ARMCI-MPI

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.

Commit Message	Contributor	Files Modified	Lines Added	Lines Removed	Code Location	Date
Jan 19, 2023 — Jan 19, 2024 Showing page 37 of 54 Search / Filter on:
Implemented buffer guarding for PutS. GetS and AccS guarding and scaling are still to-do. Don't use this yet (disabled unless you set ARMCI_STRIDED_METHOD=SUBARRAY).	James Dinan	More...				almost 14 years ago
Implemented buffer guarding for PutS. GetS and AccS guarding and scaling are still to-do. Don't use this yet (disabled unless you set ARMCI_STRIDED_METHOD=SUBARRAY).	James Dinan	More...				almost 14 years ago
Comment cleanup	James Dinan	More...				almost 14 years ago
Comment cleanup	James Dinan	More...				almost 14 years ago
Adding test for PutS and GetS that uses passed shared buffers to ARMCI.	James Dinan	More...				almost 14 years ago
Adding test for PutS and GetS that uses passed shared buffers to ARMCI.	James Dinan	More...				almost 14 years ago
Fixed the bug in converting from strides to array dimensions. This implementation works for puts, gets, and accs with no scaling. Proper guarding and processing of local buffers isn't yet implemented. Don't use this yet, should be ready within a day or so.	James Dinan	More...				almost 14 years ago
Fixed the bug in converting from strides to array dimensions. This implementation works for puts, gets, and accs with no scaling. Proper guarding and processing of local buffers isn't yet implemented. Don't use this yet, should be ready within a day or so.	James Dinan	More...				almost 14 years ago
Should have been using MPI_Type_true_extent to correctly handle types with a gap at the end. We didn't encounter these types before, but will now with subarray types.	James Dinan	More...				almost 14 years ago
Should have been using MPI_Type_true_extent to correctly handle types with a gap at the end. We didn't encounter these types before, but will now with subarray types.	James Dinan	More...				almost 14 years ago
First pass at ARMCI_STRIDED_METHOD=SUBARRAY. Works for everything except arcmi_test/test, not sure whose bug this is yet.	James Dinan	More...				almost 14 years ago
First pass at ARMCI_STRIDED_METHOD=SUBARRAY. Works for everything except arcmi_test/test, not sure whose bug this is yet.	James Dinan	More...				almost 14 years ago
Even better log2.	James Dinan	More...				almost 14 years ago
Even better log2.	James Dinan	More...				almost 14 years ago
Adding a fast log2 for Jeff.	James Dinan	More...				almost 14 years ago
Adding a fast log2 for Jeff.	James Dinan	More...				almost 14 years ago
Jeff: these are helpful	Jeff Hammond	More...				almost 14 years ago
Jeff: these are helpful	Jeff Hammond	More...				almost 14 years ago
Jeff: this is where i'm going to document how to build NWChem with ARMCI-MPI, since it is a pain in the ass	Jeff Hammond	More...				almost 14 years ago
Jeff: this is where i'm going to document how to build NWChem with ARMCI-MPI, since it is a pain in the ass	Jeff Hammond	More...				almost 14 years ago
Removed the bug I just introduced. Looks like I did something clever to figure out the number of elements in the odd group.	James Dinan	More...				almost 14 years ago
Removed the bug I just introduced. Looks like I did something clever to figure out the number of elements in the odd group.	James Dinan	More...				almost 14 years ago
Fixed an operator precedence bug in malloc call in test_groups, updated odd group creation to include all procs. Removed dead code from test_group_split	James Dinan	More...				almost 14 years ago
Fixed an operator precedence bug in malloc call in test_groups, updated odd group creation to include all procs. Removed dead code from test_group_split	James Dinan	More...				almost 14 years ago
Updated automake dependence to v1.11.1 (released 2009)	James Dinan	More...				almost 14 years ago
Updated automake dependence to v1.11.1 (released 2009)	James Dinan	More...				almost 14 years ago
Made ARMCI_Init/Finalize reference counting. Added missing check for MPI_Finalized().	James Dinan	More...				almost 14 years ago
Made ARMCI_Init/Finalize reference counting. Added missing check for MPI_Finalized().	James Dinan	More...				almost 14 years ago
Apparently you can call ARMCI_Init/Finalize as many times as you want. It doesn't seem safe to finalize on the first matching call to Finalize, seem like you ought to do it on the last (i.e. use ARMCII_GLOBAL_STATE.initialized as a counter and actually finalize when it reaches 1). We will quickly find out it any ARMCI calls are made after the first call to Finalize. Jeff, you missed the matching garbage in Finalize.	James Dinan	More...				almost 14 years ago
Apparently you can call ARMCI_Init/Finalize as many times as you want. It doesn't seem safe to finalize on the first matching call to Finalize, seem like you ought to do it on the last (i.e. use ARMCII_GLOBAL_STATE.initialized as a counter and actually finalize when it reaches 1). We will quickly find out it any ARMCI calls are made after the first call to Finalize. Jeff, you missed the matching garbage in Finalize.	James Dinan	More...				almost 14 years ago

←
1
2
…
33
34
35
36
37
38
39
40
41
…
53
54
→