Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
gaoqiong
composable_kernel
Commits
54138dc8
Commit
54138dc8
authored
Nov 25, 2020
by
Chao Liu
Browse files
update script
parent
be4e3133
Changes
2
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
27 additions
and
13 deletions
+27
-13
script/cmake-rocm3.7.sh
script/cmake-rocm3.7.sh
+4
-2
script/hipclang_opt.sh
script/hipclang_opt.sh
+23
-11
No files found.
script/cmake-rocm3.7.sh
View file @
54138dc8
...
...
@@ -8,14 +8,16 @@ MY_PROJECT_INSTALL=../install.dir
cmake
\
-D
CMAKE_INSTALL_PREFIX
=
${
MY_PROJECT_INSTALL
}
\
-D
CMAKE_BUILD_TYPE
=
Debug
\
-D
CMAKE_BUILD_TYPE
=
Release
\
-D
DEVICE_BACKEND
=
"AMD"
\
-D
CMAKE_CXX_FLAGS
=
"-O3 --amdgpu-target=gfx906 -
mllvm --amdgpu-spill-vgpr-to-agpr=0 -save-temps=
$CWD
"
\
-D
CMAKE_CXX_FLAGS
=
"-O3 --amdgpu-target=gfx906 -
gline-tables-only -save-temps=
$CWD
"
\
-D
CMAKE_CXX_COMPILER
=
/opt/rocm/bin/hipcc
\
-D
CMAKE_PREFIX_PATH
=
"/opt/rocm"
\
-D
CMAKE_VERBOSE_MAKEFILE:BOOL
=
ON
\
${
MY_PROJECT_SOURCE
}
#-D CMAKE_CXX_FLAGS="-O3 --amdgpu-target=gfx906 -save-temps=$CWD" \
#-D CMAKE_CXX_FLAGS="-O3 --amdgpu-target=gfx906 -gline-tables-only -save-temps=$CWD" \
#-D CMAKE_CXX_FLAGS="-O3 --amdgpu-target=gfx906 -mllvm --amdgpu-spill-vgpr-to-agpr=0" \
#-D CMAKE_CXX_FLAGS="-O3 --amdgpu-target=gfx906 -mllvm --amdgpu-spill-vgpr-to-agpr=0 -save-temps=$CWD" \
#-D CMAKE_CXX_FLAGS="-O3 --amdgpu-target=gfx906 -mllvm --amdgpu-enable-global-sgpr-addr -mllvm --amdgpu-spill-vgpr-to-agpr=0" \
...
...
script/hipclang_opt.sh
View file @
54138dc8
rm
*
.ll
*
.s
/opt/rocm/llvm/bin/llvm-dis driver/conv_driver-hip-amdgcn-amd-amdhsa-gfx906-optimized.bc
-o
tmp.ll
/opt/rocm/llvm/bin/opt
-S
-inline
-inline-threshold
=
104857 tmp.ll
>
inline.ll
/opt/rocm/llvm/bin/opt
-S
-O3
-sroa
inline.ll
>
o3.ll
/opt/rocm/llvm/bin/opt
-S
-O3
-sroa
o3.ll
>
o3_2.ll
/opt/rocm/llvm/bin/opt
-S
-O3
-sroa
o3_2.ll
>
o3_3.ll
/opt/rocm/llvm/bin/opt
-S
-O3
-sroa
o3_3.ll
>
o3_4.ll
/opt/rocm/llvm/bin/llc
-mcpu
=
gfx908 o3.ll
/opt/rocm/llvm/bin/llc
-mcpu
=
gfx908 o3_2.ll
/opt/rocm/llvm/bin/llc
-mcpu
=
gfx908 o3_3.ll
/opt/rocm/llvm/bin/llc
-mcpu
=
gfx908 o3_4.ll
BC_FILE
=
$1
/opt/rocm/llvm/bin/llvm-dis
$BC_FILE
-o
original.ll
/opt/rocm/llvm/bin/opt
-S
-inline
-inline-threshold
=
104857 original.ll
>
inline.ll
/opt/rocm/llvm/bin/opt
-S
-sroa
inline.ll
>
sora.ll
/opt/rocm/llvm/bin/opt
-S
-O3
sora.ll
>
o3.ll
/opt/rocm/llvm/bin/llc
-mcpu
=
gfx906 original.ll
/opt/rocm/llvm/bin/llc
-mcpu
=
gfx906 inline.ll
/opt/rocm/llvm/bin/llc
-mcpu
=
gfx906 sora.ll
/opt/rocm/llvm/bin/llc
-mcpu
=
gfx906 o3.ll
#/opt/rocm/llvm/bin/opt -S -O3 -sroa inline.ll > o3.ll
#/opt/rocm/llvm/bin/opt -S -O3 -sroa o3.ll > o3_2.ll
#/opt/rocm/llvm/bin/opt -S -O3 -sroa o3_2.ll > o3_3.ll
#/opt/rocm/llvm/bin/opt -S -O3 -sroa o3_3.ll > o3_4.ll
#/opt/rocm/llvm/bin/llc -mcpu=gfx908 opt.ll
#/opt/rocm/llvm/bin/llc -mcpu=gfx908 inline.ll
#/opt/rocm/llvm/bin/llc -mcpu=gfx908 o3.ll
#/opt/rocm/llvm/bin/llc -mcpu=gfx908 o3_2.ll
#/opt/rocm/llvm/bin/llc -mcpu=gfx908 o3_3.ll
#/opt/rocm/llvm/bin/llc -mcpu=gfx908 o3_4.ll
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment