JustinPrivitera commented on issue visit-dav/visit#20288.
I think it is any change made to it gets reverted after clicking apply….
markcmiller86 commented on issue visit-dav/visit#20288.
> Does it occur when you type to change the % or when you use the arrows?…
chu11 commented on issue flux-framework/flux-docs#295.
ooops created in wrong tab…
chu11 commented on issue flux-framework/flux-core#6694.
note for myself in current kvs-watch code…
cyrush commented on issue Alpine-DAV/ascent#1369.
resolved in https://github.com/Alpine-DAV/ascent/pull/1370…
cyrush open issue visit-dav/visit#20284.
movie wizard fails when selecting unsupported output type
### Describe the bug…View Comment
jedbrown commented on issue openjournals/joss-reviews#7845.
Hi, @rreischke! At first glance, I’m not finding automated tests or community/contributing guidelines. Can you please check the review checklist and make sure you’ve hit all items before we recruit reviewers?…
trws open issue flux-framework/flux-sched#1358.
remove use of `nodes_up` in favor of using the aggregate counts
This is, as far as I know, only used in the request_feasibility check I added not too long ago. It managed to get out of sync due to a bug in update that @milroy recently found, and should be removed so we can have less caches to worry about. My current thought is we should make it a method and in there use the aggregate filter to get the count, that way if we pivot on it later we can avoid having to track it down again….View Comment
grondo commented on issue flux-framework/flux-docs#294.
Thanks @wihobbs! I fixed your commit message nit and also a couple other issues preventing CI from passing and now I’ll set MWP….
grondo open issue flux-framework/flux-core#6692.
need a convenient way to determine the reason a job was unsuccessful
In a resilient batch job, a user may want to run a job in a loop until it is no longer feasible, or they may want to retry jobs that were unsuccessful due to node failure. Currently, there isn’t a great way to do either of these….View Comment
wihobbs commented on issue flux-framework/flux-core#6691.
Thanks for taking the time to dissect all of this @grondo! It was getting difficult in the thick of this project to figure out what exactly to do next, so thanks for getting me going. I’ll note a few things: …
jameshcorbett open issue flux-framework/flux-coral2#311.
`dws2jgf` does not apply properties to all nodes
Problem: when flux-dws2jgf writes out nodes, it neglects to output properties for all nodes that are not in DWS. For instance, on Tioga, tioga[42-43] are not associated with a rabbit, but they are listed in resource.toml and are assigned to queue mi300a. But flux dws2jgf wrote them out without assigning any properties whatsoever….View Comment
garlick commented on issue flux-framework/flux-coral2#310.
Thanks!…
adrienbernede commented on issue LLNL/radiuss-shared-ci#43.
markcmiller86 commented on issue LLNL/Silo#439.
The failing test is rocket.cxx. I just removed that test from the test suite. But apparently did so directly on 4.11RC. So, I am merging this to 4.11RC and then will cherry-pick to main….
jeremylt commented on issue CEED/libCEED#1762.
ugh, T_1D is wrong for this strategy. Pondering…
adrienbernede commented on issue spack/spack#49345.
Closed for security concerns….
hppritcha commented on issue open-mpi/ompi#13108.
closed via https://github.com/open-mpi/ompi/pull/13109…
cmoussa1 commented on issue flux-framework/flux-accounting#592.
Thanks @wihobbs! I added a small note about delete-* commands. Setting MWP here…
chu11 open issue chaos/gendersllnl#14.
compat: reitre ancient hostlist lib
copying…View Comment
chu11 closed issue chaos/genders#92.
genders.spec: does not handle python3 alone if building w/ debug packages
If building rpms w/ the spec file and only python3 is being built (vs python2 & python3) there is a logic error in the spec file IF debug packages are being built. The error will be:…View Comment
jameshcorbett commented on issue NearNodeFlash/NearNodeFlash.github.io#176.
