markcmiller86 commented on issue spack/spack-packages#2346.
I am getting this error in one of the CI checks… …
cbritopacheco open issue cbritopacheco/rodin#127.
vsoch open issue ROCm/k8s-gpu-dra-driver#7.
current_compute|memory_partition do not exist
To discover the devices, we are reliant on current_compute_partition (and the same for memory) to exist. These don’t seem to exist on the system I’m using:…View Comment
markcmiller86 commented on issue LLNL/Silo#312.
This has been started here, https://github.com/spack/spack-packages/pull/2346…
JustinPrivitera commented on issue visit-dav/visit#20152.
> Is this correct? Shouldn’t the <td></td> tags be within the <tr></tr> tags?…
tgamblin commented on issue spack/spack#51445.
Looks mostly good modulo the failing test :)…
garlick commented on issue chaos/diod#160.
LMK = Let Me Know…
BradWhitlock open issue LLNL/axom#1714.
Make printed primal primitives easier to turn into C++ code.
It is a chore to translate primal primitive printed output into a C++ program. If I’m printing a set of points, vectors, planes, or polygons, it would be nice if there was a code() method that would write the object as valid (or closer to it) C++ code. This would it easier to make export problem data to short programs for debugging….View Comment
chrundle commented on issue FLASK-LLNL/ChARGe#74.
> @chrundle Let’s put the LLNL specific files in our local project space for now and we can then add it to a gitlab internal project to keep it version controlled. …
cyrush open issue Alpine-DAV/ascent#1628.
revisit if we absoluly need cmake python introspection to be strick about python site packges dir
https://github.com/Alpine-DAV/ascent/blob/670851b87793f17f64f3346df4d1f1f251f21099/src/cmake/thirdparty/SetupPython.cmake#L57…View Comment
cyrush commented on issue visit-dav/visit#20707.
@JustinPrivitera thanks for this fix!…
becker33 commented on issue spack/spack-packages#2176.
@dguibert did you get a chance to look at this again? Let me know if you need to hand it off….
cmelone commented on issue spack/spack-packages#2290.
I can certainly add a +serial option to the package, but this is what I get when running spack spec cabana-pd
…
becker33 commented on issue spack/spack#50560.
I have not gotten a chance to debug it yet….
hppritcha commented on issue open-mpi/ompi#13280.
the arrays were’t being cleaned up anyway in the existing code. some checks for predfined datatypes resulted in not running the callback that was freeing these arrays. …
jandrej closed issue mfem/mfem#5077.
Some typos in DifferentiableOperator doxygen
See https://github.com/mfem/mfem/blob/cd4e583f9f5403ebabeb6d3dd13bc8a548c2edd8/fem/dfem/doperator.hpp#L214-L215…View Comment
v-dobrev closed issue mfem/mfem#5042.
Failed to build against hypre 3.0.0
tzanio commented on issue mfem/mfem#5097.
This PR is now under review (see the table in the PR description). To help with the review process, please do not force push to the branch….
cyrush commented on issue mfem/mfem#4307.
@JustinPrivitera this can be closed in lieu of #5097 - we can continue discussion in the PR….
garlick commented on issue flux-framework/flux-core#7112.
Thanks!…
grondo commented on issue flux-framework/flux-core#7112.
Strange. I had to re-approve and then mergify finally merged this one. :shrug: …
cyrush commented on issue LLNL/axom#1709.
thanks @kennyweiss …
adrienbernede commented on issue LLNL/Umpire#1009.
Now that I read your full question, here is the full answer: …
jeremylt commented on issue CEED/libCEED#1904.
```py …
alecbcs commented on issue spack/spack#51522.
I’m nervous about picking the “preferred” version of a package without interpreting any constraints. This works now because cmake’s preferred version is v3.31.9, but the second that flips to v4 it’ll break since clingo-bootstrap@spack <= clingo@5.7.1 which conflicts with cmake@4….
garlick closed issue flux-framework/rfc#198.
rfc14: formatting problem in tasks section
There appears to be a formatting error here …View Comment
chu11 commented on issue flux-framework/flux-core#7186.
re-pushed, throwing an extra flux-fsck cleanup in there…
grondo commented on issue flux-framework/flux-coral2#439.
