since the DIX will keep proper
refcounting.
Signed-off-by: Peter Hutterer
---
hw/xfree86/os-support/shared/sigio.c | 17 ++---
1 file changed, 2 insertions(+), 15 deletions(-)
diff --git a/hw/xfree86/os-support/shared/sigio.c
b/hw/xfree86/o
Let the dix be in charge of changing the sigprocmask so we only have one
entity that changes it.
Signed-off-by: Peter Hutterer
---
include/os.h |6 +++
os/utils.c | 54 +--
test/Makefile.am |3 +-
test/os.c| 130 +++
Drivers call xf86InstallSIGIOHandler() for their fd on DEVICE_ON. That
function does not actually enable the signal if it was blocked to begin
with. As a result, if one vt-switches away from the server (SIGIO is
blocked) and then triggers a server regeneration, the signal remains
blocked and input
I left the first patch as-is because it's easy to cherry-pick to 1.12 this
way. I also left the xf86BlockSIGIO() calls in though I don't think outside
of the ddx actually uses them. We can probably drop them, either this cycle
or the next.
Cheers,
Peter
_
Alan Coopersmith writes:
> Clear them out when needed instead of leaving whatever values were
> present in previously sent messages.
>
> Signed-off-by: Alan Coopersmith
> ---
>
> v2: Simplified as suggested by Keith to just use memset instead of a
> Duff-device style switch statement, since mos
Hopefully saves the next poor sod weird linker errors.
Signed-off-by: Peter Hutterer
---
src/Makefile.am |1 +
1 file changed, 1 insertion(+)
diff --git a/src/Makefile.am b/src/Makefile.am
index 148f1f2..0b93df0 100644
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -23,6 +23,7 @@
# SOFTWAR
XSERVER implies this has to do with the server, but these are the client
libs.
Signed-off-by: Peter Hutterer
---
configure.ac |2 +-
examples/Makefile.am |4 ++--
2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/configure.ac b/configure.ac
index 3dea940..6e9e458 100
Errors/warnings on building documentation are enough, we don't need to see
the whole spew
Signed-off-by: Peter Hutterer
---
doc/Doxyfile |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/doc/Doxyfile b/doc/Doxyfile
index 488b15c..cf58184 100644
--- a/doc/Doxyfile
+++ b/doc/D
Signed-off-by: Peter Hutterer
---
examples/Makefile.am |3 +++
1 file changed, 3 insertions(+)
diff --git a/examples/Makefile.am b/examples/Makefile.am
index b9b914b..4a3b633 100644
--- a/examples/Makefile.am
+++ b/examples/Makefile.am
@@ -29,10 +29,12 @@ check_LIBRARIES = libgtest.a libxorg
Introduced in e1c010f23272e61c28c73aa603b477ba6fbae875
Signed-off-by: Peter Hutterer
---
examples/xorg-gtest-example.cpp |2 +-
include/xorg/gtest/evemu/xorg-gtest_device.h |2 +-
include/xorg/gtest/xorg-gtest_environment.h |2 +-
include/xorg/gtest/xorg-gtest_process.h
Signed-off-by: Peter Hutterer
---
src/xorg-gtest_main.cpp | 25 ++---
1 file changed, 14 insertions(+), 11 deletions(-)
diff --git a/src/xorg-gtest_main.cpp b/src/xorg-gtest_main.cpp
index bbf232b..08927a0 100644
--- a/src/xorg-gtest_main.cpp
+++ b/src/xorg-gtest_main.cpp
@
Don't make it a separate build target, just build it.
