hppritcha commented on issue open-mpi/ompi#12796.
@mdosanjh are you still working on this? …
grondo commented on issue flux-framework/flux-core#7621.
Great! Already approved so feel free to merge when ready!…
garlick commented on issue flux-framework/flux-core#7621.
OK, just pushed an update with the following changes: …
markcmiller86 commented on issue visit-dav/visit#21034.
Will do…will have to wait ‘til monday though….
v-dobrev commented on issue spack/spack-packages#143.
Hi @jacobmerson, …
tgamblin commented on issue spack/spack#52613.
@haampie: I’ve reworked this to use rsvg, which preserves fonts properly and renders all the figures as a browser would. CI is also updated….
tzanio commented on issue mfem/mfem#5383.
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….
v-dobrev commented on issue mfem/mfem#5383.
ping: @tomstitt …
artv3 commented on issue mfem/mfem#5382.
@helloworld922 , the RAJA API takes resource objects that can be used to specify the stream, we have examples here: …
adayton1 commented on issue mfem/mfem#5383.
Thank you!…
balos1 closed issue llnl/sundials#955.
[BUG] arkLsMassSolve ignores some return values from user callbacks.
One of the folks working on the deal.II SUNDIALS interfaces, @vovannikov, has spent what I think must have been a quality afternoon tracing why one of our tests fails. The details are in https://github.com/dealii/dealii/pull/19838 but come down to code that I (mis-)wrote: If a user callback fails in user-provided linear solvers, I either returned +1 or -1, depending on whether it’s a recoverable or non-recoverable failure (and zero if everything is fine). I think that’s what the documentation of SUNDIALS at some point said I should do, but it’s possible that that was about the nonlinear solver callbacks in KINSOL and not about the SUNDIALS linear solvers. Regardless, for the linear solvers, I need to return SUNLS_PSOLVE_FAIL_REC or SUNLS_PSOLVE_FAIL_UNREC, with values +805 and -808, respectively….View Comment
jameshcorbett commented on issue llnl/scr#614.
Sorry for the delay in getting back to you! Yes we would be interested in working with you on this. I’m not sure yet how complex it would be to support this in SCR but I’d be very happy to investigate with you!…
adrienbernede open issue llnl/radiuss-shared-ci#79.
Hmm, interesting... I read the same statement that it was "gpu allocated" associated with the suggestion to use "--gpu".
Hmm, interesting… I read the same statement that it was “gpu allocated” associated with the suggestion to use “–gpu”….View Comment
JustinPrivitera open issue llnl/conduit#1637.
Device Support: Port Blueprint Transforms to the GPU
We will list transforms that we are porting and check them off here….View Comment
BradWhitlock open issue llnl/axom#1724.
Potential race condition in axom::copy when Umpire is enabled.
In the new heavily_mixed MIR example, I was calling axom::copy() in an axom::for_all<axom::OMP_EXEC>() loop to copy some data into a slice of a 3D array. This was to speed up material construction on 3D meshes since I could have many 2D slices constructed in parallel….View Comment
adayton1 commented on issue llnl/axom#1893.
MFEM is already working on a fix: https://github.com/mfem/mfem/pull/5383…
samuelkgutierrez closed issue hpc/quo-vadis#486.
Question About Device Ordinal
@eleon, what is your vision for the purpose of QV_DEVICE_ID_ORDINAL? I don’t understand its purpose. Please see https://github.com/hpc/quo-vadis/blob/master/src/qvi-hwloc.cc#L1078….View Comment
eleon commented on issue hpc/quo-vadis#486.
Yes, we could calculate the ordinal id from get_dev_id_from_name. …
grondo commented on issue flux-framework/flux-sched#1513.
@kkier and @ryanday36 asked us to bump the priority on this one. We have weekly DATs where queues need to be reconfigured and they need to understand how to avoid this issue when manually moving nodes between queues on the live system….
artv3 commented on issue llnl/RAJA#2009.
