Tests for RRCreateLease and RRFreeLease
Signed-off-by: Keith Packard <kei...@keithp.com>
---
configure.ac |8 +
test/Makefile.am |2 +-
test/randr/Makefile.am | 14 +
test/randr/randr_lease.c | 1066 ++
4 files c
as part of the default configuration.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/modes/xf86Crtc.c | 4 ++--
hw/xfree86/modes/xf86Crtc.h | 5 +
2 files changed, 7 insertions(+), 2 deletions(-)
diff --git a/hw/xfree86/modes/xf86Crtc.c b/hw/xfree86/modes/xf86Crtc.c
Here's the latest version of the DRM lease and non-desktop output code
for the X server. This time, I've gone ahead and implemented a pile of
tests for leases.
Changes since the last series:
* Merged non-desktop and lease support into a single patch
sequence. They have a dependency on
ned-off-by: Keith Packard <kei...@keithp.com>
---
configure.ac | 2 +-
hw/xfree86/drivers/modesetting/drmmode_display.c | 128 -
hw/xfree86/drivers/modesetting/drmmode_display.h | 4 +
hw/xfree86/modes/xf86Crtc.c | 24 +-
h
RRChangeOutputProperty and RRConfigureOutputProperty should not modify
their parameters, and callers may want to pass pointers to fixed data,
so declare the value pointers as const in both cases.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
randr/randrstr.h | 4 ++--
randr/rrprop
Save any value of the kernel non-desktop property in the xf86Output
structure to avoid non-desktop outputs in the default configuration.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modesetting/drmmode_display.c | 6 ++
1 file changed, 6 insertions(+)
diff
This makes sure the CRTC's cursor is hidden before we hand the CRTC
over to some other application.
Signed-off-by: Keith Packard <kei...@keithp.com>
Reviewed-by: Michel Dänzer <michel.daen...@amd.com>
Reviewed-by: Alex Deucher <alexander.deuc...@amd.com>
---
hw/xfree86/modes/
changes to DIX so it can track things as well.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modesetting/drmmode_display.c | 111 ++-
1 file changed, 86 insertions(+), 25 deletions(-)
diff --git a/hw/xfree86/drivers/modesetting/drmmode_display.c
This lets a DRM client map between X outputs and kernel connectors.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modesetting/drmmode_display.c | 24
1 file changed, 24 insertions(+)
diff --git a/hw/xfree86/drivers/modesetting/drmmode_dis
-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modesetting/vblank.c | 50 -
1 file changed, 49 insertions(+), 1 deletion(-)
diff --git a/hw/xfree86/drivers/modesetting/vblank.c
b/hw/xfree86/drivers/modesetting/vblank.c
index fe046e48b..f7d
Tracks changes to the non-desktop property so that when non-zero,
outputs will always appear to be disconnected.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
randr/randrstr.h | 3 +++
randr/rroutput.c | 26 +-
randr/rrproperty.
Alan Coopersmith writes:
> Honestly, I think the versions of Java I've used for generating the docs
> just uses fontconfig to find local fonts, but we do ship with a bunch of
> commercially licensed fonts from way back when that should be automatically
> added to
Eric Anholt writes:
> I like that it's just commits and merges instead of filter-branch.
Yeah, it makes the history really easy to see, and (if we don't move
files), the per-file history easy to track.
> I think it would make more sense if the headers matched the structure
The idea of merging all of our X protocol repositories came up again,
and Adam Jackson asked if I wanted to wait for this before merging my
randr lease/non-desktop changes.
Not really? But, I'm still interested in a merge of the protocol
repositories; we've got way too many, and they're mostly
Adam Jackson writes:
> This has been "deprecated" since 2011, but because it is still
> referenced from XORG_DEFAULT_OPTIONS nothing has ever been updated to
> get strict aliasing right. Let's fix that.
I don't understand this -- are you getting rid of this option from our
Adam Jackson <a...@nwnk.net> writes:
> On Thu, 2017-12-07 at 16:57 -0800, Keith Packard wrote:
>
>> @@ -186,14 +186,23 @@ from the CRTCs. The Monitor marked as Primary will be
>> listed first.
>>
>> 1.6. Introduction to version 1.6 of the extension
>&g
Alan Hourihane writes:
> But I've gone ahead and tested your fix, and it works too. So I'm fine
> with this.
Awesome, thanks for your review!
