<feed xmlns='http://www.w3.org/2005/Atom'>
<title>open-iov.git/tools/guest-diag/pin_gr_stream.py, branch cs-share</title>
<subtitle>Open-IOV vGPU-IO: host-side GSP emulation for unmodified NVIDIA guest drivers
</subtitle>
<id>https://git.vgpu.io/open-iov.git/atom?h=cs-share</id>
<link rel='self' href='https://git.vgpu.io/open-iov.git/atom?h=cs-share'/>
<link rel='alternate' type='text/html' href='https://git.vgpu.io/open-iov.git/'/>
<updated>2026-10-03T18:05:38+00:00</updated>
<entry>
<title>vgsp: host CUDA bridge + sema release recovery; greeneye simulator wave</title>
<updated>2026-10-03T18:05:38+00:00</updated>
<author>
<name>Arthur Hanson Rasmusson</name>
<email>arthur@vgpu.io</email>
</author>
<published>2026-10-03T18:05:38+00:00</published>
<link rel='alternate' type='text/html' href='https://git.vgpu.io/open-iov.git/commit/?id=cc79b478843f0230fe6aa8ecd822de7ba71c891e'/>
<id>urn:sha1:cc79b478843f0230fe6aa8ecd822de7ba71c891e</id>
<content type='text'>
- kernel: gsp_cuda_bridge (QMD launch capture, debugfs job/out/done, param+output resolution cached from inline writes), sema drain fixes (never hold/starve a channel), GR ring autopin + counter autoreg, per-page push reads, counter-derive gate; module builds clean
- tools/cuda-bridge: CUDA driver-API daemon executing mail/vadd kernels on the physical GPU, CUBB2/CBD2 v2 protocol, layout check + self-tests
- tools/guest-diag + tools/cuda-ab: scanner/pin fixes, victory push-only mode
- debug/risc-v-gsp-rm-vm: Phase S simulator wave (GSP device model, boot ladder, triage/re-demo harnesses, docs 30-41), GPGPU-sim scaffolding
- STATE-CURRENT/docs updates
</content>
</entry>
<entry>
<title>kernel: GR context-promote bisection modes 4-17 and server-managed-ctxbuf registry variant; complete vgsp rename; docs/tooling refresh</title>
<updated>2026-10-02T19:13:06+00:00</updated>
<author>
<name>Arthur Hanson Rasmusson</name>
<email>arthur@vgpu.io</email>
</author>
<published>2026-10-02T19:13:06+00:00</published>
<link rel='alternate' type='text/html' href='https://git.vgpu.io/open-iov.git/commit/?id=c9bf1cf098c8ea4daeb3e6c9b274a9025c03045d'/>
<id>urn:sha1:c9bf1cf098c8ea4daeb3e6c9b274a9025c03045d</id>
<content type='text'>
- gsp_phys_booter: GR promote bisection modes 4-17 (two-phase zero-field VA,
  batched entries, hVirtMemory, bare VA, KMD SKIP_CTXBUFFER_ALLOC channel,
  old-interface GPU_INITIALIZE_CTX, FBMEM context buffer, UVM-faithful PA-only,
  PATCH/PM auxiliary promotes, server-managed no-promote path); PATCH/PM
  geometry queries and buffers; pre-boot SET_REGISTRY profile selection
- gsp_phys_gpu: gsp_phys_rmcall_gr_promote clamp 0..17 and gsp_phys_srv_ctxbuf
  parameter; registry image variant with RMSetClientRMAllocatedCtxBuffer=0
- vgsp rename across kernel, scripts, tools and docs (module vgsp.ko,
  /sys/kernel/debug/vgsp/, vgsp-&lt;subsystem&gt;: log prefixes)
- risc-v-gsp-rm-debugger user-facing rename and docs/CI refresh; guest-diag
  tool updates; README/SBOM/install/Makefile sync

Live status: CE ladder and P6 guest-copy forwarding green; GR engine context
commit still open (all promote shapes accepted or cleanly refused; the channel
stays CTXNOTVALID) - see Handoff-vGPU-IO.md Addendum 21 (local).
</content>
</entry>
<entry>
<title>guest-diag: retry the guest pin-scanner startup (a transient agent exec failure left a boot without pins and the client hung)</title>
<updated>2026-09-30T14:52:12+00:00</updated>
<author>
<name>Arthur Hanson Rasmusson</name>
<email>arthur@vgpu.io</email>
</author>
<published>2026-09-30T14:52:12+00:00</published>
<link rel='alternate' type='text/html' href='https://git.vgpu.io/open-iov.git/commit/?id=84352915d3ca267044715e00b624c90bed9647b4'/>
<id>urn:sha1:84352915d3ca267044715e00b624c90bed9647b4</id>
<content type='text'>
</content>
</entry>
<entry>
<title>guest-diag: stream uvbp pushbuffer kptrs for immediate pinning</title>
<updated>2026-09-29T21:27:48+00:00</updated>
<author>
<name>Arthur Hanson Rasmusson</name>
<email>arthur@vgpu.io</email>
</author>
<published>2026-09-29T21:27:48+00:00</published>
<link rel='alternate' type='text/html' href='https://git.vgpu.io/open-iov.git/commit/?id=fcb74dfd32442af859ac45979138b40279f96553'/>
<id>urn:sha1:fcb74dfd32442af859ac45979138b40279f96553</id>
<content type='text'>
The first CE channel's push-buffer was only pinned ~1 minute after the
client's failure: uvbp fired at t+4s but semfix's trace read (guest agent,
every 5 loops) picked it up at t+70s, so the tracker wait for value 3 spun
for 16 s. The guest scanner now emits '#pb 0x&lt;kptr&gt;' from the uvbp trace
lines as soon as they appear; the host mirrors them into the sem stream and
semfix resolves/pins the push-buffer through the UVM chain on the next loop
(sub-second).
