Project

General

Profile

Activity

From 06/24/2017 to 07/23/2017

07/23/2017

10:03 AM Bug #1640: Segfault during migration for AMPI in SMP mode with "-tracemode projections"
Megampi exhibits the same problem. Matthias Diener
05:55 AM Bug #1640: Segfault during migration for AMPI in SMP mode with "-tracemode projections"
I can reproduce this, and verify that the failure only happens in SMP mode with tracing on. Additionally, it only hap... Sam White

07/22/2017

06:02 PM Bug #1640: Segfault during migration for AMPI in SMP mode with "-tracemode projections"
Going back to commit 9df608634 (which is the one before AMPI tracing changes were merged) shows the same crash, so th... Matthias Diener
05:44 PM Bug #1640 (In Progress): Segfault during migration for AMPI in SMP mode with "-tracemode projecti...
When running @make OPTS="-tracemode projections" test@ in @examples/ampi/Cjacobi3d@, the application crashes after th... Matthias Diener

07/21/2017

11:11 AM Bug #1547: Deprecate the FFT library in ck-libs in favor of Nikhil's new FFT library
Just a few questions about the way forward.
# The prior library's documentation notes that the AMPI API was 'not in ...
Justin Szaday

07/20/2017

06:25 PM Bug #1639 (Implemented): AMPI MPI_IN_PLACE support is broken
Sam White
04:52 PM Bug #1639 (In Progress): AMPI MPI_IN_PLACE support is broken
https://charm.cs.illinois.edu/gerrit/#/c/2831/ Sam White
04:03 PM Bug #1639: AMPI MPI_IN_PLACE support is broken
3 things to be careful of when handling MPI_IN_PLACE:
- error checking of send{buf,type,count} has to be aware of MP...
Sam White
02:57 PM Bug #1639 (Merged): AMPI MPI_IN_PLACE support is broken
The MPI standard says that when MPI_IN_PLACE is passed as the sendbuf, the corresponding sendtype and sendcount (and ... Sam White

07/19/2017

05:05 PM Documentation #1638 (New): Manual has incorrect populateInitial API
It still uses the old API that took number of initial elements instead of CkArrayOptions. Furthermore, the manual has... Eric Mikida
03:02 PM Bug #1409: verbs crashes on Stampede KNL and Bridges
For examples/converse/pingpong, I saw another verbs layer related crash:
[nbhat4@r135 pingpong]$ ./charmrun +p2 ....
Nitin Bhat
02:59 PM Bug #1409: verbs crashes on Stampede KNL and Bridges
I saw this crash for verbs-linux-x86_64 on Bridges as well.
Nitin Bhat

07/18/2017

06:37 PM Feature #1468 (New): Enable pre-pinning memory for the zero-copy message sends through the Entry ...
Sam White
06:37 PM Feature #1637 (Merged): Zero-copy send support in the OFI layer
Perhaps Intel can be talked into implementing the necessary support for the new OFI layer? Otherwise we'll need to do... Sam White
02:42 PM Feature #954: Update AMPI's version of MPICH test suite
I ran the mpich-3.2 test suite on mainline AMPI today. There are many more tests than the mpich-2 test suite. To summ... Sam White
10:43 AM Bug #1630: OpenAtom crashes during launch at startup on BlueWaters using master charm (again)
Run of openatom build with charm version that produced the original failure v6.8.0-beta1-295-g8b554d0, now runs to co... Eric Bohm
09:53 AM Bug #1636 (New): Update the machine layer template configuration in src/arch/template
The template directory src/arch/template serves as a template for developing a new machine layer. The configuration s... Nitin Bhat
09:39 AM Bug #1634: HDF5 issues in AMPI
'-tlsglobals' currently requires static linking: https://charm.cs.illinois.edu/redmine/issues/1220
For migration, ...
Sam White