> Quick poll @llnl/raja-core , do we want to call it RAJA::mask (
adayton1 commented on issue llnl/RAJA#2009.
Could you update the title of this MR?…
trws commented on issue llnl/RAJA#2006.
Honestly I think it does, at least in as much as when iter + stride is executed we want the result to be of the same type as iter. I suppose it wouldn’t have to be, but if not it would have to be convertible to the type of iter and that seems a bit harder to reason about. Also even if we actually do re-order them based on the sign of the stride, in a logical sense I think of it more like:
…
adayton1 closed issue llnl/CHAI#353.
Add STL compliant typedefs to chai::ManagedArray
It would be useful if the chai::ManagedArray provided some of the standard typedefs that ordinary C++ STL containers generally provide. I was thinking a nice set would be at least:…View Comment
bgunnar5 open issue hyperlight-dev/hyperlight-sandbox#141.
Wheels for glibc 2.28
The MADA team from LLNL is using Microsoft Agent Framework. We are currently unable to install this on our HPCs due to the hyperlight-sandbox-backend-wasm only having build wheels for glibc 2.34:…View Comment
grondo commented on issue flux-framework/rfc#526.
Thanks! I’ve set MWP….
chu11 commented on issue flux-framework/flux-core#7681.
> Thanks @chu11 - I appreciate the review comments! I’ll start working through those starting with the design doc. …
cmoussa1 closed issue flux-framework/flux-accounting#858.
job usage: redesign job usage decay to make it more configurable post-DB deployment
I had an offline conversation with @ryanday36 about the possibility of redesigning how job usage is decayed in flux-accounting to make it more configurable post-deployment….View Comment
becker33 commented on issue spack/spack#52602.
Closing this for now because #52603 and #52604 both resolve it without reverting….
mej closed issue mej/nhc#169.
NHC fails on Rocky 9
NHC complains about sshd not running on Rocky Linux 9.6:…View Comment
cbritopacheco commented on issue cbritopacheco/rodin#302.
Closing: this was an optimization/uniformity change (MatAXPY preassembled-merge fast path + OpenMP buffer hoisting), not a bugfix. The fast path regressed value-Dirichlet assembly on PETSc 3.19 (MatZeroRowsColumns missing-diagonal). Not worth the risk
artv3 commented on issue llnl/RAJAPerf#684.
@michaelmckinsey1 take a look here: https://github.com/llnl/RAJA/blob/develop/benchmark/ltimes.cpp, it would be cool to also have a GPU shared memory version as a tuning! …
hppritcha closed issue open-mpi/ompi#12080.
MPI 4.1 - need to implement MPI_GET_HW_RESOURCE_INFO
See https://github.com/mpi-forum/mpi-issues/issues/154 …View Comment
tzanio commented on issue mfem/mfem#5358.
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….
JustinPrivitera commented on issue llnl/conduit#1620.
Closing in lieu of #1627 and #1625…
artv3 commented on issue llnl/RAJAPerf#693.
Nice @michaelmckinsey1 ! What would be neat is to add a batching parameter and allow for different batch sizes, maybe 4 is good GPU X or 7 is good for GPU Y type of thing …
artv3 commented on issue llnl/RAJA#2009.
@llnl/raja-core , building on the comments and suggestions, the latest push explores the following function to be used in RAJA::launch : …
garlick commented on issue flux-framework/flux-core#7681.
Updates: …
grondo commented on issue flux-framework/flux-core#7674.
I’ve set MWP here….
markcmiller86 commented on issue visit-dav/visit#20971.
@cyrush if you a min, pls take a look. Artifact handling follows newest guidance from MS and other changes may reduce (I am still evaluating LFS budget usage reports to be sure) LFS BW usage….
balos1 commented on issue llnl/sundials#961.
@BarrySmith this should fix the issue you were having in #935 …
cyrush commented on issue llnl/conduit#1625.
