Sent out this week were more Intel Xe kernel graphics driver changes targeting the upcoming Linux 7.4 kernel cycle. The theme this week was around dealing with video memory health and also being able to better deal with hardware degraded video RAM.

The Intel Xe driver is rolling out a basic vRAM health check routine focused on sanity testing the last vRAM page. Intel engineer Matthew Auld explained of this vRAM health check for the Xe driver with this patch:

"During early probe, use the last page as a canary for BAR sizing, CCS sizing, identity map setup etc. If something is wrong the last page is where we will likely find it. Hit it with everything we have. For now this is gated behind a debug config option, so shouldn't trigger on production.

Main motivation is around CCS sizing where on some BMG cards the CCS offset is programmed misaligned, for whatever reason, and our handling of that was busted, as found by Linus, leading to some amount of CCS storage getting pulled into the allocator as normal VRAM."

For now at least this vRAM health check option is gated by the CONFIG_DRM_XE_DEBUG_MEM kernel build flag.

Separately this week's Xe pull request also brings in memory page offlining support after going through 21 rounds of review. This is a "significant step" toward making the Xe kernel driver gracefully handle hardware memory degradation. With this work the Xe driver can permanently carve out faulty video memory so it isn't reused by subsequent allocations. The bad vRAM pages are also exposed via DebugFS for user-space to be aware of the faulty memory areas.

The Intel Xe driver with these newest queued changes also now exposes per-engine idle status via DebugFS.

The full list of Intel Xe kernel driver changes submitted this week to DRM-Next for Linux 7.4 can be found via this pull.