drm/fb-helper: don't sleep for screen unblank when an oopps is in progress
authorDaniel Vetter <daniel.vetter@ffwll.ch>
Thu, 24 Jan 2013 15:42:07 +0000 (16:42 +0100)
committerDaniel Vetter <daniel.vetter@ffwll.ch>
Wed, 13 Feb 2013 23:08:33 +0000 (00:08 +0100)
commit1b1d5397058f06bc5bd87d43ed93f34b28546ea4
tree8202c1fb814b335e319bd2e5a048db0d80be6f66
parent207fd32970b1def91b11ae28f6bebffc792db714
drm/fb-helper: don't sleep for screen unblank when an oopps is in progress

Otherwise the system will burn even brighter and worse, leave the user
wondering what's going on exactly.

Since we already have a panic handler which will (try) to restore the
entire fbdev console mode, we can just bail out. Inspired by a patch
from Konstantin Khlebnikov. The callchain leading to this, cut&pasted
from Konstantin's original patch:

callstack:
panic()
bust_spinlocks(1)
unblank_screen()
vc->vc_sw->con_blank()
fbcon_blank()
fb_blank()
info->fbops->fb_blank()
drm_fb_helper_blank()
drm_fb_helper_dpms()
drm_modeset_lock_all()
mutex_lock(&dev->mode_config.mutex)

Note that the entire locking in the fb helper around panic/sysrq and
kdbg is ... non-existant. So we have a decent change of blowing up
everything. But since reworking this ties in with funny concepts like
the fbdev notifier chain or the impressive things which happen around
console_lock while oopsing, I'll leave that as an exercise for braver
souls than me.

v2: Drop the -EBUSY return value I've copied, we don't need it since
the we'll take care of things ourselves anyway.

Cc: Konstantin Khlebnikov <khlebnikov@openvz.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
References: https://patchwork.kernel.org/patch/1878181/
Reviewed-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drivers/gpu/drm/drm_fb_helper.c