Ok, I kind of doubt that is where the time in the sysadmin prolog (/etc/flux/system/prolog) is being spent anyway, since we know the dws_environment event is already available by the time it starts running….
cmoussa1 commented on issue flux-framework/flux-accounting#772.
It’s also worth mentioning that #752 added a continuous re-check of held jobs after every flux account-priority-update, so at least there is the potential for the plugin to sort itself out on a periodic basis….
grondo commented on issue flux-framework/flux-accounting#772.
Looking at the current job-manager code, a newly loaded plugin is sent the list of added jobs using zhashx_values() on the active job list:…
JustinPrivitera commented on issue LLNL/conduit#1475.
You are welcome. Please let us know if you have further questions….
artv3 open issue LLNL/RAJAPerf#578.
FEM Kernel Update
- [ ] Add an atomic variant for the mass PA kernel …View Comment
camierjs commented on issue mfem/mfem#4333.
Tested on macos with latest Caliper - works fine!…
wihobbs commented on issue flux-framework/flux-core#7043.
Yes, let’s check this off!…
jameshcorbett commented on issue flux-framework/flux-coral2#439.
> Oh, I see now, there’s zstop_nnf_services.sh in /etc/flux/system/prolog-job-manager.d and zstop_nnf_clientmount.sh in /usr/libexec/flux/prolog.d. Do we need both of them?…
artv3 commented on issue LLNL/RAJA#1832.
@bechols97 , can we bring this up to date?…
tomstitt open issue openai/codex#6224.
fallback when landlock isn't available
### What feature would you like to see?…View Comment
wihobbs commented on issue flux-framework/flux-pmix#113.
Hmm, it looks like I don’t have write access here to add labels….
trws commented on issue flux-framework/flux-core#7181.
That sounds right to me, but we should probably check and document. I think hwloc logical cores is exactly what we want it to be, should mean we can mock the topology however we want, but best to be sure….
eleon commented on issue flux-framework/flux-core#7182.
> > Another approach to this PR would be to move the adjustment of the cores and gpus strings into the shell itself so that the values returned to plugins are the adjusted values.
…
markcmiller86 open issue LLNL/Silo#479.
Memory leak in PDB Silo driver
I found this memory leak in the Silo PDB driver with valgrind and VisIt….View Comment
tgamblin commented on issue spack/spack#51165.
I think this is subsumed by #51162, but let me know if there is something here that is not covered there….
jedbrown commented on issue openjournals/joss-reviews#9333.
Hi @lindsayad @mprobson
tzanio commented on issue mfem/mfem#5023.
> I also saw that an ex41 is already in-review (#4289); please let me know if I should update this to be ex42. …
grondo commented on issue flux-framework/flux-core#7182.
I’ve force pushed a new version of this PR that fixes up the core/gpu assignments in the shell proper, so that flux_shell_rank_info_unpack(3) reflects the update core and gpu sets. This allows mpibind to just work:
…
tzanio commented on issue visit-dav/visit#19046.
@cyrush – I am happy to elaborate further on this if it will be useful…
cyrush commented on issue visit-dav/visit#19046.
Yes this does help - thanks for this explanation. …
cyrush commented on issue spack/spack-packages#2222.
The concretizer can be unpredictable, frustrating aspect of all of this …
imguoguo commented on issue radxa-pkg/radxa-firmware#29.
>
imguoguo commented on issue radxa-pkg/qualcomm-profiles#15.
>
hppritcha commented on issue open-mpi/ompi#13280.
all good ideas. i’ll work on these next week….
tomstitt commented on issue mfem/mfem#4333.
> @tomstitt, is this backward compatible, and if not, should we check for the Caliper version? …
v-dobrev commented on issue mfem/mfem#4993.
Re-merged in next for testing……
cjvogl commented on issue mfem/mfem#4496.
> I’m happy to see this through @cjvogl. Do I have your permission to commit and push to your branch? :) …
dylan-copeland commented on issue mfem/mfem#5087.
Thank you @chunyu79 for the update, and congratulations on finding a solution! I was able to run your test and reproduce the issue, but did not have time yet to investigate deeply. This was labeled as an AMR issue, but it seems more closely related to contact problems with constraints to couple boundaries of different domains. There has been recent work on this topic in PR 4996, and @psocratis may have some advice….
jeremylt commented on issue freeCodeCamp/contribute#1008.
