Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[MI300] add CI test stages #2396

Merged
merged 66 commits into from
Oct 19, 2023
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
66 commits
Select commit Hold shift + click to select a range
b3e5ed4
disable non-applicable igemm
Jul 20, 2023
ea74377
add gfx94x to list in ThisSolverIsDeprecatedStatic::IsDisabled
cderb Jul 20, 2023
0302cc8
change enabled tests in test/CMakeLists
cderb Jul 20, 2023
46cc5a9
update maketest
cderb Jul 21, 2023
2646a2d
add ci test
cderb Jul 21, 2023
073a5d5
disable gtest for convbiasactive
cderb Jul 24, 2023
e06d80b
Merge remote-tracking branch 'origin/develop' into cderb/94x_maketest
cderb Jul 24, 2023
fb706bd
disable test
cderb Jul 24, 2023
5788ebf
Promote from public (#42)
cderb Aug 1, 2023
966a338
update git repo for Jenkinsfile
cderb Aug 1, 2023
8706a9f
update Jenkinsfile
cderb Aug 1, 2023
2a48b0b
rename machine
cderb Aug 2, 2023
ba745f4
update regex
cderb Aug 3, 2023
54f590f
Merge remote-tracking branch 'public/develop' into cderb/pub_mergres2
cderb Aug 3, 2023
549dfbe
Revert "update regex"
cderb Aug 3, 2023
aaa7040
add logging for rocminfo in cmake
cderb Aug 3, 2023
8e72ce1
add arch to build list in Jenkinsfile
cderb Aug 3, 2023
83e4c35
bump rocm version
cderb Aug 4, 2023
b56f7c7
dockerfile update
cderb Aug 4, 2023
cf0885d
fix repo
cderb Aug 4, 2023
dd456d0
Merge pull request #43 from ROCmSoftwarePlatform/cderb/pub_mergres2
junliume Aug 5, 2023
58c739d
docker rocm apt version
cderb Aug 5, 2023
078508c
disable miopengemm from apt
cderb Aug 7, 2023
87b7f3e
Merge remote-tracking branch 'origin/develop' into cderb/94x_maketest
cderb Aug 7, 2023
7b18b51
clang-tidy disables
cderb Aug 9, 2023
acfdaeb
fix clang tidy errors
cderb Aug 10, 2023
152ee66
format
cderb Aug 10, 2023
ef365f3
swap rocm install for newer version
cderb Aug 11, 2023
e12968e
swap llvm
cderb Aug 11, 2023
039acef
disable llvm check
cderb Aug 12, 2023
77ba655
clang ignore list
cderb Aug 12, 2023
b7ab032
Merge remote-tracking branch 'public/develop' into cderb/pub_mergres3
cderb Aug 14, 2023
fd0cf2d
Merge pull request #44 from ROCmSoftwarePlatform/cderb/pub_mergres3
junliume Aug 14, 2023
08cf3ab
avoid index bug in hipoc_kernel.hpp sensitized by debug flag -fsanitize
cderb Aug 14, 2023
7d5c8fa
test
cderb Aug 14, 2023
bc79f1d
use rocm5.7
cderb Aug 15, 2023
0e5388a
silence warning
cderb Aug 15, 2023
489de4e
add warning disable
cderb Aug 15, 2023
c56bb25
fix for fsanitize
cderb Aug 18, 2023
3ca7a4b
Merge branch 'develop' into cderb/94x_maketest
cderb Aug 21, 2023
771c427
Merge remote-tracking branch 'public/develop' into cderb/pub_mergres4
cderb Aug 21, 2023
373c66f
#defines expansion
cderb Aug 22, 2023
e9d51ca
Merge branch 'cderb/pub_mergres4' into cderb/94x_maketest
cderb Aug 23, 2023
feab6fd
Fastforward Public Develop (#45)
cderb Aug 25, 2023
126631b
change docker syntax to correct https error.
cderb Aug 29, 2023
54fb391
bugfix
cderb Aug 30, 2023
57f1407
maketest disables
cderb Sep 6, 2023
05a485b
cmake test update
cderb Sep 8, 2023
4ea5056
Merge remote-tracking branch 'origin/develop' into cderb/94x_maketest
cderb Sep 11, 2023
068d5ae
flip default value for 94x test
cderb Sep 14, 2023
e626b44
match dockerfile to public
cderb Sep 14, 2023
ba5ca7e
Merge remote-tracking branch 'public/develop' into cderb/94x_ci_stages
cderb Sep 18, 2023
a152edb
update jenkins label
cderb Sep 18, 2023
c22189a
cleanup
cderb Sep 18, 2023
32ef210
Merge remote-tracking branch 'origin/develop' into cderb/94x_ci_stages
cderb Sep 28, 2023
7b5b7ff
default 94x stage to false
cderb Oct 2, 2023
85b07d0
Merge remote-tracking branch 'origin/develop' into cderb/94x_ci_stages
cderb Oct 3, 2023
4e11bd9
Merge remote-tracking branch 'origin/develop' into cderb/94x_ci_stages
cderb Oct 3, 2023
e309911
Merge remote-tracking branch 'origin/develop' into cderb/94x_ci_stages
cderb Oct 3, 2023
8cd3e71
Merge remote-tracking branch 'origin/develop' into cderb/94x_ci_stages
cderb Oct 4, 2023
1370523
Merge branch 'develop' into cderb/94x_ci_stages
cderb Oct 5, 2023
36a107f
Merge remote-tracking branch 'origin/develop' into cderb/94x_ci_stages
cderb Oct 11, 2023
df3ae58
Merge branch 'develop' into cderb/94x_ci_stages
cderb Oct 13, 2023
4a73ee3
remove gfx94x from gfx10 test
