Seeking review of xserver patches

[Adam Jackson]

On Tue, 2016-04-05 at 17:35 +0900, Michel Dänzer wrote:
> I haven't received any feedback on the patches listed below. Please let
> me know if you want me to look at any of your patches in return.
> 
> 
> os: Use strtok instead of xstrtokenize in ComputeLocalClient
> https://patchwork.freedesktop.org/patch/78645/
> 
> Fixes a memory leak every time a client connects.

r-b. Seems like xstrtokenize might want a matching free API too, all
the callers are rolling their own.

> The following two patches make the xf86Crtc field
> driverIsPerformingTransform actually usable. Once these land, the
> xf86-video-{ati,amdgpu} drivers can use the HW cursor with rotation and
> TearFree enabled.
> 
> xfree86/modes: Fix HW cursor clipping for driverIsPerformingTransform (v2)
> https://patchwork.freedesktop.org/patch/78321/

r-b.

> xfree86/modes: Disambiguate driverIsPerformingTransform
> https://patchwork.freedesktop.org/patch/78319/

r-b if you also fix the omap driver (I'm assuming you have patches
ready to go for ati and amdgpu).

> xfree86/modes: Simplify in_range logic in xf86_crtc_set_cursor_position
> https://patchwork.freedesktop.org/patch/78318/
> 
> Minor simplification on top of the previous two patches.

r-b.

> xfree86/modes: Remove xf86_reload_cursors
> https://patchwork.freedesktop.org/patch/78320/
> 
> No longer needed in favour of xf86CursorResetCursor.

r-b, though if you're not willing to patch up the other drivers
(amdgpu, armsoc, ati, freedreno, geode, nouveau, qxl, and intel and
vmware which I think both prefer patches than drivebys) then please add
#define xf86_reload_cursors(s) do {} while (0) in the appropriate
place.

In exchange, either of these two series could use some love:

https://patchwork.freedesktop.org/patch/72365/
https://patchwork.freedesktop.org/patch/72364/
https://patchwork.freedesktop.org/patch/72366/

Fixes how the resource extension accounts pixmap bytes, it was both
over-accounting for redirected windows and not accounting for GLXPixmap
references.

Alternatively, the resend I just did to remove the xf86CreateRootWindow special case (and incidentally fix a memory leak):

https://lists.freedesktop.org/archives/xorg-devel/2016-April/049289.htm
l

- ajax