07/17/2017

02:06 PM Bug #1635 (New): Cleanup and Activate the Charm++ Topology Examples
The 3D examples under charm++/topology (jacobi3d and matmul3d) fail due to Arithmetic/Floating point errors in the Co... Justin Szaday
11:36 AM Bug #1239: Cleanup reduction uses in the runtime
I cleaned up all of the uses of `setReductionClient` in the tests except for those in `tests/charm++/megatest/reducti... Justin Szaday
10:57 AM Bug #1627 (Closed): preserving OPTS_CXX breaks application builds on different machines
Phil Miller
09:41 AM Bug #1630 (Rejected): OpenAtom crashes during launch at startup on BlueWaters using master charm ...
I managed to halfway automate the bisect process and have reproduced Juan's finding. Binaries produced by current ch... Eric Bohm
11:07 PM Bug #1630: OpenAtom crashes during launch at startup on BlueWaters using master charm (again)
I built latest charm master branch on BW (up to 58f6f16077535e24f5c474cc1ee7ea8aaa059bb2: Bug #1632: ST_RecursivePart... Juan Galvez

07/16/2017

02:55 PM Bug #1634: HDF5 issues in AMPI
The issues to test and improve are:
* -Test if applications work with the shared library (currently, only the stat...
Matthias Diener
02:52 PM Bug #1634 (New): HDF5 issues in AMPI
The HDF5 library is available for AMPI at
https://charm.cs.illinois.edu/gerrit/#/admin/projects/hdf5-ampi
This b...
Matthias Diener

07/15/2017

04:51 PM Bug #1448 (Implemented): Potential buffer overflows in fscanf()
Patch here: https://charm.cs.illinois.edu/gerrit/2796 Matthias Diener
04:31 PM Feature #1573 (Merged): Make HDF5 library available for AMPI
Matthias Diener
04:20 PM Feature #1573 (Implemented): Make HDF5 library available for AMPI
The current version of hdf5-ampi is usable and passes the PlasCom2 tests, which exercises a substantial part of the A... Matthias Diener

07/14/2017

05:08 PM Bug #1630: OpenAtom crashes during launch at startup on BlueWaters using master charm (again)
Sounds reasonable, I agree. Juan Galvez
05:06 PM Bug #1630: OpenAtom crashes during launch at startup on BlueWaters using master charm (again)
Also, one hopes that fixing #1632 will make debugging this issue faster Phil Miller
05:06 PM Bug #1630: OpenAtom crashes during launch at startup on BlueWaters using master charm (again)
The relationship linkage is their discovery - a link in the comments here doesn't indicate over there where #1632 cam... Phil Miller
05:00 PM Bug #1630: OpenAtom crashes during launch at startup on BlueWaters using master charm (again)
Like I said, these shouldn't be related at all, since the error on Edison would not appear on Blue Waters. Juan Galvez
01:23 PM Bug #1630: OpenAtom crashes during launch at startup on BlueWaters using master charm (again)
Opened new issue for that here: https://charm.cs.illinois.edu/redmine/issues/1632
Fix coming soon.
Juan Galvez
01:10 PM Bug #1630: OpenAtom crashes during launch at startup on BlueWaters using master charm (again)
That wouldn't happen on Blue Waters. Juan Galvez
01:08 PM Bug #1630: OpenAtom crashes during launch at startup on BlueWaters using master charm (again)
Edison dies with a topology error at higher node counts. Unclear if this is actually a related bug, or a different i... Eric Bohm
04:55 PM Bug #1632 (Merged): ST_RecursivePartition (spanning tree) crashes during init with large node cou...
Phil Miller
02:32 PM Bug #1632: ST_RecursivePartition (spanning tree) crashes during init with large node counts on sy...
patch tested and found to address the issue correctly for large (>32node) openatom runs on edison. Eric Bohm
01:28 PM Bug #1632 (Implemented): ST_RecursivePartition (spanning tree) crashes during init with large nod...
Juan Galvez
01:28 PM Bug #1632: ST_RecursivePartition (spanning tree) crashes during init with large node counts on sy...
Fix here: https://charm.cs.illinois.edu/gerrit/#/c/2789/ Juan Galvez
01:23 PM Bug #1632 (Merged): ST_RecursivePartition (spanning tree) crashes during init with large node cou...
As documented by Eric B:
"Edison dies with a topology error at higher node counts. Unclear if this is actually a rel...
Juan Galvez
02:26 PM Bug #1633: NodeGroup Broadcasts creates many copies of the message for point to point sends
This all boils down to @src/arch/util/machine-broadcast.c@ and the setting of @CMK_BROADCAST_USE_CMIREFERENCE@ in @sr... Phil Miller
02:14 PM Bug #1633 (In Progress): NodeGroup Broadcasts creates many copies of the message for point to poi...
Broadcast to a nodegroup results in the message being copied once for each destination node. These are then sent in ... Eric Bohm
11:05 AM Bug #1201: SMP builds segfault on NULL lock in tests/charm++/chkpt
As for Sam's note, Ck*Array*ReductionMgr got removed - CkNodeReductionMgr is still very much there, to support NodeGr... Phil Miller
10:59 AM Bug #1201: SMP builds segfault on NULL lock in tests/charm++/chkpt
I just checked - this failure wasn't seen in Jenkins Nightly-Build for mpi-linux-x86_64-smp or its successors since t... Phil Miller
09:10 AM Bug #1627: preserving OPTS_CXX breaks application builds on different machines
By setting __INTEL_PRE_CFLAGS rather than OPTS_CXX I am able to accomplish what I want. Jim Phillips
10:33 PM Bug #1507: ckio test failure on gni-crayxc
I'm getting this failure again on Blue Waters using charm version 5fd855b. It seems to only be happening with large ... Thomas Quinn
09:51 PM Feature #1631 (Merged): AMPI lacks API to clear LB database
Sam White