Closing because it is a Flux issue….
artv3 commented on issue LLNL/RAJA#1760.
> > I like this idea: it makes permutations and layouts much easier to use. However, I think it should be called something like ViewMaker, or something else. When I see Reshape, I think that this will take an existing View I’ve already defined, and will alter its dimensions. Instead, this PR is creating a wrapper that returns a new View, and can be used to construct Views from scratch. I realize that ReShape is used in a similar way in MFEM as well, so if everyone else is on board with this then that’s fine with me too.
…
camierjs commented on issue mfem/mfem#3658.
> Is there a convenient way to modify T_MAX in the kernels?
…
jedbrown commented on issue coreylowman/cudarc#340.
I’m not sure what’s being proposed here, as memcpy_dtoh_async is unsafe, and any safe dtoh_copy_pinned would take a &mut PinnedHostSlice<T>. But it’s true that it’s necessary to prevent passing the same slice as destination for multiple copies (or as immutable arguments) on different streams….
wihobbs open issue codecov/feedback#665.
Bug Report: a 90-day lifetime of comments on deleted branches or closed PRs seems to cause a race condition
Describe the bug…View Comment
cmoussa1 commented on issue LLNL/cmvl#8.
thanks! merging this…
jedbrown commented on issue EnzymeAD/Enzyme#2025.
Is it appropriate to add a test that pulls in rsmpi? That would require building a Rust toolchain using this Enzyme and then cargo test (which would build rsmpi, assuming an MPI implementation is available somewhere in the CI environment). I’m supposing that “soon”, people will be getting their Enzyme-enable Rust from rustup, and that will increase interest in support for differentiable rsmpi, etc.
…
jedbrown commented on issue CEED/libCEED#281.
That’s a different sort of test harness. It would be a good project to create it, but it’s going to be more intrusive than something we already know how to measure. Also, constants matter, so having a test that enforces “no algorithm in PETSc should exhibit $\omega(n \log^k n)$ complexity” ($\omega$ notation is an open lower bound) isn’t assurance that code is efficient. This issue is about something practical for performance, not that other thing (which would be useful, but not a replacement for this)….
BradWhitlock open issue visit-dav/visit#20273.
Blueprint reader can mess up fields for point meshes.
### Describe the bug…View Comment
jedbrown commented on issue openjournals/joss#1411.
I agree with @sneakers-the-rat, and would go further, that people should know (and thus be defensive toward) the use of Copilot and similar products can automate what would assuredly be copyright violation if a human did it. For example, it reproduces substantial portions of SuiteSparse (an LGPL library) with copyright stripped and only trivial cosmetic changes (article from 2022), and this question of whether slight obfuscation is defense against copyright infringement is the subject of ongoing class-action litigation. (Obfuscation is not a defense against copyright infringement by a human, which is why clean-room design exists.) Companies are banking on use of such products being sufficiently ubiquitous they can obtain a legislative solution if courts rule against them. They may be successful, and yet I object on ethical grounds. …
correaa commented on issue mull-project/mull#1090.
right, I think in my case it combines with the problem of having to run the whole ctest for each mutation (for some reason that I can’t understand completely, perhaps because my code it heavily templated and different test instantiate different functions)….
hppritcha commented on issue mpi-forum/mpi-issues#916.
I will see about reorganizing the chapter without using items….
tzanio commented on issue mfem/mfem#4730.
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….
brendankeith commented on issue mfem/mfem#4702.
@tzanio, should I remove the modified file ex40.cpp to prepare for merging? …
milroy commented on issue flux-framework/flux-sched#1353.
Thanks for the review and feedback @jameshcorbett and @trws! Setting MWP….
alecbcs commented on issue flux-framework/flux-core#6667.
From the user perspective, I think any of these proposed solutions work for me as long as we can enforce exclusivity between users at the admin level to meet the security requirements to turn this on. (E.g. so users can’t disable the policy and get on a node with another user.)…
tpatki commented on issue flux-framework/PerfFlowAspect#182.
