Project

General

Profile

Activity

From 12/30/2018 to 01/28/2019

01/29/2019

07:39 PM Bug #2049 (Merged): Virtual inheritance off multiple base classes fails to compile
Sam White

01/28/2019

02:23 PM Feature #2039 (Implemented): Update our copy of the Boost fcontext ULTs to Boost v1.69
https://charm.cs.illinois.edu/gerrit/#/c/charm/+/4928/ Sam White
02:12 PM Cleanup #2028 (Merged): Replace use of deprecated std::mem_fun with lambdas
https://charm.cs.illinois.edu/gerrit/#/c/charm/+/4797/ Sam White
02:11 PM Documentation #2044 (Implemented): Improve CkCallback documentation examples
https://charm.cs.illinois.edu/gerrit/#/c/charm/+/4927/ Sam White
12:44 PM Bug #2049 (Implemented): Virtual inheritance off multiple base classes fails to compile
https://charm.cs.illinois.edu/gerrit/#/c/charm/+/4926/ Sam White

01/27/2019

02:15 PM Bug #2049: Virtual inheritance off multiple base classes fails to compile
We should be able to have the fix in 6.9.1, thanks Sam White
06:21 PM Bug #2049: Virtual inheritance off multiple base classes fails to compile
I'm hoping this could make it into 6.9.1 since it's a very small fix :) Nils Deppe
06:15 PM Bug #2049 (Merged): Virtual inheritance off multiple base classes fails to compile
I have a base class... Nils Deppe

01/24/2019

03:07 PM Bug #2048 (New): Examine the converse header fields for all layers and remove/reduce fields where...
While implementing the ZC Bcast Send API, we came across several fields that could be potentially reduced/removed for... Nitin Bhat
02:36 PM Bug #2047 (New): Potential GCC 8 Hang
A NAMD user has emailed the NAMD mailing list with the following:
> So you can take pure NAMD/Charmm++, build it w...
Ronak Buch
02:21 PM Bug #2045: Chare array broadcast messages don't appear to be freed by the runtime
I agree with Jim's suggestion for moving away from a timer-based spring cleaning mechanism. I don't think the documen... Sam White
01:45 PM Bug #2046 (Merged): TRAM doesn't appear to support higher dimensional chare arrays
Some digging suggests that TRAM tries to invoke CkLocMgr::updateLocation with 2 parameters, whereas the version that ... Venkatasubrahmanian Narayanan

01/22/2019

10:12 AM Bug #2045: Chare array broadcast messages don't appear to be freed by the runtime
Perhaps spring cleaning could be triggered by the number of accumulated broadcasts (and/or amount of memory used by a... Jim Phillips
09:09 PM Bug #2045: Chare array broadcast messages don't appear to be freed by the runtime
Yes, and in that scenario you really want the zero copy API for broadcasts and reductions in order to avoid intermedi... Sam White
09:02 PM Bug #2045: Chare array broadcast messages don't appear to be freed by the runtime
One case that occurs to me that could be problematic is a typical implementation of Stochastic Gradient Descent commo... Juan Galvez
08:58 PM Bug #2045 (In Progress): Chare array broadcast messages don't appear to be freed by the runtime
Sam White
08:58 PM Bug #2045: Chare array broadcast messages don't appear to be freed by the runtime
Yeah, I agree with everything you said. Timer defaults are always hacky, but 1 or 5 seconds seems better than 1 minut... Sam White
08:44 PM Bug #2045: Chare array broadcast messages don't appear to be freed by the runtime
I'm just running with default values, so spring cleaning option seems to be the default. I didn't know about the othe... Juan Galvez

01/21/2019

04:05 PM Bug #2045: Chare array broadcast messages don't appear to be freed by the runtime
That's true. AMPI (and possibly other applications) sets '_isAnytimeMigration = false;' and '_isStaticInsertion = tru... Sam White
03:46 PM Bug #2045: Chare array broadcast messages don't appear to be freed by the runtime
I haven't looked at the array broadcast internals to know what the issue is, but it's possible that this works this w... Juan Galvez
02:31 PM Bug #2045: Chare array broadcast messages don't appear to be freed by the runtime
This should only affect parameter marshalled entry methods and [nokeep] broadcasts, where the user doesn't take owner... Sam White
02:16 PM Bug #2045 (In Progress): Chare array broadcast messages don't appear to be freed by the runtime
Need to debug further, but it seems like doing many broadcasts we run out of memory eventually do to the messages not... Sam White

01/18/2019

10:40 AM Documentation #2044 (Merged): Improve CkCallback documentation examples
"CkCallback(CkIndex_Foo::bar(), thisProxy[thisIndex])" should be the primary example shown. Sam White

01/17/2019

01:17 PM Feature #1987: Take advantage of streamable reductions inside CkMulticast
Could be done in 6.9.1 but we won't block 6.9.1 on it being done Sam White
01:16 PM Feature #1116 (Merged): AMPI broadcasts should be [nokeep]
Sam White
06:00 AM Support #2022 (Resolved): Add examples for Kokkos interoperability
Merged https://charm.cs.illinois.edu/gerrit/#/c/charm/+/4735/ Jaemin Choi

01/16/2019

10:47 AM Feature #1236 (Implemented): Avoid receiver-side copy for large contiguous messages. API for char...
Gerrit: https://charm.cs.illinois.edu/gerrit/c/charm/+/4820 Nitin Bhat

01/14/2019

11:44 AM Bug #1968 (Merged): Some libraries are built as shared libraries even without --build-shared
Evan Ramos

01/13/2019

01:13 PM Feature #2043 (New): Add +isomalloc_sync build-time options
Add two options: one to run with '+isomalloc_sync' by default, another to disable the warning about it completely.
T...
Sam White

01/11/2019

04:25 PM Bug #1968 (Implemented): Some libraries are built as shared libraries even without --build-shared
https://charm.cs.illinois.edu/gerrit/c/charm/+/4893 Evan Ramos
02:11 PM Bug #2030: tests/ampi/megampi crashes in MPI_Comm_free
I think this showed up in a multicore-win-x86_64 build today, in addition to mpi-win-x86_64-smp:
http://charm.cs.i...
Evan Ramos

01/10/2019

01:57 PM Bug #2026 (Merged): charm-6.8.2 breaks linking process of NAMD if it's patched with shared versio...
Evan Ramos
10:45 AM Documentation #1927 (Implemented): Evaluate other documentation possibilities
Patch to convert Latex to RST here:
https://charm.cs.illinois.edu/gerrit/c/charm/+/3624
Matthias Diener

01/07/2019

02:16 PM Bug #2026: charm-6.8.2 breaks linking process of NAMD if it's patched with shared version of PLUMED
Any confirmation from upstream about whether my patch resolves the issue? Evan Ramos

01/04/2019

04:35 PM Bug #2026 (Implemented): charm-6.8.2 breaks linking process of NAMD if it's patched with shared v...
Evan Ramos
03:39 PM Bug #2026: charm-6.8.2 breaks linking process of NAMD if it's patched with shared version of PLUMED
How's this? https://charm.cs.illinois.edu/gerrit/c/charm/+/4884 Evan Ramos
02:06 PM Bug #2026: charm-6.8.2 breaks linking process of NAMD if it's patched with shared version of PLUMED
Ping. Jim Phillips

01/03/2019

05:14 PM CharmDebug Bug #1942: CkStartQD never triggered, even though all all entry methods have returned.
Is there a short .ci and .C example of this problem? Evan Ramos
03:16 PM Support #2041: charmrun with mpirun instead of srun?
For the mpi-... builds of Charm++ you can run the binary with mpirun/mpiexec/srun directly rather than via charmrun. ... Jim Phillips
12:46 PM Support #2041: charmrun with mpirun instead of srun?
No, it's not. What MPI library are you using? And could you post the output of what happens when you try the "++mpiex... Sam White
11:08 AM Support #2041: charmrun with mpirun instead of srun?
It's v6.8.0. We're working on adding support for v6.9.0 to our code base (https://github.com/sxs-collaboration/spectr... Geoffrey Lovelace
12:38 PM Feature #2042: Add [whenidle] to simplify speculative execution
Basic pattern (without templates or dependencies on more than C based trickery) is:... Eric Bohm
12:21 PM Feature #2042: Add [whenidle] to simplify speculative execution
If you have an example of the ugly Ccd code to do this, could you post it here? I anticipate needing to implement thi... Sam White
10:49 AM Feature #2042 (New): Add [whenidle] to simplify speculative execution
There are various cases where some computation can be done in the absence of other work, but would be of a low priori... Eric Bohm

12/31/2018

04:37 PM Support #2041: charmrun with mpirun instead of srun?
Is that v6.8.0 or v6.8.2? Can you try with v6.9.0? You shouldn't need to build Charm++ any differently. Sam White

12/30/2018

07:07 AM Support #2041: charmrun with mpirun instead of srun?
I’m using charm++ v6.8. Geoffrey Lovelace
06:47 AM Support #2041: charmrun with mpirun instead of srun?
I tried this, and it still called srun. Might I need to build charm++ differently? I used
./build charm++ mpi-lin...
Geoffrey Lovelace
 

Also available in: Atom