diff options
| author | Arthur Hanson Rasmusson <arthur@vgpu.io> | 2026-10-06 13:47:57 -0400 |
|---|---|---|
| committer | Arthur Hanson Rasmusson <arthur@vgpu.io> | 2026-10-06 13:47:57 -0400 |
| commit | 20e0fbcfb258dfc067265694ddae9c5250f353ce (patch) | |
| tree | 8871dc6af7590eadcbd145e421276ae07526dc89 /kernel/gsp_phys_booter.h | |
| parent | 14e5418b24b4dff131c311213cfa17a4aadd813d (diff) | |
| download | open-iov-20e0fbcfb258dfc067265694ddae9c5250f353ce.tar.gz open-iov-20e0fbcfb258dfc067265694ddae9c5250f353ce.zip | |
vgsp: install the Tier-2 scratch-pool window spine via PDE_UPDATE + IQ2M sentinel diagnostic
- The Tier-2 shadow VAs (0x200000 + page<<21 + line<<16) had no PTE chain
in the engine's table view on the shared-management VAS: pdb_build's
client PD0 spine for the guest scratch PT pool is discarded by the
server's root migration, so every param/program/release/alias leg the
launch rewrites to a shadow VA was unreachable by the engine. Install
each scratch pool window lazily through DMA_UPDATE_PDE_2 (same control
shape as the Tier-1 installer, pool chunk as the leaf PT) the first
time a shadow line in it is consumed; shared VAS only, idempotent via
rmcall_guest_pt_win_mask. m17 confirmed the install is accepted
(pdeIndex=1 VA 0x200000 status 0x0).
- IQ2M sentinel: write 0xdeadbeef to the inline-QMD destination word 0
after the prewrite and report iq2m=ran/not-run in the post-mortem so
'engine stalled before the inline window' is distinguishable from 'QMD
deposited but not executed' without a server capture.
Diffstat (limited to 'kernel/gsp_phys_booter.h')
| -rw-r--r-- | kernel/gsp_phys_booter.h | 9 |
1 files changed, 9 insertions, 0 deletions
diff --git a/kernel/gsp_phys_booter.h b/kernel/gsp_phys_booter.h index 9a0331f..67d5d5c 100644 --- a/kernel/gsp_phys_booter.h +++ b/kernel/gsp_phys_booter.h @@ -335,6 +335,15 @@ void gsp_phys_backend_relay_counters(u64 *requested, u64 *ok, u64 *verify_fail, #define GSP_PHYS_GR_MAX_INLINE 8u #define GSP_PHYS_GR_INLINE_BYTES 64u #define GSP_PHYS_GR_QMD_WORDS 64u +/* + * IQ2M sentinel: written to the inline-QMD destination's word 0 after the + * prewrite. If the engine processes the inline window, the copy overwrites + * the page (word 0 -> the inline value, 0) and the post-mortem sees the page + * changed; if the sentinel survives, the engine stalled before the inline + * window and the page was never touched. Distinguishes "IQ2M never ran" from + * "QMD deposited but not executed" without a server-side capture. + */ +#define GSP_PHYS_GR_QMD_SENTINEL 0xdeadbeefu #define GSP_PHYS_GR_QMD_WORD_NONE 0xffffffffu #define GSP_PHYS_GR_RAW_WORDS 256u |