Repository navigation

glm cuda: TF_GLM_KV_HOST keeps the KV pages' latent rows in pinned host memory - #1795

Merged

ashhart merged 7 commits into

ashhart:mainfrom

russellmm:glm-kv-host

Nov 3, 2026

Merged

glm cuda: TF_GLM_KV_HOST keeps the KV pages' latent rows in pinned host memory#1795
ashhart merged 7 commits into
ashhart:mainfrom
russellmm:glm-kv-host

Commits on Nov 3, 2026
- glm cuda: TF_GLM_KV_HOST keeps the KV pages' latent rows in pinned host memory
  
  russellmm
  authored and
  ashhart
  committed
  063ce37
- glm cuda: the last rank to map TF_GLM_KV_HOST_SHARED's page file unlinks it
  
  russellmm
  authored and
  ashhart
  committed
  4ae9f09
- glm cuda: TF_GLM_KV_HOST_SHARED's /dev/shm pages build on Linux only
  
  russellmm
  authored and
  ashhart
  committed
  76612bb
- glm cuda: TF_GLM_KV_HOST, _SHARED and _GIB join the ranks' fingerprint
  
  russellmm
  authored and
  ashhart
  committed
  3a132d6
- glm cuda: TF_GLM_KV_HOST's gather kernels join the contract, and a call past its slots is refused
  
  russellmm
  authored and
  ashhart
  committed
  1ae7acb
- glm cuda: a TF_GLM_KV_HOST call past its slots is refused by name, with a test
  
  russellmm
  authored and
  ashhart
  committed
  6da7ba6
- changelog: #1795 under Unreleased
  ashhart
  committed
  adc001b