Project

General

Profile

Activity

From 09/01/2017 to 09/30/2017

09/30/2017

04:28 PM Bug #1701 (New): Cannot have non-copyable types in constructor arguments
If I try to pass a non-copyable type to a chare constructor my only option is to move an lvalue reference, which is n... Nils Deppe
11:37 AM Bug #1700 (New): Overloaded reduction targets result in compilation error
In my ci file I have:... Nils Deppe
10:52 AM Bug #1699 (New): [inline] entry methods should use perfect forwarding in C++11
Currently inline entry methods are not as useful as they could be in C++11 because they do not use perfect forwarding... Nils Deppe
10:42 AM Bug #1698 (New): [inline] template entry methods do not work
I'm trying to call an @[inline]@ template entry method, and it fails because the generated code calls:... Nils Deppe

09/29/2017

04:27 PM Cleanup #567 (Merged): Charm++ barnes-charm example cleanup
Phil Miller
04:27 PM Bug #1437 (Merged): CkLoop worker traces to previous entry on pe rather than caller
Phil Miller
02:02 PM Feature #1695: Allow "type aliases" for explicit instantiations of member function templates
Alright, cool. I'm not sure how far along you are with replacing interface files, but I have managed to eliminate the... Nils Deppe
12:18 PM Support #1534 (Closed): Print out the job output, error and log for failed BGQ autobuild jobs and...
Phil Miller
10:12 AM Bug #829: CkLoop projections tracing doesn't reflect the origin/traceback of work
Is this a direct duplicate of #1437? Or is there something more subtle here still to be addressed beyond that? Phil Miller

09/28/2017

05:04 PM Bug #1437: CkLoop worker traces to previous entry on pe rather than caller
https://charm.cs.illinois.edu/gerrit/#/c/3084/ Seonmyeong Bak
05:03 PM Bug #1437 (Implemented): CkLoop worker traces to previous entry on pe rather than caller
Seonmyeong Bak
02:28 PM Bug #78: AMPI failure with migration under Cray compiler due to tcmalloc bugs or incompatibility
I don't think there's going to be urgency for fixing this any time soon, so pulling it off any release target. Phil Miller
12:39 PM Bug #1696: Allow default arguments in group entry methods
Alright, thanks for the clarification! Nils Deppe
12:20 PM Bug #1696: Allow default arguments in group entry methods
I'd call this a bug, indeed. Default arguments to entry methods are supported somewhat intentionally, and it's incorr... Phil Miller
12:39 PM Support #1681: Document Exception Handling in Charm++
Thanks for the thorough explanation! I look forward to the Charm++ changes resulting from 1) and 2). These might even... Nils Deppe
12:06 PM Support #1681: Document Exception Handling in Charm++
Thanks for bringing these questions.
1. @std::terminate@: We should probably modify the runtime system itself to m...
Phil Miller
12:37 PM Bug #1685: charmc Chokes on @explicit@ constructors
I'm trying to maintain consistency between my C++ code and the CI file, that is all. I would not expect marking the c... Nils Deppe
12:11 PM Bug #1685 (Feedback): charmc Chokes on @explicit@ constructors
I'm not sure what it would mean to mark a chare class constructor @explicit@ at all, let alone declaring it in the .c... Phil Miller
12:26 PM Bug #1669: Add machine layer files as Makefile dependencies for building the machine layer library
Could this be addressed by running something equivalent to the current @make depends@ and putting the output directly... Phil Miller
12:18 PM Feature #1695: Allow "type aliases" for explicit instantiations of member function templates
Along the path to prospectively eliminating the need for .ci files and just using straight C++, we're going to try to... Phil Miller
12:14 PM Feature #1682: Expose Arrays' Index Type as a Type Alias
Sounds simple enough. Any suggestions for the name of the typedef? Evan Ramos
12:07 PM Feature #1682: Expose Arrays' Index Type as a Type Alias
Good suggestion. We'll try to get it into the next release. Phil Miller
12:13 PM Bug #1692: charmc Dies on Array Indices with Template Parameters
The long-range direction on this is to get away from .ci files entirely, so we don't need to run on the 'parsing C++'... Phil Miller
12:11 PM Bug #1686: Use a namespace for Charm++ code
When we take an API break with the next major release, addressing this will be part of it. Phil Miller
11:12 AM Bug #1686: Use a namespace for Charm++ code
Looking through more of the Charm++ code, this ticket should be made a higher priority than I initially thought. Char... Nils Deppe
12:01 PM Bug #1679 (In Progress): Do Not Require Default Constructors for Serializable Classes
We actually ended up implementing something that should satisfy this request for a contract customer who wanted it AS... Phil Miller
11:53 AM Bug #1687 (Merged): charmc Strips Empty Template Parameter Lists
Phil Miller
11:14 AM Support #1534 (In Progress): Print out the job output, error and log for failed BGQ autobuild job...
Fix to print out the error, output and scheduler log: https://charm.cs.illinois.edu/gerrit/#/c/3083/ Nitin Bhat
11:01 PM Bug #1441 (Merged): Lies at startup about "The comm. thread both sends and receives messages"
Phil Miller
07:32 PM Bug #1690 (Merged): Verbs autobuild crashes for rdma example zerocopy/simpleZeroCopy on Golub in ...
Phil Miller