Those linked PRs are from staff. Staff are not external contributors….
wihobbs open issue flux-framework/flux-sched#1415.
provide a reason a job was unsatisfiable
A job can fail on CORAL-2 systems because it requests ‘too many’ gpus (too many is italicized because the job might also request CPX or other modes that create more gpus.) A job can also fail for the standard reason that more nodes were requested than are available in the cluster. …View Comment
garlick commented on issue chaos/diod#158.
Oh this changed the github workflow so might need a manual merge. I’ll press the button….
adrienbernede commented on issue LLNL/RAJA#1934.
(cancelled the CI pipeline to restart and force Matrix to run with variable)…
adayton1 commented on issue LLNL/CARE#351.
@adrienbernede thank you!…
adrienbernede commented on issue LLNL/CARE#351.
@adayton1 This is ready ;)…
markcmiller86 commented on issue visit-dav/visit#20682.
Wow…great catch @biagas! No idea that logic existed….
becker33 commented on issue spack/spack#51490.
Looking at the optimization criteria for variant_default_not_used and variant_not_default, it looks like we might have to swap the ordering some in addition to removing the one for variant_not_default for already built packages….
alecbcs commented on issue spack/spack#51433.
Trying this PR I ran into a couple of errors on edge cases. Happy to help work on fixing these if needed. …
gardner48 commented on issue mfem/mfem#5073.
> Should the calls be added to the examples? …
jandrej commented on issue mfem/mfem#5073.
Should the calls be added to the examples?…
bgunnar5 commented on issue LLNL/merlin#587.
@lucpeterson I think they are but then again I’m biased cause I wrote them…
adayton1 closed issue LLNL/RAJA#1618.
Support assignment operator for layouts (and views)
RAJA/util/TypedViewBase.hpp:481:23: …View Comment
JustinPrivitera commented on issue visit-dav/visit#20666.
@IdoAkkerman please see continued discussion in #19046…
imguoguo commented on issue radxa-pkg/radxa-system-config#54.
> etc
wihobbs commented on issue flux-framework/flux-core#7153.
For reference, here’s how mpibind prints affinity in verbose mode:…
cmoussa1 commented on issue flux-framework/flux-accounting#778.
Thanks @jameshcorbett!…
tpatki commented on issue flux-framework/PerfFlowAspect#191.
Closing with PR #199 and #200 …
tpatki commented on issue LLNL/variorum#590.
@slabasan and I are in the process of testing ESMI+HSMP to obtain CCD-only (CPU) power on the APU. …
cmelone open issue LLNL/hubcast#180.
Hubcast for Spack CI
I went through the Spack CI Github-Gitlab sync script to see what features might be good to include in Hubcast, and which would be better served as plugins….View Comment
artv3 closed issue LLNL/RAJA#1664.
Improving RAJA integration with perf tools
PR: https://github.com/LLNL/RAJA/pull/1662 propagates the kernel naming API to all back-ends. For CUDA and HIP it is well defined as we can use the ROCTX and NVTX tools to name our kernels. We currently don’t have tool support for the other back-ends. An area of future investigation is how we can better integrate with tools and provide kernel naming when running on other backends. …View Comment
emily-howell closed issue Alpine-DAV/ascent#1515.
visit camera mode helper
We should explore a unified helper to accept visit style camera and convert it to a ascent camera. …View Comment
trws commented on issue ACCBDD/reclamation-dev#276.
Not to say it’s “good” but a create deployer with buckets of water will fill the cauldron. Still trying to work out if I can automate this thing the rest of the way, but it’s something….
JustinPrivitera commented on issue LLNL/conduit#1471.
put this in the blueprint spec docs…
mplegendre commented on issue LLNL/Spindle#43.
Closing since Nick is putting together updated containers….
vladotomov commented on issue CEED/Laghos#199.
