diff options
| author | Arthur Hanson Rasmusson <arthur@vgpu.io> | 2026-10-05 16:04:33 -0400 |
|---|---|---|
| committer | Arthur Hanson Rasmusson <arthur@vgpu.io> | 2026-10-05 16:04:33 -0400 |
| commit | cc1028dffcbf9e028938555827f02b18ceba1665 (patch) | |
| tree | 106cb12ed468d1df87725cbbcd965fb2f7cf9a1a /kernel/gsp_phys_gpu.h | |
| parent | 9a8a48b71ddc868f2c52e36e8e4e2f622236054d (diff) | |
| download | open-iov-cc1028dffcbf9e028938555827f02b18ceba1665.tar.gz open-iov-cc1028dffcbf9e028938555827f02b18ceba1665.zip | |
vgsp: FWCLIENT entry-selection mask and golden-replay hardening
- gsp_phys_rmcall_fwclient_mask (default 0x3f): per-entry selection for the
mode-24 FWCLIENT promote set; bit6 selects the MAIN UVM VA-only shape.
- golden replay: tolerate the client-local system-memory PB step and continue
with hObjectBuffer=0; on-ladder TURING_A 3D object retained best-effort.
Live bisection on the KMD channel: MAIN alone reaches the per-buffer commit
and stops at the engine-context memdesc state; PM-only and PATCH-only are
accepted; any multi-entry control containing MAIN is refused 0x1f. All
cycles release clean.
Diffstat (limited to 'kernel/gsp_phys_gpu.h')
| -rw-r--r-- | kernel/gsp_phys_gpu.h | 7 |
1 files changed, 7 insertions, 0 deletions
diff --git a/kernel/gsp_phys_gpu.h b/kernel/gsp_phys_gpu.h index 8322725..9e809e3 100644 --- a/kernel/gsp_phys_gpu.h +++ b/kernel/gsp_phys_gpu.h @@ -99,6 +99,13 @@ int gsp_phys_srv_ctxcommit_mode(void); int gsp_phys_rmcall_channel_value(void); /* + * gsp_phys_rmcall_fwclient_mask (default 0x3f): bitmask selecting the mode-24 + * FWCLIENT promote entries to emit (bit0 MAIN, bit1 PM, bit2 PATCH, + * bit3 FECS_EVENT, bit4 PRIV_ACCESS_MAP, bit5 UNRESTRICTED_PRIV_ACCESS_MAP). + */ +u32 gsp_phys_rmcall_fwclient_mask_value(void); + +/* * gsp_phys_rmcall_golden (default 0): replay the stock client's golden-image * channel allocations in stage 3 before the ladder's own channel when set and * gsp_phys_gpu=1. With 0, the ladder is byte-identical to before. |