Project

General

Profile

Activity

From 12/18/2017 to 01/16/2018

01/16/2018

05:19 PM Feature #973: multicore: spawn a thread per core by default
Bring it up at the next Core meeting, I don't think there's a consensus Sam White
05:03 PM Feature #973: multicore: spawn a thread per core by default
Bump. Is this something we want to proceed with, and is it important for 6.9.0? Evan Ramos
05:16 PM Bug #1442: CkLoop fixed tree limits helper recruitment
https://charm.cs.illinois.edu/gerrit/#/c/3517/ Sam White
04:54 PM Bug #1442 (Implemented): CkLoop fixed tree limits helper recruitment
Added API to set scheduling policy for CkLoop: CkLoop_SetSchedPolicy(sched schedPolicy);
Three options available: ...
Seonmyeong Bak
05:11 PM Feature #1098 (Merged): Better CMake support
I don't think so. Sam White
05:00 PM Feature #1098: Better CMake support
Bump. These patches are merged: https://charm.cs.illinois.edu/gerrit/#/q/project:charm+topic:cmake-wrapper
Is ther...
Evan Ramos
05:10 PM Bug #1220: AMPI: Support tlsglobals with dynamically linked objects
It's not blocking anything, so I'll move it off 6.9.0, but it is important. Sam White
05:06 PM Bug #1220: AMPI: Support tlsglobals with dynamically linked objects
Is this feature blocking 6.9.0 or should we reschedule it for a future version? Evan Ramos
05:09 PM Bug #979 (Closed): charmxi: mismatched module braces leads to confusing error messages.
The above message seems a good enough solution to me. I'm closing this. Feel free to re-open if there is disagreement. Evan Ramos
05:07 PM Bug #736 (Closed): charmxi: #endif terminates any enclosing comment
Closing this until/unless I hear otherwise. Evan Ramos
04:38 PM Bug #1776 (Merged): netlrts-darwin-x86_64 autobuild failing due changes in CmiTmpAlloc
Sam White
04:21 PM Bug #1776 (Implemented): netlrts-darwin-x86_64 autobuild failing due changes in CmiTmpAlloc
This patch is at fault: https://charm.cs.illinois.edu/gerrit/#/c/3397/
Fix is here: https://charm.cs.illinois.edu/...
Sam White
08:48 PM Bug #1776 (Merged): netlrts-darwin-x86_64 autobuild failing due changes in CmiTmpAlloc
The past few days netlrts-darwin-x86_64 autobuild has failed in ping pong or other basic tests. Not sure why, so inve... Sam White
04:23 PM Bug #1689 (New): CMK_TIMER_USE_PPC64 inaccurate with variable clock speeds
Not sure why this was marked "implemented"... Sam White
03:58 PM Bug #1547 (Merged): Deprecate the FFT library in ck-libs in favor of Nikhil's new FFT library
Sam White
12:08 PM Feature #1670: Migration of Taskqueue from Ck to Converse layer
As noted in Gerrit, this broke Windows builds. Jim Phillips
06:51 AM Documentation #1777 (Merged): Document message types in Charm's message hander (_processHandler i...
Sanjay's description of NewChareMsg, NewVChareMsg & ForVidMsg (given on slack).
These are both related to singlet...
Nitin Bhat

01/15/2018

05:50 PM Bug #1307 (Merged): AMPI_Comm_free should free the ampi instance
Sam White

01/14/2018

03:17 PM Feature #1480 (Merged): API to control whether a PE helps other threads that generate CkLoop/Open...
Sam White
03:17 PM Documentation #1771 (Merged): Update on documentation of Loop-level parallelism
Sam White
02:23 PM Feature #1723 (Merged): Rebase the OpenMP version onto the latest version of LLVM runtime library
Sam White

01/13/2018

01:18 PM Feature #1763: Generate libmpi.{a,so} for AMPI configure scripts
Which application does that? Would it make sense to patch those manually (similar to what we do with hdf5)? Matthias Diener
01:05 PM Bug #1746 (Merged): build --help needs information from src/scripts/configure
Sam White
12:46 PM Feature #1433 (Merged): lambda syntax for CkLoop
Sam White

