aboutsummaryrefslogtreecommitdiff
path: root/build.mk
diff options
context:
space:
mode:
authorOnur Gündüz <ogunduz@gmail.com>2026-07-02 19:44:08 +0200
committerOnur Gündüz <ogunduz@gmail.com>2026-07-02 19:44:08 +0200
commit6a10d3577232f1fd646a20f22a4a1ab4ec29f28a (patch)
treee4a077b064876ddb39144935fa256a4ab0692488 /build.mk
parent98a713e86c23bc6098bf7a5558507a9f27a7a1c1 (diff)
downloadffwd-6a10d3577232f1fd646a20f22a4a1ab4ec29f28a.tar.gz
ffwd-6a10d3577232f1fd646a20f22a4a1ab4ec29f28a.zip
cuda: add cuDNN fused attention path with runtime fallback
Diffstat (limited to 'build.mk')
-rw-r--r--build.mk28
1 files changed, 26 insertions, 2 deletions
diff --git a/build.mk b/build.mk
index a49cebb..73eefc4 100644
--- a/build.mk
+++ b/build.mk
@@ -120,9 +120,33 @@ MLX_BACKEND_LDFLAGS = -framework Metal -framework Foundation \
-L$(MLXC_PREFIX)/lib -lmlxc \
-L$(MLX_PREFIX)/lib -lmlx \
-Wl,-rpath,$(MLX_PREFIX)/lib -Wl,-rpath,$(MLXC_PREFIX)/lib
+# cuDNN fused attention (libffwd/cuda_sdpa.cu): compiles against the vendored
+# cudnn-frontend headers and dlopens libcudnn at runtime, so only cudnn.h is
+# needed and only at build time. CUDNN_INC overrides the probe below (point it
+# at the directory containing cudnn.h, e.g. a pip nvidia-cudnn-cu12 wheel's
+# include/); when no cudnn.h is found the CUDA backend builds without the
+# fused path and keeps the built-in kernels.
+CUDNN_INC ?=
+ifeq ($(strip $(CUDNN_INC)),)
+ CUDNN_H := $(firstword $(wildcard /usr/include/cudnn.h \
+ /usr/include/x86_64-linux-gnu/cudnn.h $(CUDA_HOME)/include/cudnn.h))
+ ifneq ($(strip $(CUDNN_H)),)
+ CUDNN_INC := $(patsubst %/,%,$(dir $(CUDNN_H)))
+ endif
+endif
+ifneq ($(strip $(CUDNN_INC)),)
+ CUDNN_SDPA_CFLAGS = -I$(ROOT)/deps/cudnn-frontend/include -I$(CUDNN_INC) \
+ -DFFWD_CUDA_SDPA=1 -DNV_CUDNN_FRONTEND_USE_DYNAMIC_LOADING
+else
+ CUDNN_SDPA_CFLAGS =
+endif
+
CUDA_BACKEND_CFLAGS = -DUSE_BLAS -DUSE_OPENBLAS -DUSE_GPU $(OPENBLAS_CFLAGS) \
- -I$(CUDA_HOME)/include
-CUDA_BACKEND_LDFLAGS = -L$(CUDA_HOME)/lib64 $(OPENBLAS_LDFLAGS) -lcudart -lcublas -lcublasLt
+ -I$(CUDA_HOME)/include $(CUDNN_SDPA_CFLAGS)
+# -lstdc++: cuda_sdpa.o uses the C++ runtime (cudnn-frontend), and the tools
+# link with the C driver.
+CUDA_BACKEND_LDFLAGS = -L$(CUDA_HOME)/lib64 $(OPENBLAS_LDFLAGS) -lcudart -lcublas -lcublasLt \
+ -ldl -lstdc++
ifeq ($(BACKEND),blas)
BACKEND_CFLAGS = $(BLAS_BACKEND_CFLAGS)