Signed-off-by: Peter Hutterer
---
configure.ac|1 +
doc/Makefile.am |4
2 files changed, 5 insertions(+)
diff --git a/configure.ac b/configure.ac
index 7d7b2e1..3dea940 100644
--- a/configure.ac
+++ b/configure.ac
@@ -23,6 +
Signed-off-by: Peter Hutterer
---
include/xorg/gtest/xorg-gtest_environment.h |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/xorg/gtest/xorg-gtest_environment.h
b/include/xorg/gtest/xorg-gtest_environment.h
index 723223f..80982c6 100644
--- a/include/xorg/gtest/xo
Signed-off-by: Peter Hutterer
---
src/environment.cpp |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/environment.cpp b/src/environment.cpp
index ca499ee..60a5190 100644
--- a/src/environment.cpp
+++ b/src/environment.cpp
@@ -163,9 +163,9 @@ void xorg::testing::Envi
Peter Hutterer writes:
> Chase Douglas (15):
> Add global variable inSignalContext
> Save log file file descriptor for signal context logging
> Add FormatUInt64{,Hex}() for formatting numbers in a signal safe manner
> Add LogMessageVerbSigSafe() for logging messages while
Alan Coopersmith writes:
> Adds new function padding_for_int32() and uses existing pad_to_int32()
> depending on required results.
Whoa. Even tests! Totally awesome!
Reviewed-by: Keith Packard
--
keith.pack...@intel.com
pgp7WPwqVTc9i.pgp
Description: PGP signature
_
Adds new function padding_for_int32() and uses existing pad_to_int32()
depending on required results.
Signed-off-by: Alan Coopersmith
---
On 07/ 2/12 10:58 AM, Keith Packard wrote:
> (also, we should get rid of the padlength array and just use
> '-count & 3' instead :-)
Here you go. I'm not t
Mixing dashes and spaces in xorg-gtest_foobar.h is a painful naming
convention. Use dashes only instead, and provide compat headers plus a
warning for those using the old header files.
Those users should be using xorg-gtest.h anyway.
Signed-off-by: Peter Hutterer
---
If the diff looks confusing,
On Mon, Jun 18, 2012 at 11:17:19AM -0700, Chase Douglas wrote:
> GNU automake puts -I. -I.. at the beginning of every invocation of g++.
> This breaks xorg-gtest compilation if there is a file of the same name
> as an xorg-gtest implementation file in ../src/ or ../../src. For
> example, if compili
Ian Romanick writes:
> From: Ian Romanick
>
> The server does not want GL extension prototypes. It never links with
> anything that could possibly provide implementations of these functions. It
> *is* the provide, and it does not provde these symbols. All this does is
> create hundreds of war
Clear them out when needed instead of leaving whatever values were
present in previously sent messages.
Signed-off-by: Alan Coopersmith
---
v2: Simplified as suggested by Keith to just use memset instead of a
Duff-device style switch statement, since most calls won't need this.
os/io.c |6
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 07/ 2/12 10:13 AM, Keith Packard wrote:
> Alan Coopersmith writes:
>
>> Ensures padding bytes are zero-filled
>
> Yeah, we've wanted to do this for a long time for compression. Is this
> really all of the replies allocated with malloc?
It's all
Changes to v1:
- 32 bit issues fixed by squashing in
http://patchwork.freedesktop.org/patch/10876/ and
http://patchwork.freedesktop.org/patch/10877/
The following changes since commit 8dc70acbf3d82611ac9ec1ec2a52edcc01934850:
Merge remote-tracking branch 'alanc/master' (2012-06-14 13:04:29 -070
Adam Jackson writes:
> I'm not sure this makes sense? The list of slave outputs is the list of
> outputs from other screens connected to _this_ screen's CRTC list,
> right? So _this_ screen's CRTC list is all you need to walk.
Nope, the slave screens keep their own OUTPUTs and CRTCs, there's
Dave Airlie writes:
> From: Dave Airlie
>
> Add the simple passthrough interface for drivers to use,
> so they can set scanout pixmaps.
Yeah, looks good.
Reviewed-by: Keith Packard
--
keith.pack...@intel.com
pgpIdcsMST2Wq.pgp
Description: PGP signature
Dave Airlie writes:
> From: Dave Airlie
>
> When randr notices a crtc configuration request for a slave device,
> it checks if the slave allocated pixmap exists and is suitable,
> if not it allocates a new shared pixmap from the master, shares
> it to the slave, and starts the master tracking da
Dave Airlie writes:
> +void
> +AttachOutputGPU(ScreenPtr pScreen, ScreenPtr new)
> +{
> +assert(new->isGPU);
> +xorg_list_add(&new->output_head, &pScreen->output_slave_list);
> +new->current_master = pScreen;
> +}
I'd love an assert that new isn't already on a list somehow?