01/12/2018

05:08 PM Feature #1113: smart-build.pl should detect supercomputers with specialized software environments...
Any opinions on the following one-liner as a Cray environment detection test?... Evan Ramos
04:46 PM Bug #1668 (Merged): Ensure that all libraries/modules will build as dynamic/shared objects (.so/....
I tried the build and grep commands in the issue description and no items were listed. I agree that it is complete. Evan Ramos
12:38 PM Bug #1668: Ensure that all libraries/modules will build as dynamic/shared objects (.so/.dylib vs .a)
This enables shared builds of ROMIO: https://charm.cs.illinois.edu/gerrit/#/c/3499/
I think this issue can be clos...
Sam White
04:34 PM Bug #1442 (In Progress): CkLoop fixed tree limits helper recruitment
Seonmyeong Bak
08:57 PM Bug #1442: CkLoop fixed tree limits helper recruitment
Actually, the overhead (number of messages sent and received) is at worst identical. The latency is only higher if a... Jim Phillips
04:33 PM Bug #1165: avoid -lm with Intel compiler
Jim Phillips wrote:
> Doing a little experimenting, I only see the link error for gcc on a C file, not for g++ on a ...
Evan Ramos
03:16 PM Feature #1670 (Merged): Migration of Taskqueue from Ck to Converse layer
Sam White
02:39 PM Bug #1770 (Merged): when cross-compiling for KNL, charmc -host uses -xMIC-AVX512 command-line arg...
Sam White
07:41 PM Bug #1770: when cross-compiling for KNL, charmc -host uses -xMIC-AVX512 command-line argument to ...
Fix submitted: https://charm.cs.illinois.edu/gerrit/#/c/3507/ Jim Phillips
01:59 PM Bug #1775: chkpt test hangs for pami-linux-ppc64le-smp & pamilrts-linux-ppc64le-smp build
Similar hang is seen for pamilrts-linux-ppc64le-smp build as well. (https://charm.cs.illinois.edu/gerrit/#/c/3141/) Nitin Bhat
10:53 AM Bug #1774: Thread migration fails on ppc64le builds
Seeing the hang even with '+isomalloc_sync'.
Similar hang is seen for pamilrts-linux-ppc64le-smp build as well. (...
Nitin Bhat
08:12 AM Bug #1774: Thread migration fails on ppc64le builds
Looks ilke you need to run with '+isomalloc_sync', but there might be other issues too Sam White
09:48 AM Feature #1236: Avoid receiver-side copy for large contiguous messages. API for charm and converse...
With the Direct API getting merged for 6.9.0, I think there is no pressing need to have this in 6.9.0. Since there is... Nitin Bhat

01/11/2018

05:18 PM Bug #1775 (New): chkpt test hangs for pami-linux-ppc64le-smp & pamilrts-linux-ppc64le-smp build
On Summitdev, chkpt test (tests/charm++/chkpt) with pami-linux-ppc64le-smp build hangs with the following output:
<p...
Nitin Bhat
05:10 PM Bug #1774 (In Progress): Thread migration fails on ppc64le builds
On Summitdev, megampi test (tests/ampi/megampi) with pami-linux-ppc64le-smp build hangs with the following output:
...
Nitin Bhat
05:02 PM Bug #1773 (New): Zerocopy examples fail on pami-linux-ppc64le-smp due to a low level assertion fa...
... Nitin Bhat
04:56 PM Bug #1442: CkLoop fixed tree limits helper recruitment
I think this patch can increase the overhead of CkLoop initialization by eliminating the use of implicit tree in CkLo... Seonmyeong Bak
01:21 PM Bug #1772 (New): Programs built on pami-linux-ppc64le-async-smp fail a low level assertion at run...
On running any test/example with the pami-linux-ppc64le-async-smp build on SummitDev, the following error is seen:
...
Nitin Bhat
09:51 AM Bug #1770: when cross-compiling for KNL, charmc -host uses -xMIC-AVX512 command-line argument to ...
This is where the AVX512 option is showing up in conv-mach-opt.sh:... Jim Phillips
09:13 PM Bug #1259 (Merged): AMPI missing built-in MPI_INFO_ENV
Matthias Diener
06:31 PM Bug #1746 (Implemented): build --help needs information from src/scripts/configure
Please test: https://charm.cs.illinois.edu/gerrit/3503 Evan Ramos

