Open Hub
Open Hub

 Settings |  Report Duplicate

1
I Use This!
Activity Not Available

Commits : Listings

Analyzed about 1 year ago. based on code collected about 1 year ago.
Jan 19, 2023 — Jan 19, 2024
Commit Message Contributor Files Modified Lines Added Lines Removed Code Location Date
Cleanup -pedantic
James Dinan
More... over 13 years ago
Cleanup -pedantic
James Dinan
More... over 13 years ago
Extracted buffer flushed into ARMCII_Flush_local() so we can place elsewhere if needed.
James Dinan
More... over 13 years ago
Extracted buffer flushed into ARMCII_Flush_local() so we can place elsewhere if needed.
James Dinan
More... over 13 years ago
This patch adds support for calls to ARMCI_Malloc where all processes pass an allocation size of 0 bytes. Before it was detected only when calling ARMCI_Free and reported as an error. Not that it was ever actually detected since no application makes a call like this.
James Dinan
More... almost 14 years ago
This patch adds support for calls to ARMCI_Malloc where all processes pass an allocation size of 0 bytes. Before it was detected only when calling ARMCI_Free and reported as an error. Not that it was ever actually detected since no application makes a call like this.
James Dinan
More... almost 14 years ago
Replaced return on error with a call to ARMCI_Error(). This was causing tests to hang on BG/P after falling over dead with datatype bug induced memory corruption.
James Dinan
More... almost 14 years ago
Replaced return on error with a call to ARMCI_Error(). This was causing tests to hang on BG/P after falling over dead with datatype bug induced memory corruption.
James Dinan
More... almost 14 years ago
Create each ARMCI mutex in its own window. This fixes a performance bug (that we knew about but ignored because it seemed like not a big deal) reported by Alex Dickson. All ARMCI tests pass; still needs to be tested against GA and on a real network.
James Dinan
More... almost 14 years ago
Create each ARMCI mutex in its own window. This fixes a performance bug (that we knew about but ignored because it seemed like not a big deal) reported by Alex Dickson. All ARMCI tests pass; still needs to be tested against GA and on a real network.
James Dinan
More... almost 14 years ago
yeah...
Jeff Hammond
More... almost 14 years ago
yeah...
Jeff Hammond
More... almost 14 years ago
useful on cray
Jeff Hammond
More... almost 14 years ago
useful on cray
Jeff Hammond
More... almost 14 years ago
this is a script to run make check on Cray
Jeff Hammond
More... almost 14 years ago
this is a script to run make check on Cray
Jeff Hammond
More... almost 14 years ago
Turned off MODE_SET by default. Give CFLAGS=-DMODE_SET to enable.
James Dinan
More... almost 14 years ago
Turned off MODE_SET by default. Give CFLAGS=-DMODE_SET to enable.
James Dinan
More... almost 14 years ago
Use double precision accumulate in contiguous-bench.
James Dinan
More... almost 14 years ago
Use double precision accumulate in contiguous-bench.
James Dinan
More... almost 14 years ago
Bugfix: int passed, double expected.
James Dinan
More... almost 14 years ago
Bugfix: int passed, double expected.
James Dinan
More... almost 14 years ago
Relaxing version requirement for automake.
James Dinan
More... almost 14 years ago
Relaxing version requirement for automake.
James Dinan
More... almost 14 years ago
Verbose output for those intrepid enough to enable NO_SEATBELTS.
James Dinan
More... almost 14 years ago
Verbose output for those intrepid enough to enable NO_SEATBELTS.
James Dinan
More... almost 14 years ago
Prioritize MPICC over CC for the compiler to pick.
Pavan Balaji
More... almost 14 years ago
Prioritize MPICC over CC for the compiler to pick.
Pavan Balaji
More... almost 14 years ago
Refactored mutex_grp to mutex_hdl. Fixed spinlocks, these are working again but I've figured out a better (non-livelock) algorithm that needs to be implemented. Will need to add a configure option to select betwee spin vs queue locks.
James Dinan
More... almost 14 years ago
Refactored mutex_grp to mutex_hdl. Fixed spinlocks, these are working again but I've figured out a better (non-livelock) algorithm that needs to be implemented. Will need to add a configure option to select betwee spin vs queue locks.
James Dinan
More... almost 14 years ago