On Tue, 11 Feb 2020, Daniel Vetter <daniel@xxxxxxxx> wrote: > On Tue, Feb 11, 2020 at 04:47:53PM +0200, Jani Nikula wrote: >> There is no real reason to require drivers to set and use >> dev->dev_private. Indeed, the current recommendation, as documented in >> drm_device.h, is to embed struct drm_device in the per-device struct >> instead of using dev_private. >> >> Remove the requirement for dev_private to have been set to indicate >> driver initialization. > > Yeah this is nonsense. Also, drm_irq_install is purely optional > semi-midlayer (it's not really a midlayer for the legacy drivers, but > whatever, who cares about those). > > Now there might be some hilarious races this papers over, but: > > - Proper drivers should only call drm_dev_register once everything is set > up, including this stuff here. No race possible with anything else > really. > > - Slightly more wobbly drivers, including the legacy ones, all use > drm_global_mutex. This was the former BKL, which means that it was > impossible for soeone to go through the load/unload/reload (between > lastclose and firstopen) paths and also run the ioctl. But the ioctl had > to be made unlocked because blocking there killed X: > > commit 8f4ff2b06afcd6f151868474a432c603057eaf56 > Author: Ilija Hadzic <ihadzic@xxxxxxxxxxxxxxxxxxxxxx> > Date: Mon Oct 31 17:46:18 2011 -0400 > > drm: do not sleep on vblank while holding a mutex > > The even more legacy DRM_CONTROL ioctl stayed fully locked. But the file > open/close paths are still fully locked, and that's the only place > legacy drivers should call drm_irq_install/uninstall, so should all > still be fully ordered and protected and happy. > > Feel free to quote or not quote the above in the commit message. > >> Cc: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> >> Cc: Daniel Vetter <daniel@xxxxxxxx> >> Signed-off-by: Jani Nikula <jani.nikula@xxxxxxxxx> >> >> --- >> >> Any ideas for something else drm_irq_install() could/should check to >> ensure "Driver must have been initialized"? >> >> There are only a few instances of dev_private uses in i915, also to be >> removed, and we could stop initializing dev_private altogether. We could >> in fact do that without this patch too, as we don't use >> drm_irq_install(). But it would be cleaner to not have any checks for >> driver private stuff outside of drivers. > > I hope my review above answers your question here. Patch, as-is: > > Reviewed-by: Daniel Vetter <daniel.vetter@xxxxxxxx> Many thanks, pushed to drm-misc-next with the details addded to commit message. BR, Jani. > >> --- >> drivers/gpu/drm/drm_irq.c | 4 ---- >> 1 file changed, 4 deletions(-) >> >> diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c >> index 03bce566a8c3..588be45abd7a 100644 >> --- a/drivers/gpu/drm/drm_irq.c >> +++ b/drivers/gpu/drm/drm_irq.c >> @@ -111,10 +111,6 @@ int drm_irq_install(struct drm_device *dev, int irq) >> if (irq == 0) >> return -EINVAL; >> >> - /* Driver must have been initialized */ >> - if (!dev->dev_private) >> - return -EINVAL; >> - >> if (dev->irq_enabled) >> return -EBUSY; >> dev->irq_enabled = true; >> -- >> 2.20.1 >> -- Jani Nikula, Intel Open Source Graphics Center _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/dri-devel