01/10/2018

04:57 PM Bug #1274 (Merged): Tracemode utilization produces incorrect results for EP utilization metric
Ronak Buch
07:25 PM Bug #1274 (Implemented): Tracemode utilization produces incorrect results for EP utilization metric
https://charm.cs.illinois.edu/gerrit/#/c/3491/ Sam White
02:52 PM Feature #858 (In Progress): improve efficiency of exclusive entry methods
Kavitha Chandrasekar
02:48 PM Feature #1433 (Implemented): lambda syntax for CkLoop
https://charm.cs.illinois.edu/gerrit/#/c/3498/
This patch adds Jim's lambda syntax prototype to CkLoop. I will loo...
Kavitha Chandrasekar
02:46 PM Support #1674 (In Progress): Add 'ofi' target to autobuild
Autobuild for OFI is not passing yet. I don't even think it has actually gotten to building charm yet. Here's last ni... Sam White
02:44 PM Support #1674 (Resolved): Add 'ofi' target to autobuild
Jaemin Choi
02:45 PM Bug #965: ampicc -swapglobals is broken for ld v2.24+
The integration work, and then validation that it's a practical path forward for deployment. Phil Miller
02:41 PM Bug #965: ampicc -swapglobals is broken for ld v2.24+
I'm not exactly sure what's left here, is it integration/distribution of the patched libbfd with AMPI? Sam White
02:39 PM Bug #1566 (In Progress): Parameter marshalled entry methods mostly don't set group dependence in ...
Nitin Bhat
02:39 PM Bug #1565 (In Progress): Non-group entry methods don't respect envelope group dependence
Nitin Bhat
09:21 AM Bug #1565 (New): Non-group entry methods don't respect envelope group dependence
Nitin Bhat
02:39 PM Bug #1563 (In Progress): Chare Array construction doesn't respect setGroupDepID in CkEntryOptions...
Nitin Bhat
02:39 PM Bug #1563 (Implemented): Chare Array construction doesn't respect setGroupDepID in CkEntryOptions...
Nitin Bhat
12:57 PM Charm-NG Feature #1333: Eliminate need for .ci files
I just learned of something that may be of use to work in this area: the @__COUNTER__@ macro - each expansion gives a... Phil Miller

01/09/2018

05:14 PM Documentation #1771: Update on documentation of Loop-level parallelism
https://charm.cs.illinois.edu/gerrit/#/c/3495/ Seonmyeong Bak
05:13 PM Documentation #1771 (Implemented): Update on documentation of Loop-level parallelism
Seonmyeong Bak
04:21 PM Documentation #1771 (Merged): Update on documentation of Loop-level parallelism
Our current OpenMP integration supports more pragmas over the initial version.
So, the documentation should be upda...
Seonmyeong Bak
03:17 PM Bug #1566 (New): Parameter marshalled entry methods mostly don't set group dependence in message ...
Nitin Bhat
03:10 PM Bug #1564 (In Progress): Inline entry methods don't respect group dependence from CkEntryOptions
Is there a test/example that can reproduce this bug? (just to be sure that the fix works correctly).
I am not able ...
Nitin Bhat
02:50 PM Bug #1501 (In Progress): charmxi fails to deal with rvalue references
Evan Ramos
02:45 PM Feature #988 (Closed): charmrun should not ignore +ppn
Evan Ramos
02:43 PM Feature #988 (Resolved): charmrun should not ignore +ppn
If an error is sufficient, I am marking this resolved. Evan Ramos
06:01 PM Feature #988: charmrun should not ignore +ppn
Producing an error is sufficient. Silently ignoring an option is not.
The confusion is that platforms that don't ...
Jim Phillips
09:53 AM Feature #1742: Send large readonly variables using the Nocopy API to reduce memory footprint and ...
I think it'd be good to get started on this sooner than later since it may serve as a motivating case for the differe... Sam White

01/08/2018

05:50 PM Bug #1770: when cross-compiling for KNL, charmc -host uses -xMIC-AVX512 command-line argument to ...
OK, this seems to be the critical difference between the MPI build (which works) and the multicore build (which doesn... Jim Phillips
03:52 PM Bug #1770 (Merged): when cross-compiling for KNL, charmc -host uses -xMIC-AVX512 command-line arg...
Related to https://charm.cs.illinois.edu/redmine/issues/1745 as fixed by https://charm.cs.illinois.edu/gerrit/c/3339/... Jim Phillips
05:09 PM Bug #1501: charmxi fails to deal with rvalue references
https://charm.cs.illinois.edu/gerrit/c/3490
This patch adds rvalue references to charmxi's parsing grammar. I'm no...
Evan Ramos
04:36 PM Feature #1401: Converting OpenMP test suite for the OpenMP integration.
https://charm.cs.illinois.edu/gerrit/#/c/2245/
Fixed inherent issues in the tests, which lead to failure in the co...
Seonmyeong Bak
02:46 PM Feature #988: charmrun should not ignore +ppn
The convention for command line arguments appears to be that options with a single plus "+" are reserved for the Char... Evan Ramos
01:39 PM Bug #1560: icc build fails on NASA Pleiades
Wanted to update with two alternative solutions, neither of which preserves the options in the build config files.
...
Jim Phillips

