util: Update ROCm to 1.6.4 in gcn Dockerfile, install HIP by .deb

Previously, we were using ROCm 1.6.2 as there were issues with some of
the machine learning applications that weren't present on 1.6.2.
However, after re-running them we've found that they, and all other
applications previously tested, run to completion.

Additionally, there have been patches to enable BLIT kernels which made
it so we no longer need to build HIP and MIOpen differently for APU and
DGPU code. This allows us to install HIP directly from the .deb packages
instead of from source. Installing from the .deb packages also avoid the
hipDeviceSynchronize() bug. Finally, this makes it so most GPU programs
can be run as-is without modifications to remove hipMalloc/hipMemcpy
calls as was done previously.

Change-Id: Ic61b09ed200b19f759d891487cde874abd607537
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/37675
Reviewed-by: Bobby R. Bruce <bbruce@ucdavis.edu>
Maintainer: Bobby R. Bruce <bbruce@ucdavis.edu>
Tested-by: kokoro <noreply+kokoro@google.com>

diff --git a/util/dockerfiles/gcn-gpu/Dockerfile b/util/dockerfiles/gcn-gpu/Dockerfile
index e13367fcde2e3b41b980570f2f70d1fe7dddb7b1..dad41b92c73de6d691b54cef492c7ca39cd9d72a 100644 (file)
--- a/util/dockerfiles/gcn-gpu/Dockerfile
+++ b/util/dockerfiles/gcn-gpu/Dockerfile
@@ -54,20 +54,21 @@ RUN pip install -U setuptools scons==3.1.2 six
 ARG gem5_dist=http://dist.gem5.org/dist/develop
 
 # Install ROCm 1.6 binaries
-RUN wget -qO- ${gem5_dist}/apt_1.6.2.tar.bz2 \
+RUN wget -qO- ${gem5_dist}/apt_1.6.4.tar.bz2 \
     | tar -xjv \
-    && cd apt_1.6.2/pool/main/ \
+    && cd apt_1.6.4/pool/main/ \
     && dpkg -i h/hsakmt-roct-dev/* \
     && dpkg -i h/hsa-ext-rocr-dev/* \
     && dpkg -i h/hsa-rocr-dev/* \
     && dpkg -i r/rocm-utils/* \
     && dpkg -i h/hcc/* \
     && dpkg -i r/rocm-opencl/* \
-    && dpkg -i r/rocm-opencl-dev/*
+    && dpkg -i r/rocm-opencl-dev/* \
+    && dpkg -i h/hip_base/* \
+    && dpkg -i h/hip_hcc/*
 
 # Get ROCm libraries we need to compile from source (and ROCm-profiler)
-RUN git clone --single-branch https://github.com/ROCm-Developer-Tools/HIP/ && \
-    git clone --single-branch https://github.com/ROCmSoftwarePlatform/hipBLAS/ && \
+RUN git clone --single-branch https://github.com/ROCmSoftwarePlatform/hipBLAS/ && \
     git clone --single-branch https://github.com/ROCmSoftwarePlatform/rocBLAS/ && \
     git clone --single-branch https://github.com/ROCmSoftwarePlatform/MIOpenGEMM/ && \
     git clone --single-branch https://github.com/ROCmSoftwarePlatform/MIOpen/ && \
@@ -77,12 +78,10 @@ RUN git clone --single-branch https://github.com/ROCm-Developer-Tools/HIP/ && \
 # Apply patches to various repos
 RUN mkdir -p /patch && cd /patch && \
     wget ${gem5_dist}/rocm_patches/hipBLAS.patch && \
-    wget ${gem5_dist}/rocm_patches/hip.patch_v2 && \
     wget ${gem5_dist}/rocm_patches/miopen-conv.patch && \
     wget ${gem5_dist}/rocm_patches/rocBLAS.patch
 
-RUN git -C /HIP/ checkout 0e3d824e && git -C /HIP/ apply /patch/hip.patch_v2 && \
-    git -C /hipBLAS/ checkout ee57787e && git -C /hipBLAS/ apply /patch/hipBLAS.patch && \
+RUN git -C /hipBLAS/ checkout ee57787e && git -C /hipBLAS/ apply /patch/hipBLAS.patch && \
     git -C /rocBLAS/ checkout cbff4b4e && git -C /rocBLAS/ apply /patch/rocBLAS.patch && \
     git -C /rocm-cmake/ checkout 12670acb && \
     git -C /MIOpenGEMM/ checkout 9547fb9e && \
@@ -97,17 +96,13 @@ ENV PATH ${ROCM_PATH}/bin:${HCC_HOME}/bin:${HSA_PATH}/bin:${HIP_PATH}/bin:${PATH
 ENV HCC_AMDGPU_TARGET gfx801
 
 # Create build dirs for machine learning ROCm installs
-RUN mkdir -p /HIP/build && \
-    mkdir -p /rocBLAS/build && \
+RUN mkdir -p /rocBLAS/build && \
     mkdir -p /hipBLAS/build && \
     mkdir -p /rocm-cmake/build && \
     mkdir -p /MIOpenGEMM/build && \
     mkdir -p /MIOpen/build
 
 # Do the builds, empty build dir to trim image size
-WORKDIR /HIP/build
-RUN cmake .. && make -j$(nproc) && make install && rm -rf *
-
 WORKDIR /rocBLAS/build
 RUN CXX=/opt/rocm/bin/hcc cmake -DCMAKE_CXX_FLAGS="--amdgpu-target=gfx801" .. && \
     make -j$(nproc) && make install && rm -rf *
@@ -144,7 +139,7 @@ RUN CXX=/opt/rocm/hcc/bin/hcc cmake \
     -DMIOPEN_CACHE_DIR=/.cache/miopen \
     -DMIOPEN_AMDGCN_ASSEMBLER_PATH=/opt/rocm/opencl/bin \
     -DHALF_INCLUDE_DIR=/MIOpen/half-1.12.0/include \
-    -DCMAKE_CXX_FLAGS="-isystem /usr/include/x86_64-linux-gnu" .. && \
+    -DCMAKE_CXX_FLAGS="-isystem /usr/include/x86_64-linux-gnu -DDGPU" .. && \
     make -j$(nproc) && make install && rm -rf *
 
 # Re-set defaults
@@ -166,4 +161,5 @@ RUN dpkg -i package/rocm-profiler_4.0.6036_amd64.deb
 # Always use python3 and create a link to config command for gem5 to find
 RUN ln -sf /usr/bin/python3 /usr/bin/python
 RUN ln -sf /usr/bin/python3.9-config /usr/bin/python3-config
+
 WORKDIR /