The simplified templates look good – Note: think we want to keep dispatch with the functor as a just in case. …
cmoussa1 commented on issue flux-framework/flux-accounting#883.
Thanks a bunch for the feedback here @jameshcorbett! Setting MWP here…
cyrush commented on issue Alpine-DAV/ascent#1735.
To test, we can use: …
JustinPrivitera commented on issue visit-dav/visit#19544.
reopening so that we see this since it was marked un-reviewed….
imguoguo closed issue sipeed/picoclaw#3070.
[Security] OneBot inbound media URL handling allows host-side arbitrary fetch in PicoClaw
### Advisory Details…View Comment
camierjs commented on issue mfem/mfem#5230.
Some failures are hit on GPU with ./pgpu_unit_tests [ParticleSet] triggered by the CHECK(code_N_count == 0); of tests/unit/fem/test_particleset.cpp with expansion: 1 == 0….
trws commented on issue flux-framework/flux-sched#1509.
Yeah, that hardcoded list of subsystems should never have existed. I’ll add it to the list of things to fix for subsystems….
wihobbs commented on issue flux-framework/flux-sched#1502.
> so I guess the docs are not being built by default? …
milroy commented on issue flux-framework/flux-sched#1505.
> In the resource graphs used to repro the issue (e.g. hetchy_jgf_formatted.json) there are many (32 I think?) ssd vertices per chassis, so even if the Fluxion can’t match multiple slots on a single SSD vertex, the job should still match. Am I missing something? …
garlick commented on issue flux-framework/flux-sched#1502.
Nice! Content-wise this seems like a good start. …
wihobbs closed issue flux-framework/flux-docs#130.
reorganize top-level site
We have some good documentation on the docs site but the ad-hoc growth of the documentation has resulted in lack of coherence and flow of the overall site. Also given that Flux is starting to be installed as a system RM at this point, we may get many more users that are most interested in “how do I use a cluster running Flux” vs “how do I build this new thing from source and poke at it” …View Comment
chu11 commented on issue flux-framework/flux-core#7668.
haven’t looked at the code yet, but it looks like in this proposal we would be changing the database schema. If we go forward with this general approach, would we like to consider other schema changes while we are at it? I’m mostly thinking of:…
garlick commented on issue chaos/diod#171.
Np! Thanks for the link!…
sam-maloney commented on issue Vanille-N/meander.typ#6.
> if you call par explicitly then the new paragraph splitting logic should be capable of correctly indenting it….
adrienbernede commented on issue llnl/CARE#378.
@lc-hubcast run pipeline…
grondo commented on issue flux-framework/rfc#523.
As I commented in flux-framework/flux-core#7615, this won’t work because it would break backwards compatibility: the scheduler and resource modules of subinstances fetch the R key from the parent. This change would break launching a previous version of Flux under the new version….
tpatki commented on issue flux-framework/flux-multi-cluster-utilities#8.
Closing as the other PRs supersede this one….
jameshcorbett commented on issue flux-framework/flux-multi-cluster-utilities#24.
Hmmm, mergify config not yet complete. Will just merge manually while I figure it out….
sam-maloney commented on issue flux-framework/flux-core#7677.
At a basic housekeeping level, eventually the commit messages will have to follow the formatting requirements: https://flux-framework.readthedocs.io/projects/flux-rfc/en/latest/spec_1.html#patch-requirements…
emily-howell commented on issue Alpine-DAV/ascent#1494.
@jfavre I have made a PR #1736 that addresses a bug I found relating to this ticket when trigger actions being passed via ascent_actions.yaml. I wanted to confirm if this is the same bug you encountered here. If not, could you provide more information for me to be able to replicate this bug….
sam-maloney commented on issue spack/spack-packages#5177.
