From 22f36e5b592b587af5ee50324271df55eff879ab Mon Sep 17 00:00:00 2001 From: "Jan \"yenda\" Trmal" Date: Mon, 16 Sep 2024 15:32:12 +0200 Subject: [PATCH] make sure nvtx3 is used (#4937) --- src/cudadecoder/batched-threaded-nnet3-cuda-online-pipeline.cc | 2 +- src/cudadecoder/batched-threaded-nnet3-cuda-pipeline.cc | 2 +- src/cudadecoder/batched-threaded-nnet3-cuda-pipeline2.cc | 2 +- src/cudadecoder/cuda-decoder.cc | 2 +- src/cudadecoder/cuda-fst.cc | 2 +- src/cudadecoderbin/batched-wav-nnet3-cuda-online.cc | 2 +- src/cudadecoderbin/batched-wav-nnet3-cuda.cc | 2 +- src/cudadecoderbin/batched-wav-nnet3-cuda2.cc | 2 +- src/cudafeat/feature-online-batched-spectral-cuda-kernels.cu | 2 +- src/cudafeat/feature-spectral-cuda.cu | 2 +- src/cudafeat/feature-window-cuda.cu | 2 +- src/cudafeat/online-batched-feature-pipeline-cuda.cc | 2 +- src/cudafeat/online-ivector-feature-cuda.cc | 2 +- src/cudafeatbin/compute-online-feats-batched-cuda.cc | 2 +- src/cudafeatbin/compute-online-feats-cuda.cc | 2 +- src/cudamatrix/cu-common.h | 2 +- src/makefiles/cuda_64bit.mk | 2 +- 17 files changed, 17 insertions(+), 17 deletions(-) diff --git a/src/cudadecoder/batched-threaded-nnet3-cuda-online-pipeline.cc b/src/cudadecoder/batched-threaded-nnet3-cuda-online-pipeline.cc index ed0c0a2f5e9..bec20cb9e07 100644 --- a/src/cudadecoder/batched-threaded-nnet3-cuda-online-pipeline.cc +++ b/src/cudadecoder/batched-threaded-nnet3-cuda-online-pipeline.cc @@ -26,7 +26,7 @@ #include "hipify.h" #else -#include +#include #endif #include diff --git a/src/cudadecoder/batched-threaded-nnet3-cuda-pipeline.cc b/src/cudadecoder/batched-threaded-nnet3-cuda-pipeline.cc index 23d0ca283a2..32d7ac40e12 100644 --- a/src/cudadecoder/batched-threaded-nnet3-cuda-pipeline.cc +++ b/src/cudadecoder/batched-threaded-nnet3-cuda-pipeline.cc @@ -31,7 +31,7 @@ #include "hipify.h" #else -#include +#include #endif #include "base/kaldi-utils.h" diff --git a/src/cudadecoder/batched-threaded-nnet3-cuda-pipeline2.cc b/src/cudadecoder/batched-threaded-nnet3-cuda-pipeline2.cc index 01d6b1165e7..4b30c568e73 100644 --- a/src/cudadecoder/batched-threaded-nnet3-cuda-pipeline2.cc +++ b/src/cudadecoder/batched-threaded-nnet3-cuda-pipeline2.cc @@ -28,7 +28,7 @@ #include "hipify.h" #else -#include +#include #endif namespace kaldi { diff --git a/src/cudadecoder/cuda-decoder.cc b/src/cudadecoder/cuda-decoder.cc index 056d563a791..15f29d27122 100644 --- a/src/cudadecoder/cuda-decoder.cc +++ b/src/cudadecoder/cuda-decoder.cc @@ -44,7 +44,7 @@ #include "hipify.h" #else #include -#include +#include #endif #include "base/kaldi-utils.h" diff --git a/src/cudadecoder/cuda-fst.cc b/src/cudadecoder/cuda-fst.cc index 682485f6ce4..6b0d34f81b7 100644 --- a/src/cudadecoder/cuda-fst.cc +++ b/src/cudadecoder/cuda-fst.cc @@ -29,7 +29,7 @@ #include "hipify.h" #else #include -#include +#include #endif namespace kaldi { diff --git a/src/cudadecoderbin/batched-wav-nnet3-cuda-online.cc b/src/cudadecoderbin/batched-wav-nnet3-cuda-online.cc index 2bc0a483a0f..a47ea2e2300 100644 --- a/src/cudadecoderbin/batched-wav-nnet3-cuda-online.cc +++ b/src/cudadecoderbin/batched-wav-nnet3-cuda-online.cc @@ -30,7 +30,7 @@ #else #include #include -#include +#include #endif #include diff --git a/src/cudadecoderbin/batched-wav-nnet3-cuda.cc b/src/cudadecoderbin/batched-wav-nnet3-cuda.cc index 0e4a719bc75..06aac47b5e0 100644 --- a/src/cudadecoderbin/batched-wav-nnet3-cuda.cc +++ b/src/cudadecoderbin/batched-wav-nnet3-cuda.cc @@ -24,7 +24,7 @@ #else #include #include -#include +#include #endif #include #include "cudadecoder/batched-threaded-nnet3-cuda-pipeline.h" diff --git a/src/cudadecoderbin/batched-wav-nnet3-cuda2.cc b/src/cudadecoderbin/batched-wav-nnet3-cuda2.cc