cderb Oct 16, 2023
c8c6e92
Merge branch 'develop' into cderb/94x_ci_stages
cderb Oct 16, 2023
859e221
Merge branch 'develop' into cderb/94x_ci_stages
cderb Oct 18, 2023
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions Dockerfile
Original file line number Diff line number Diff line change
Expand Up @@ -89,8 +89,9 @@ ARG CCACHE_DIR="/tmp"
RUN env
# RUN cget -p $PREFIX install https://github.com/ccache/ccache/archive/7f1572ae9ca958fa923a66235f6a64a360b03523.tar.gz -DZSTD_FROM_INTERNET=ON -DHIREDIS_FROM_INTERNET=ON
ARG CCACHE_COMMIT=7f1572ae9ca958fa923a66235f6a64a360b03523
RUN rm -rf /tmp/ccache* && mkdir /tmp/ccache && wget https://github.com/ccache/ccache/archive/${CCACHE_COMMIT}.tar.gz -O /tmp/ccache.tar.gz && \
tar zxvf /tmp/ccache.tar.gz -C /tmp/ && mkdir /tmp/ccache-${CCACHE_COMMIT}/build && \
RUN rm -rf /tmp/ccache* && mkdir /tmp/ccache
ADD https://github.com/ccache/ccache/archive/${CCACHE_COMMIT}.tar.gz /tmp/ccache.tar.gz
RUN tar zxvf /tmp/ccache.tar.gz -C /tmp/ && mkdir /tmp/ccache-${CCACHE_COMMIT}/build && \
cd /tmp/ccache-${CCACHE_COMMIT}/build && \
cmake -DZSTD_FROM_INTERNET=ON -DHIREDIS_FROM_INTERNET=ON .. && make -j install && rm -rf /tmp/*
RUN ccache -s
Expand Down
82 changes: 82 additions & 0 deletions Jenkinsfile
Original file line number Diff line number Diff line change
Expand Up @@ -469,6 +469,10 @@ pipeline {
name: "TARGET_GFX90A",
defaultValue: true,
description: "")
booleanParam(
name: "TARGET_GFX94X",
defaultValue: false,
description: "")
booleanParam(
name: "TARGET_NAVI21",
defaultValue: false,
Expand Down Expand Up @@ -678,6 +682,19 @@ pipeline {
buildHipClangJobAndReboot(build_type: 'debug', config_targets: Smoke_targets)
}
}
stage('Fp32 Hip Debug gfx94X') {
when {
beforeAgent true
expression { params.TARGET_GFX94X }
}
options {
retry(2)
}
agent{ label rocmnode("gfx94X") }
steps{
buildHipClangJobAndReboot(build_type: 'debug', config_targets: Smoke_targets)
}
}
}
}
stage("Smoke Aux 1") {
Expand Down Expand Up @@ -863,6 +880,32 @@ pipeline {
buildHipClangJobAndReboot(setup_flags: Bf16_flags, config_targets: Smoke_targets)
}
}
stage('Fp16 Hip gfx94X') {
when {
beforeAgent true
expression { params.TARGET_GFX94X && params.DATATYPE_FP16 }
}
options {
retry(2)
}
agent{ label rocmnode("gfx94X") }
steps{
buildHipClangJobAndReboot( setup_flags: Fp16_flags, config_targets: Smoke_targets)
}
}
stage('Bf16 Hip gfx94X') {
when {
beforeAgent true
expression { params.TARGET_GFX94X && params.DATATYPE_BF16 }
}
options {
retry(2)
}
agent{ label rocmnode("gfx94X") }
steps{
buildHipClangJobAndReboot(setup_flags: Bf16_flags, config_targets: Smoke_targets)
}
}
}
}
stage("Full Tests") {
Expand Down Expand Up @@ -914,6 +957,19 @@ pipeline {
buildHipClangJobAndReboot(setup_flags: Bf16_flags + Full_test, build_install: "true")
}
}
stage('Bf16 Hip Install All gfx94X') {
when {
beforeAgent true
expression { params.TARGET_GFX94X && params.DATATYPE_BF16 }
}
options {
retry(2)
}
agent{ label rocmnode("gfx94X") }
steps{
buildHipClangJobAndReboot(setup_flags: Bf16_flags + Full_test, build_install: "true")
}
}
stage('Fp16 Hip All gfx1030') {
when {
beforeAgent true
Expand Down Expand Up @@ -963,6 +1019,19 @@ pipeline {
// buildHipClangJobAndReboot(setup_flags: Full_test, enforce_xnack_on: true)
// }
// }
stage('Fp32 Hip All gfx94X') {
when {
beforeAgent true
expression { params.TARGET_GFX94X && params.DATATYPE_FP32 }
}
options {
retry(2)
}
agent{ label rocmnode("gfx94X") }
steps{
buildHipClangJobAndReboot(setup_flags: Full_test)
}
}
stage('Fp16 Hip Install All Vega20') {
when {
beforeAgent true
Expand Down Expand Up @@ -1028,6 +1097,19 @@ pipeline {
buildHipClangJobAndReboot(setup_flags: Full_test + Fp16_flags, build_install: "true")
}
}
stage('Fp16 Hip All Install gfx94X') {
when {
beforeAgent true
expression { params.TARGET_GFX94X && params.DATATYPE_FP16 }
}
options {
retry(2)
}
agent{ label rocmnode("gfx94X") }
steps{
buildHipClangJobAndReboot(setup_flags: Full_test + Fp16_flags, build_install: "true")
}
}
}
}
stage("Performance Tests - gfx90a") {
Expand Down
3 changes: 1 addition & 2 deletions src/include/miopen/sqlite_db.hpp
Original file line number Diff line number Diff line change
Expand Up @@ -230,8 +230,7 @@ class SQLiteBase
dbInvalid = true;
return;
}