07/13/2017

04:57 PM Bug #1627: preserving OPTS_CXX breaks application builds on different machines
Actually, I think I can do what I need using the __INTEL_PRE_CFLAGS environment variable. Jim Phillips
04:53 PM Bug #1627: preserving OPTS_CXX breaks application builds on different machines
I don't think there's anywhere that we 'support' building applications in one environment against a build of Charm++ ... Phil Miller
04:46 PM Bug #1627: preserving OPTS_CXX breaks application builds on different machines
My automated builds use the same (old, picky) environment for Charm++ and NAMD. The problem arises when people try t... Jim Phillips
04:19 PM Bug #1627 (Feedback): preserving OPTS_CXX breaks application builds on different machines
Sam White
03:40 PM Feature #1434 (Implemented): optimize degenerate CkLoop cases
https://charm.cs.illinois.edu/gerrit/#/c/2782/ Sam White
10:06 PM Feature #1237: Onesided sender side implementation for GNI layer
Related patches merged:
Allocate machine specific info along with the charm message: https://charm.cs.illinois.edu...
Sam White
10:04 PM Cleanup #1203 (Merged): AMPI forces builds to be serial for ROMIO
https://charm.cs.illinois.edu/gerrit/#/c/2761/ Sam White
10:03 PM Feature #952 (In Progress): Update AMPI's version of ROMIO
Marking "In Progress" to reflect the fact that we've updated to a more recent version, though not the most recent ver... Sam White
09:56 PM Feature #1631 (Implemented): AMPI lacks API to clear LB database
https://charm.cs.illinois.edu/gerrit/#/c/2778/ Sam White

07/12/2017

05:36 PM Feature #1631 (Merged): AMPI lacks API to clear LB database
AMPI has APIs for starting and stopping LB instrumentation, but no method to clear or reset the LB database. This cou... Sam White
05:18 PM Bug #1630 (Rejected): OpenAtom crashes during launch at startup on BlueWaters using master charm ...
OpenAtom ran fine on 75ea9d902febb0552170ab61ff79d14588494d7c
But crashes during application startup with current ...
Eric Bohm
11:12 AM Bug #1629: Valgrind shows invalid writes at GNI_SmsgSendWTag calls for charm programs built on gn...
A third possibility that I neglected up in comment 2 - all of this space is properly allocated, just through a functi... Phil Miller
09:25 AM Bug #1331: Isomalloc hangs in startup for Clang non-SMP builds
os-isomalloc is now the default version of isomalloc used on all non-SMP builds: https://charm.cs.illinois.edu/gerrit... Sam White
09:24 AM Bug #1375: os-isomalloc failures during startup on SMP builds
os-isomalloc is now the default version of isomalloc used on all non-SMP builds: https://charm.cs.illinois.edu/gerrit... Sam White

