[Mesa-dev] [PATCH 1/3] anv/device: Don't even try to map memory with a size of 0

[Nanley Chery]

On Tue, Nov 08, 2016 at 01:50:01PM -0800, Jason Ekstrand wrote:
> On Tue, Nov 8, 2016 at 1:46 PM, Nanley Chery <nanleychery at gmail.com> wrote:
> 
> > On Mon, Nov 07, 2016 at 05:28:12PM -0800, Jason Ekstrand wrote:
> > > Signed-off-by: Jason Ekstrand <jason at jlekstrand.net>
> > > Cc: "12.0 13.0" <mesa-dev at lists.freedesktop.org>
> > > ---
> > >  src/intel/vulkan/anv_device.c | 5 +++++
> > >  1 file changed, 5 insertions(+)
> > >
> > > diff --git a/src/intel/vulkan/anv_device.c
> > b/src/intel/vulkan/anv_device.c
> > > index 5393144..8055893 100644
> > > --- a/src/intel/vulkan/anv_device.c
> > > +++ b/src/intel/vulkan/anv_device.c
> > > @@ -1258,6 +1258,11 @@ VkResult anv_MapMemory(
> > >     if (size == VK_WHOLE_SIZE)
> > >        size = mem->bo.size - offset;
> > >
> > > +   if (size == 0) {
> >
> > The user isn't allowed to make such a call. Does this fix a CTS test?
> >
> 
> Heh, so they aren't.  It doesn't fix anything, it just ensures that you
> never hit the ioctl with a size of zero.  How about I replace it with an
> assert?
> 

An assert or no assert is fine. The validation layers technically should
catch this for us.

> 
> > > +      *ppData = NULL;
> > > +      return VK_SUCCESS;
> > > +   }
> > > +
> > >     /* FIXME: Is this supposed to be thread safe? Since vkUnmapMemory()
> > only
> > >      * takes a VkDeviceMemory pointer, it seems like only one map of the
> > memory
> > >      * at a time is valid. We could just mmap up front and return an
> > offset
> > > --
> > > 2.5.0.400.gff86faf
> > >
> > > _______________________________________________
> > > mesa-dev mailing list
> > > mesa-dev at lists.freedesktop.org
> > > https://lists.freedesktop.org/mailman/listinfo/mesa-dev
> >