01/05/2018

01:41 PM Feature #988 (New): charmrun should not ignore +ppn
Evan Ramos
08:23 PM Feature #988: charmrun should not ignore +ppn
Rejected why? Does charmrun silently ignore +ppn? It should either accept it or throw an error. Jim Phillips
06:21 PM Feature #1480: API to control whether a PE helps other threads that generate CkLoop/OpenMP/Parall...
https://charm.cs.illinois.edu/gerrit/#/c/3481/ Sam White
06:21 PM Feature #1670: Migration of Taskqueue from Ck to Converse layer
https://charm.cs.illinois.edu/gerrit/#/c/2988/ Sam White
06:19 PM Feature #1727: Make Boost uFcontext the default ULT implementation on supported platforms
https://charm.cs.illinois.edu/gerrit/#/c/3482/ Sam White

01/04/2018

05:40 PM Feature #1727 (Implemented): Make Boost uFcontext the default ULT implementation on supported pla...
Seonmyeong Bak
05:15 PM Feature #988 (Rejected): charmrun should not ignore +ppn
Evan Ramos
03:54 PM Feature #1480 (Implemented): API to control whether a PE helps other threads that generate CkLoop...
Seonmyeong Bak
03:48 PM Bug #872 (Merged): smp standalone inconsistency
Evan Ramos
02:49 PM Bug #1165: avoid -lm with Intel compiler
Doing a little experimenting, I only see the link error for gcc on a C file, not for g++ on a C++ file:... Jim Phillips
10:14 PM Bug #1165: avoid -lm with Intel compiler
You would need to link something that calls a math function in libm such as sin, cos, etc. Jim Phillips

01/03/2018

04:38 PM Bug #979 (Feedback): charmxi: mismatched module braces leads to confusing error messages.
The output of charmxi with the given test case is currently:... Evan Ramos
03:59 PM Bug #1165 (In Progress): avoid -lm with Intel compiler
https://charm.cs.illinois.edu/gerrit/3477 is the first half of a fix for this, removing `-lm` from link line. However... Evan Ramos
03:23 PM Bug #736 (Feedback): charmxi: #endif terminates any enclosing comment
I cannot replicate this problem. Note that the provided test case has a mismatched number of braces. When that is fix... Evan Ramos

12/30/2017

09:45 PM Bug #1701: Cannot have non-copyable types in constructor arguments
It looks like with all the changes the generated code is @std::move(impl_noname_0).t@ which has the type @T@, not @T&... Nils Deppe