Should I merge it or do you plan more additions?…
chu11 open issue flux-framework/flux-core#7173.
libkvs: function to asynchronously lookup all blobrefs in a kvs treeobj
Per discussion in https://github.com/flux-framework/flux-core/pull/7158 and https://github.com/flux-framework/flux-core/pull/6828, it’d be convenient if there was a function to handle parallel lookups of all blobrefs in a valref treeobj. This could be used in flux-dump and flux-fsck as they both do these parallelized lookups already….View Comment
balos1 commented on issue LLNL/sundials#789.
What compiler and version are you using? Can you attach your CMakeCache.txt file?…
gardner48 commented on issue LLNL/sundials#787.
@drreynolds I’ve updated the correction modification to address the roundoff issues and resolved the open comments. However, this PR also includes updates from #788 (makes adding logging to the constraint checks easier), so hold off on taking another pass over this PR until that one is merged….
jameshcorbett commented on issue LLNL/scr#604.
I fixed this up so all the actions pass, I think this is ready to go in. Will check in next team meeting. Actions are now passing: https://github.com/jameshcorbett/scr/actions/runs/18811341893…
hppritcha commented on issue open-mpi/ompi#9837.
this PR can be closed. we moved well beyond this as part of the big count work. and there’s more to come sometime later after the ‘c’ ABI stuff is complete and we want to venture into a fortran ABI world….
camierjs commented on issue mfem/mfem#3658.
> mpirun -np 4 pmesh-optimizer -m kershaw_6x6x6.mesh -o p -rs rl -mid 321 -tid 5 -ls 3 -nor -rtol 1e-8 -vl 1 -lc 0.5 -pa
…
v-dobrev commented on issue mfem/mfem#5085.
Some thoughts: some of the methods in ARKStepODE will be common with other similar classes. How do you envision avoiding repeated definitions? Also, when you inherit from both TimeDependentOperator and ARKStepODE, you get Mult from both and you have Width/Height (from the former) and Size from the latter which is not ideal.
…
garlick commented on issue chaos/diod#147.
I’ve got it, thanks!…
grondo commented on issue chaos/diod#147.
Yeah, want me to press the button?…
gardner48 commented on issue LLNL/sundials#788.
The main logging updates are in tools/suntools/logs.py. The changes to the source files and outputs are to update logging region labels to the new convention of <begin|end>-<region-name>[-list] (mostly adding -list where needed)….
markcmiller86 commented on issue LLNL/Silo#477.
Am recording here a smoking gun reproducer of the issue in error stack error messages. Attached is an intentionally corrupted HDF5 file. 1.14.4 produces an error message indicating a checksum failure. 1.14.6 does not. …
tgamblin commented on issue spack/spack#51461.
Per discussion in Slack, @alalazo’s going to make a PR that fixes the docs and does this: …
tzanio commented on issue mfem/mfem#5086.
Merged in next for testing……
artv3 commented on issue mfem/mfem#5082.
This is great! I’ll give it a spin these next few days. The one suggestion I have is lets add a constructor that enables us to specify the device memory type — I just pushed something up for quadrature function here: https://github.com/mfem/mfem/pull/5086 …
vladotomov commented on issue mfem/mfem#3658.
@camierjs …
jandrej commented on issue mfem/mfem#5022.
> Are there some GPU Tests enabled ?
…
cjvogl commented on issue mfem/mfem#4977.
This is now out of scope for this PR (given what was discussed at the technical meeting today), but if anyone is interested in future work into how refactoring TimeDependentOperator into modular interfaces might look (as @bslazarov had suggested at one point), I’ve started a WIP PR that showcases how an ARKStepSolver-specific interface could work: https://github.com/mfem/mfem/pull/5085…
chakshinglee commented on issue liruipeng/ml2#37.
The PR looks good to me too. I only have a design question around the Chebyshev frequencies (see above). Another thing, which is beyond the scope of this PR, is that how do you guys think about using config files instead of relying on command line options to pass in values of various parameters?…
grondo commented on issue flux-framework/flux-core#7168.
Well, I don’t actually see anything in RFC 44 that says an undrain event can only be posted for ranks that were previously drained, so one idea would be to add a new “mode” to the undrain RPC like “update” which simply posts an extra undrain event to the eventlog with the new reason. A tool would then have to process the entire resource eventlog to be sure the most updated undrain reason event for any given rank has been processed (instead of assuming the first undrain event after the corresponding drain event has the correct reason)….