trunk/521f9d784733cb2f4571804b0f6e4687fb7a91be: [inductor] Unify threads-per-wave (warp_size) extraction (#181112)
🔒
https://github.com
«Route threads-per-wave through a single source of truth
(DeviceProperties.warp_size, from torch.cuda.get_device_properties) and
thread it through the config-generation helpers, replacing _NUM_THREADS_PER_WARP,
cc_warp_si...»
Automatische Weiterleitung...
1.5s