[Intel-gfx] [PATCH] drm: Use ENOENT consistently for the error return for an unmatched handle.

Chris Wilson chris at chris-wilson.co.uk
Thu Aug 5 10:25:32 CEST 2010


On Thu, 05 Aug 2010 08:46:31 +1000, Dave Airlie <airlied at redhat.com> wrote:
> Have you verified no userspace relies on this return value? since this
> technically an ABI change.
> 
> >From what I can see probably only libdrm tests care.

I haven't found any other instances of code checking return values, more
often the error code is simply reported. Even those tests show that we can
expect EINVAL, ENOENT or EBADF for an invalid buffer handle.

It's the reporting that I want clarified as the "invalid fd" is misleading
for bug reporters. (Doubly so when this gets confused with a genuine
EBADF!)
-- 
Chris Wilson, Intel Open Source Technology Centre



More information about the Intel-gfx mailing list