Re: [petsc-dev] regression in master
OK, this comes from '--with-viewfromoptions=0' On Sun, Jun 26, 2016 at 2:15 PM, Mark Adams <[email protected]> wrote:
This is puzzling but this seems to be a problem from Feb 28. How could this be broken for this long ...
14:10 (27b0f28...)|BISECTING ~/Codes/petsc$ git bisect bad 27b0f280e5bf9c639a7764debf82fd1b76d8a199 is the first bad commit commit 27b0f280e5bf9c639a7764debf82fd1b76d8a199 Author: Barry Smith <[email protected]> Date: Sun Feb 28 23:09:06 2016 -0600
Added configure option --with-viewfromoptions=0 which will speed up code with many small solves substantially
Reported-by: Amneet Pal S Bhalla <[email protected]>
:040000 040000 46be6ed2948b221a6ea65b9639782ba839557306 010eacea0d6a858466b21641de11a1ee3089763b M config :040000 040000 52698dd9a2c4e4a4ac740457407730108ed9a0b7 202bf4b8bba01b04e74da4db3f5abf6e23ef4662 M src
On Sun, Jun 26, 2016 at 1:12 PM, Mark Adams <[email protected]> wrote:
I am finding ksp_[view|monitor|converged_reason] are not working in master. I am trying bisection search now, looking for a good place.
12:49 master= ~/Codes/petsc/src/ksp/ksp/examples/tutorials$ make runex1 1,6d0 < 0 KSP Residual norm 0.707107 < 1 KSP Residual norm 0.316228 < 2 KSP Residual norm 0.188982 < 3 KSP Residual norm 0.129099 < 4 KSP Residual norm 0.0953463 < 5 KSP Residual norm < 1.e-11 24a19,21
WARNING! There are options you set that were not used! WARNING! could be spelling mistake, etc! Option left: name:-ksp_monitor_short (no value) /Users/markadams/Codes/petsc/src/ksp/ksp/examples/tutorials Possible problem with ex1_1, diffs above
participants (1)
-
Mark Adams