[Mesa-dev] Odd issue caused by commit d030a3404ca ("linker: Sort shader I/O variables into a canonical order")

Ilia Mirkin imirkin at alum.mit.edu
Mon Jul 21 08:08:04 PDT 2014


Hi Ian,

A user has bisected an issue with the color of the rendered plane in
fgfs to your commit d030a3404ca (and has verified that reverting it on
top of 10.2.x resolves the problem). The issue also disappears on
master starting with commit 1d9f74eda7 ("glsl: Rebalance expression
trees that are reduction operations."), however that's probably
because it just rejiggers the IR enough to not trigger whatever issue
is going on. There's an apitrace in the bug as well:

https://bugs.freedesktop.org/show_bug.cgi?id=81500#c22

Any ideas what it could be?

  -ilia


More information about the mesa-dev mailing list