--
-keith
signature.asc
Description: PGP signature
___
xorg-devel@lists.x.org:
This makes sure the CRTC's cursor is hidden before we hand the CRTC
over to some other application.
Signed-off-by: Keith Packard <kei...@keithp.com>
Reviewed-by: Michel Dänzer <michel.daen...@amd.com>
Reviewed-by: Alex Deucher <alexander.deuc...@amd.com>
---
hw/xfree86/modes/
This lets a DRM client map between X outputs and kernel connectors.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modesetting/drmmode_display.c | 24
1 file changed, 24 insertions(+)
diff --git a/hw/xfree86/drivers/modesetting/drmmode_dis
-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modesetting/vblank.c | 50 -
1 file changed, 49 insertions(+), 1 deletion(-)
diff --git a/hw/xfree86/drivers/modesetting/vblank.c
b/hw/xfree86/drivers/modesetting/vblank.c
index fe046e48b..f7d
ain.
This required splitting the DIX level lease termination code
into two pieces, one to remove the lease from the system
(RRLeaseTerminated) and a new function that frees the lease
data structure (RRLeaseFree).
Signed-off-by: Keith Packard <kei...@keithp.com>
---
con
This is an updated version of this series, rebased on current master
and including v3 of the main patch. That final patch has a description
of the changes from v2, but the main thing is a lot of changes needed
to keep the server from doing anything with the leased objects or
including them in any
Keith Packard <kei...@keithp.com> writes:
> We were updating the link-status property when a uevent came in, but
> we also want to update the non-desktop property, and potentially
> others as well. This patch just updates every property provided by the
> kernel, sending change
changes to DIX so it can track things as well.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modesetting/drmmode_display.c | 115 ++-
1 file changed, 90 insertions(+), 25 deletions(-)
diff --git a/hw/xfree86/drivers/modesetting/drmmode_display.c
Save any value of the kernel non-desktop property in the xf86Output
structure to avoid non-desktop outputs in the default configuration.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modesetting/drmmode_display.c | 6 ++
1 file changed, 6 insertions(+)
diff
Non-desktop monitors are described in RandR protocol version 1.6 to be
those which "shouldn't" normally be part of a desktop
environment. Monitors like head-mounted displays or the Apple Touch
Bar.
Linux detects these using EDID quirks and sets a connector
property. We reflect that property up to
Tracks changes to the non-desktop property so that when non-zero,
outputs will always appear to be disconnected.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
randr/randrstr.h | 3 +++
randr/rroutput.c | 26 +-
randr/rrproperty.
as part of the default configuration.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/modes/xf86Crtc.c | 4 ++--
hw/xfree86/modes/xf86Crtc.h | 5 +
2 files changed, 7 insertions(+), 2 deletions(-)
diff --git a/hw/xfree86/modes/xf86Crtc.c b/hw/xfree86/modes/xf86Crtc.c
RRChangeOutputProperty and RRConfigureOutputProperty should not modify
their parameters, and callers may want to pass pointers to fixed data,
so declare the value pointers as const in both cases.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
randr/randrstr.h | 4 ++--
randr/rrprop
We were updating the link-status property when a uevent came in, but
we also want to update the non-desktop property, and potentially
others as well. This patch just updates every property provided by the
kernel, sending changes to DIX so it can track things as well.
Signed-off-by: Keith Packard
Delete output grabs
Add LeaseNotify events
Add FreeLease with option to terminate
Signed-off-by: Keith Packard <kei...@keithp.com>
---
configure.ac | 2 +-
randr.h| 15 +++--
randrproto.h | 57 ++
ran
Here's a two patch series which adds leases and support for
'non-desktop' outputs.
Leases are a way to take a set of X server resources and give them
over to another application for a while. The X server implementation
for this uses Linux changes already heading upstream.
non-desktop outputs
this information and have it reflected to X applications
through this extension.
v2: fix puncutation and duplicated 'the'.
v3: switch to 32-bit property named non-desktop to match Linux
Signed-off-by: Keith Packard <kei...@keithp.com>
---
randr.h| 1 +
randrproto.tx
Adam Jackson writes:
> Also, git://people.freedesktop.org/~keithp/newproto appears to contain
> the script used to generate the merged repo.
Right, that's probably more useful today. The trick was to get the
headers merged without losing any of the history.