if(!is_system && !filename.empty())
junliume marked this conversation as resolved.
Show resolved Hide resolved
else if(!is_system)
{
auto file = boost::filesystem::path(filename_);
const auto directory = file.remove_filename();
Expand Down
1 change: 1 addition & 0 deletions src/kernel_warnings.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -68,6 +68,7 @@ static std::vector<std::string> HipKernelWarnings()
"-Wno-double-promotion",
"-Wno-exit-time-destructors",
"-Wno-extra-semi",
"-Wno-extra-semi-stmt",
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

[Q] This looks like a workaround of some other problem (albeit non-critical). Can we resolve it right here?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'll need to regain access to a gfx94x machine before I can reassess.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@cderb [Recommendation] Fix kernel build warnings right here and revert this change.


As usual, "recommendations" are up to the author discretion. The author can actually resolve or just mark as resolved -- no explanation needed.

"-Wno-float-conversion",
"-Wno-gnu-anonymous-struct",
"-Wno-gnu-zero-variadic-macro-arguments",
Expand Down
4 changes: 2 additions & 2 deletions src/sqlite_db.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -78,12 +78,12 @@ class SQLite::impl
{
sqlite3* ptr_tmp = nullptr;
int rc = 0;
#ifdef __clang__
#if defined(__clang__) || defined(__llvm__)
#pragma clang diagnostic push
#pragma clang diagnostic ignored "-Wcast-function-type-strict"
#endif
sqlite3_auto_extension(reinterpret_cast<void (*)(void)>(miopen_sqlite3_memvfs_init));
#ifdef __clang__
#if defined(__clang__) || defined(__llvm__)
#pragma clang diagnostic pop
#endif
// Open an in-memory database to use as a handle for loading the memvfs extension
Expand Down
Loading