On Thu, Apr 11, 2013 at 9:15 AM, Mika Kuoppala <mika.kuopp...@linux.intel.com> wrote: > Kees Cook <keesc...@chromium.org> writes: > >> On Thu, Apr 11, 2013 at 6:22 AM, Mika Kuoppala >> <mika.kuopp...@linux.intel.com> wrote: >>> commit 647416f9eefe7699754b01b9fc82758fde83248c >>> Author: Kees Cook <keesc...@chromium.org> >>> Date: Sun Mar 10 14:10:06 2013 -0700 >>> >>> drm/i915: use simple attribute in debugfs routines >>> >>> made i915_next_seqno debugfs entry to crop it's output >>> if returned value was large enough. Using simple_attr >>> will limit the output to 24 bytes. Fix this by returning >>> only the value and nothing else. >>> >>> Cc: Kees Cook <keesc...@chromium.org> >>> Signed-off-by: Mika Kuoppala <mika.kuopp...@intel.com> >> >> Oh! Thanks for catching that. What a weird limitation. >> >> What about max freq, min freq, and wedged? Do those run the risk of >> truncation too? > > max and min freq should be safe, and wedged too on 32bit platforms. > But if gpu is declared wedged on host with 64bit atomic_t, > it will crop the output.
Should we consider proposing an increase in the size of the simple attr buffer? It seems silly to provide an arbitrary format string but leave the buffer so small. -Kees -- Kees Cook Chrome OS Security _______________________________________________ Intel-gfx mailing list Intel-gfx@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/intel-gfx
