On Wed, May 14, 2014 at 08:51:10PM +0200, Daniel Vetter wrote:
We need to start somewhere ... With this the only places left in i915 where we use pipe integers is in the interrupt handling code. And there it actually makes some amount of sense.
Very much welcome addition. Some minor comments below.
Signed-off-by: Daniel Vetter daniel.vetter@ffwll.ch
drivers/gpu/drm/drm_irq.c | 81 ++++++++++++++++++++++++++++++++++++ drivers/gpu/drm/i915/intel_display.c | 22 +++++-----
Perhaps move the i915 changes into a separate commit?
diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
[...]
/**
- drm_crtc_vblank_get - get a reference count on vblank events
- @dev: drm device
- @crtc: which CRTC to own
- Acquire a reference count on vblank events to avoid having them disabled
- while in use.
- This is the native kms version of drm_vblank_off().
- Returns:
- Zero on success, nonzero on failure.
- */
+int drm_crtc_vblank_get(struct drm_device *dev, struct drm_crtc *crtc) +{
- return drm_vblank_get(dev, drm_crtc_index(crtc));
+} +EXPORT_SYMBOL(drm_crtc_vblank_get);
This seems slightly backwards. Since drm_vblank_get() is what's being deprecated here, wouldn't it make more sense to write drm_crtc_vblank_get() in terms of struct drm_crtc and make drm_vblank_get() call that instead? I can't seem to find a helper to get the CRTC from an index, but it seems like that wouldn't be hard to do.
I guess it doesn't matter all that much either way, though, since we could equally well make that change when drm_vblank_get() is dropped, in which case at least there's no longer a need for the reverse lookup.
I'd still prefer to have i915 changes in a separate commit, but otherwise:
Reviewed-by: Thierry Reding treding@nvidia.com