Trac October 2018

trac@lists.einsteintoolkit.org

1 participants
40 discussions

[Einstein Toolkit] #2205: hwloc in OpenMPI self-built code uses CUDA and OpenCL
by Einstein Toolkit 08 Oct '18

08 Oct '18

#2205: hwloc in OpenMPI self-built code uses CUDA and OpenCL ---------------------------------+----------------------------------- Reporter: Roland Haas | Type: defect Status: new | Priority: major Milestone: | Component: EinsteinToolkit thorn Version: development version | Keywords: MPI ---------------------------------+----------------------------------- This causes problems since we do not then link against CUDA or OpenCL. This happens even if one passes the {{{--with-cuda=no}}} option to OpenMPI's configure script. Possibly one has to follow the fix in OpenMPI's 2.x and 3.x branches to fix this: https://github.com/open-mpi/ompi/issues/4248 -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2205> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 0

[Einstein Toolkit] #2165: Cactus: output actual path to config-info on failure
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2165: Cactus: output actual path to config-info on failure ---------------------------------+------------------------- Reporter: Roland Haas | Type: enhancement Status: new | Priority: minor Milestone: | Component: Cactus Version: development version | Keywords: ---------------------------------+------------------------- Pull request is here: https://bitbucket.org/cactuscode/cactus/pull-requests/51/cactus-output- actual-path-to-config-info/diff -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2165> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 5

[Einstein Toolkit] #2151: thorn NSTracker cannot track unequal mass neutron star systems
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2151: thorn NSTracker cannot track unequal mass neutron star systems ---------------------------------+------------------------- Reporter: Roland Haas | Type: enhancement Status: new | Priority: optional Milestone: | Component: Other Version: development version | Keywords: NSTracker ---------------------------------+------------------------- NSTracker makes the hard-coded assumption that if there are two NS then one is at (x,y,z) and the other one at (-x,-y,z). It (and the code in Hydro_Analysis) should support an arbitrary number of NS with arbitrary symmetries imposed. Note that NSTracker is not part of the ET but used by one of the gallery examples. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2151> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 1

[Einstein Toolkit] #2126: SimFactory should choose an appropriate cfg on a Mac.
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2126: SimFactory should choose an appropriate cfg on a Mac. ------------------------+--------------------------------------------------- Reporter: sbrandt | Owner: Type: defect | Status: new Priority: unset | Milestone: ET_2018_08 Component: SimFactory | Version: development version Keywords: | ------------------------+--------------------------------------------------- generic.cfg does not work on a mac because gcc doesn't exist (one needs to use gcc-7 instead, for example). Simfactory should detect and use the correct cfg file from the set generic.cfg, osx-homebrew.cfg, and osx- macports.cfg. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2126> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 7

[Einstein Toolkit] #2199: Cleanup Distros
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2199: Cleanup Distros ---------------------------------+------------------------ Reporter: Steven R. Brandt | Type: defect Status: new | Priority: unset Milestone: | Component: SimFactory Version: development version | Keywords: ---------------------------------+------------------------ Get rid of obsolete/problematic configs: centos.cfg, debian.cfg, etc. Branch: cleanupdistros -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2199> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 1

[Einstein Toolkit] #2193: centos.cfg disables OpenMP
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2193: centos.cfg disables OpenMP -------------------------+--------------------------------- Reporter: Roland Haas | Owner: Steven R. Brandt Type: defect | Status: assigned Priority: minor | Milestone: Component: SimFactory | Version: development version Keywords: | -------------------------+--------------------------------- Git commit 2e827ac "Set MPI_DIR so that centos doesn't have to build MPI. Update the recommended yum packages so that libtldl is present." commented out all OpenMP settings in centos.cfg which given that the configure script says > # determine OPENMP_MODE from the OPENMP option > # if this option isn't set OPENMP_MODE will default to 'no' this would mean that the code is not compiled with OpenMP making it impossible to test OpenMP parallelisation using this option list. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2193> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 12

[Einstein Toolkit] #2156: responsibilities of each maintainer page is outdated
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2156: responsibilities of each maintainer page is outdated ---------------------------------+------------------------------------- Reporter: Roland Haas | Type: defect Status: new | Priority: minor Milestone: | Component: EinsteinToolkit website Version: development version | Keywords: ---------------------------------+------------------------------------- The wiki page https://docs.einsteintoolkit.org/et-docs/Organization_and_Responsibilities is outdated since it eg lists Frank but not Steve. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2156> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 1

[Einstein Toolkit] #2186: PUGH: support zero sized grid arrays
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2186: PUGH: support zero sized grid arrays ---------------------------------+----------------------------------- Reporter: Roland Haas | Type: defect Status: new | Priority: minor Milestone: | Component: EinsteinToolkit thorn Version: development version | Keywords: PUGH ---------------------------------+----------------------------------- if a grid array has size 0 we ignore it during communication by setting its ghost size and overlap region to zero, cctk_lsh will show up a zero as well This fixes the bug that Erik observed in https://bitbucket.org/cactuscode/cactustest/commits/8a6601f about 14 years ago. As far as I can tell this test has been failing since then and only was not caught b/c the testing system actually does not look at the Cactus return code. Pull request is here: https://bitbucket.org/cactuscode/cactuspugh/pull- requests/2/pugh-support-zero-sized-grid-arrays/diff -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2186> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 4