--
kei
Adam Jackson writes:
> On 7/2/12 6:13 AM, Dave Airlie wrote:
>> From: Dave Airlie
>>
>> This bumps totalPixmapSize in all attached screens.
>
> Aieee. totalPixmapSize is now monotonically increasing every time you
> plug something in. Until it wraps, of course, and then bang. Not sure
> wha
Adam Jackson writes:
> On 7/2/12 6:13 AM, Dave Airlie wrote:
> Aieee. totalPixmapSize is now monotonically increasing every time you
> plug something in. Until it wraps, of course, and then bang. Not sure
> what a better solution looks like, but IWBNI host storage was shared if
> possible.
Dave Airlie writes:
> From: Dave Airlie
>
> These asserts stop us increasing the size of privates after create screen
> resources has been called, however for dynamic screens we need to be able
> to increase private size as new drivers are loaded.
>
> In theory any object allocated by an older d
Dave Airlie writes:
> From: Dave Airlie
>
> This driver is currently really the only choice for a hotplug
> at the moment, we can refine this in the future to try and
> pick the driver names.
>
> We have to reload this as it can get unloaded it not used
> in the original probe.
Should xf86platf
Adam Jackson writes:
> On 7/2/12 6:13 AM, Dave Airlie wrote:
> I think anywhere you see a screen number on the wire the safe thing to
> do is index protocolscreens not gpuscreens.
DGAAvailable is also called from things like xf86SetScrnInfoModes, which
seem likely to be called with non-protoco
Dave Airlie writes:
> From: Dave Airlie
>
> Just check for GPU screens, in DGAAvailable,
> should probably overhaul DGA interface at some point,
> or remove it.
Yuck, but it seems to do the trick.
Reviewed-by: Keith Packard
Does this need to be merged well ahead of the other changes in this
Dave Airlie writes:
>
> if (pScrPriv->provider)
> total_providers++;
> +xorg_list_for_each_entry(iter, &pScreen->unattached_list,
> unattached_head) {
> +pScrPriv = rrGetScrPriv(iter);
> +total_providers += pScrPriv->provider ? 1 : 0;
> +}
>
> pScrP
Dave Airlie writes:
> I'm not sure if putting it in right now is necessary.
Sounds good then.
Reviewed-by: Keith Packard
--
keith.pack...@intel.com
pgpjJdm8soBDD.pgp
Description: PGP signature
___
xorg-devel@lists.x.org: X.Org development
Archive
Dave Airlie writes:
> On Mon, Jul 2, 2012 at 8:11 PM, Keith Packard wrote:
> Well an unbound gpu you probably don't know its a slave, its just
> unbound.
It's about to become a slave though
--
keith.pack...@intel.com
pgpLuUPZWcpA4.pgp
Description: PGP signature
Dave Airlie writes:
> On Mon, Jul 2, 2012 at 8:20 PM, Keith Packard wrote:
> though a void * that just cases an fd is probably okay.
That was my thinking. void * covers a lot of sins.
--
keith.pack...@intel.com
pgpYRB2zYjbn0.pgp
Description: PGP signature
__
On 07/02/2012 03:22 AM, Dave Airlie wrote:
Hi,
Just wondering about what sort of things are considered ABI breaks in reality,
From my understanding, I should be able to
a) add new members to end of ScreenRec.
b) add new members to xf86ScrnInfoRec in the reserved sections.
Can I though add m
Dave Airlie writes:
> No they can all do it, just haven't the code to do the picking yet of
> a driver, and since
> USB is really the only thing I can hotplug so far.
It seems like more driver ABI changes would be required to let them know
about the new device then?
> the udev interface is fine
On 06/29/2012 11:11 PM, Alan Coopersmith wrote:
Fixes leaks in error paths found by Parfait 1.0.0:
Error: X Resource Leak
Leaked X Resource pix
at line 62 of CrBFData.c in function 'XCreateBitmapFromData'.
pix initialized at line 60 with XCreatePixmap
Error: X Resource Le
On Mon, Jul 2, 2012 at 6:13 AM, Dave Airlie wrote:
> From: Dave Airlie
>
> For DRI2 in some offload cases we need to set a new pixmap on the crtc,
> this hook allows dri2 to call into randr to do the necessary work to set
> a pixmap as the scanout pixmap for the crtc the drawable is currently on.