09/27/2017

05:30 PM Bug #1441 (Implemented): Lies at startup about "The comm. thread both sends and receives messages"
Did a rundown of the LRTS layers to confirm that the displayed message is correct. The default mode is COMM_THREAD_SE... Nitin Bhat
03:04 PM Bug #902 (In Progress): Projections shows garbage for the source PE of a chare array insertion event
The straight-forward fix for this messes up AMPI intercommunicator creation. Deferring for now, since the full fix wi... Phil Miller
02:59 PM Bug #529 (New): LrtsInitCpuTopo() operates asynchronously and overlaps with group constructors
I'm pretty sure this bug has been solved because the above mentioned patches prevent Charm init from progressing on a... Juan Galvez
02:47 PM Documentation #1611: Document network dependent rdma thresholds, above which benefits of the zero...
Since the memory registration API and registration cache optimization hasn't been done for the Rdma Nocopy Entry meth... Nitin Bhat
01:29 PM Feature #1697 (New): Distinguishing local object calls from entry method calls
I would like to be able to check if a method invocation was made locally via a @ckLocalBranch()@ or as a remote call ... Nils Deppe
07:42 AM Bug #1696 (New): Allow default arguments in group entry methods
I'm not sure this is a bug or a feature request since that depends on whether or not default arguments are supposed t... Nils Deppe
01:01 AM Bug #1437 (In Progress): CkLoop worker traces to previous entry on pe rather than caller
This is because CkLoop uses converse msgs by default. The msg doesn't contain info of the caller.
Figure out how t...
Seonmyeong Bak

09/26/2017

04:37 PM Feature #1695 (New): Allow "type aliases" for explicit instantiations of member function templates
I have a class template that has a non-negligible number of template parameters, which also has an entry method templ... Nils Deppe
12:23 PM Bug #902 (Implemented): Projections shows garbage for the source PE of a chare array insertion event
https://charm.cs.illinois.edu/gerrit/3079 Phil Miller
10:52 AM Bug #1694 (In Progress): Projections shows garbage for indices of 4d, 5d, 6d array elements
Projections seems to come with an implicit assumption that chare arrays have (at most) 3 dimensions, and hence elemen... Phil Miller
10:39 AM Feature #1693 (New): Print template parameters when missing explicit instantiation
First, thanks for now reporting missing explicit instantiations of template entry methods! I have a minor feature req... Nils Deppe
10:04 AM Bug #1692 (New): charmc Dies on Array Indices with Template Parameters
If the index to a chare array has template parameters then charmc chokes. A workaround is defining a type alias to th... Nils Deppe
09:24 AM Bug #1661 (Merged): Building charm with papi support is successful even when papi is not found (-...
Phil Miller
11:32 PM Bug #1687: charmc Strips Empty Template Parameter Lists
I haven't tested yet, but thanks for getting to this so quickly! Nils Deppe

