On Wed, Feb 11, 2015 at 6:42 AM, Daniel Vetter <daniel.vetter@xxxxxxxx> wrote:
> Well just core drm. All the other callers in there that still use
> direct calls to ioctl have some custom retry logic already, so should
> be good already.
>
> freedreno/kgsl ahas all the other bare ioctl calls, dunnot what to do
> about that.
fwiw, the freedreno/kgsl bare ioctls are actually against the non-drm
device file.. freedreno/msm is the stuff for the upstream drm/msm
driver.
background: downstream android kernel w/ kgsl kernel driver has this
shim drm driver which is really just for getting flink names + non drm
device for everything else. It's kinda bonghits, really just exists
to give folks stuck w/ android drivers a fighting chance. And due to
different versions of kgsl floating around on the many different
android kernel branches, odds are still pretty grim (which is why kgsl
backend is disabled by default)
I guess technically they could still use drmIoctl() but it seemed a bit strange
BR,
-R
> Signed-off-by: Daniel Vetter <daniel.vetter@xxxxxxxxx>
> ---
> xf86drm.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/xf86drm.c b/xf86drm.c
> index fb673b58afb3..263d6835c29a 100644
> --- a/xf86drm.c
> +++ b/xf86drm.c
> @@ -2543,12 +2543,12 @@ void drmCloseOnce(int fd)
>
> int drmSetMaster(int fd)
> {
> - return ioctl(fd, DRM_IOCTL_SET_MASTER, 0);
> + return drmIoctl(fd, DRM_IOCTL_SET_MASTER, 0);
> }
>
> int drmDropMaster(int fd)
> {
> - return ioctl(fd, DRM_IOCTL_DROP_MASTER, 0);
> + return drmIoctl(fd, DRM_IOCTL_DROP_MASTER, 0);
> }
>
> char *drmGetDeviceNameFromFd(int fd)
> --
> 2.1.4
>
> _______________________________________________
> dri-devel mailing list
> dri-devel@xxxxxxxxxxxxxxxxxxxxx
> http://lists.freedesktop.org/mailman/listinfo/dri-devel
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