- @loudsun1997 : Please run the check-c-code-format.sh in the scripts directory to format the code.
…
JustinPrivitera closed issue LLNL/conduit#1377.
Help about defining the connectivity
I am trying to construct mesh using conduit. I checked the documentation (https://llnl-conduit.readthedocs.io/en/latest/blueprint_mesh.html#examples) and the examples but I could not figure out how connectivity needs to be defined. In this case I could have either tri or quad type elements in my mesh. So, I am trying to construct like following,…View Comment
jandrej commented on issue mfem/mfem#4730.
@mfem/editors whoever picks this up, this can be considered reviewed and ready to merge as long as CI passes. …
wihobbs commented on issue flux-framework/rfc#450.
I’m livid that the Oxford English Dictionary is behind a paywall now
trws commented on issue flux-framework/flux-core#6667.
That’s really good to know. If we go with the property approach, that should be pretty trivial (as would arbitrary other ones) if we plan it as a more general mechanism. Our label implementation is a generic string=>string map, and we can do matches on labels and their values, so that would be a pretty good way to start (I could see reasons not to do it that way longer term but I think it would get us pretty far). Maybe we could add something to jobspec or in the generic attributes for “ephemeral labels” that are associated with a job and apply to a resource or the node? Then it’s just two additions to fluxion:…
milroy commented on issue flux-framework/flux-sched#1352.
@trws can you give this a look when you get a chance? I think it’s near merge-able but it would be good to get your feedback….
grondo commented on issue flux-framework/flux-core#6670.
Thanks! setting MWP….
garlick commented on issue flux-framework/flux-core#6616.
Well, I’ll go ahead and set MWP here to keep things moving along. Thanks!…
jedbrown open issue coreylowman/cudarc#340.
soundness with async copies
The non-sync copy interfaces do not hold onto the host memory, so subsequent modification of the host memory can race with the DMA transfer to device. This is an example with htod_copy_pinned #336. Small sizes will spuriously pass this test due to what looks like eager copying, but N=10000 here fails every time in my tests (cuda-12.8, sm_89)….View Comment
JustinPrivitera open issue visit-dav/visit#20268.
Sort by date in open file dialog
https://teams.microsoft.com/l/message/19:f2ed7be3682d40d1b8e038744e500a09@thread.skype/1740678263001?tenantId=a722dec9-ae4e-4ae3-9d75-fd66e2680a63&groupId=70162982-9587-4bcc-ad53-20178c76fe11&parentMessageId=1740678263001&teamName=VisIt&channelName=General&createdTime=1740678263001…View Comment
markcmiller86 commented on issue visit-dav/visit#20263.
By the way, the “current solution” I refer to is available only via adjusting plugin code. That is problematic in two ways. First, it doesn’t help if a user just happens to open a database of a given plugin type with a baziilion variables. Second, every plugin needs to be adjusted in the same way to handle the bzillion variable cases….
jedbrown commented on issue rsmpi/rsmpi#204.
This would be great to support, especially since MSMPI appears to be unmaintained. I don’t have a test environment, but it should be doable. You might need to start with build-probe-mpi to detect Intel MPI….
jandrej commented on issue mfem/mfem#4728.
If you want to keep this branch as a testbed, sure….
vsoch commented on issue flux-framework/spack#303.
This will be closed by https://github.com/spack/spack/pull/49230…
grondo commented on issue flux-framework/flux-sched#1347.
Thanks! I’ll set MWP….
trws commented on issue flux-framework/flux-core#6667.
It depends on what you mean by policy in this context. We don’t support different match policies, currently, but we do support different queueing policies. We’ve given some thought to how to do something like this in the past, because it would be an efficiently improvement certainly. It ended up on the back burner largely because we get a somewhat similar effect by running a user-level flux instance that isn’t node-exclusive under the node-exclusive system scheduler. This would let us be much more flexible though….
chu11 commented on issue flux-framework/flux-core#6581.
saw this today …
vsoch open issue converged-computing/state-machine-operator#6.
