11:41 <Soroush> zmike: I think your MR !15906 is introducing a bug in KHR-GL33.shaders.uniform_block.random.all_per_block_buffers.3

11:41 <Soroush> The spirv zink produces for it sets the ubo array size to 92 ints, but it is accessing indices outside of that like 112.

11:41 <Soroush> I see from the nir dump that the remove_bo_access pass is incorrectly sizing that ubo.

12:19 <zmike> Soroush: what platform are you observing this on?

12:23 <Soroush> It's on a powervr platform. The uniforms in general are aggressively optimised, so any out of bound accesses for ubo show up.

12:23 <zmike> ah I see

12:23 <zmike> can you file a ticket so I don't lose track of this?

12:24 <zmike> will try to look into it in the next day or two

12:24 <zmike> kusma: any chance you have an idea what's going on with gallium hud?

12:25 <kusma> zmike: I can take a look on Monday. I'm off on the cabin right now.

12:26 <zmike> oh cool

12:26 <zmike> enjoy it

12:26 <zmike> Soroush: also while I think of it does this platform have ShaderInt64?

12:27 <kusma> But I'm off again for a couple of weeks from Tuesday, so if I don't figure it out in time... Well, I won't manage I guess

12:27 <zmike> eh, not like it's a super urgent issue

12:27 <Soroush> zmike: sure, I'll file a ticket. thanks

12:27 <zmike> just thought you'd have an idea since you've worked on hud

20:12 <zmike> Soroush: reasonably sure https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17217 should take care of it