enricobuehler d381cdf7f4 fix(host): NVENC open-failure resilience — backoff, failed-open hygiene, self-diagnosis ...

Field report (Linux direct NVENC): after a codec switch, every session open
failed with NV_ENC_ERR_INVALID_VERSION until the host process was restarted —
so the poisoned state is per-process, not a driver install issue. On-hardware
investigation (RTX 5070 Ti, 610.43.03) could not reproduce it with clean codec
cycles, dirty teardowns, or open/destroy storms, but established the failure
class: the driver enforces a per-process concurrent-session cap (12 there,
status INCOMPATIBLE_CLIENT_KEY; other branches report differently) whose
exhaustion is exactly this signature — persistent open failures healed only by
a process restart. Harden every path that can feed or mask that state:

* Rebuild backoff: the in-place encoder-rebuild retries slept one frame
  interval, so all 5 attempts burned within ~40 ms at 120 Hz — no driver-side
  transient (deferred teardown of the previous session, engine reset) can
  clear that fast. Exponential backoff 100 ms → 1.6 s (~3 s total) so
  transients heal instead of killing the session.
* Destroy-on-failed-open (Linux + Windows, all four open sites): the NVENC
  docs require NvEncDestroyEncoder even when OpenEncodeSessionEx FAILS — the
  driver may have allocated the session slot before erroring. Without it a
  retry burst against a transient leaks slots toward the cap, converting the
  transient into permanent exhaustion.
* Teardown: a destroy_encoder failure (a session slot the driver may keep) is
  now logged with its status instead of silently discarded.
* One-shot self-diagnosis on a failed session open (Linux): retry the raw open
  on a fresh dedicated CUDA context and log which of the three causes applies
  — shared-context poisoned (fresh works), driver-level skew/exhaustion/GPU
  loss (fresh fails the same way), or CUDA itself unhealthy (no fresh context)
  — so the next field report pinpoints the root cause with zero reporter
  effort.

On-hardware regression tests (RTX box .21, all green): codec-switch reopen
cycle (H265→AV1→H265→H264→H265), dirty teardown with in-flight encodes, and
the full open-failure→diagnosis→in-place-recovery path via real session-cap
exhaustion. Existing RFI/reconfigure/4:4:4 smokes still pass; clippy clean.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

2026-07-16 01:57:36 +02:00

..

pf-client-core

feat(gamepad): SC2 Puck-dongle passthrough with the native 28DE:1304 topology

2026-07-15 20:47:52 +02:00

pf-console-ui

feat: PyroWave Phase 3 — pinned rate, all-intra silencing, opt-in UI, notices, docs

2026-07-15 09:45:49 +02:00

pf-driver-proto

feat(gamepad): Windows Steam Deck backend — Steam-Input-promoted UMDF virtual Deck

2026-07-14 12:36:26 +02:00

pf-ffvk

feat(clients/windows): port the Vulkan session client to Windows — session-always

2026-07-08 23:21:36 +02:00

pf-presenter

feat(host,clients): PyroWave ships in default builds; NVIDIA hosts advertise it

2026-07-15 13:53:15 +02:00

punktfunk-core

feat(gamepad): SC2 Puck-dongle passthrough with the native 28DE:1304 topology

2026-07-15 20:47:52 +02:00

punktfunk-host

fix(host): NVENC open-failure resilience — backoff, failed-open hygiene, self-diagnosis

2026-07-16 01:57:36 +02:00

punktfunk-tray

refactor(client): extract UI-agnostic plumbing into pf-client-core (phase 0)

2026-07-07 23:13:16 +02:00

pyrowave-sys

fix(pyrowave-sys): link user32 on Windows (Granite breadcrumbs MessageBoxA)

2026-07-15 01:02:19 +02:00