09/25/2017

04:53 PM Bug #902 (In Progress): Projections shows garbage for the source PE of a chare array insertion event
I'll try to take care of this quickly. Phil Miller
03:33 PM Feature #1691 (New): GPU Handler PEs
It is often the case that the execution of heterogeneous tasks on the same PE deters the execution of GPU tasks:
1) ...
Jaemin Choi
03:15 PM Bug #1690 (Implemented): Verbs autobuild crashes for rdma example zerocopy/simpleZeroCopy on Golu...
This bug caused the verbs-smp crash in autobuild. The crash occurred because of the default send tokens in verbs bein... Nitin Bhat
02:59 PM Bug #1690 (Merged): Verbs autobuild crashes for rdma example zerocopy/simpleZeroCopy on Golub in ...
... Nitin Bhat
02:41 PM Bug #1687 (Implemented): charmc Strips Empty Template Parameter Lists
Fix here: https://charm.cs.illinois.edu/gerrit/3077 Phil Miller
10:16 AM Bug #1687: charmc Strips Empty Template Parameter Lists
Here is the simplest ci file I can think of that shows the problem:... Nils Deppe
10:05 AM Bug #1687: charmc Strips Empty Template Parameter Lists
Could you please post a complete .ci file that exhibits this? Phil Miller
01:01 PM Bug #1683 (Implemented): Charmc ends up deleting .C/.cpp file in case you forget to specify the o...
https://charm.cs.illinois.edu/gerrit/3076 Phil Miller
12:06 PM Bug #1544 (Merged): CMK_TIMER_USE_PPC64 inaccurate with variable clock speeds
Phil Miller
10:00 AM Bug #1544: CMK_TIMER_USE_PPC64 inaccurate with variable clock speeds
Switch to just use getrusage for now posted here: https://charm.cs.illinois.edu/gerrit/3075 Phil Miller
12:06 PM Bug #1689 (Implemented): CMK_TIMER_USE_PPC64 inaccurate with variable clock speeds
The pami-linux-ppc64le machine layer is defaulting to CMK_TIMER_USE_PPC64
src/arch/pami-linux-ppc64le/conv-mach.h:...
Phil Miller
10:48 PM Bug #1675: OFI replica crashes
Replicas never share a process, they simply distribute the processes among the replicas. Jim Phillips

09/24/2017

06:28 PM Feature #1321: multiple communication threads per process
Hello team, is there any update on that work?
The latest versions of PSM2 and OFI have improved support for multi-th...
Mikhail Shiryaev

09/23/2017

05:37 PM Bug #1281: Bigsim crash in SDAG code in examples/charm++/load_balancing/stencil3d
Pushed my change here: https://charm.cs.illinois.edu/gerrit/3074
As I said before, no idea whether this is right or ...
Phil Miller

09/22/2017

05:01 PM Bug #1675: OFI replica crashes
In a single instance, 1 thread is created. But I'm not sure how exactly replicas work. Will 2 replicas (2 charm insta... Nitin Bhat
01:27 PM Bug #1675: OFI replica crashes
But are there actually multiple threads launched, or is there just a single thread making pthread calls? Jim Phillips
09:39 AM Bug #1675: OFI replica crashes
Yes. It is a non-smp run. I'm guessing these are pthreads created by libfabric. I've asked the folks from Intel about... Nitin Bhat
08:00 PM Bug #1675: OFI replica crashes
This is a non-smp run, right? Are there other threads created by libfabric? Does pthreads require initialization? Jim Phillips
12:17 PM Bug #1688 (New): Core Dump file not available unless `--disable-charmdebug` is used while building.
While building charm with debug symbols without production mode, CMK_CHARMDEBUG is turned on and this leads to Charm ... Nitin Bhat
12:10 PM Bug #1441: Lies at startup about "The comm. thread both sends and receives messages"
Sure. I'll do that. Nitin Bhat

09/21/2017

04:07 PM Bug #1675: OFI replica crashes
I was able to replicate the hang for a smaller case with following command on Stampede 2:... Karthik Senthil
07:46 PM Bug #1675: OFI replica crashes
On 8 nodes of Stampede 2 with OFI layer (non-smp), the following command results in a hang at the end of simulation:
...
Karthik Senthil
12:16 PM Bug #1687 (Merged): charmc Strips Empty Template Parameter Lists
If I have @Klass<>@ in a ci file the generated code is @Klass@, which has a completely different meaning and results ... Nils Deppe
11:29 AM Bug #1686 (New): Use a namespace for Charm++ code
Charm++ code that is not generated should be in a namespace. Currently name collisions arise (eg Group). In general, ... Nils Deppe
10:44 AM Bug #1685 (Closed): charmc Chokes on @explicit@ constructors
If I mark a constructor explicit using the explicit keyword as follows:... Nils Deppe