On Mon, Jul 2, 2012 at 6:13 AM, Dave Airlie wrote:
> From: Dave Airlie
>
> Current USB devices have no hw rendered cursors, so we need the
> master GPU to render the cursor, so whenever we plug in a
> slave device, fallback to sw rendered cursors.
What about when we have a non-USB output slave t
On 7/2/12 6:13 AM, Dave Airlie wrote:
From: Dave Airlie
These asserts stop us increasing the size of privates after create screen
resources has been called, however for dynamic screens we need to be able
to increase private size as new drivers are loaded.
In theory any object allocated by an o
On 7/2/12 6:13 AM, Dave Airlie wrote:
@@ -158,3 +158,94 @@ PixmapPtr PixmapShareToSlave(PixmapPtr pixmap, ScreenPtr
slave)
return spix;
}
+
+Bool
+PixmapStartDirtyTracking(PixmapPtr src,
+ PixmapPtr dst_slave,
+ int x, int y)
+{
+Scre
On 7/2/12 6:13 AM, Dave Airlie wrote:
From: Dave Airlie
Current code constrains the cursor to the crtcs on the master
device, for slave outputs to work we have to include their crtcs
in the constrain calculations.
Signed-off-by: Dave Airlie
I'm not sure this makes sense? The list of slave
On Mon, Jul 2, 2012 at 6:13 AM, Dave Airlie wrote:
> From: Dave Airlie
>
> These asserts stop us increasing the size of privates after create screen
> resources has been called, however for dynamic screens we need to be able
> to increase private size as new drivers are loaded.
>
> In theory any
Dave Airlie writes:
> On Mon, Jul 2, 2012 at 8:54 PM, Keith Packard wrote:
>> Dave Airlie writes:
>>
>>
>>> Yes thats going to be the standard on switchable GPU machines, two masters
>>> and the ability to jump between them. In that case max master is one, and
>>> you'd
>>> have to set the pro
On 7/2/12 6:13 AM, Dave Airlie wrote:
From: Dave Airlie
This bumps totalPixmapSize in all attached screens.
Aieee. totalPixmapSize is now monotonically increasing every time you
plug something in. Until it wraps, of course, and then bang. Not sure
what a better solution looks like, but I
On 7/2/12 6:12 AM, Dave Airlie wrote:
+void
+RemoveGPUScreen(ScreenPtr pScreen)
+{
+int idx, j;
+if (!pScreen->isGPU)
+return;
+
+idx = pScreen->myNum - GPU_SCREEN_OFFSET;
+for (j = idx; j < screenInfo.numGPUScreens - 1; j++) {
+screenInfo.gpuscreens[j] = screenIn
On 7/2/12 6:13 AM, Dave Airlie wrote:
diff --git a/hw/xfree86/common/xf86DGA.c b/hw/xfree86/common/xf86DGA.c
index 6416372..4fe12f2 100644
--- a/hw/xfree86/common/xf86DGA.c
+++ b/hw/xfree86/common/xf86DGA.c
@@ -523,10 +523,15 @@ DGAChangePixmapMode(int index, int *x, int *y, int mode)
Bool
D
On Mon, Jul 2, 2012 at 8:54 PM, Keith Packard wrote:
> Dave Airlie writes:
>
>
>> Yes thats going to be the standard on switchable GPU machines, two masters
>> and the ability to jump between them. In that case max master is one, and
>> you'd
>> have to set the provider roles. If maxmaster > 1 t
Dave Airlie writes:
> Yes thats going to be the standard on switchable GPU machines, two masters
> and the ability to jump between them. In that case max master is one, and
> you'd
> have to set the provider roles. If maxmaster > 1 then xinerama
> emulation is available.
