On Wed, 2024-10-16 at 15:47 +0200, Boris Brezillon wrote:
> On Wed, 16 Oct 2024 15:16:22 +0200
> Erik Faye-Lund <erik.faye-lund@xxxxxxxxxxxxx> wrote:
>
> > On Thu, 2024-02-29 at 17:22 +0100, Boris Brezillon wrote:
> > > +/**
> > > + * enum drm_panthor_sync_op_flags - Synchronization operation
> > > flags.
> > > + */
> > > +enum drm_panthor_sync_op_flags {
> > > + /** @DRM_PANTHOR_SYNC_OP_HANDLE_TYPE_MASK:
> > > Synchronization
> > > handle type mask. */
> > > + DRM_PANTHOR_SYNC_OP_HANDLE_TYPE_MASK = 0xff,
> > > +
> > > + /** @DRM_PANTHOR_SYNC_OP_HANDLE_TYPE_SYNCOBJ:
> > > Synchronization object type. */
> > > + DRM_PANTHOR_SYNC_OP_HANDLE_TYPE_SYNCOBJ = 0,
> > > +
> > > + /**
> > > + * @DRM_PANTHOR_SYNC_OP_HANDLE_TYPE_TIMELINE_SYNCOBJ:
> > > Timeline synchronization
> > > + * object type.
> > > + */
> > > + DRM_PANTHOR_SYNC_OP_HANDLE_TYPE_TIMELINE_SYNCOBJ = 1,
> > > +
> > > + /** @DRM_PANTHOR_SYNC_OP_WAIT: Wait operation. */
> > > + DRM_PANTHOR_SYNC_OP_WAIT = 0 << 31,
> > > +
> > > + /** @DRM_PANTHOR_SYNC_OP_SIGNAL: Signal operation. */
> > > + DRM_PANTHOR_SYNC_OP_SIGNAL = (int)(1u << 31),
> >
> > Why do we cast to int here? 1u << 31 doesn't fit in a 32-bit signed
> > integer, so isn't this undefined behavior in C?
> >
> > I'm asking, because Coverity complains about overflows when we
> > assign
> > the value to drm_panthor_sync_op::flags in Mesa, and looking at
> > it...
> > it seems to be right in complaining!
> >
> > For reference, here's the Coverity issue (requires login,
> > unfortunately):
> > https://scan5.scan.coverity.com/#/project-view/59795/10037?selectedIssue=1605083
> >
>
> The cast was originally suggested here [1]. If we don't want to rely
> on gcc/llvm implementation of this UB, I guess the only option is to
> redefine those enums as #defines.
>
> [1]
> https://lore.kernel.org/dri-devel/89be8f8f-7c4e-4efd-0b7b-c30bcfbf1d23@xxxxxxx
> /
>
I think we could either do the defines, or use "-1 << 31", with a
comment to explain what's going on. Either works for me.