12/29/2017

01:31 PM Bug #901 (Merged): Threads awoken by CthAwaken don't let Projections trace back to the event that...
Phil Miller
12:19 PM Feature #111: object location services: scalable cache pre-population
With the move to identify array elements internally with 64-bit IDs, it's now possible for applications to use arbitr... Phil Miller
12:12 PM Cleanup #12: Factor out massive duplication in reductions
Per Eric B - reassign to Juan as supervisor of the collectives subgroup. It could be a good way to get a new person i... Phil Miller
12:08 PM Cleanup #221: Delete deprecated FEM library
Another possible reason not to rush this: the Fractography3D code may be based on FEM and nor ParFUM. Phil Miller
11:52 AM Bug #1205: AMPI's -tlsglobals option is only supported by GCC
Mass re-assign AMPI-related issues on my plate to Sam, for subsequent redistribution.
For real this time.
Phil Miller
11:51 AM Bug #1205: AMPI's -tlsglobals option is only supported by GCC
Mass re-assign AMPI-related issues on my plate to Sam, for subsequent redistribution. Phil Miller
11:52 AM Bug #965: ampicc -swapglobals is broken for ld v2.24+
Mass re-assign AMPI-related issues on my plate to Sam, for subsequent redistribution.
For real this time.
Phil Miller
11:51 AM Bug #965: ampicc -swapglobals is broken for ld v2.24+
Mass re-assign AMPI-related issues on my plate to Sam, for subsequent redistribution. Phil Miller
11:52 AM Feature #896: AMPI support for compiler-automated static variable privatization
Mass re-assign AMPI-related issues on my plate to Sam, for subsequent redistribution.
For real this time.
Phil Miller
11:51 AM Feature #896: AMPI support for compiler-automated static variable privatization
Mass re-assign AMPI-related issues on my plate to Sam, for subsequent redistribution. Phil Miller
11:52 AM Feature #283: TLS-based AMPI variable privatization source-to-source translation tool
Mass re-assign AMPI-related issues on my plate to Sam, for subsequent redistribution.
For real this time.
Phil Miller
11:51 AM Feature #283: TLS-based AMPI variable privatization source-to-source translation tool
Mass re-assign AMPI-related issues on my plate to Sam, for subsequent redistribution. Phil Miller
11:48 AM Bug #1501: charmxi fails to deal with rvalue references
Reassign to Evan based on discussion with Eric B. Confer on how important this work is to schedule in the near future. Phil Miller
11:48 AM Bug #979: charmxi: mismatched module braces leads to confusing error messages.
Reassign to Evan based on discussion with Eric B. Confer on how important this work is to schedule in the near future. Phil Miller
11:48 AM Bug #736: charmxi: #endif terminates any enclosing comment
Reassign to Evan based on discussion with Eric B. Confer on how important this work is to schedule in the near future. Phil Miller
11:48 AM Bug #676: Compiling ci files that use "typename T" in argument list fails
Reassign to Evan based on discussion with Eric B. Confer on how important this work is to schedule in the near future. Phil Miller
11:46 AM Bug #1566: Parameter marshalled entry methods mostly don't set group dependence in message from C...
Per discussion with Eric B, assigning to Nitin, based on recent work around CkEntryOptions handling Phil Miller
11:46 AM Bug #1565: Non-group entry methods don't respect envelope group dependence
Per discussion with Eric B, assigning to Nitin, based on recent work around CkEntryOptions handling Phil Miller
11:46 AM Bug #1564: Inline entry methods don't respect group dependence from CkEntryOptions
Per discussion with Eric B, assigning to Nitin, based on recent work around CkEntryOptions handling Phil Miller
10:44 AM Bug #1564 (New): Inline entry methods don't respect group dependence from CkEntryOptions
This says 'Implemented', but I can't find code where I've actually done it. Shouldn't be too hard, but it's not somet... Phil Miller
11:46 AM Bug #1563: Chare Array construction doesn't respect setGroupDepID in CkEntryOptions (or any other...
Per discussion with Eric B, assigning to Nitin, based on recent work around CkEntryOptions handling Phil Miller

12/28/2017

08:47 PM Bug #1668: Ensure that all libraries/modules will build as dynamic/shared objects (.so/.dylib vs .a)
The ROMIO issue mentioned above is not solved yet, so AMPI applications still require at least the static ROMIO libra... Matthias Diener

12/27/2017

04:06 PM Bug #1668 (Implemented): Ensure that all libraries/modules will build as dynamic/shared objects (...
With https://charm.cs.illinois.edu/gerrit/3469 I get:... Evan Ramos

12/26/2017

12:08 PM Bug #1668 (In Progress): Ensure that all libraries/modules will build as dynamic/shared objects (...
hwloc builds a dynamic library, but it's not presently usable on Mac. I needed to uncomment the @cp@ command in @src/... Phil Miller
11:48 AM Bug #1668: Ensure that all libraries/modules will build as dynamic/shared objects (.so/.dylib vs .a)
QuickThreads is also still kinda messed up, at least on Mac builds Phil Miller

12/22/2017

04:41 PM Feature #1178: Support automated launch/spawn/bind when using charmrun ++mpiexec
Are there any more details available about what this entails? Evan Ramos
04:38 PM Feature #1098: Better CMake support
Is there anything more than https://charm.cs.illinois.edu/gerrit/#/q/project:charm+topic:cmake-wrapper that this issu... Evan Ramos
09:59 AM Bug #1653 (Implemented): NeighborLB segfaults during startup in SMP/multicore builds
https://charm.cs.illinois.edu/gerrit/#/c/3458/ Sam White

12/21/2017

03:28 PM Bug #1665 (In Progress): DDT needs to reference count and free its type objects
Initial support for freeing types in DDT: https://charm.cs.illinois.edu/gerrit/#/c/3461/
Charmworks intern Nimay i...
Sam White