[Einstein Toolkit] #2171: PAPI Compile Issues with Intel 19 beta on Ubuntu 18.04
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2171: PAPI Compile Issues with Intel 19 beta on Ubuntu 18.04 ---------------------------------+-------------------- Reporter: zachetie@… | Type: defect Status: new | Priority: unset Milestone: | Component: Other Version: development version | Keywords: ---------------------------------+-------------------- I get this error message when attempting to compile the latest ETK master (Jul 27, 2018) with the default ThornList using Intel 19.0.0.070 20180524 on Ubuntu 18.04: PAPI: Building... In file included from pfmlib_common.c(32): /usr/include/stdlib.h(140): error: identifier "_Float32" is undefined extern _Float32 strtof32 (const char *__restrict __nptr, ^ In file included from pfmlib_common.c(32): /usr/include/stdlib.h(146): error: identifier "_Float64" is undefined extern _Float64 strtof64 (const char *__restrict __nptr, ^ In file included from pfmlib_common.c(32): /usr/include/stdlib.h(158): error: identifier "_Float32x" is undefined extern _Float32x strtof32x (const char *__restrict __nptr, ^ In file included from pfmlib_common.c(32): /usr/include/stdlib.h(164): error: identifier "_Float64x" is undefined extern _Float64x strtof64x (const char *__restrict __nptr, ^ In file included from pfmlib_common.c(32): /usr/include/stdlib.h(233): error: identifier "_Float32" is undefined _Float32 __f) ^ In file included from pfmlib_common.c(32): /usr/include/stdlib.h(239): error: identifier "_Float64" is undefined _Float64 __f) ^ In file included from pfmlib_common.c(32): /usr/include/stdlib.h(251): error: identifier "_Float32x" is undefined _Float32x __f) ^ In file included from pfmlib_common.c(32): /usr/include/stdlib.h(257): error: identifier "_Float64x" is undefined _Float64x __f) ^ In file included from pfmlib_common.c(32): /usr/include/stdlib.h(316): error: identifier "_Float32" is undefined extern _Float32 strtof32_l (const char *__restrict __nptr, ^ In file included from pfmlib_common.c(32): /usr/include/stdlib.h(323): error: identifier "_Float64" is undefined extern _Float64 strtof64_l (const char *__restrict __nptr, ^ In file included from pfmlib_common.c(32): /usr/include/stdlib.h(337): error: identifier "_Float32x" is undefined extern _Float32x strtof32x_l (const char *__restrict __nptr, ^ In file included from pfmlib_common.c(32): /usr/include/stdlib.h(344): error: identifier "_Float64x" is undefined extern _Float64x strtof64x_l (const char *__restrict __nptr, ^ compilation aborted for pfmlib_common.c (code 2) -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2171> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 2

[Einstein Toolkit] #2185: PAPI's compilation produces warnings
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2185: PAPI's compilation produces warnings ---------------------------------+----------------------------------- Reporter: Roland Haas | Type: defect Status: new | Priority: minor Milestone: | Component: EinsteinToolkit thorn Version: development version | Keywords: PAPI ---------------------------------+----------------------------------- PAPI's build procedure currently contains two failing parts, one that shows up as error messages along the lines of "libthorn_PAPI.a_tests cannot be build" which is caused by Cactus setting a NAME variable in MAKEFLAGS which overrides what PAPI's Makefile wants to use and once that is finished, the tests actually fail to run since eg OpenMP libraries are missing. The attached patch gets rid of NAME and builds only the parts we need. The $(pwd)/libpfm4/lib/libpfm.a path is copied from the Makefile (Rules.pfm4_pe I believe). -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2185> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

