Skip to content

prov/verbs: Automatic GPU-NIC affinity #12203

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.

Sign up for GitHub

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

Open
gad-arbel wants to merge 5 commits into
base: main
Choose a base branch
from
Open

prov/verbs: Automatic GPU-NIC affinity #12203

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
79d8007
prov/verbs: Add infrastructure for GPU-NIC affinity
gad-arbel Mar 25, 2026
e0b5e90
prov/verbs: Add manual policy to GPU-NIC affinity feature
gad-arbel Mar 30, 2026
c087066
prov/verbs: Add auto policy to GPU-NIC affinity feature
gad-arbel Mar 30, 2026
40c38ce
prov/verbs: Add hwloc to build for GPU-NIC affinity
gad-arbel Mar 30, 2026
d4755d3
prov/verbs: Add tests for GPU-NIC affinity
gad-arbel Apr 30, 2026
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
7 changes: 7 additions & 0 deletions fabtests/Makefile.am
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,6 +65,7 @@ bin_PROGRAMS = \
unit/fi_getinfo_test \
unit/fi_setopt_test \
unit/fi_check_hmem \
unit/fi_nic_affinity_test \
ubertest/fi_ubertest \
multinode/fi_multinode \
multinode/fi_multinode_coll \
Expand Down Expand Up @@ -496,6 +497,11 @@ unit_fi_check_hmem_LDADD = libfabtests.la
unit_fi_check_hmem_CFLAGS = \
$(AM_CFLAGS)

unit_fi_nic_affinity_test_SOURCES = \
unit/nic_affinity_test.c \
$(unit_srcs)
unit_fi_nic_affinity_test_LDADD = libfabtests.la

ubertest_fi_ubertest_SOURCES = \
ubertest/fabtest.h \
ubertest/uber.c \
Expand Down Expand Up @@ -678,6 +684,7 @@ dummy_man_pages = \
man/man1/fi_eq_test.1 \
man/man1/fi_getinfo_test.1 \
man/man1/fi_mr_test.1 \
man/man1/fi_nic_affinity_test.1 \
man/man1/fi_flood.1 \
man/man1/fi_rdm_multi_client.1 \
man/man1/fi_ubertest.1 \
Expand Down
12 changes: 12 additions & 0 deletions fabtests/include/windows/osd.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -802,4 +802,16 @@ OFI_COMPLEX_OPS(float)
OFI_COMPLEX_OPS(double)
OFI_COMPLEX_OPS(long_double)

static inline int setenv(const char *name, const char *value, int overwrite)
{
if (!overwrite && getenv(name) != NULL)
return 0;
return _putenv_s(name, value);
}

static inline int unsetenv(const char *name)
{
return _putenv_s(name, "");
}

#endif /* _WINDOWS_OSD_H_ */
3 changes: 3 additions & 0 deletions fabtests/man/fabtests.7.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -226,6 +226,9 @@ testing scope is limited.
*fi_mr_cache_evict*
: Tests provider MR cache eviction capabilities.

*fi_nic_affinity_test*
: Validates that fi_getinfo returns correct output when the GPU-NIC affinity feature is enabled.

## Multinode

This test runs a series of tests over multiple formats and patterns to help
Expand Down
1 change: 1 addition & 0 deletions fabtests/man/man1/fi_nic_affinity_test.1
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
.so man7/fabtests.7
1 change: 1 addition & 0 deletions fabtests/scripts/runfabtests.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -246,6 +246,7 @@ unit_tests=(
"fi_mr_test"
"fi_cntr_test"
"fi_setopt_test"
"fi_nic_affinity_test"
)

regression_tests=(
Expand Down
Loading