</content>
</entry>
<entry>
<title>sema-svc: persist GR counter slot per channel; stable slot layout; bulk pin queue 4k</title>
<updated>2026-09-29T20:14:12+00:00</updated>
<author>
<name>Arthur Hanson Rasmusson</name>
<email>arthur@vgpu.io</email>
</author>
<published>2026-09-29T20:14:12+00:00</published>
<link rel='alternate' type='text/html' href='https://git.vgpu.io/open-iov.git/commit/?id=6d5131e87a2c9cb5c20437669c1917f7be6252b5'/>
<id>urn:sha1:6d5131e87a2c9cb5c20437669c1917f7be6252b5</id>
<content type='text'>
- counter_pa autoreg result is written back to the channel table so later
  passes cannot re-target a different slot when the live GR count changes
  during teardown; re-registration of the same channel preserves the slot.
- counter slot layout uses the max GR count ever seen in the generation
  instead of the shrinking live count.
- sema_map_add queue 64 -&gt; 4096 entries so a full-window pin burst drains
  without ENOSPC retries.
- pin_gr_stream: one batched sudo pass per flush (was one process per run,
  ~4 s for 400 runs), retrying only queue-rejected runs.
- semfix_uvmt: SEMFIX_SKIP_PB=1 skips the blocking push-buffer chain walk
  when the vsock window stream already supplies those pages.
</content>
</entry>
<entry>
<title>guest-diag: scan 512KB per 2MB region and start window at 0x200000000 (cover GR GPFIFO rings)</title>
<updated>2026-09-29T19:42:58+00:00</updated>
<author>
<name>Arthur Hanson Rasmusson</name>
<email>arthur@vgpu.io</email>
</author>
<published>2026-09-29T19:42:58+00:00</published>
<link rel='alternate' type='text/html' href='https://git.vgpu.io/open-iov.git/commit/?id=32446284021c44f46767ebffd1c64ffddfed1ed3'/>
<id>urn:sha1:32446284021c44f46767ebffd1c64ffddfed1ed3</id>
<content type='text'>
</content>
</entry>
<entry>
<title>guest-diag: stream uvmt sems over vsock (no guest-agent race); dump channel last release VA; larger push log budget</title>
<updated>2026-09-29T19:24:31+00:00</updated>
<author>
<name>Arthur Hanson Rasmusson</name>
<email>arthur@vgpu.io</email>
</author>
<published>2026-09-29T19:24:31+00:00</published>
<link rel='alternate' type='text/html' href='https://git.vgpu.io/open-iov.git/commit/?id=f4803e303e1e4bfde9f241ec3b6226399d01825d'/>
<id>urn:sha1:f4803e303e1e4bfde9f241ec3b6226399d01825d</id>
<content type='text'>
</content>
</entry>
<entry>
<title>sema-svc: never apply swept-page releases with no live channel (guest boot corruption); tooling guards</title>
<updated>2026-09-29T19:19:34+00:00</updated>
<author>
<name>Arthur Hanson Rasmusson</name>
<email>arthur@vgpu.io</email>
</author>
<published>2026-09-29T19:19:34+00:00</published>
<link rel='alternate' type='text/html' href='https://git.vgpu.io/open-iov.git/commit/?id=40411a6837620097efe2cf15f322f0526ae45385'/>
<id>urn:sha1:40411a6837620097efe2cf15f322f0526ae45385</id>
<content type='text'>
</content>
</entry>
<entry>
<title>guest-diag: vsock pin stream (guest pagemap -&gt; host sema_map_add)</title>
<updated>2026-09-29T18:52:41+00:00</updated>
<author>
<name>Arthur Hanson Rasmusson</name>
<email>arthur@vgpu.io</email>
</author>
<published>2026-09-29T18:52:41+00:00</published>
<link rel='alternate' type='text/html' href='https://git.vgpu.io/open-iov.git/commit/?id=c07584101ebd7a8cdabbd463a093ae53756b4c4b'/>
<id>urn:sha1:c07584101ebd7a8cdabbd463a093ae53756b4c4b</id>
<content type='text'>
The guest-agent exec path is too slow under driver-init load (a 15k-page
window scan took 17-35 s to complete), so GR window pins landed after the
channels' entries had been consumed and after the client's cuCtxCreate
timeout. The guest now scans /proc/&lt;pid&gt;/pagemap locally and streams only
delta runs over vsock (no agent on the hot path); the host listener pins
them as they arrive. The scanner writes a pidfile so restarts do not kill
the new instance.
</content>
</entry>
</feed>