[Einstein Toolkit] #2190: make sim-ThornGuide fails
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2190: make sim-ThornGuide fails ---------------------------------+-------------------- Reporter: Roland Haas | Type: defect Status: new | Priority: minor Milestone: | Component: Cactus Version: development version | Keywords: ---------------------------------+-------------------- One cannot build the ThornGuide using the current ET master (90d12c0b): {{{ rhaas@ekohaes8:.../cactus/ET_vanilla$ make VERBOSE=yes sim-ThornGuide echo ________________________________________________________________________ ________________________________________________________________________ Creating ThornGuide for configuration sim cd /data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim; \ mkdir -p doc rm -rf /data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/doc/build mkdir /data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/doc/build if test -r /data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/ThornList ; then \ cd /data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/doc/build; \ make -f /data/rhaas/postdoc/gr/cactus/ET_vanilla/doc/ThornGuide/Makefile THORNLIST=/data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/ThornList MASTER_FILE=ThornGuide-sim DOCBUILDDIR=/data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/doc/build; \ if test -e "/data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/doc/build /ThornGuide-sim.pdf"; then \ mv "/data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/doc/build /ThornGuide-sim.pdf" /data/rhaas/postdoc/gr/cactus/ET_vanilla/doc /ThornGuide-sim.pdf; \ echo " ThornGuide-sim.pdf created in doc directory."; \ echo " Done."; \ fi \ else \ echo " Error: /data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/ThornList not found."; \ fi make[1]: Entering directory '/data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/doc/build' THORN_DEPENDENCIES=`/usr/bin/perl -s /data/rhaas/postdoc/gr/cactus/ET_vanilla/lib/sbin/ThornList.pl -arrangements_dir=/data/rhaas/postdoc/gr/cactus/ET_vanilla/arrangements -thornlist=/data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/ThornList` Gathering documentation from thorn ccl files Processing.... param.ccl.... /usr/bin/perl -s /data/rhaas/postdoc/gr/cactus/ET_vanilla/lib/sbin/ParamLatex.pl -cctk_home=/data/rhaas/postdoc/gr/cactus/ET_vanilla -directory=/data/rhaas/postdoc/gr/cactus/ET_vanilla/arrangements -thornlist=/data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/ThornList -outdir=./ -section -grouping=bythorn -sort=scope > LOG_PARAMLATEX_MSGS 2>&1 make[1]: *** [/data/rhaas/postdoc/gr/cactus/ET_vanilla/doc/ThornGuide/Makefile:91: ThornGuide-sim.tex] Error 255 make[1]: Leaving directory '/data/rhaas/postdoc/gr/cactus/ET_vanilla/configs/sim/doc/build' }}} and the LOG file contains {{{ Undefined subroutine &main::parse_ccl called at /data/rhaas/postdoc/gr/cactus/ET_vanilla/lib/sbin/parameter_parser.pl line 46. CTThorns }}} The line was introduced in git hash 618ef008 "Speedup branch implements caching of CCL files." about 1.25 years ago. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2190> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 4

[Einstein Toolkit] #2191: parameter file download seems wrong in Gallery: Binary black hole GW150914
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2191: parameter file download seems wrong in Gallery: Binary black hole GW150914 ---------------------------------+-------------------- Reporter: anonymous | Type: defect Status: new | Priority: unset Milestone: | Component: Other Version: development version | Keywords: ---------------------------------+-------------------- I use curl -Lf http://einsteintoolkit.org/gallery/bbh/GW150914.rpar -o par/GW150914.rpar to down the parameter file and run the simulation, it give me error like: Error: Error while executing parameter file script /home/someone/simulations/GW150914_1/output-0000/GW150914.rpar -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2191> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

[Einstein Toolkit] #2143: trac statistics link is broken
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2143: trac statistics link is broken -------------------------+--------------------------------- Reporter: Roland Haas | Owner: Steven R. Brandt Type: defect | Status: assigned Priority: minor | Milestone: Component: Other | Version: development version Keywords: trac | -------------------------+--------------------------------- The link "Ticket statistics" ie https://trac.einsteintoolkit.org/stats/tickets on the front page of trac leads to an error page {{{ No handler matched request to /stats/tickets }}} -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2143> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 1

[Einstein Toolkit] #2195: Multipole Thorn: Why is max_vars set to 10?
by Einstein Toolkit 03 Oct '18

03 Oct '18

#2195: Multipole Thorn: Why is max_vars set to 10? ---------------------------------+-------------------- Reporter: zachetie@… | Type: defect Status: new | Priority: unset Milestone: | Component: Other Version: development version | Keywords: ---------------------------------+-------------------- arrangements/EinsteinAnalysis/Multipole/src/multipole.cc , line 22 reads: static const int max_vars = 10; Occasionally I'd like to output a multipole decomposition, e.g., for 12 variables. When I attempt this, because of the above limit, I get the mysterious error message that appears due to the assert on line 42 of this same file. First, I think this should be a freely specifiable parameter. Second, the error message should be clearer about what is wrong. Third, by just setting the above to 30, I was able to perform a run with 12 output variables. Is this the correct fix? (The run no longer errors out.) -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2195> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 1

[Einstein Toolkit] #2198: simfactory's user level run command does not record jobid in a segment's properties.ini file
by Einstein Toolkit 01 Oct '18

01 Oct '18

#2198: simfactory's user level run command does not record jobid in a segment's properties.ini file ---------------------------------+------------------------ Reporter: Roland Haas | Type: defect Status: new | Priority: minor Milestone: | Component: SimFactory Version: development version | Keywords: ---------------------------------+------------------------ The (user level) run command does not record a jobid in a segment's properties.ini file which means one cannot eg restart from a checkpoint since simfactory complains about an unset jobid: {code} simfactory/bin/sim create-run test1 --parfile par/tov_ETK_2018_lisbon.par --procs 1 --walltime 0:5:0 grep jobid ~/simulations/test1/output-0000/SIMFACTORY/properties.ini {code} shows the issue (but not producing any output). And the error obtained when trying to restart is along the lines of > Simulation name: binarybh9 > Error: job id is negative > Aborting Simfactory This was reported by Qingwen Wang in http://lists.einsteintoolkit.org/pipermail/users/2018-September/006524.html A workaround is to not use run as a user but always submit or to use the cleanup command after which simfactory is fine with there not being a job id. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/2198> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 2

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Trac October 2018