Luc Berger
Luc Berger
There is a job on blake going on for 6hrs in the PV queue, we should cancel that so we can get an interactive sessions and investigate a bit what's...
That last failure could have been a network issue, it's worth waiting for a second round of testing, even though my sense is this would be fine to merge as...
I am still not quite sure how/where to document this type of hierarchical BLAS thing, or whatever we want to call it. Also I believe we might need to run...
Starting to look into this. Might want to change the handle and the `no_reuse` version of the TPL implementation layer. I don't think these are callable from the spgemm implementation...
Yes, it was good timing you reminded me about this, I was able to "jump on it" effectively during the meeting. Since we are the biggest user of this container...
Yeah, some of the logic around how the roc TPLs are detected in Kokkos Kernels was changed recently to accommodate Trilinos but I suspect it was not tested with spack...
Thanks for reporting, we will have a look at this
Okay, thanks for digging a bit into this, I will run the code in valgrind / gdb as well and hopefully can reproduce and report my observation. The algorithm indeed...
The PR above, #2342, has a fix for the rectangular matrices and introduces more tests for the Serial QR feature. The tests are not fully implemented yet but the fix...
Okay, I will try to wrap up the PR and get that tested and merged, then I can move on to trsv, hopefully it's not more complicated than the QR...