In those machines, isn
On Mon, Jul 2, 2012 at 7:52 PM, Keith Packard wrote:
> Dave Airlie writes:
>
>
>> +struct _rrProvider {
>> +RRProvider id;
>> +ScreenPtr pScreen; /* gpu screen more than likely */
>> +int current_role;
>
> uint32_t current_role
indeed,
>
>> diff --git a/randr/rrprovider.c b/randr/rr
Alan Coopersmith writes:
>> Alan Coopersmith (4):
>> Add dixGetGlyphs to replace GetGlyphs from libXfont to simplify linking
>> Build tda9885.c functions into fi1236 to avoid libtool relinking
>> problems
>> Add pixman to module dependencies for pixman_region_* functions
>>
On Mon, Jul 2, 2012 at 8:07 PM, Keith Packard wrote:
> Dave Airlie writes:
>
>> +xf86ProviderPtr
>> +xf86ProviderCreate(ScrnInfoPtr scrn,
>> + const xf86ProviderFuncsRec *funcs, const char *name)
>> +{
>> +xf86CrtcConfigPtr xf86_config = XF86_CRTC_CONFIG_PTR(scrn);
>> +x
On Mon, Jul 2, 2012 at 8:12 PM, Keith Packard wrote:
> Dave Airlie writes:
>
>> From: Dave Airlie
>>
>> This is the default attachment, unbound gpu screens get
>> attached to the 0 protocol screen.
>
> This code doesn't check to see if the GPU is unbound; should it be
> looking at GPUScreens[i]-
On Mon, Jul 2, 2012 at 8:20 PM, Keith Packard wrote:
> Dave Airlie writes:
>
>> From: Dave Airlie
>>
>> This is a hooks for pixmap sharing and tracking.
>>
>> The pixmap sharing ones get an integer handle for the pixmap
>> and use a handle to be the backing for a pixmap.
>>
>> The tracker interf
On Mon, Jul 2, 2012 at 8:11 PM, Keith Packard wrote:
> Dave Airlie writes:
>
>> From: Dave Airlie
>>
>> This list is meant for attaching unbound gpu screens to initially,
>> before the client side rebinds them.
>
> You'll need this before 10/36 can be applied. Seems like simply
> reordering the
Dave Airlie writes:
> From: Dave Airlie
>
> This just adds exa interfaces for mixed exa so drivers can
> share and set shared pixmaps up correctly.
Reviewed-by: Keith Packard
--
keith.pack...@intel.com
pgpIHAiNKRN0a.pgp
Description: PGP signature
___
Dave Airlie writes:
> From: Dave Airlie
>
> This is a hooks for pixmap sharing and tracking.
>
> The pixmap sharing ones get an integer handle for the pixmap
> and use a handle to be the backing for a pixmap.
>
> The tracker interface is to be used when a GPU needs to
> track pixmaps to be updat
Dave Airlie writes:
> From: Dave Airlie
>
> This just hooks up a simple role change to go from unattached
> to slave output and back.
>
> Signed-off-by: Dave Airlie
> ---
> hw/xfree86/modes/xf86RandR12.c | 27 ++-
> 1 file changed, 26 insertions(+), 1 deletion(-)
>
>
Dave Airlie writes:
> From: Dave Airlie
>
> This is the default attachment, unbound gpu screens get
> attached to the 0 protocol screen.
This code doesn't check to see if the GPU is unbound; should it be
looking at GPUScreens[i]->pScreen->current_master before calling?
--
keith.pack...@intel
Dave Airlie writes:
> From: Dave Airlie
>
> This list is meant for attaching unbound gpu screens to initially,
> before the client side rebinds them.
You'll need this before 10/36 can be applied. Seems like simply
reordering the patches should work, or possibly merging them together.