12/20/2017

05:46 PM Feature #1763: Generate libmpi.{a,so} for AMPI configure scripts
This gets us AMPI's MPI_* routines to not be renamed when building on non-MPI layers: https://charm.cs.illinois.edu/g... Sam White
02:46 PM Bug #529 (Resolved): LrtsInitCpuTopo() operates asynchronously and overlaps with group constructors
Eric Bohm
02:33 PM Feature #1655 (In Progress): Enable use of shm transport for regular messages in LRTS
Gerrit Patch: https://charm.cs.illinois.edu/gerrit/#/c/3457/ Nitin Bhat
02:19 PM Feature #1769: AMPI implementation of Buffered sends
Should also be able to optimize for ready sends (Rsend, Irsend, etc). Sam White
02:16 PM Feature #1769 (New): AMPI implementation of Buffered sends
Now that AMPI supports in-place sends via RDMA, we should actually support buffered sends (Bsend, Ibsend, etc) and th... Sam White
11:39 AM Documentation #1768 (New): document CkIO
CkIO documentation is currently limited to a header file and a test. Jim Phillips

12/19/2017

03:35 PM Bug #1668: Ensure that all libraries/modules will build as dynamic/shared objects (.so/.dylib vs .a)
All of the above changes have been merged, so the only remaining issue here should be ROMIO. Sam White
02:48 PM Feature #952: Update AMPI's version of ROMIO
We'd still like this to be in 6.9.0 if possible. I don't expect 6.9.0 to be released in the next 3-4 weeks. Sam White
01:42 PM Feature #973: multicore: spawn a thread per core by default
Yes, this would be trivial to implement with hwloc integrated. Has there been a consensus on whether or not multicore... Evan Ramos
01:38 PM Bug #1641: charmrun with nodelist option (++nodelist) fails on campus cluster
All this error means is that the ssh connection cannot be established. I don't have access to these machines to speci... Evan Ramos
01:14 PM Bug #872 (Implemented): smp standalone inconsistency
https://charm.cs.illinois.edu/gerrit/3452 Evan Ramos
12:08 PM Feature #1173 (In Progress): Automatic process launching, thread spawning, and hardware binding
Evan Ramos
12:00 PM Feature #1173 (Implemented): Automatic process launching, thread spawning, and hardware binding
Evan Ramos
12:07 PM Feature #1180 (Implemented): Support automated thread spawn/bind for standalone runs
Evan Ramos
12:06 PM Feature #1181 (Implemented): Support automated process launch on a single host for standalone runs
`++processPerSocket 1` should work with the hwloc patchsets both with charmrun and with standalone SMP executables. Evan Ramos
12:03 PM Cleanup #621 (Merged): Convert Converse from C to C++
Evan Ramos
12:03 PM Cleanup #622 (Merged): Convert MPI Machine layer from C to C++
Evan Ramos
12:03 PM Cleanup #623 (Merged): Convert Verbs machine layer from C to C++
Evan Ramos
12:03 PM Cleanup #624 (Merged): Convert netlrts machine layer from C to C++
Evan Ramos
12:03 PM Cleanup #625 (Merged): Convert uGNI machine layer from C to C++
Evan Ramos
12:02 PM Cleanup #626 (Merged): Convert PAMI-LRTS machine layer from C to C++
Evan Ramos
12:02 PM Cleanup #627 (Merged): Convert common LRTS code from C to C++
Evan Ramos
12:01 PM Feature #988 (Feedback): charmrun should not ignore +ppn
Evan Ramos
12:00 PM Feature #1176 (Implemented): Detect unsupported non-uniformity of processes/threads in charmrun, ...
Evan Ramos
12:00 PM Feature #1178 (New): Support automated launch/spawn/bind when using charmrun ++mpiexec
Evan Ramos
12:00 PM Feature #1178 (Implemented): Support automated launch/spawn/bind when using charmrun ++mpiexec
Evan Ramos
12:00 PM Bug #1174 (Implemented): Use hwloc data from compute host, rather than assuming they're identical...
Evan Ramos

12/18/2017

05:45 PM Bug #1668: Ensure that all libraries/modules will build as dynamic/shared objects (.so/.dylib vs .a)
hwloc will build a shared object with https://charm.cs.illinois.edu/gerrit/3450 Evan Ramos
 

Also available in: Atom