07/11/2017

03:29 PM Projections Feature #1525: Bracketed User Events Track Numerical Values
Agreed, being able to provide data and use a vector would be useful. In the mean time, we have a tracing method calle... Ronak Buch
03:25 PM Documentation #1588 (In Progress): FAQ : How do I use GPGPUs in Charm++?
Michael Robson
03:25 PM Documentation #1596 (In Progress): FAQ: 2 . 0 . 7 How do I specify the processors I want to use?
Michael Robson
02:34 PM Feature #1626 (Merged): Modify the keyword "rdma" to "nocopy"
Changed "rdma" .ci file keyword to "nocopy" and the "rdma()" send wrapper routine to "CkSendBuffer()". Sam White
02:11 PM Bug #1572 (In Progress): Improve pup_stl performance
Eric Bohm
01:06 PM Documentation #1603 (Merged): FAQ: 5 . 0 . 4 How do I use gdb with Charm++ programs?
Sam White
01:06 PM Documentation #1597 (Merged): FAQ: 3 . 0 . 2 What is an ``entry method''?
Sam White
01:06 PM Documentation #1595 (Merged): FAQ: 2 . 0 . 3 How do I compile Charm++?
Sam White
11:55 AM Documentation #1592 (Merged): FAQ: 1 . 0 . 7 What is the future of Charm++?
Eric Bohm
11:47 AM Bug #1625 (Merged): Rdma example rdma/simpleRdma crashes on pamilrts-bluegeneq for within physica...
Phil Miller
11:14 AM Bug #1045: Array pointers in messages can be overwritten by generated code in the message class c...
My original notes from 2011-01-07:
> Shortly before Christmas, Akhil and I isolated a bug in Charm++ code
> that ...
Phil Miller

07/10/2017

06:26 PM Bug #1629: Valgrind shows invalid writes at GNI_SmsgSendWTag calls for charm programs built on gn...
And the addresses being touched in this run:... Phil Miller
06:18 PM Bug #1629: Valgrind shows invalid writes at GNI_SmsgSendWTag calls for charm programs built on gn...
In @converse/commbench@, I get the initial report from ~10 instruction addresses in @GNII_SmsgSend@, and the followin... Phil Miller
05:29 PM Bug #1629: Valgrind shows invalid writes at GNI_SmsgSendWTag calls for charm programs built on gn...
I know we want to make a comparison to other errors that Valgrind reports, but this one isn't a leak. It's an outrigh... Phil Miller
04:14 PM Bug #1629: Valgrind shows invalid writes at GNI_SmsgSendWTag calls for charm programs built on gn...
Are there other leaks that look like this that don't come from the Charm initialization phases? This particular leak,... Sam White
03:37 PM Bug #1629 (New): Valgrind shows invalid writes at GNI_SmsgSendWTag calls for charm programs built...
1. Build charm with target gni-crayxc and flags -g -O0
2. Build examples/charm++/hello/1darray with -g -O0
3. Run t...
Nitin Bhat
01:14 AM Bug #1628 (New): Error while compiling LogAnalyzer tool in BigNetSim
Hello,
I am trying to convert the trace logs produced by the BigSim Emulator to be able to analyze them and find a...
Ashmita Raju

