Trac June 2014

trac@lists.einsteintoolkit.org

1 participants
124 discussions

[Einstein Toolkit] #1506: compiling Cactus on Windows
by Einstein Toolkit 17 May '15

17 May '15

#1506: compiling Cactus on Windows --------------------------------+------------------------------------------- Reporter: jtao | Owner: Type: defect | Status: new Priority: minor | Milestone: Component: Cactus | Version: development version Keywords: compiling, windows | --------------------------------+------------------------------------------- Hi, In the process to create the cactusBSSN benchmark for SPEC, I worked with Mat from PGI(now NVIDIA) to help resolve some compilation issues on Windows. I don't have patches for each one of them, but attemptive solutions could be found in the email below. These are not critical issues but it would be good to be resolved to help future windows users if there is any. Regards, Jian Mathew COLGROVE wrote: > Hi Jian, > > I've been working on porting cactusBSSN to Windows. After working through some header file configuration, I encountered several undefined references. I've been able to work around some of them, but a few others I need some help with. > > 1) "STDOUT_FILENO" and "STDERR_FILENO" in Cactus/main/Warnlevel.c > These symbols are found in "unistd.h" which is not available on Windows. I just commented these lines out since I don't think this section is executed. Though, the proper fix would be to define then in > the cctk_Config.h file if not "HAVE_UNISTD_H" is not. The function isatty is defined in unistd.h. A conditional macro will be necessary for that function as well in this case. #if HAVE_UNISTD_H if (!isatty (STDOUT_FILENO)) val = ""; #endif > 2) "hypot" and "copysign" in Cactus/main/Complex.c > > Windows does support "hypot" and "copysign", however they spell them "_hypot" and "_copysign". To work around, I created two versions of each macro that use these symbols. I will probably do the same. > 3) "regexec", "regfree", and "regcomp" in Cactus/main/Parameters.c and Cactus/utils/misc.c > > Windows does not support these functions. We may be able to integrate a port from glibc into the benchmark but I would rather not. Since previous versions of Cactus did not used these symbols, can we use an older version other these files? As a matter of fact, Cactus includes files from the src of the GNU C library. You can find them under Your_Cactus/src/gnu. It is likely that you pre-configured Cactus on a linux box or other systems which provide the library. Cactus will then use the system default instead. To compiler regex.c getopt.c distributed with Cactus, you need to go to Your_Cactus/configs/spec/config-data/ and set both BUILD_GETOPT BUILD_REGEX to yes. # GNU stuff BUILD_GETOPT = yes BUILD_REGEX = yes This will solve #4 below as well. > 4) "getopt_long_only", "optarg", and "optind" in Cactus/main/ProcessCommandLine.c > > Again, there is no Windows equivalent for these symbols. We may be able to integrate something from http://gnuwin32.sourceforge.net/packages/libgw32c.htm, but I would rather not. Since previous versions of Cactus did not used these symbols, can we use an older version other this file? > > 5) "gethostname" and "gethostbyname" in Cactus/utils/Network.c > > Windows does have these functions but we'd need to link with a DLL which I'd rather not have to do (the user should have the option on how to link). Though, I think we'd be safe to remove these functions. Yes, the prototypes of gethostname and gethostbyname can be found in winsock2.h, and Cactus knows how to deal with them on Windows. --------------------------------------- Network.c : ... #elif defined HAVE_WINSOCK2_H #include <winsock2.h> #endif /* HAVE_WINSOCK2_H */ ... --------------------------------------- If linking DLL will be an issue, I would suggest you keep Util_GetHostName but make it a dummy function since Util_GetHostName serves as the interface to gethostname for Cactus. Regards, Jian -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/1506> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 6

[Einstein Toolkit] #1211: CarpetIOScalar should write file info for restart files
by Einstein Toolkit 22 Apr '15