09/20/2017

04:19 PM Bug #1684 (New): AMPI message sequence numbers are susceptible to overflow
AMPI uses sequence numbers that are ints, which may overflow if more than INT_MAX messages are sent from one rank to ... Sam White
03:44 PM Cleanup #567: Charm++ barnes-charm example cleanup
Removing the entire folder; Barnes-Hut mini-app should be the one used as it is well documented and works fine. Jaemin Choi
03:31 PM Bug #903: ckexit with interop hangs sometimes
This issue was reproducible using examples/charm++/mpi-coexist - multirun_time code. Nikhil Jain
02:54 PM Bug #1680: ci file compilation fails with no details to debug when the module name has a hyphen
Two tasks
* Document what characters are allowed (should be the same as C++ identifiers)
* Update Charmxi to print ...
Michael Robson
02:52 PM Bug #1683: Charmc ends up deleting .C/.cpp file in case you forget to specify the output file whi...
We should catch the lack of input file the way gcc (or similar) does Michael Robson
02:45 PM Bug #1683 (Merged): Charmc ends up deleting .C/.cpp file in case you forget to specify the output...
Repro:
1. touch test.cpp
2. charmc -o test.cpp
This will end up deleting test.cpp.
Note: g++ does not have ...
Samarth Kulshreshtha
02:34 PM Feature #1682: Expose Arrays' Index Type as a Type Alias
Retargeting new feature request to next major version Eric Bohm
09:14 AM Feature #1682 (Merged): Expose Arrays' Index Type as a Type Alias
It would be really convenient if the template parameter to the base class of @CBase_@ (specifically the @T@ passed to... Nils Deppe
12:57 PM Bug #1510 (In Progress): Hang in tests/charm++/chkpt when using -tracemode perfReport
I was able to reproduce this on Linux, it looks like it's because it the PICS code tries to exit prematurely, which c... Ronak Buch
09:07 AM Support #1681 (New): Document Exception Handling in Charm++
I've encountered a few things I couldn't find in the documentation with regards to error handling. In no particular o... Nils Deppe

09/19/2017

01:53 PM Cleanup #567: Charm++ barnes-charm example cleanup
Using @std::isnan@ failed on titan, so replaced all @isnan@ occurences with @std::isnan@. The input file specified in... Jaemin Choi
11:29 AM Bug #1680 (Merged): ci file compilation fails with no details to debug when the module name has a...
If you name a module with a hyphen and try to compile it then the compilation fails with -
Fatal Error by charmc i...
Samarth Kulshreshtha

09/18/2017

05:06 PM Cleanup #567: Charm++ barnes-charm example cleanup
Removing @ref_src@ folder entirely. Jaemin Choi
04:42 PM Bug #1661 (Implemented): Building charm with papi support is successful even when papi is not fou...
https://charm.cs.illinois.edu/gerrit/#/c/3033/ Ronak Buch
04:22 PM Bug #1661 (In Progress): Building charm with papi support is successful even when papi is not fou...
Ronak Buch
01:48 PM Bug #1679 (In Progress): Do Not Require Default Constructors for Serializable Classes
I've set the target to 7 because this could be implemented as a breaking change.
Currently Charm++ requires all cl...
Nils Deppe
01:29 PM Feature #1677: improved topology-aware partitioner
Yes, the code in Charm++ that handles the +partitions (+replicas) setup. Jim Phillips
09:37 AM Feature #1677: improved topology-aware partitioner
I assume this issue refers to code inside Charm++?
I don't quite understand your description of the issue with the...
Juan Galvez
12:59 PM Bug #921: Entry tag [inline] is unable to optimize away most of the overhead
If I've understood the above discussion correctly then this is the right place (otherwise I can create a new ticket).... Nils Deppe
11:08 AM Bug #1331: Isomalloc hangs in startup for Clang non-SMP builds
We ended up needing to revert the above patch because it broke autobuild targets for uth-linux-x86_64, mpi-crayxc, an... Sam White

09/16/2017

12:32 PM Bug #1281: Bigsim crash in SDAG code in examples/charm++/load_balancing/stencil3d
All of the non-AMPI tests pass with that change baked into charmxi. But I have no real understanding why/how Phil Miller
11:44 AM Bug #1281: Bigsim crash in SDAG code in examples/charm++/load_balancing/stencil3d
We've just started seeing this become relatively common in some automatic builds: http://ppl-jenkins:8080/job/Nightly... Phil Miller