07/07/2017

02:59 PM Bug #1627: preserving OPTS_CXX breaks application builds on different machines
Why are the runtime system and application being built on two different machines, with such distinct environments? Co... Phil Miller
02:30 PM Bug #1625 (Implemented): Rdma example rdma/simpleRdma crashes on pamilrts-bluegeneq for within ph...
The machine specific information is now allocated with the charm message (containing the buffer and the md message) t... Nitin Bhat
02:27 PM Documentation #1595 (Implemented): FAQ: 2 . 0 . 3 How do I compile Charm++?
https://charm.cs.illinois.edu/gerrit/#/c/2757/ Sam White
02:27 PM Documentation #1597 (Implemented): FAQ: 3 . 0 . 2 What is an ``entry method''?
https://charm.cs.illinois.edu/gerrit/#/c/2757/ Sam White
02:27 PM Documentation #1603 (Implemented): FAQ: 5 . 0 . 4 How do I use gdb with Charm++ programs?
https://charm.cs.illinois.edu/gerrit/#/c/2757/ Sam White
01:40 PM Bug #1331: Isomalloc hangs in startup for Clang non-SMP builds
After upgrading prowess to Xenial (default compiler: gcc 5.4.0), Jenkins nightly build failed with mpi-linux-x86_64 i... Sam White
12:02 PM Feature #1573 (In Progress): Make HDF5 library available for AMPI
Patch to merge hdf5 into the charm/ampi distribution was abandoned in favor of hosting a separate repo for our modifi... Sam White
11:28 AM Bug #1240 (Merged): MPI_COMM_SELF handling issues in AMPI
Sam White

07/06/2017

06:32 PM Bug #1240: MPI_COMM_SELF handling issues in AMPI
Related: make collectives optimizations for MPI_COMM_SELF apply to all comms with size of 1: https://charm.cs.illinoi... Sam White
12:59 PM Bug #1240: MPI_COMM_SELF handling issues in AMPI
It would probably be good to get this into 6.8.0 Sam White
12:59 PM Bug #1240 (Implemented): MPI_COMM_SELF handling issues in AMPI
https://charm.cs.illinois.edu/gerrit/#/c/2753/ Sam White
10:53 AM Bug #1240: MPI_COMM_SELF handling issues in AMPI
Yes, I'm working on the full support now Sam White
05:44 PM Documentation #1592: FAQ: 1 . 0 . 7 What is the future of Charm++?
Where was this implemented? Sam White
05:43 PM Documentation #1590 (Merged): FAQ: 1 . 0 . 5 Does anybody actually use Charm++?
https://charm.cs.illinois.edu/gerrit/#/c/2718/ Sam White
05:42 PM Documentation #1587 (Merged): FAQ: 1 . 0 . 4 Will Charm++ run on my machine?
Sam White
05:04 PM Bug #1560: icc build fails on NASA Pleiades
Jim, could you please elaborate? Phil Miller
04:56 PM Bug #1560: icc build fails on NASA Pleiades
When I pass OPTS_CXX="-gxx-name=g++44" to the build script it is preserved for future charmc invocations, which break... Jim Phillips
04:52 PM Bug #1627 (Closed): preserving OPTS_CXX breaks application builds on different machines
After https://charm.cs.illinois.edu/gerrit/gitweb?p=charm.git;a=commit;h=cf9d5d1f4ca0f4d3c27727f8ab30e665e24b9154 com... Jim Phillips
04:17 PM Bug #1155: AMPI's non-blocking collectives are not sequenced
I updated the above patches to support all kinds of bcasts (non-blocking, intercomm). Sam White
03:48 PM Feature #1626 (Implemented): Modify the keyword "rdma" to "nocopy"
Fix: https://charm.cs.illinois.edu/gerrit/#/c/2754/ Nitin Bhat
11:49 AM Feature #1626: Modify the keyword "rdma" to "nocopy"
As the Zero Copy API supports within process transfers and non-RDMA layers, it was decided over a slack discussion on... Nitin Bhat
11:47 AM Feature #1626 (Merged): Modify the keyword "rdma" to "nocopy"
Nitin Bhat

07/05/2017

06:09 PM Bug #1625 (Merged): Rdma example rdma/simpleRdma crashes on pamilrts-bluegeneq for within physica...
This only happens when the two processes are within the same physical node. The crash is not because of the newly add... Nitin Bhat
02:09 PM Documentation #1590 (Implemented): FAQ: 1 . 0 . 5 Does anybody actually use Charm++?
Eric Bohm
02:08 PM Documentation #1592 (Implemented): FAQ: 1 . 0 . 7 What is the future of Charm++?
Eric Bohm
01:36 PM Bug #1624: missing fences in CkLoop implementation
Are CmiMemoryReadFence and CmiMemoryWriteFence somehow not portable and sufficient? Otherwise 6.8.0 CkLoop is broken ... Jim Phillips

