diff options
author | David S. Miller <davem@sunset.davemloft.net> | 2007-07-27 22:31:46 -0700 |
---|---|---|
committer | David S. Miller <davem@sunset.davemloft.net> | 2007-07-30 00:27:33 -0700 |
commit | c7f439b99efbea74c70a5531f92566db5a6731f2 (patch) | |
tree | 7053ceffa23d54670862e14d0bc2eec9d5d42427 /drivers/video/riva/riva_hw.c | |
parent | a0afaa6ab12cf696d170c22a8fdfd88c3e33555c (diff) | |
download | kernel_goldelico_gta04-c7f439b99efbea74c70a5531f92566db5a6731f2.zip kernel_goldelico_gta04-c7f439b99efbea74c70a5531f92566db5a6731f2.tar.gz kernel_goldelico_gta04-c7f439b99efbea74c70a5531f92566db5a6731f2.tar.bz2 |
[VIDEO]: Fix OOPS in all SBUS framebuffer drivers.
All of these drivers use a silly:
struct all_info {
struct fb_info info;
struct foo_par par;
};
struct all_info *all = kzalloc(sizeof(*all), GFP_KERNEL);
all->info.par = &all->par;
etc. etc. code sequence, basically replicating the provided
framebuffer_alloc()/framebuffer_release(), and doing it badly.
Not only is this massive code duplication, it also caused a
bug in that we weren't setting the fb_info->device pointer
which results in an OOPS when fb_is_primary_device() runs.
Fix all of this by using framebuffer_{alloc,release}() and
passing in "&of_device->dev" as the device pointer.
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'drivers/video/riva/riva_hw.c')
0 files changed, 0 insertions, 0 deletions