sethwatts open issue visit-dav/visit#19469.
Incorrect rendering of piecewise constant pseudocolor apears in VisIt 3.4.1
Hello - …View Comment
alexrichert opened a pull request to spack/spack
sungho91 open issue mfem/mfem#4267.
How to prevent negative value with Legendre basis?
Hello all, …View Comment
colmenaresj commented on issue mfem/mfem#4258.
thank you for clarifying…
jandrej commented on issue mfem/mfem#4258.
If the mesh is actually distributed, the element numbers don’t match on each MPI rank. That part of the algorithm is basically missing. gslib knows which element it found on what MPI rank but the gather/scatter requires a custom data package to be communicated. We’re currently trying to get gslib to accommodate that….
mergify[bot] closed issue flux-framework/flux-core#5896.
flux-job: check for stopped queue in `flux job attach` and note this in the status line
When flux job attach is waiting for a job to start it emits a status line to give the user some feedback about what’s happening:
…View Comment
Steven-Roberts open issue LLNL/sundials#461.
Incorrect Format Specifies for Error Messages using Double Precision
I noticed that some ARKODE error messages where getting cut off, .e.g, At t = 0.001, mxstep steps taken before reaching tou. When SUNDIALS is built in double precision, some format specifiers seem to be for long double, which I think is the issue:
…View Comment
BradWhitlock opened a pull request to LLNL/axom
markcmiller86 pushed to visit-dav/visit
update opts
brugger1 pushed to visit-dav/visit
Update the last test suite pass on pascal.
JustinPrivitera open issue visit-dav/visit#19468.
Remove `MV2_ENABLE_TOPO_AWARE_COLLECTIVES=0` from custom launcher
We should remove SETENV("MV2_ENABLE_TOPO_AWARE_COLLECTIVES", 0) from customlauncher now that LC has added this to the default SLURM environment….View Comment
eugeneswalker pushed to spack/spack
[TAU Package] Updates for rocm (#43790)
- Updates for rocm
Updated for rocm@6 Added conflict between rocprofiler and roctracer. Request either +rocprofiler or +roctracer when +rocm. In this case, it automatically builds for one, instead of displaying the message. Request +rocm when using either +rocprofiler or +roctracer. In this case, it automatically builds with +rocm, instead of displaying the message.
Disabled the tests. Will update them with the new test method.
- [@spackbot] updating style on behalf of jordialcaraz
Co-authored-by: jordialcaraz jordialcaraz@users.noreply.github.com</small>
jrood-nrel opened a pull request to spack/spack
dependabot[bot] opened a pull request to spack/spack
jordialcaraz commented on issue spack/spack#43790.
Any other comments? …
spackbot-app[bot] commented on issue spack/spack#43830.
…
jack-morrison commented on issue spack/spack#43830.
@spackbot fix style…
Integrate changes not caught in merge
add sample run to getsubmesh
Fixed single precision compilation by using std namespace math functions.
fix time step alg
tomstitt commented on issue mfem/mfem#3844.
Thank you @sebastiangrimberg for fixing the unit tests! I completely neglected those =| …
v-dobrev commented on issue mfem/mfem#4172.
> Perhaps we could put the commit (2e37f2c) fixing (p)minamal-surface.cpp here? 🤔
…
mlstowell commented on issue mfem/mfem#4254.
Hi, @Raghavendra0117, …
liruipeng pushed to hypre-space/hypre
update tioga saved
oseikuffuor1 reviewed a hypre-space/hypre pull request
@victorapm, I made a few comments for your review. Otherwise, the overall idea looks good to me. Thanks….
mergify[bot] closed issue flux-framework/flux-sched#1108.
Rv1 `attributes` section will be deprecated
flux-framework/rfc#402 proposes to remove the optional attributes section from Rv1. Fluxion currently uses this section to store the queue in attributes.system.scheduler.queue, though it isn’t clear why this is needed since this duplicates the queue set in current jobspec.
…View Comment
milroy commented on issue flux-framework/flux-sched#1176.
