Skip to content

intel: Compute the optimal preferred SLM size per subslice and other fixes around SLM

José Roberto de Souza requested to merge zehortigoza/mesa:xe2-preferred-slm into main

What does this MR do and why?

Up to now preferred SLM size was being set to maximum preferred SLM size for GFX 12.5 platforms and to 0kb for Xe2, causing GPU hangs in Xe2 platforms and over-allocation in some cases for GFX 12.5. This MR adds a optimal calculation for preferred SLM size and other fixes around SLM.

Merge request reports