> +Attach
Dave Airlie writes:
> +xf86ProviderPtr
> +xf86ProviderCreate(ScrnInfoPtr scrn,
> + const xf86ProviderFuncsRec *funcs, const char *name)
> +{
> +xf86CrtcConfigPtr xf86_config = XF86_CRTC_CONFIG_PTR(scrn);
> +xf86ProviderPtr provider;
> +int len;
> +
> +if (xf86_co
On Mon, Jul 2, 2012 at 7:37 PM, Keith Packard wrote:
> Dave Airlie writes:
>
>> +int
>> +xf86platformAddDevice(int index)
>> +{
>> +int i, old_screens, scr_index;
>> +DriverPtr drvp = NULL;
>> +int entity;
>> +screenLayoutPtr layout;
>> +
>> +for (i = 0; i < xf86NumDrivers; i+
Dave Airlie writes:
> +struct _rrProvider {
> +RRProvider id;
> +ScreenPtr pScreen; /* gpu screen more than likely */
> +int current_role;
uint32_t current_role
> diff --git a/randr/rrprovider.c b/randr/rrprovider.c
> new file mode 100644
> index 000..549db5d
> --- /dev/null
>
Dave Airlie writes:
> +/* if autoaddgpu devices is enabled then go find a few more and add them
> as GPU screens */
> +if (xf86Info.autoAddGPU == FALSE)
> +return foundScreen;
> +
> +if (numDevs) {
> +for (j = 0; j < xf86_num_platform_devices; j++) {
> +pr
On Mon, Jul 2, 2012 at 7:19 PM, Keith Packard wrote:
> Dave Airlie writes:
>
>
>> +PROVIDER { XID }
>> +
>> +MAXMASTER { CARD16 }
>> + Maximum number of master GPUs server can support.
>> + 1 unless dynamic xinerama is enabled.
>
> You don't need a type for this; it's just a count, right?
Dave Airlie writes:
> +int
> +xf86platformAddDevice(int index)
> +{
> +int i, old_screens, scr_index;
> +DriverPtr drvp = NULL;
> +int entity;
> +screenLayoutPtr layout;
> +
> +for (i = 0; i < xf86NumDrivers; i++) {
> +if (!xf86DriverList[i])
> +continue;
>
Dave Airlie writes:
> v3: introduce MAXGPUSCREENS, fix return value check
Reviewed-by: Keith Packard
--
keith.pack...@intel.com
pgpdolhPk7zkO.pgp
Description: PGP signature
___
xorg-devel@lists.x.org: X.Org development
Archives: http://lists.x.org
Dave Airlie writes:
> v2: fixup int check as per Keith's review.
Reviewed-by: Keith Packard
--
keith.pack...@intel.com
pgpxT6MrPGSI0.pgp
Description: PGP signature
___
xorg-devel@lists.x.org: X.Org development
Archives: http://lists.x.org/archives
Dave Airlie writes:
> v10: address most of Keith's concerns.
Looks good now.
Reviewed-by: Keith Packard
--
keith.pack...@intel.com
pgpZqtwTti1WG.pgp
Description: PGP signature
___
xorg-devel@lists.x.org: X.Org development
Archives: http://lists.x
Dave Airlie writes:
> +PROVIDER { XID }
> +
> +MAXMASTER { CARD16 }
> + Maximum number of master GPUs server can support.
> + 1 unless dynamic xinerama is enabled.
You don't need a type for this; it's just a count, right?
> +
> +FLAGS { Dynamic }
> + Informational flags about accep
On 06/28/2012 10:30 PM, Peter Hutterer wrote:
Make %u and %x sizeof(unsigned int), %p sizeof(void*). This is printf
behaviour and we can't guarantee that void* is uint64_t anyway.
Signed-off-by: Peter Hutterer
---
This needs to be squashed into:
Add LogMessageVerbSigSafe() for logging messag
Oops, repushed now with correct reviewed-by tags that I'd forgotten the first
time.
The following changes since commit 212b9803238d2de2e77cbe5de62d3f616ae50daf:
randr: Fix REQUEST vs. REQUEST_SIZE_MATCH mismatch (2012-06-29 23:34:30 -0700)
are available in the git repository at:
git+ssh://
Alan Coopersmith writes:
> Clear them out when needed instead of leaving whatever values were
> present in previously sent messages.
Given that nearly all of the calls to WriteToClient have already aligned
their data, I'd suggest that this should probably just use memset under
an if
if
On Mon, Jul 2, 2012 at 11:13 AM, Dave Airlie wrote:
> From: Dave Airlie
>
> This adds the initial provider object and provider property
> support to the randr dix code.