A few TODO for state machine operator
These are from my personal notes - not high priority so putting them here….View Comment
markcmiller86 commented on issue betterscientificsoftware/bssw.io#2225.
@rinkug as per phone conversation, I followed your lead and adjusted example to encryption and softened ending remarks a tad. I think I’ve captured the spirit of what you asked. Have a look….
jeremylt commented on issue CEED/libCEED#1734.
I think we want to merge this before the review so we can have libCEED 0.13 and Ratel 0.4 with this? Is the question of the sync call the big blocker right now?…
vsoch commented on issue singularityhub/shpc-registry#301.
I think we would want to make sure that the path is derived as simply the digest with sif. If you want to open a PR to work on it I’d be happy to review….
jedbrown commented on issue openjournals/joss-reviews#7484.
@editorialbot recommend-accept…
tzanio commented on issue mfem/mfem#4452.
Re-merged in next for testing……
garlick commented on issue flux-framework/flux-security#203.
Also need reapproval on this one :-)…
trws commented on issue flux-framework/flux-sched#1348.
Looks like it, I’ll reopen then….
jameshcorbett commented on issue flux-framework/flux-sched#1348.
> Bah, just noticed you added a commit for this to #1346, I’ll close this one. …
wihobbs commented on issue flux-framework/flux-core#6667.
Ah, my apologies @alecbcs, I might have misled you yesterday – I thought we supported different match policies per queue. Although, as @grondo said, it might be possible to overcome that limitation if the only policy difference between the ci queue and batch or debug is node exclusivity….
jameshcorbett commented on issue flux-framework/flux-coral2#298.
Thanks for checking on our standard formats! I would never have thought of that….
wihobbs commented on issue flux-framework/flux-coral2#298.
Thanks @jameshcorbett!…
chu11 closed issue chaos/genders#70.
genders: slow on elcap
The admins have noticed that nodeattr (and by consequence libgenders) is a bit slow on elcap. The size and the buildup of data structures within libgenders is probably just eating up too much time. As an example with this genders db:…View Comment
cyrush commented on issue LLNL/conduit#1377.
Connectivity is local for each processor, actually local to each domain….
JustinPrivitera commented on issue LLNL/conduit#1377.
Hello @uturuncoglu! We provide example meshes within the Blueprint library. You can see how to create them and view examples with labeled renders for unstructured meshes in our docs (for triangles) (for quads)….
tgamblin closed issue spack/spack#44289.
`python+tkinter` cannot be concretized with --fresh
### Steps to reproduce…View Comment
trws commented on issue spack/spack#46416.
I’m not really fond of this as an idea, and especially not as a name for the variant. It’s building LLVM the package, but doesn’t provide LLVM, which prevents the use of LLVM in the same package DAG if this variant is selected. It would be less disruptive if it were a separate package, though I do understand that’s a harder rework. If this needs to go in this way, please make the variant self-explanatory. The current name standalone implies to me that it builds a standalone LLVM (I would expect that to mean self-hosted on the llvm libc or similar). Perhaps openmp_libraries_only?…
hppritcha commented on issue open-mpi/ompi#12226.
@jsquyres don’t forget to add this to your queue
chakshinglee commented on issue mfem/mfem#4704.
yes, changing the curvature works…
camierjs commented on issue mfem/mfem#4723.
> * I am trying to recreate the results seen in this paper, but I’m having issues running the benchmarks in bench_ceed.cpp. In particular, in BP2/1, I get an assertion error …
wihobbs commented on issue flux-framework/rfc#448.
> Sometimes it is impossible to restrict a commit message title to 50 characters, I think the pr-validator does set a hard limit at some point (I think it is 70 characters). That is why this one’s a SHOULD not a MUST. Same with the body line length….
cmoussa1 commented on issue flux-framework/flux-accounting#498.