> I would be entirely
Daniel Martin writes:
> Hi,
>
> I've ever wondered why are the proto headers split up into distinct
> repos? (It takes "ages" to just copy (install) a few files with
> autotools.)
They were split as part of the great dis-aggregation. I had a prototype
of them merged
receiving events from other clients would
> be required.
>
> v2:
> * Also restrict events sent to additional windows to the presenting
> client
> * Don't shorten line lengths
Looks good to me.
Reviewed-by: Keith Packard <kei...@keithp.com>
--
Adam Jackson writes:
> You added that comment when you added XFIXES to XFree86, presumably you
> know what it means better than I do.
Never mind -- the original version of the code had the assignment to the
local DisplayCursor variable below the comment while you've moved it up
Adam Jackson writes:
> +/*
> + * Not a simple Unwrap/Wrap as this isn't called along the DisplayCursor
> + * wrapper chain.
> + */
> +pScreen->DisplayCursor = as->DisplayCursor;
> +(void) (*pScreen->DisplayCursor) (dev, pScreen, ac->elts[elt].pCursor);
>
Michel Dänzer writes:
> Turns out it's way worse than that. The PresentPixmap and
> PresentNotifyMSC requests don't take a PRESENTEVENTID parameter.
> Instead, present_send_idle/complete_notify send every event to every
> client which called PresentSelectInput for the window
instead of 'seq'. That only happens in
an error case, which probably means approximately never.
For your patch:
Reviewed-by: Keith Packard <kei...@keithp.com>
--
-keith
signature.asc
Description: PGP signature
___
xorg-devel@lists.x.org: X.Org devel
Michel Dänzer writes:
> Looking at the Mesa code, it uses xcb_generate_id to generate the event
> ID. AFAICT the XCB code is properly checking and only queuing special
> events with a matching event ID, so the only possible explanation seems
> that xcb_generate_id ends up
"Keith Packard" <kei...@keithp.com> writes:
> I'm *not* saying this isn't a good patch in practice, I just want to
> understand whether the system was designed wrong, or if we've
> implemented it wrong.
I spent some time this afternoon on IRC chatting with Pierre-Loup.
Giuseppe Bilotta writes:
> Potentially stupid question, but any chances that some “higher level”
> clients (window manager, screensaver) might be interested in the other
> clients' notifications?
Well, Present isn't the only way to perform this operation, so anything
Giuseppe Bilotta writes:
> (Alternatively, one could lease that output and then maybe start an X
> server just for it, but that seems a bit … overkill?)
Oh, that's a completely plausible plan. Yes, that would work and would
offer a completely separate environment for
Giuseppe Bilotta writes:
> (Still, I've been fancying the idea of dynamic X Screens for a while
> now, and the use case I mentioned seemed to fit the bill pretty
> nicely.)
Yes it does. When you have two or three years to spare, let us know!
--
-keith
Michel Dänzer writes:
> From: Michel Dänzer
>
> We were sending the events to all clients listening for them on the
> window. But clients can get confused by events from another client, and
> I can't imagine any case where reciving events from other
Aaron Plattner writes:
> I think this makes sense. Comments below.
>
> Reviewed-by: Aaron Plattner
Thanks, Aaron. Have you also looked at the leasing changes on the same
branch? I'd like to know what you think of that plan for implementing
the Vulkan
Giuseppe Bilotta writes:
> (I actually wish I had some free time to put my coding hands where my
> mouth is 8-P).
I'd say it's more about waiting until we find a situation that actually
requires the additional work. Right now, the goal is to make a leased
output
Giuseppe Bilotta writes:
> FWIW, some Razer laptop models have a touchpad that doubles as a
> secondary display. Not sure if it's worth mentioning (it does predate
> the Apple Touch Bar, has some very patchy support under Linux, none that
> I know of under Xorg).
this information and have it reflected to X applications
through this extension.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
randrproto.txt | 84 +++---
1 file changed, 81 insertions(+), 3 deletions(-)
diff --git a/randrproto
I know, Present was supposed to support all of the GL vblank modes, and
it mostly does, except for the most common case...
In any case, here's a summary of four presentation modes from Vulkan,
how you'd get that with GL and Present and what they mean:
Vulkan GL
Andy Ritger writes:
> If the NVIDIA X driver finds an HMD display, it:
>
> (a) Marks it as disconnected.
> (b) Does not make its EDID available to RandR clients.