09/15/2017

04:33 PM Bug #1676: Replicas slower than separate jobs on GNI systems
Correct, as far as I know this is a GNI issue.
I've only tested on Blue Waters. It may or may not affect Titan, Eos...
Jim Phillips
12:50 PM Bug #1676: Replicas slower than separate jobs on GNI systems
Just to be clear, the only context in which this bug has been observed is GNI on Blue Waters?
MPI there is unaffec...
Phil Miller
09:56 AM Bug #1676: Replicas slower than separate jobs on GNI systems
verbs layer does not appear to be affected. Jim Phillips
01:02 PM Bug #1441: Lies at startup about "The comm. thread both sends and receives messages"
Nitin, could you follow up with checks on the other layers and note whether/what needs to be done here? Phil Miller
09:54 AM Support #1534: Print out the job output, error and log for failed BGQ autobuild jobs and relaunch...
There are spurious build failures which occur because of the job booting too many times and job not starting after bo... Nitin Bhat

09/14/2017

11:28 AM Feature #1677 (New): improved topology-aware partitioner
The current topology-aware recursive-bisection partitioner (described in Phillips et al., SC14) was based on an algor... Jim Phillips
10:45 AM Bug #1676: Replicas slower than separate jobs on GNI systems
The bug does not affect the MPI layer on Blue Waters.
Still need to test verbs.
Jim Phillips
09:59 PM Bug #1676: Replicas slower than separate jobs on GNI systems
I see the exact same performance for v6.7.0-574-g7d61794-namd-charm-6.8.0-build-2017-Jan-23-80737 and v6.7.0-0-g46f86... Jim Phillips
09:33 PM Bug #1676: Replicas slower than separate jobs on GNI systems
6.8.0 from Sept 5 (v6.8.0-0-ga36028e-namd-charm-6.8.0-build-2017-Sep-05-28093).
No observed performance difference o...
Jim Phillips
07:20 PM Bug #1676: Replicas slower than separate jobs on GNI systems
What commit of charm are you using? We recently merged changes to make broadcasts and reductions topology-aware. Sam White
10:33 PM Bug #1544: CMK_TIMER_USE_PPC64 inaccurate with variable clock speeds
What I was seeing in projections sure looked like it was just counting clock cycles. Jim Phillips
10:31 PM Support #1674: Add 'ofi' target to autobuild
Bridges doesn't use 2-factor authentication. Jim Phillips

09/13/2017

04:40 PM Bug #1676 (New): Replicas slower than separate jobs on GNI systems
On Blue Waters, 200 nodes, 50 replicas (4 nodes per replica) non-smp running apoa1 is uniformly and significantly slo... Jim Phillips
04:36 PM Bug #1675 (Merged): OFI replica crashes
Testing NAMD replicas (8 nodes, 2 replicas) on Bridges with the non-smp ofi layer hangs or crashes most of the time i... Jim Phillips
03:02 PM Cleanup #567 (Implemented): Charm++ barnes-charm example cleanup
Sam White
02:55 PM Bug #1048 (Rejected): Verbs on Bridges at PSC crashes or hangs
OFI should be used on Bridges. Sam White
02:55 PM Bug #1381 (Rejected): Crash in LrtsInitCpuTopo() on Quartz with verbs layer
OFI should be used on Quartz. Sam White
02:55 PM Bug #1409 (Rejected): verbs crashes on Stampede KNL and Bridges
OFI should be used on Stampede2 and Bridges. Sam White
02:46 PM Support #1674: Add 'ofi' target to autobuild
Bridges at PSC would be an ideal candidate for autobuild, if we can set it up to run without 2-factor authentication. Sam White
02:42 PM Support #1674: Add 'ofi' target to autobuild
ofi builds could be added on the same target machine (campus cluster) as verbs, so assigning to Jaemin. Sam White
02:43 PM Bug #1649 (Merged): NullLB shouldn't wait for LB period
Sam White
02:36 PM Bug #1544: CMK_TIMER_USE_PPC64 inaccurate with variable clock speeds
Intel fixed their timer instruction (rdtsc/rdtscp) to fix this kind of issue years ago, by having the counter run on ... Phil Miller

09/12/2017

09:09 AM Bug #635 (Merged): all trees should be pe/node/physnode/network topology aware
Phil Miller
09:09 AM Bug #1663 (Merged): Fix CMK_PAGESIZE definitions
Phil Miller
09:09 AM Feature #1494 (Merged): Broadcast trees are not topology-aware
Phil Miller