22 Apr '15

#1211: CarpetIOScalar should write file info for restart files --------------------+------------------------------------------------------- Reporter: knarf | Owner: eschnett Type: defect | Status: new Priority: minor | Milestone: Component: Carpet | Version: Keywords: | --------------------+------------------------------------------------------- Currently Carpet doesn't write file info for files created by a restarted simulation (from a checkpoint, writing into a new file). It should instead write the header if it created the file - which means it need to check whether it appends or created the file new. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/1211> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

[Einstein Toolkit] #1282: Enable HDF5 compression by default in Carpet
by Einstein Toolkit 22 Apr '15

22 Apr '15

#1282: Enable HDF5 compression by default in Carpet -------------------------+-------------------------------------------------- Reporter: hinder | Owner: eschnett Type: enhancement | Status: new Priority: major | Milestone: Component: Carpet | Version: Keywords: | -------------------------+-------------------------------------------------- The current default for CarpetIOHDF5::compression_level is 0 (no compression). I have been using compression in most of my HDF5 files for years, and have never run into any problem. CPUs are typically much faster than storage nowadays. I propose that the compression level should default to 9. This would affect output and checkpoint files, and could lead to huge space savings. Apart from the checkpoint files being written and read quicker and taking less disk space, the user should not notice, as the HDF5 library handles compression transparently. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/1282> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 8

[Einstein Toolkit] #1617: Calls to Accelerator_NotifyDataModified should be paired with calls to Accelerator_RequireInvalidData
by Einstein Toolkit 25 Mar '15

25 Mar '15

#1617: Calls to Accelerator_NotifyDataModified should be paired with calls to Accelerator_RequireInvalidData -----------------------------------+---------------------------------------- Reporter: sbrandt | Owner: Type: defect | Status: new Priority: minor | Milestone: Cactus_4.3.0 Component: EinsteinToolkit thorn | Version: development version Keywords: | -----------------------------------+---------------------------------------- Calls to Accelerator_NotifyDataModified should be paired with calls to Accelerator_RequireInvalidData, otherwise there's a risk of seg fault in Accelerator_NotifyDataModified due to a missing accelerator data structure. MoL fails to do this in Operators.c. Patch attached. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/1617> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

[Einstein Toolkit] #1612: reace condition when building Cactus utilities
by Einstein Toolkit 24 Mar '15

24 Mar '15

#1612: reace condition when building Cactus utilities --------------------+------------------------------------------------------- Reporter: rhaas | Owner: Type: defect | Status: new Priority: minor | Milestone: Component: Cactus | Version: development version Keywords: | --------------------+------------------------------------------------------- I just compiled on bluewaters and received output {{{ Done creating cactus_simO3. All done ! Building utilities for simO3 Building utilities for simO3 ... Creating hdf5_recombiner in /mnt/a/u/sciteam/rhaas/ET_trunk/exe/simO3 from /mnt/a/u/sciteam/rhaas/ET_trunk/configs/simO3/build/CarpetIOHDF5/hdf5_recombiner.o mkdir: cannot create directory `/tmp/1399318629': File exists mkdir: cannot create directory `/tmp/1399318629': File exists }}} with the full (last part of) the log in the attached file. So it seems as if we have a race condition in the make system that causes it to try and build the utilities twice (in parallel). On bluewaters simfactory (which I used) builds Cactus using 16 processes. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/1612> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 7

[Einstein Toolkit] #1613: provide CCTK_VParamWarm
by Einstein Toolkit 22 Mar '15

22 Mar '15

#1613: provide CCTK_VParamWarm -------------------------+-------------------------------------------------- Reporter: rhaas | Owner: Type: enhancement | Status: new Priority: optional | Milestone: Component: Cactus | Version: development version Keywords: | -------------------------+-------------------------------------------------- it would be nice if there was a CCTK_VParamWarn along with CCTK_PARAMWARN, analogous to CCTK_VWarn and CCTK_WARN. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/1613> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 1