07/03/2017

04:02 PM Bug #1624: missing fences in CkLoop implementation
The group discussed this in core at some point. The code is apparently fine on x86 with its strict memory ordering se... Phil Miller
01:12 PM Bug #1624 (Merged): missing fences in CkLoop implementation
Reported by Sameer Kumar, CkLoop reportFinished() needs a write fence to guarantee that results are visible to the th... Jim Phillips
04:00 PM Bug #1623 (Merged): Fix autobuild failure in gni-crayxc-smp caused by examples/rdma/simpleRdma be...
Phil Miller
02:06 PM Bug #1076: failure to exit after long run on PSC Bridges
The hangs went away after switching from Intel MPI to OpenMPI. I'm assuming it was a bug in Intel MPI on OmniPath. Jim Phillips
08:22 AM Bug #1076: failure to exit after long run on PSC Bridges
It would be helpful to know if this is still an issue. MPI_Test is not supposed to be a blocking operation, though th... Phil Miller
08:13 AM Bug #1076: failure to exit after long run on PSC Bridges
Format description Phil Miller
08:12 AM Bug #1001 (Closed): netlrts with replicas leaves zombie processes behind
Clearing out bugs like this. Phil Miller

06/30/2017

04:50 PM Bug #1623 (Implemented): Fix autobuild failure in gni-crayxc-smp caused by examples/rdma/simpleRd...
Fix: https://charm.cs.illinois.edu/gerrit/#/c/2743/
Copied the received ack message into a new message before the ...
Nitin Bhat
03:38 PM Bug #1623 (Merged): Fix autobuild failure in gni-crayxc-smp caused by examples/rdma/simpleRdma be...
Autobuild failure: http://charm.cs.illinois.edu/autobuild/old.2017_06_30__01_01/gni-crayxc-smp.txt
The crash has t...
Nitin Bhat
02:14 PM Bug #529: LrtsInitCpuTopo() operates asynchronously and overlaps with group constructors
I'm referring to *this* bug (529), not 1381. Juan Galvez
02:13 PM Bug #529: LrtsInitCpuTopo() operates asynchronously and overlaps with group constructors
I still get the exact same "Length mismatch!!" abort with verbs-linux-x86_64 on Quartz on today's master version on c... Sam White
02:02 PM Bug #529: LrtsInitCpuTopo() operates asynchronously and overlaps with group constructors
This particular bug has probably been solved by the recent cputopology patches:
https://charm.cs.illinois.edu/gerrit...
Juan Galvez

06/29/2017

03:17 PM Bug #1620 (Rejected): Fatal error on PE 0> Converse zero handler executed while running with +bglog
As Jeff Hammond said on the email thread for this, the abort lists the reason for the failure:... Sam White
09:42 PM Bug #1621 (Merged): Fix autobuild failure in net-linux-x86_64 caused by examples/rdma/stencil3d
https://charm.cs.illinois.edu/gerrit/#/c/2736/ Sam White
09:41 PM Bug #1616 (Merged): Race condition during ckRestart causes hang after merging patch for bug #1576
Sam White