Thanks for the detailed and helpful feedback @trws and @garlick! Setting MWP….
trws commented on issue flux-framework/flux-sched#1176.
Bounced the go test runner, looks like it failed to fetch the image for some reason. Feel free to set MWP @milroy, great stuff!…
wihobbs reviewed a flux-framework/flux-core pull request
Sorry, user error on my part when running the previous reproducer. Looks great, @grondo! …
codecov[bot] commented on issue flux-framework/flux-core#5911.
## Codecov Report…
grondo commented on issue flux-framework/flux-core#5911.
Thanks! And thanks for catching the original error @wihobbs. I’ll set MWP now….
garlick commented on issue flux-framework/flux-core#5899.
Thanks! Setting MWP….
gardner48 pushed to LLNL/sundials
Feature: Support Kokkos Kernels v4 (#460)
drreynolds pushed to LLNL/sundials
Finished first pass at revised ARKODE documentation
balos1 opened a pull request to LLNL/sundials
tong10 pushed to LLNL/psuade
Update PsuadeInterpreter.cpp
JustinPrivitera pushed to LLNL/conduit
fields is done but no tests
Add host device annotations and device unit tests for Polygon
Add quotation marks for Windows case
Co-authored-by: Chris White white238@llnl.gov</small>
davidbeckingsale pushed to LLNL/Umpire
Use spack environments (#875)
Updates packages and compiler wrappers</small>
nicolemarsaglia pushed to Alpine-DAV/ascent
add docs. add fixed point unit tests
cyrush open issue Alpine-DAV/ascent#1278.
all reuse of past actions
if user knows actions won’t change, provide a path to avoid checking for ascent_actions.{yaml,json}, and use last set of actions….View Comment
biagas commented on issue visit-dav/visit#19465.
@vijaysm that might be helpful, but perhaps other developers more familiar with the MOAB plugin could say for certain….
vijaysm commented on issue visit-dav/visit#19465.
> Moab requires lapack, which wasn’t installed in the container from which I built the distro….
jacksono1762 commented on issue visit-dav/visit#19210.
Fresh install of 3.4.1 on Ubuntu 22 has fixed this issue…
BenWibking commented on issue visit-dav/visit#19413.
Just after I posted that, I did some more digging and I think it’s a CMake environment variable issue. Testing a possible fix now….
iulian787 commented on issue visit-dav/visit#19465.
Thank you Kathleen, …
ulrikeyang pushed to hypre-space/hypre
fixed GMRES convergence issue
garlick commented on issue flux-framework/flux-sched#1186.
I’ll go ahead and set MWP then, thanks….
S-o-T commented on issue LLNL/zfp#230.
> We early on anticipated the potential to vary compression parameters spatially, perhaps even from one block to the next, and in that case the overhead of storing compression parameters becomes large. …
lindstro commented on issue LLNL/zfp#230.
As mentioned above, one rationale is that the consumer may not want to organize the data the same way the producer does. In fact, I cannot think of a case where the consumer, which processes the data, does not know how it wants the data to be organized. Can you think of a scenario where it would be beneficial to have the producer dictate the data layout for the consumer? In the case of a code processing 2D vector fields, the consumer needs to know if the data layout is float field[ny][nx][2] or float field[2][ny][nx] (or some other permutation) so it can index the multidimensional field properly. If the code is written with one of these conventions, it will fail if the data producer mandates the other convention. While one can write such a code using strides (e.g. field[stride_x * x + stride_y * y + stride_c * c] to access vector component c at (x, y)), oftentimes you want to use some container class like a NumPy array whose strides are given by the container, not the data producer.
…
ilumsden opened a pull request to LLNL/hatchet
adayton1 commented on issue LLNL/axom#1307.
Looks like a lot of the differences disappear with a really small tolerance (EPS = 1e-50)….
kab163 pushed to LLNL/Umpire
rearranging types
adrienbernede commented on issue LLNL/Umpire#875.