I’m going to close this in favor of re-opening a more specific in the future if it pops up again. I’ve just made some changes in #590 to add more specific error messages to jobspec-updates in the case where an exception is raised….
hppritcha commented on issue mpi-forum/mpi-issues#892.
I am also going to leave the split example problem for any final changes to the doc since the break may move if we do additional edits….
v-dobrev closed issue mfem/mfem#4683.
AddMultTransposeMF inverted in SumIntegrator
Hi MFEM developers,…View Comment
brendankeith commented on issue mfem/mfem#4460.
> @brendankeith, a friendly reminder to please take another look! …
gardner48 commented on issue erf-model/ERF#2143.
@skang67 this will fix your build issues with FFTW on Mac…
garlick commented on issue flux-framework/flux-coral2#283.
Thanks! Setting MWP….
cjekel commented on issue cjekel/piecewise_linear_fit_py#126.
Thanks for the report! I believe i just fixed this, but let me know if you still have issues!…
alecbcs commented on issue spack/spack#48507.
@nhanford looks like the mixins have been added. If you’ve got a moment, it’d be great to get a final review….
vsoch commented on issue rootless-containers/usernetes#242.
I got this fully working in rootless mode - I’ll put together a quick writeup soon….
tzanio commented on issue mfem/mfem#4689.
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….
acfisher commented on issue mfem/mfem#4689.
> The CI is failing, @acfisher can you take a look? …
milroy commented on issue flux-framework/flux-sched#1337.
After further thought, I don’t think the above observation is related to the incident on tioga. In the off chance it is I’ll leave the comment posted….
grondo commented on issue flux-framework/flux-sched#1219.
Oh, wow! Thanks for that extra detail @trws!…
trws commented on issue flux-framework/flux-sched#1345.
I’m surprised to say I’m on the fence. If we can do it with existing facilities and without having to add a new state to track, that would be nice, and it sounds like the partial cancel plus remove should do that. That said, it seems like the lost state might be useful for long-term down nodes. Maybe worth getting some feedback from Ryan or someone on if that would be a thing admins would want?…
grondo open issue flux-framework/flux-core#6658.
missing `flux-restart` event in job eventlog after restart
There was an interesting difference in a couple job eventlogs on elcap after a Flux restart. Both of these jobs should have contained a flux-restart event at around the same time:…View Comment
garlick commented on issue flux-framework/flux-core#6656.
OK I did the latter quick test and verified that sending STOPPING=1 transitions the unit state to “deactivating”….
vsoch commented on issue converged-computing/flux-usernetes#21.
Note that we will need to update GKE to use n1-standard-8 with 1 GPU node, I just don’t see that we are going to get enough capacity of GPU to get 4/node….
JustinPrivitera open issue visit-dav/visit#20250.
`build_visit` invocation fails with mpich enabled on LC
I figured this would work now that #19600 had been addressed. mpich builds but we fail to build VisIt itself….View Comment
trws commented on issue spack/spack#49065.
I would be ok with this, but is there a reason not to add it back to the base? Does any other package use the result?…
jedbrown commented on issue openjournals/joss-reviews#7484.
@editorialbot generate pdf…
hppritcha commented on issue open-mpi/ompi#13076.
its a problem in the prte data server code. looking at a way to best address the issue….
v-dobrev commented on issue mfem/mfem#4403.
This issue is addressed in #4691. Please use that PR for further discussion….
cjvogl commented on issue mfem/mfem#4673.
@nmnobre , thanks for the improvements! I’ve added some finishing touches to the defaults and documentation… @jandrej and @helloworld922 , I believe this PR is ready for another look….
jeremylt open issue freeCodeCamp/freeCodeCamp#58931.
Book Organizer Misleading Instructions
### Describe the Issue…View Comment
wihobbs commented on issue flux-framework/flux-core#6655.
> It is interesting we don’t see this on other clusters….