06/28/2017

04:03 PM Bug #1202: Memory leaks in converse's cldb
Here's the new output:... Sam White
03:55 PM Documentation #1515 (Merged): Document [inline] entry method optimization for const& args
Sam White
03:24 PM Bug #1616 (Implemented): Race condition during ckRestart causes hang after merging patch for bug ...
Actually, the option of doing p2p send for each physical node is probably not robust, and instead a message by each P... Juan Galvez
10:02 AM Bug #1616: Race condition during ckRestart causes hang after merging patch for bug #1576
If that's the case about the p2p sends, it may be a perfectly acceptable option for the time being. I guess we just n... Phil Miller
09:58 AM Bug #1616: Race condition during ckRestart causes hang after merging patch for bug #1576
I agree that p2p sends is not the best option, although we are talking one message per physical node (not per PE or p... Juan Galvez
09:53 AM Bug #1616: Race condition during ckRestart causes hang after merging patch for bug #1576
The all-to-PE0 p2p messages definitely won't be acceptable, because this is code that does run at large scale. It cou... Phil Miller
09:40 AM Bug #1616: Race condition during ckRestart causes hang after merging patch for bug #1576
Have been investigating and have much better understanding of what's going on.
In the original code (before implem...
Juan Galvez
11:48 AM Feature #1622 (New): Optimize AMPI RMA routines for shared memory
Currently, AMPI does nothing to optimize RMA routines for shared memory. This patch optimizes RMA for PE-local calls,... Sam White
10:14 AM Bug #1621 (Merged): Fix autobuild failure in net-linux-x86_64 caused by examples/rdma/stencil3d
The non-rdma case (!CMK_ONESIDED_IMPL) is failing when migration occurs. This issue was not seen while testing on my ... Nitin Bhat
01:53 AM Bug #1620 (Rejected): Fatal error on PE 0> Converse zero handler executed while running with +bglog
I tried executing the test program given in the distribution (jocobi3d under ampi)
While running without +bglog, ...
Ashmita Raju
10:34 PM Documentation #1589 (Merged): FAQ: how do I use Charm++ on Xeon Phi?
Phil Miller
10:33 PM Documentation #1196 (Merged): Document and exemplify chare class inheritance
Phil Miller
10:33 PM Documentation #1265 (Merged): Document LLVM OpenMP runtime integration
Phil Miller
10:31 PM Bug #1617 (Merged): OpenAtom experiences significant slowdown (observed up to 25%) after changing...
Phil Miller

06/27/2017

03:45 PM Bug #1614 (Merged): SDAG Rdma example /examples/rdma/stencil3D failing on nightly build because o...
Sam White
03:38 PM Bug #1240: MPI_COMM_SELF handling issues in AMPI
To fix this properly, we might have to implement full support for MPI_COMM_SELF in AMPI. For example, how would MPI_C... Matthias Diener
03:32 PM Bug #1240 (In Progress): MPI_COMM_SELF handling issues in AMPI
This patch has some improvements, but is still broken for the ROMIO MPI_COMM_SELF MPI_Comm_dup case: https://charm.cs... Sam White
02:12 PM Feature #1573 (Implemented): Make HDF5 library available for AMPI
Sam White
10:30 AM Bug #1616: Race condition during ckRestart causes hang after merging patch for bug #1576
On Jenkins Nightly-Build job, the hang in question still appears, and quite consistently. Phil Miller
09:48 PM Bug #1202: Memory leaks in converse's cldb
I see memory leaks from the new topology code merge in the last couple weeks. I'm not sure if they are really new, or... Sam White
09:40 PM Bug #1618 (Merged): AMPI's Waitany is broken
Sam White

06/26/2017

04:46 PM Documentation #1611 (In Progress): Document network dependent rdma thresholds, above which benefi...
Performance Evaluation before 6.8.0 release: https://docs.google.com/spreadsheets/d/1dL2AYO-kKZYGeM1PLDRX3k9z_uL95eDI... Nitin Bhat
11:43 AM Bug #1618 (Implemented): AMPI's Waitany is broken
Fix is to not decrement the ampiParent's numBlockedReqs if it is already zero. This makes sense because numBlockedReq... Sam White
11:40 AM Bug #1618 (Merged): AMPI's Waitany is broken
AMPI_Waitany falsely aborts with the following message when 2+ messages for reqs that are involved in the Waitany are... Sam White

06/24/2017

02:01 PM Support #1391 (Resolved): Add an SMP/multicore build test to Jenkins
Sam White
08:50 AM Support #1391: Add an SMP/multicore build test to Jenkins
I think this is done? Phil Miller
01:36 PM Bug #1616: Race condition during ckRestart causes hang after merging patch for bug #1576
Maybe it's been solved for Windows too but there are lingering issues from previous chkpt hangs? Maybe dexterity shou... Juan Galvez
08:46 AM Bug #1616: Race condition during ckRestart causes hang after merging patch for bug #1576
Given that checkpoint-restart really doesn't get used on Windows (which is mostly for NAMD users at present), and the... Phil Miller
 

Also available in: Atom