On Thu, 11 Jan 2024, Jonathan Gray <jsg@xxxxxxxxx> wrote:
> use ATOMIC64_INIT() not ATOMIC_INIT() for atomic64_t
>
> Fixes: 3f09a0cd4ea3 ("drm: Add common fdinfo helper")
> Signed-off-by: Jonathan Gray <jsg@xxxxxxxxx>
Reviewed-by: Jani Nikula <jani.nikula@xxxxxxxxx>
> ---
> drivers/gpu/drm/drm_file.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/drm_file.c b/drivers/gpu/drm/drm_file.c
> index 446458aca8e9..d3b10dd91584 100644
> --- a/drivers/gpu/drm/drm_file.c
> +++ b/drivers/gpu/drm/drm_file.c
> @@ -149,7 +149,7 @@ bool drm_dev_needs_global_mutex(struct drm_device *dev)
> */
> struct drm_file *drm_file_alloc(struct drm_minor *minor)
> {
> - static atomic64_t ident = ATOMIC_INIT(0);
> + static atomic64_t ident = ATOMIC64_INIT(0);
I think we should probably redefine both ATOMIC_INIT and ATOMIC64_INIT
to cast the result to the correct type to avoid this problem once and
for all.
If we had
#define ATOMIC_INIT(i) (atomic_t){ (i) }
and
#define ATOMIC64_INIT(i) (atomic64_t){ (i) }
we'd get a build failure using them incorrectly.
BR,
Jani.
> struct drm_device *dev = minor->dev;
> struct drm_file *file;
> int ret;
--
Jani Nikula, Intel