Realized I was missing a colon in the when clause of flux_sched/package.py at L102, because I had copied the previous line which was also missing it :laughing: and there were 2 lines with when="@0.19.0" so I changed the second one to when="@0.20.0" as that seems to make sense with the release dates and the commit when it was added….
adayton1 commented on issue llnl/axom#1885.
Is there a way to iterate over the objects in the variant array without first converting to a vector?…
sam-maloney commented on issue flux-framework/flux-sched#1504.
> Works for me, may need to check the spack package for consistency after this. …
chu11 closed issue chu11/freeipmi-mirror#81.
groff warnings in manpages
Hi, QA checks on Debian spotted groff warnings in manpages, I did a fast patch in Debian I used to fix them, if you want, look at it:…View Comment
jedbrown commented on issue rsmpi/rsmpi#224.
Thanks, can you add the # Standard section(s) like all the other public functions?
…
mplegendre closed issue llnl/Spindle#189.
Careless mistake in commpath PR
A reminder that I need to be compiling with -Wall -Wextra -Werror….View Comment
tzanio commented on issue mfem/mfem#5352.
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….
jameshcorbett commented on issue flux-framework/flux-sched#1490.
> @jameshcorbett can comment more specifically, but my understanding is that the flux-coral2 module modifies the jobspec after the feasibility check is already done, so the job gets accepted then the jobspec is modified to contain the rabbit request. If you try that same command with --dry-run you should get the original jobspec….
trws commented on issue flux-framework/flux-sched#1490.
@jameshcorbett can comment more specifically, but my understanding is that the flux-coral2 module modifies the jobspec after the feasibility check is already done, so the job gets accepted then the jobspec is modified to contain the rabbit request. If you try that same command with --dry-run you should get the original jobspec….
grondo commented on issue flux-framework/flux-core#7660.
…
benzwick closed issue A2B-Technology-Corporation/Talk2View-Writer#33.
macOS: make install-oxt fails
JustinPrivitera closed issue visit-dav/visit#4976.
Revolve operator crashes with curvilinear meshes
### Describe the bug …View Comment
markcmiller86 commented on issue visit-dav/visit#21000.
@brugger1 is this ok now?…
hppritcha commented on issue spack/spack-packages#5169.
i thought these issues were addressed via https://github.com/spack/spack-packages/pull/5148 ….
v-dobrev open issue mfem/mfem#5363.
[CUDA] nvcc / cicc memory exhaustion (OOM) compiling element assembly kernels with -arch=sm_120 (CUDA 12.8)
# cicc memory ballooning / OOM compiling element-assembly kernels for sm_120 (Blackwell) with CUDA 12.8…View Comment
adrienbernede closed issue llnl/radiuss-spack-configs#175.
[spack-packages-tracker:shared-ci] Update spack-packages reference to 599874d0ccdf
Automated weekly check detected relevant changes in spack/spack-packages for shared-ci….View Comment
JustinPrivitera open issue llnl/conduit#1614.
Conduit Device Support Ongoing Development
## High-Level Goals…View Comment
markcmiller86 commented on issue llnl/H5Z-ZFP#168.
> From my reading it looks like the test is just trying to corrupt a bit of the raw data towards the start of the dataset’s data (now the first chunk in this case)? If so, this looks fine to me. …
grondo commented on issue flux-framework/rfc#521.
Sure. I had considered doing that but figured someone looking a job eventlog might naturally think to look at the “Job States and Events” RFC. I didn’t see anywhere mentioned that this was the job manager RFC. Easy enough to add a new RFC….
garlick commented on issue flux-framework/rfc#521.
I’m surprised we haven’t done this already! …
milroy commented on issue flux-framework/flux-sched#1503.
> In v0.51.0 the jobspec matches with firstnodex….
imguoguo closed issue sipeed/picoclaw#3080.
[Security] PicoClaw launcher `allowed_cidrs` can be bypassed through same-host loopback proxying during first-run setup
picoclaw: PicoClaw launcher allowed_cidrs` can be bypassed through same-host loopback proxying during first-run setup…View Comment
cjvogl commented on issue mfem/mfem#4905.
