refactor(rocm): remove the gpu_iface layer and flatten the ROCm headers - #329

Merged

demandal25 merged 13 commits into

amd-integrationfrom

rocm-evacuate-per-op

Aug 29, 2026

Commits on Aug 29, 2026

refactor(rocm): move the torch.compile custom-op machinery out of utils.py
demandal25
committed
e9708b6
refactor(rocm): move csrc_rocm to flashinfer/csrc/rocm
demandal25
committed
8fa4517
fix(rocm): merge the duplicate dispatch and exception headers
demandal25
committed
2c093de
fix(rocm): stop the ROCm sampling path pulling upstream headers
demandal25
committed
126e615
refactor(rocm): make the HIP backend headers self-contained
demandal25
committed
e8c3c75
refactor(rocm): flatten gpu_iface into include/flashinfer/rocm
demandal25
committed
b48b587
refactor(rocm): drop the gpu_iface namespace
demandal25
committed
e29465b
docs(rocm): replace the gpu_iface convention with directory/rocm/
demandal25
committed
30bf194
chore(rocm): give the fork headers their own include guards
demandal25
committed
579f04c
fix(rocm): tripwire the six headers the guard sweep skipped
demandal25
committed
6131aba
fix(rocm): pin macros.hpp first and restore the allocator diagnostic
demandal25
committed
d14c973
chore(rocm): drop a duplicate macro and relabel the coverage snapshot
demandal25
committed
4b9ceea
fix(rocm): pin macros.hpp ahead of the hip/ includes in the last four headers
demandal25
committed
f6e4606