diff options
| author | Arthur Hanson Rasmusson <arthur@vgpu.io> | 2026-10-05 14:02:31 -0400 |
|---|---|---|
| committer | Arthur Hanson Rasmusson <arthur@vgpu.io> | 2026-10-05 14:02:31 -0400 |
| commit | 6e2e0965c3e268acddd12bcc4c983c36c573f67e (patch) | |
| tree | 7ad4fdfcbacbe9939399c46a24caa4472d7eaed4 /tools | |
| parent | e2fbb2756cca242f8ee14e1106d43b7d18195dd1 (diff) | |
| download | open-iov-6e2e0965c3e268acddd12bcc4c983c36c573f67e.tar.gz open-iov-6e2e0965c3e268acddd12bcc4c983c36c573f67e.zip | |
vgsp: V6 client-owned ctx-buffer VA commit path (srvctx)
Source-confirmed: with the SKIP_CTXBUFFER_ALLOC channel flag the server does
not allocate MAIN, does not map/book it, and its internal init/promote
helpers return early for KMD virtual contexts. The client therefore owns the
buffer. When no server MAIN is reported on a V6 channel, map the prepared
client MAIN at the ladder ctx VA (sysmem PTEs), send GPU_INITIALIZE_CTX to
install the memdesc, then the UVM-shape VA-only promote; with no server
pre-booking the vaListSetManaged gate should pass and the per-buffer commit
program the engine context pointer. V5 keeps the historical PA-only
fallback. Build clean; not hardware-tested (GPU wedged).
Diffstat (limited to 'tools')
0 files changed, 0 insertions, 0 deletions