Closing this PR in lieu of #5362 that will include all the reconstruction work here alongside the SAMRAI-MFEM coupling work….
gardner48 commented on issue llnl/sundials#957.
When using an implicit method with the default nonlinear solver (Newton’s method) you need to create and attach a linear solver before evolving the problem in time. The lines starting here in the ark_analytic.c example show how to do this for a dense linear solver (and matrix). There should be a runtime error in this case, but it seems there is a missing check….
jeremylt commented on issue jeremylt/personal_website#3.
I need to read the Leiden Declaration and see if it applies here…
wihobbs commented on issue flux-framework/flux-sched#1493.
I want to look into that failure, so removing MWP for now….
sam-maloney commented on issue flux-framework/flux-sched#1501.
> If we require >= flux-core v0.78.0, then we could drop the old rc1 style scripts in etc/. In fact, there could be an implicit dependency on modprobe for the sched-fluxion-feasibility module already….
garlick commented on issue flux-framework/flux-core#7467.
It might actually be better to just add a FLUX_MSGGFLAG_COMPOUND flag to indicate that the payload is compound (steal FLUX_MSGFLAG_USER1), and rather than requiring JSONL, use size prefixed segments. Then a message could contain multiple raw payloads, which could potentially avoid base64 encoding in KVS commit messages, and allow batching of content load/store payloads….
cbritopacheco commented on issue cbritopacheco/rodin#294.
@copilot resolve the merge conflicts in this pull request…
jeremylt commented on issue CEED/libCEED#1982.
hmm, this should be compared to the operator application kernel to make sure we are not breaking HipStar support …
hppritcha commented on issue open-mpi/ompi#13956.
Yes I do agree….
balos1 open issue llnl/sundials#955.
[BUG] arkLsMassSolve ignores some return values from user callbacks.
One of the folks working on the deal.II SUNDIALS interfaces, @vovannikov, has spent what I think must have been a quality afternoon tracing why one of our tests fails. The details are in https://github.com/dealii/dealii/pull/19838 but come down to code that I (mis-)wrote: If a user callback fails in user-provided linear solvers, I either returned +1 or -1, depending on whether it’s a recoverable or non-recoverable failure (and zero if everything is fine). I think that’s what the documentation of SUNDIALS at some point said I should do, but it’s possible that that was about the nonlinear solver callbacks in KINSOL and not about the SUNDIALS linear solvers. Regardless, for the linear solvers, I need to return SUNLS_PSOLVE_FAIL_REC or SUNLS_PSOLVE_FAIL_UNREC, with values +805 and -808, respectively….View Comment
samuelkgutierrez commented on issue hpc/quo-vadis#467.
From today’s meeting with Guillaume….
cmoussa1 commented on issue flux-framework/flux-accounting#872.
Thanks @jameshcorbett! These were all great suggestions. I’ve gone ahead and force-pushed up some changes to this PR based on your feedback. I will set MWP here shortly…
hppritcha commented on issue eschnett/mpif#1.
I will set up a poll and invite some folks so we can set up a good time for a call….
alecbcs commented on issue spack/spack-packages#5104.
@spackbot run pipeline…
hppritcha commented on issue open-mpi/ompi-www#563.
@edgargabriel oops forgot to add this file…
alecbcs commented on issue llnl/hubcast-test#6.
@lc-hubcast help…
artv3 commented on issue llnl/camp#205.
Used internally: https://github.com/llnl/camp/blob/56b194f3ea3d0e9d6e2f236de39a4886063e4968/include/camp/resource/hip.hpp#L187C5-L187C37 …
adayton1 commented on issue llnl/Umpire#1072.
> Update: Hubcast doesn’t understand jobs that are allowed to fail. We can either fix the job failure or try to fix within Hubcast somehow.. stay tuned …