>
> So, unless I'm mistaken, RandR clients will see the HMD as an RandR
> output. But, perhaps the problem is
James Jones writes:
> However, I think direct enumeration is generally useful functionality,
> if nothing else just for things like vkinfo or DRM equivalents.
Sure, I've enabled direct enumeration when the process has access to
that information from the kernel. Right now,
Dave Airlie writes:
> If they use acquire extensions they get display enumeration, the
> problem is display enumeration before the acquire would need special
> permission elevation.
And the problem here is that the nVidia driver hides the HMD displays
from X, forcing the
James Jones writes:
> Understood. In that case, no displays should be enumerated on the
> "rendering hardware" device in Vulkan, unless the driver vendor opts for
> heroics.
We try to avoid software heroics where possible.
> Right, but normal processes don't need the
James Jones writes:
> I think at a lower level, we have different views of how
> vkGetPhysicalDeviceDisplayPropertiesKHR/VK_KHR_display works.
> vkGetPhysicalDeviceDisplayPropertiesKHR() is suppose to enumerate all
> the displays on a given device.
In many devices, the
"Uecker, Martin" writes:
> The question is what security risks all the different extensions
> may expose.
For things like Render and Present, they're no more serious than the
existing core protocol. I'm pretty sure we could list the extensions
which provide
Alex Goins writes:
> Thanks, Adam.
>
> Here's an updated version of the spec:
This is looking very good. I don't have any architectural concerns at
this point, just some editorial comments.
> Rendering to DeepColor windows using the core protocol, however, is loosely
>
I've implemented this extension in DRM and have run into a conflict
between the spec and the Linux architecture.
The VkDisplayKHR parameter for VK_EXT_acquire_xlib_display can be
found in two different ways:
1) Enumerate displays using vkGetPhysicalDeviceDisplayPropertiesKHR
2) Enumerate
he necessary
functionality. I assume someone might review the Solaris patch, but is
there anyone who can review the AIX patch?
For the pair:
Acked-by: Keith Packard <kei...@keithp.com>
--
-keith
signature.asc
Description: PGP signature
___
xorg-devel@
se to me. No other concerns from me, this is
Reviewed-by: Keith Packard <kei...@keithp.com>
--
-keith
signature.asc
Description: PGP signature
___
xorg-devel@lists.x.org: X.Org development
Archives: http://lists.x.org/archives/xorg-devel
Eric Anholt writes:
> +/* Include the enums here for the moment, to keep from needing to bump
> epoxy. */
> +#ifndef GL_TILE_RASTER_ORDER_FIXED_MESA
> +#define GL_TILE_RASTER_ORDER_FIXED_MESA 0x8BB8
> +#define GL_TILE_RASTER_ORDER_INCREASING_X_MESA 0x8BB9
> +#define
Michel Dänzer writes:
> No, but mode changes are also irrelevant for whether or not page
> flipping can be used.
What about changing the size of the CRTC that the application is
displayed on? From fitting the screen to not fitting seems like it would
make page flipping no
Michel Dänzer writes:
> Or maybe it would be sufficient for clients to simply re-query the
> modifiers in response to PresentConfigureNotify events?
Would that also get sent when modes are changed?
--
-keith
signature.asc
Description: PGP signature
Here's a second complete series with two issues addressed:
1) Missing '|' when computing old ioctl type param
2) Moved a bit of code from the first patch to the second (thanks Michel
Dänzer)
Daniel Vetter asked if I might demonstrate the new kernel vblank API
I've proposed by implementing
This provides an API wrapper around the kernel interface for queueing
a vblank event, simplifying all of the callers.
v2: Fix missing '|' in computing vbl.request.type
v3: Remove spurious bit of next patch (thanks, Michel Dänzer)
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/x
drmCrtcGetSequence returns the current vblank sequence and time.
drmCrtcQueueSequence queues an event for delivery at a specified
vblank sequence.
Use these (when available) in preference to drmWaitVBlank.
v2: Move bit of code from first patch (thanks, Michel Dänzer)
Signed-off-by: Keith
Michel Dänzer writes:
> The DRM_IOCTL_CRTC_QUEUE_SEQUENCE part belongs in patch 2.
Yeah, it's harmless (aside from a compiler warning) in patch 1, but of
course it belongs in the second patch. Thanks!