Ready for review….
johnbowen42 pushed to LLNL/RAJA
Update to BLT v0.6.2.
daboehme created a new tag, v2.11.0 at LLNL/Caliper
zslttaxx open issue XBraid/xbraid#96.
How to call the Xbraid library in my serial code？
What are the steps to call the xbraid library? What commands do I need to input on the terminal, or what things do I need to add to my serial code to successfully call? What information should I check to obtain a solution method…View Comment
tbennun opened a pull request to LLNL/lbann
jvwilliams23 commented on issue LLNL/lbann#2433.
Quick question, how much extra overhead do you expect the environment variable LBANN_NO_INPLACE=1 to create? I am seeing a significantly lower performance compared to pytorch (approx 3.5 days compared to 1 day for 1M iterations through my GAN)….
Merge branch ‘develop’ into artv3/sycl-thread-grid-layout-fix
adrienbernede pushed to LLNL/RAJA
fixup fraction static_assert
MrBurmark open issue LLNL/RAJA#1631.
Scan Accumulation Type Consistency
Among our different back-end implementations the type used to accumulate the intermediate values of a scan differs. …View Comment
polymorphic object host->device copy testing; Working on getting SharedPtrManager copies working correctly with umpire/chai.
bmhan12 pushed to GEOS-DEV/LvArray
Bump gcc version for C++20 warnings
PhilipDeegan closed issue LLNL/SAMRAI#177.
error: an attribute list cannot appear here
when trying to compile with umpire and raja …View Comment
CusiniM pushed to GEOS-DEV/LvArray
update tpls path on LC to 2024-04-19 (#320)
stgeke commented on issue LLNL/conduit#1275.
In this case we can close it. I was just using what Ascent develop is using by default. …
cyrush commented on issue LLNL/conduit#1275.
@stgeke …
dtaller open issue LLNL/CARE#258.
Add mask to care::IntersectArrays
Add an input argument for care::IntersectArrays that allows the algorithm to skip array elements in the intersection…View Comment
adrienbernede commented on issue LLNL/CARE#255.
@adayton1 this is ready for review !…
fiedorowicz1 pushed to LLNL/lbann
Fix to PeriodicPadding to match pytorch circular padding (#2435)
-
Fix to PeriodicPadding to match pytorch circular padding
-
Minor changes to docstring and kwargs
As batch size is implicit, remove from docstring. Remove unused “name” kwarg in function.
-
Fix test by comparing against pytorch F.pad with mode=”circular”
-
Skip test if pytorch isn’t available
Co-authored-by: Pier Fiedorowicz fiedorowicz1@llnl.gov</small>
robinson96 opened a pull request to LLNL/CARE
adrienbernede pushed to LLNL/CHAI
Apply changes required by LC (token handling)
bgunnar5 created a new tag, 1.12.1 at LLNL/merlin
adrienbernede created a new branch, woptim/shared-alloc-mpi-tests at LLNL/SAMRAI
ammarwa opened a pull request to LLNL/Caliper
rhornung67 reviewed a LLNL/CHAI pull request
Thanks @adayton1 …
slabasan commented on issue LLNL/hatchet#123.
Try running with CALI_LOG_VERBOSITY=1 CALI_CONFIG=hatchet-sample-profile ./app and see if you are getting any errors. If you do hatchet-region-profile and run cali-query on it, is the behavior the same?…
PhilipDeegan commented on issue LLNL/hatchet#123.
Hi @slabasan …
cielling commented on issue LLNL/Caliper#536.
Hi @daboehme, I wanted to check in to see when you/the Caliper team will have time to look into the append option? Thanks!…
ltaylor16 merged a pull request to LLNL/shroud
Merge pull request #589 from jameshcorbett/flux-jobscripts
Add example Flux jobscripts</small>
jameshcorbett commented on issue LLNL/scr#589.
Ok it took a lot of trial and error (some nodes just wouldn’t die on command) but I finally got a solid test of this and it seems to work! Ready for review now @mcfadden8 …
