Search results

SIGSEGV in InputBatch::add_request serving a text completion on a synthetic DeepSeek-V4 fixture, ~50% of runs

Status: Open.

#3027 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026

A PNG or http(s) image is HTTP 500 on the Qwen3-VL and dots3-note chat seams, not 400

Status: Open.

#3026 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026

A2-5 is blocked: lifting the veto falsifies the device-mirror arm's own staleness justification, on the GB10 default

Status: Open.

#3025 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026

SPEC-DFLASH2 A2-5: lift the async input-combine veto, the wave that captures the idle

Status: Open.

#3023 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026

ROCm split-phase random sample: phase B launches 1024 threads onto 128-entry shared arrays

Status: Open.

#3022 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026

No CI lane compiles a .hip: every src/vt/rocm change reaches main with zero build coverage, and #2957 landed that way

Status: Open.

#3021 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026

matmul_bt_quant_grouped's 'the loader pre-filters, so reaching here is a bug' assertion is reachable: GLM-5.3-Flash hits it mid-forward on ROCm

Status: Open.

#3019 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026

gfx1151 falls back to the legacy QuantizeQ8KK: the cooperative Q8_K activation quantizer is gated on a strict gfx1100 prefix

Status: Open.

#3018 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026

DotQ6K is scalar where DotQ4K uses hardware dot4, and materializes a 256-byte per-thread array; it serves ffn_down, 14.5% of Qwen3.8-27B's decode byte budget

Status: Open.

#3017 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026

gfx1151: 61.7% of Qwen3.8-27B's per-token byte budget runs KQuantGemmK with 8-12 of 32 lanes idle, and #2086's cooperative arm is Q6_K-only

Status: Open.

#3016 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026

gfx1151 decode: the 71.3% board-idle figure is a whole-leg average; windowed to the generate interval the board is 100% busy at 2872 MHz

Status: Open.

#3015 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026

The hybrid-placement speed axis is untracked: its campaign issue #149 was deleted, and the fleet has no device that can measure it

Status: Open.

#3013 In mudler/vllm.cpp;

· localai-org-maint-bot opened on Sep 6, 2026