--
-keith
signature.asc
Description: PGP signature
This provides an API wrapper around the kernel interface for queueing
a vblank event, simplifying all of the callers.
v2: Fix missing '|' in computing vbl.request.type
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modesetting/dri2.c
drmCrtcGetSequence returns the current vblank sequence and time.
drmCrtcQueueSequence queues an event for delivery at a specified
vblank sequence.
Use these (when available) in preference to drmWaitVBlank.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modes
This provides an API wrapper around the kernel interface for queueing
a vblank event, simplifying all of the callers.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modesetting/dri2.c| 74 +++-
hw/xfree86/drivers/modesetting/dr
Daniel Vetter asked if I might demonstrate the new kernel vblank API
I've proposed by implementing support in the modesetting driver. It
turned out to be a good idea because the easiest way was to wrap
drmWaitVBlank in a new function that provided a simpler API and then
go add support for the new
Manasi Navare writes:
> Thanks for the patch. Does this take care of the cases where
> MST monitors might get unplugged, but not generate the
> hotplug event?
Nope, this is strictly for devices which have disappeared from the
kernel API, but which the X server is
Adam Jackson writes:
>> Martin Peres (1):
>> modesetting: re-set the crtc's mode when link-status goes BAD
>
> Keith had part of this change disabled locally at XDC, apparently he'd
> seen crashes from it. Hopefully we can figure that one out relatively
> quickly...
Outputs may have NULL mode_output (connector) pointers if the
connector disappears while the server is running. Skip these when
resetting outputs with BAD link status.
Signed-off-by: Keith Packard <kei...@keithp.com>
---
hw/xfree86/drivers/modesetting/drmmode_display.c | 5 -
1 file c
Aaron Plattner writes:
>>> 3. Colorspace/Encoding Window Properties
>>>
>>> Windows with DeepColor visuals must rely on window properties, as opposed to
>>> colormaps, to determine the relationship between pixel values and colors.
>>> These
>>> properties must specify
n VT switch, while the root window itself may
> not contain a valid image we won't try to touch it, but GetImage from a
> redirected window will now work even when switched away.
>
> Signed-off-by: Adam Jackson <a...@redhat.com>
Acked-by: Keith Packard <kei...@
Adam Jackson writes:
> The reason I suggested moving it to exa is that it's the one remaining
> in-tree renderer that has that problem.
Yeah, I think that should be sufficient then.
--
-keith
signature.asc
Description: PGP signature
Emmanuel Gil Peyrot writes:
> This effectively fixes screenshotting in Xwayland without GLAMOR, which
> was previously always hitting that check since the borderClip was never
> set to a non-zero area.
We chatted about this issue on IRC this morning. The essential
Alex Goins writes:
> The DeepColor Extension provides a new visual class, DeepColor, designed for
> handling visuals that are incompatible with the existing core X visual
> classes:
> StaticGray, StaticColor, TrueColor, GrayScale, PseudoColor, or
> DirectColor.
It seems like
to prevent
> looping.
> No animation played after such non-looped animation played once.
Reviewed-by: Keith Packard <kei...@keithp.com>
--
-keith
signature.asc
Description: PGP signature
___
xorg-devel@lists.x.org: X.Org development
Arch
is that the case
here? I'm just trying to make sure we don't have pixmaps larger than
32767 sneaking through somehow, which would be bad in all kinds of
places.
In any case, we still need this check to avoid the problem:
Reviewed-by: Keith Packard <kei...@keithp.com>
(someday we may ju
Michel Dänzer writes:
> Is assigning an unsigned value with the MSB set to a signed variable
> well-defined in C?
I have no idea. And I just spent a few hours wading through the N1570
draft of the C standard on a related issue. In particular, this is
worrying:
6.3.1.3
3
e -fwrapv
whenever using gcc so that you get correct behavior, instead of
compiler-writers-potluck. I hope the gcc authors agree with your reading
of the spec in this case; it's getting very hard to trust them to ever
do the right thing though.
Reviewed-by: Keith Packard <kei...@keithp.com>
--
Adam Jackson writes:
> Consistency and cleanliness? They're the only out-of-lined reply swap
> functions which puts them solidly in the minority (seven out of GLX's
> 32ish replies), and glxserver.h is currently kind of a misc-and-other
> header that I was trying to give some
Adam Jackson <a...@redhat.com> writes:
> Signed-off-by: Adam Jackson <a...@redhat.com>
Reviewed-by: Keith Packard <kei...@keithp.com>
--
-keith
signature.asc
Description: PGP signature
___
xorg-devel@lists.x.org: X.Org devel
Adam Jackson <a...@redhat.com> writes:
> Those are xlib spellings, we say TRUE/FALSE pretty consistently
> elsewhere in the server.
