On 12/12/2013 02:19 AM, Juha-Pekka Heikkilä wrote:
On Thu, Dec 12, 2013 at 3:14 AM, Brian Paul <bri...@vmware.com> wrote:
On 12/11/2013 02:05 AM, Juha-Pekka Heikkila wrote:
Change save_attrib_data() to return true/false depending on success.
Signed-off-by: Juha-Pekka Heikkila <juhapekka.heikk...@gmail.com>
I'd like to reconsider these changes from scratch.
The basic issue is just null-checking of the malloc calls in
glPush[Client]Attrib(). Why can't we just do something like the (partial)
patch below?
The patch below is really similar to my original try at this. Ian
pointed out from my first try if save_attrib_data would fail we'd lose
"attr" thus I now have evolved the patch to the way it is.
save_attrib_data() could free the attr data if the
malloc(gl_attrib_node) fails. But that's pretty unlikely anyway since
it's such a small allocation.
I think now
it would not in any case leak anything. The original patch for this
issue with later comments was here:
https://urldefense.proofpoint.com/v1/url?u=http://lists.freedesktop.org/archives/mesa-dev/2013-December/049628.html&k=oIvRg1%2BdGAgOoM1BIlLLqw%3D%3D%0A&r=lGQMzzTgII0I7jefp2FHq7WtZ%2BTLs8wadB%2BiIj9xpBY%3D%0A&m=zqSh7DDzSfIFtJicUG1X47XPh1zbuQpq8eL5erODtM8%3D%0A&s=fe0885203b966f88a03e9669530943c79908cd339f961ae50137a400f4145fd9
As a small bonus using helper in glPushAttrib() make the function more
clean looking and easier to follow.
OK, if you really want to got that way, I don't have a strong opinion.
But I do have more review comments then...
-Brian
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