>
> v2: destroy provider in screen close
> v2.1: fix whitespace
>
> v3: update for latest rev of protocol + renumber after 1.4 t
Alan Coopersmith writes:
Ok, I like the general plan here, but there are a mish-mash of styles
which doesn't really seem neater than the original code.
If we can, indeed, use this form:
> +
> +reply = (xQueryTreeReply) {
> +.type = X_Reply,
> +.sequenceNumber = client->seque
Alan Coopersmith writes:
> Let the compiler worry about 0-filling the rest of the fields,
> instead of memsetting the whole struct and then going back to
> overwrite some of the fields.
Reviewed-by: Keith Packard
--
keith.pack...@intel.com
pgpLbKGMSfeEy.pgp
Description: PGP signature
__
Alan Coopersmith writes:
> Ensures padding bytes are zero-filled
Yeah, we've wanted to do this for a long time for compression. Is this
really all of the replies allocated with malloc?
Reviewed-by: Keith Packard
--
keith.pack...@intel.com
pgpYfRzMpbGeE.pgp
Description: PGP signature
__
The following changes since commit 212b9803238d2de2e77cbe5de62d3f616ae50daf:
randr: Fix REQUEST vs. REQUEST_SIZE_MATCH mismatch (2012-06-29 23:34:30 -0700)
are available in the git repository at:
git://people.freedesktop.org/~alanc/xserver.git master
for you to fetch changes up to 0e6710d0c
Alan Coopersmith writes:
> #define _WriteQueryAdaptorsReply(_c,_d) \
>if ((_c)->swapped) SWriteQueryAdaptorsReply(_c, _d); \
> - else WriteToClient(_c, sz_xvQueryAdaptorsReply, (char*)_d)
> + else WriteToClient(_c, sz_xvQueryAdaptorsReply, _d)
I checked the callers of all of these macros
On 6/28/12 11:17 AM, Michal Srb wrote:
LoaderSymbol calls dlsym with RTLD_DEFAULT pseudo handle making it search in
every loaded library. In addition glibc adds NODELETE flag to the library
containing the symbol.
It's used in doLoadModule to locateModuleData symbol, the
module's library gets the
On Mon, Jul 2, 2012 at 5:31 PM, walter harms wrote:
>
>
> Am 02.07.2012 12:13, schrieb Dave Airlie:
>> From: Dave Airlie
>>
>> This list is meant for attaching unbound gpu screens to initially,
>> before the client side rebinds them.
>>
>> Signed-off-by: Dave Airlie
>> ---
>> dix/dispatch.c
Am 02.07.2012 12:13, schrieb Dave Airlie:
> From: Dave Airlie
>
> This list is meant for attaching unbound gpu screens to initially,
> before the client side rebinds them.
>
> Signed-off-by: Dave Airlie
> ---
> dix/dispatch.c | 19 +++
> include/screenint.h |5 ++
Hello
On 22 June 2012 03:55, Peter Hutterer wrote:
> On Fri, Jun 08, 2012 at 04:19:17PM +0200, Michal Suchanek wrote:
>> Hello,
>>
>> sending the patch witch should fix issue with unloading sibling
>> modules along with a couple of patches that allow actually unloading
>> modules.
>
> in the futu
i would suggest to split even further to improve readability.
would you bother to place that #ifdef ARGB_CURSOR stuff into a separate if() ?
re,
wh
Am 02.07.2012 12:12, schrieb Dave Airlie:
> From: Dave Airlie
>
> This fixes some really ugly code that got mangled by the indenting.
>
Hi,
Just wondering about what sort of things are considered ABI breaks in reality,
>From my understanding, I should be able to
a) add new members to end of ScreenRec.
b) add new members to xf86ScrnInfoRec in the reserved sections.
Can I though add members to GC, Pixmap, Picture structs or does
From: Dave Airlie
This adds the initial prime support for dri2 offload. The main thing is
when we get a connection from a prime client, we stored the information
and mark all drawables from that client as prime. We then create all
buffers for that drawable on the prime device dri2screen.
Then DR
1 - 100 of 134 matches
Mail list logo