>
> Signed-off-by: Adam Jackson <a...@redhat.com>
Reviewed-by: Keith Packard <kei...@keithp.com>
--
-keith
signature.
Peter Harris writes:
> HAVE_OSPOLL is only defined inside this file, not by autoconf/meson, so
> it's always defined to 1.
>
> The rest of os/ospoll.c uses #if instead of #ifdef everywhere. I'd be
> inclined to update all of them if I were to paint that particular
>
Peter Harris writes:
> The epoll code depends on epoll_create1, not epoll_create.
>
> Signed-off-by: Peter Harris
Thanks!
> -#if !HAVE_OSPOLL && HAVE_EPOLL_CREATE1
> +#if !HAVE_OSPOLL && defined(HAVE_EPOLL_CREATE1)
Should be using defined for
Olivier Fourdan writes:
> @@ -722,11 +723,11 @@ glamor_compute_transform_clipped_regions(PixmapPtr
> pixmap,
> temp_box.x2 = MIN(temp_box.x2, pixmap->drawable.width);
> temp_box.y2 = MIN(temp_box.y2, pixmap->drawable.height);
> }
> -/* Now copy
sing CARD64 to try to actually store a
> uint64_t. Now that stdint.h exists, let's just use that here,
> instead.
>
> v2: Fix alarm delta changes.
5.5/7 and 6 of 7 are
Reviewed-by: Keith Packard <kei...@keithp.com>
--
-keith
signature.asc
Description: PGP signature
Eric Anholt <e...@anholt.net> writes:
> Keith Packard <kei...@keithp.com> writes:
>
>> [ Unknown signature status ]
>> Eric Anholt <e...@anholt.net> writes:
>>
>>> The extension was using the name CARD64 to represent 64-bit values,
>
Eric Anholt writes:
> So I GetInputFocus, force the reply, free the reply, then poll for
> events in a little loop that's basically "if anything happened,
> exit(1)"?
Yeah, you can be assured that any events generated by the previous
requests will have been queued when the
s (large enough to cover rounded window corners, but probably
> not xeyes) to avoid overhead on desktop GL.
>
> No performance difference on i965 with x11perf -rect1 -repeat 1 -reps
> 1 (n=50)
>
> v2: Clamp rectangle bounds addition.
For the series:
Rev
eah, for your CARD64 changes, these tests seem sufficient. Of course,
it'd be great to add tests for event generation and blocking clients at
some point too.
Reviewed-by: Keith Packard <kei...@keithp.com>
--
-keith
signature.asc
Description: PGP signature
Eric Anholt writes:
> I want to be able to call client tests with simple-xinit, so assertion
> failures should be an error.
> +if (WIFSIGNALED(wstatus))
> +return 1;
> +
> +if (WCOREDUMP(wstatus))
> +return 1;
> +
> +
n.sh \
> scripts/run-piglit.sh \
> - ddxstubs.c \
> $(NULL)
Nice. All we did was ship the file.
Reviewed-by: Keith Packard <kei...@keithp.com>
--
-keith
signature.asc
Description: PGP signature
___
xorg-devel@lists.x.org:
Eric Anholt writes:
> I couldn't find any, and I was modifying the implementation, so I had
> to write some. I would like the test to end with a "make sure there
> weren't any stray unchecked errors", but I didn't figure out how to do
> that.
Xlib used XSync for this, which
Eric Anholt <e...@anholt.net> writes:
> Signed-off-by: Eric Anholt <e...@anholt.net>
Reviewed-by: Keith Packard <kei...@keithp.com>
--
-keith
signature.asc
Description: PGP signature
___
xorg-devel@lists.x.org: X.Org devel
Eric Anholt writes:
> The extension was using the name CARD64 to represent 64-bit values,
> with a #define from CARD64 to XSyncValue, a struct with a pair of
> 32-bit values representing a signed 64-bit value. Now that stdint.h
> exists, let's just use that instead.
Really,
201 - 300 of 4634 matches
Mail list logo