[Einstein Toolkit] #1326: running loopcontrol on strange number of threads fails
by Einstein Toolkit 22 Mar '15

22 Mar '15

#1326: running loopcontrol on strange number of threads fails -----------------------------------+---------------------------------------- Reporter: rhaas | Owner: Type: defect | Status: new Priority: minor | Milestone: Component: EinsteinToolkit thorn | Version: Keywords: LoopControl | -----------------------------------+---------------------------------------- my machine has 8 cores (according to /proc/cpuinfo). Running eg the trigger test with 3 threads fails inside of loopcontrol. To reproduce: {{{ export OMP_NUM_THREADS=3 mpirun -n 2 exe/cactus_bns_all arrangements/AEIThorns/Trigger/test/trigger.par }}} -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/1326> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 1

[Einstein Toolkit] #1361: disable hyperthreading in loopcontrol by default
by Einstein Toolkit 22 Mar '15

22 Mar '15

#1361: disable hyperthreading in loopcontrol by default -----------------------------------+---------------------------------------- Reporter: rhaas | Owner: Type: defect | Status: new Priority: major | Milestone: Component: EinsteinToolkit thorn | Version: Keywords: LoopControl | -----------------------------------+---------------------------------------- when running with both openmp and sufficiently many threads that hyperthreading threads are used, many tests using LoopControl (Cartoon, RotatingSymmety180, RotatingSymmetry90) fail. This can be tracked down to disabling hyperthreading support in LoopControl (ie. turning of hyperthreading makes things work). In particular on bethe with smt and 8 physical cores: The Cartoon/test_cartoon_2.par test shows differences from the recorded results when run with 16 threads (but not with 8 threads). If I then go ahead and disable OMP in all ML source files but ML_BSSN_enforce *and* comment out the #include "loopcontrol.h", then the difference goes away. Adding back #include "loopcontrol.h" brings back the error. Some further experimenting with LoopControl's options shows that indeed the use_smt_threads option is what causes problems. If I turn it off things work fine even with a vanilla source tree. Otherwise relative differences are on the order 1e-7 and absolute 1e-11 (in momx_z_[2][2].xg). Without smt the results are identical to the stored values. The issue only occurs in combination of OpenMP, vectorization and hyperthreading. The issue is independent of the compiler (both intel 13 and gcc 4.4 show the same behaviour), and vectorization (sse2) and many threads (up to 4 times the number of physical cores) works fine on non-smt machines. I propose to disable LoopControl::use_smt_threads by default. Note that we cannot completely remove it since apparently for Vesta (a Blue Gene/Q) smt is required to get and multi-threading at all. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/1361> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 4

[Einstein Toolkit] #1204: carpet bug
by Einstein Toolkit 22 Mar '15

22 Mar '15

#1204: carpet bug -------------------------------------+-------------------------------------- Reporter: abdik@… | Owner: eschnett Type: defect | Status: new Priority: major | Milestone: Component: Carpet | Version: Keywords: | -------------------------------------+-------------------------------------- The latest version of carpet seems to contain a bug that affects AMR+multipatch runs. My stderr and stdout and par file are attached. Found by Roland and Ernazar. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/1204> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

[Einstein Toolkit] #997: problem in appending output after recovery
by Einstein Toolkit 22 Mar '15

22 Mar '15

#997: problem in appending output after recovery ----------------------------------------+----------------------------------- Reporter: corvino.giovanni@… | Owner: eschnett Type: defect | Status: new Priority: major | Milestone: Component: Carpet | Version: Keywords: | ----------------------------------------+----------------------------------- I have a problem in appending output files from Carpet. I used to produce 3d HDF5 output of grid variables and write the output in the same directory also after recovery from checkpoint. The new output was automatically appended to the existing one. Now I am producing h5 output also on 2D slices but in this case the output is overwritten so I lost the data for all but the last recovery. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/997> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

← Newer
1
...
5
6
7
8
9
10
11
12
13
Older →

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Trac June 2014