09/11/2017

04:36 PM Support #1674 (In Progress): Add 'ofi' target to autobuild
Sam White
04:24 PM Bug #1409: verbs crashes on Stampede KNL and Bridges
(Updating all of #1048 #1381 #1409 )
Since these machines are all ostensibly going to be using the ofi network lay...
Phil Miller
04:24 PM Bug #1381: Crash in LrtsInitCpuTopo() on Quartz with verbs layer
(Updating all of #1048 #1381 #1409 )
Since these machines are all ostensibly going to be using the ofi network lay...
Phil Miller
04:24 PM Bug #1048: Verbs on Bridges at PSC crashes or hangs
(Updating all of #1048 #1381 #1409 )
Since these machines are all ostensibly going to be using the ofi network lay...
Phil Miller
04:09 PM Feature #1673 (Merged): Basic OFI (libfabric) network layer using PSM transport
Creating this sub-task, so that the parent can be re-targetted to 6.9 Phil Miller
01:43 PM Bug #1664 (New): Port Sameer's PAMI changes for POWER8 to PAMILRTS
Some of the things in this patch probably need to be ported for PAMILRTS: https://charm.cs.illinois.edu/gerrit/#/c/89... Sam White
09:54 AM Feature #1672 (New): LLDB support for charmrun
Since gdb support on the latest macOs is minimal and has issues related to codesigning, it'll be good to have lldb su... Nitin Bhat

09/07/2017

03:01 PM Cleanup #567: Charm++ barnes-charm example cleanup
Patch: https://charm.cs.illinois.edu/gerrit/#/c/2995/
Wondering if we should get rid of this example altogether and ...
Jaemin Choi
02:56 PM Cleanup #567 (In Progress): Charm++ barnes-charm example cleanup
The two folders @src@ and @barnes@ are exactly the same.
From looking at the git logs, Pretish tried to rename @barn...
Jaemin Choi
02:24 PM Support #1450 (In Progress): Clean up and add CUDA example programs
Current example programs in @examples/cuda@ should be cleaned up, and some new ones such as stencil2d and AMR should ... Jaemin Choi
02:22 PM Support #1456: Add more stream callbacks for use after HToD transfer and kernel execution
Patch: https://charm.cs.illinois.edu/gerrit/#/c/2994/ Jaemin Choi
02:21 PM Support #1454: GPUManager API change
Patch: https://charm.cs.illinois.edu/gerrit/#/c/2994/ Jaemin Choi
02:21 PM Feature #1451: NVTX integration for profiling
Patch: https://charm.cs.illinois.edu/gerrit/#/c/2994/ Jaemin Choi
02:20 PM Feature #1393: Redesign of Hybrid API (GPU Manager) to support concurrent kernel execution
Patch: https://charm.cs.illinois.edu/gerrit/#/c/2994/ Jaemin Choi

09/06/2017

03:51 PM Feature #1393: Redesign of Hybrid API (GPU Manager) to support concurrent kernel execution
Includes 2 schemes: CUDA event-based & CUDA callback-based, but the default is CUDA event-based due to better perform... Jaemin Choi
02:47 PM Cleanup #567: Charm++ barnes-charm example cleanup
Rationalize this with the one in its own repository. Either there should only be one, or there should be a clear wel... Eric Bohm
01:32 PM Feature #1670 (Implemented): Migration of Taskqueue from Ck to Converse layer
Seonmyeong Bak
12:56 PM Feature #1670 (Implemented): Migration of Taskqueue from Ck to Converse layer
The task queue is currently implemented in ck-core even though it doesn't use any constructs on ck-layer.
Taskqueue...
Seonmyeong Bak
01:25 PM Bug #1671: Verbs memory pool may leak pinned memory when message is deleted on a PE different fro...
Note that this is specific to a mempool in src/arch/verbs/machine-ibverbs.c, not the generic mempool that we have for... Sam White
01:12 PM Bug #1671 (New): Verbs memory pool may leak pinned memory when message is deleted on a PE differe...
Receive a message on a node group
Section multicast a pointer to that message to the pe's local to that host.
Handl...
Eric Bohm
11:56 PM Feature #1666: Standalone OpenMP implementation based on Converse User-level threads
Basically, it works with a few lines of codes in converse and OpenMP runtime. Seonmyeong Bak

09/01/2017

02:07 PM Bug #1162: tracing runs segfault while writing logs
Still an issue? Sam White
02:04 PM Bug #1409: verbs crashes on Stampede KNL and Bridges
This is probably not a real issue, since MPI and OFI will be in 6.8.1 and OFI should be the preferred comm layer on b... Sam White
02:01 PM Bug #1661: Building charm with papi support is successful even when papi is not found (-lpapi)
Assigning to Ronak b/c tracing... Sam White
 

Also available in: Atom