index b2ad9254c67..b7a9d463214 100644 --- a/src/cudadecoderbin/batched-wav-nnet3-cuda2.cc +++ b/src/cudadecoderbin/batched-wav-nnet3-cuda2.cc @@ -27,7 +27,7 @@ #else #include #include -#include +#include #endif #include diff --git a/src/cudafeat/feature-online-batched-spectral-cuda-kernels.cu b/src/cudafeat/feature-online-batched-spectral-cuda-kernels.cu index 856d2acab81..bc06ea32d69 100644 --- a/src/cudafeat/feature-online-batched-spectral-cuda-kernels.cu +++ b/src/cudafeat/feature-online-batched-spectral-cuda-kernels.cu @@ -25,7 +25,7 @@ #include "hipify.h" #else #include -#include +#include #endif #include "cudafeat/lane-desc.h" diff --git a/src/cudafeat/feature-spectral-cuda.cu b/src/cudafeat/feature-spectral-cuda.cu index d8fc215b80b..7b514010562 100644 --- a/src/cudafeat/feature-spectral-cuda.cu +++ b/src/cudafeat/feature-spectral-cuda.cu @@ -24,7 +24,7 @@ #include "hipify.h" #else -#include +#include #include #endif diff --git a/src/cudafeat/feature-window-cuda.cu b/src/cudafeat/feature-window-cuda.cu index 60fe113d402..e001eb0790f 100644 --- a/src/cudafeat/feature-window-cuda.cu +++ b/src/cudafeat/feature-window-cuda.cu @@ -22,7 +22,7 @@ #include "hipify.h" #else -#include +#include #endif #include "matrix/matrix-functions.h" diff --git a/src/cudafeat/online-batched-feature-pipeline-cuda.cc b/src/cudafeat/online-batched-feature-pipeline-cuda.cc index 7736f525237..e03fda01ca7 100644 --- a/src/cudafeat/online-batched-feature-pipeline-cuda.cc +++ b/src/cudafeat/online-batched-feature-pipeline-cuda.cc @@ -25,7 +25,7 @@ #include "hipify.h" #else -#include +#include #endif namespace kaldi { diff --git a/src/cudafeat/online-ivector-feature-cuda.cc b/src/cudafeat/online-ivector-feature-cuda.cc index f96b2a81ce2..daf1c7dfbf9 100644 --- a/src/cudafeat/online-ivector-feature-cuda.cc +++ b/src/cudafeat/online-ivector-feature-cuda.cc @@ -26,7 +26,7 @@ #endif #define CUBLAS_FILL_MODE_LOWER HIPSOLVER_FILL_MODE_LOWER #else -#include +#include #endif #endif diff --git a/src/cudafeatbin/compute-online-feats-batched-cuda.cc b/src/cudafeatbin/compute-online-feats-batched-cuda.cc index 2cd6bbb6a93..e3f2ed75d30 100644 --- a/src/cudafeatbin/compute-online-feats-batched-cuda.cc +++ b/src/cudafeatbin/compute-online-feats-batched-cuda.cc @@ -18,7 +18,7 @@ #if HAVE_CUDA #ifndef __IS_HIP_COMPILE__ #include -#include +#include #endif #endif diff --git a/src/cudafeatbin/compute-online-feats-cuda.cc b/src/cudafeatbin/compute-online-feats-cuda.cc index 70380f8ccad..d54ba56be84 100644 --- a/src/cudafeatbin/compute-online-feats-cuda.cc +++ b/src/cudafeatbin/compute-online-feats-cuda.cc @@ -17,7 +17,7 @@ #if HAVE_CUDA == 1 #ifndef __IS_HIP_COMPILE__ -#include +#include #endif #endif #include "base/kaldi-common.h" diff --git a/src/cudamatrix/cu-common.h b/src/cudamatrix/cu-common.h index 3206fe7e7f4..f7f45b8043a 100644 --- a/src/cudamatrix/cu-common.h +++ b/src/cudamatrix/cu-common.h @@ -44,7 +44,7 @@ #include #include #include -#include +#include #define GPU_WARP_SIZE 32 #define GPU_MAX_THREADS_PER_BLOCK 1024 diff --git a/src/makefiles/cuda_64bit.mk b/src/makefiles/cuda_64bit.mk index 2931881d2bd..713d8982f9e 100644 --- a/src/makefiles/cuda_64bit.mk +++ b/src/makefiles/cuda_64bit.mk @@ -16,4 +16,4 @@ CUDA_FLAGS = --compiler-options -fPIC --machine 64 -DHAVE_CUDA -DOPENFST_VER=$(O CUDA_LDFLAGS += -L$(CUDATKDIR)/lib64/stubs -L$(CUDATKDIR)/lib64 -Wl,-rpath,$(CUDATKDIR)/lib64 CUDA_LDFLAGS += -L$(CUDATKDIR)/lib/stubs -L$(CUDATKDIR)/lib -Wl,-rpath,$(CUDATKDIR)/lib -CUDA_LDLIBS += -lcuda -lcublas -lcusparse -lcusolver -lcudart -lcurand -lcufft -lnvToolsExt +CUDA_LDLIBS += -lcuda -lcublas -lcusparse -lcusolver -lcudart -lcurand -lcufft