Trac September 2011

trac@lists.einsteintoolkit.org

1 participants
184 discussions

[Einstein Toolkit] #566: Support compressed (tar.gz) testsuite data
by Einstein Toolkit 26 Sep '11

26 Sep '11

#566: Support compressed (tar.gz) testsuite data -------------------------+-------------------------------------------------- Reporter: knarf | Owner: Type: enhancement | Status: new Priority: minor | Milestone: ET_2011_11 Component: Cactus | Version: development version Keywords: | -------------------------+-------------------------------------------------- This patch to Cactus adds support for compressed testsuite data. It assumes that if no directory is found within the thorn containing the reference data for a particular test, that it might be in a correspondingly named file (test.par with a missing test/ as output_dir might actually have test.tar.gz). Cactus would, when running the testsuites, decompress the tarfile to TEST/config/thorn/test.orig, and compare that against the new results (in TEST/config/thorn/test, as usual). The test.orig directory will remain within TEST for easier debugging/comparing later, but will be overwritten for each new re-run of the test. Using ET_2011_05 as example, compressing all testsuites this way, and disregarding the svn metadata, this reduces the number of files within arrangements (following symlinks) from 16k to 4.8k, and the size from 220MB to 112MB. The metadata from svn is likely to about double these numbers. These numbers mean that using this would not only save space, but also a lot of time when syncing between machines. I propose to add support for this before the next ET release, but to only actually use it in thorns after this. Commands you may find useful when testing this: Compress all testsuite data (from within arrangements/): for i in `find -L -maxdepth 3 -mindepth 3 -type d -name test`; do cd $i; find -maxdepth 1 -mindepth 1 -not -name .svn -type d | awk '{print $1".tar.gz "$1;}' | xargs -n 2 -P 4 tar -czf; cd -; done After that: remove the test data directories (because they have higher priority than the tar.gz files): find -L -maxdepth 4 -mindepth 4 -type d | grep -v .svn | grep -e '\/test\/' | xargs rm -rf -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/566> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 9

[Einstein Toolkit] #577: Formaline: Add thorn tarballs to thorns
by Einstein Toolkit 26 Sep '11

26 Sep '11

#577: Formaline: Add thorn tarballs to thorns -----------------------------------+---------------------------------------- Reporter: eschnett | Owner: Type: enhancement | Status: new Priority: major | Milestone: Component: EinsteinToolkit thorn | Version: Keywords: | -----------------------------------+---------------------------------------- Add the thorns' tarballs to the individual thorns' libraries instead of building a library with all tarballs. This should reduce build time, since it avoids regenerating a potentially large library for every source code change. Instead, only those thorns which changed would have their library rebuilt. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/577> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 0

[Einstein Toolkit] #559: Cactus may use wrong test config file
by Einstein Toolkit 26 Sep '11

26 Sep '11

#559: Cactus may use wrong test config file --------------------+------------------------------------------------------- Reporter: tbode | Owner: Type: defect | Status: new Priority: minor | Milestone: ET_2011_11 Component: Cactus | Version: Keywords: | --------------------+------------------------------------------------------- If for some reason there exists both a test.ccl and a file 'config' in a thorn's test directory, Cactus uses the config file while telling the user that it's using the test.ccl. Config is so outdated this shouldn't affect anybody. Quick fix. I'm attaching a patch. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/559> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 4

[Einstein Toolkit] #444: Code documentation for Carpet
by Einstein Toolkit 25 Sep '11

25 Sep '11

#444: Code documentation for Carpet -------------------------+-------------------------------------------------- Reporter: hinder | Owner: eschnett Type: enhancement | Status: new Priority: minor | Milestone: Component: Carpet | Version: Keywords: | -------------------------+-------------------------------------------------- Doxygen is a program which generates HTML documentation describing all the files, functions and classes of a program based on the C++ source. It uses comments in a particular format in the code. There is a search box and a tree view for the classes, as well as links to HTML versions of the source files. An example of the output is at http://damiana2.aei.mpg.de/~ianhin/carpet-doc/ To generate the documentation, go into the root Carpet directory and type "doxygen", assuming that doxygen is installed on the local system. The documentation will be generated in a new carpet-doc directory. Open the file carpet-doc/html/index.html to browse the documentation. It would be convenient to have the documentation generated in this way made available on the Carpet web site and updated after each commit to the repository. The attached patch provides a doxygen configuration file and a header file for the main documentation page. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/444> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

[Einstein Toolkit] #528: Use "andnot" instruction when vectorising
by Einstein Toolkit 25 Sep '11

25 Sep '11

#528: Use "andnot" instruction when vectorising -----------------------------------+---------------------------------------- Reporter: eschnett | Owner: Type: enhancement | Status: new Priority: minor | Milestone: Component: EinsteinToolkit thorn | Version: Keywords: | -----------------------------------+---------------------------------------- Use the "andnot" instruction to reduce the number of different bit masks that are required. Using fewer different bit masks may require fewer registers to hold them, or fewer load instructions to access them, thus potentially improving performance. Do not scalarize ifpos when SSE 4.1 is not available; instead, use logical operations to create a bit mask. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/528> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

[Einstein Toolkit] #451: Implement "dd" tensor type in symmetry conditions
by Einstein Toolkit 25 Sep '11

25 Sep '11

#451: Implement "dd" tensor type in symmetry conditions -------------------------+-------------------------------------------------- Reporter: eschnett | Owner: Type: enhancement | Status: new Priority: major | Milestone: Component: Cactus | Version: Keywords: | -------------------------+-------------------------------------------------- The attached patch implements the "dd" tensor type (a full 3x3 tensor without symmetries) in the rotating symmetries. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/451> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

[Einstein Toolkit] #540: Change case in internal auto-generated file
by Einstein Toolkit 25 Sep '11

25 Sep '11

#540: Change case in internal auto-generated file -------------------------+-------------------------------------------------- Reporter: eschnett | Owner: Type: enhancement | Status: new Priority: optional | Milestone: Component: Cactus | Version: Keywords: | -------------------------+-------------------------------------------------- Cactus auto-generates header files from CCL files into the bindings/include directory. One of these files is called ${thorn}_arguments.h, which stands out because "a" is lower case. Other auto-generated files are called e.g. ${thorn}_Schedule.h with an upper case "S". I suggest to change "arguments" to upper case "Arguments". -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/540> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 5

[Einstein Toolkit] #479: speed up VisIt's CarpetHDF5 plugin, set 2d masks
by Einstein Toolkit 25 Sep '11

25 Sep '11

#479: speed up VisIt's CarpetHDF5 plugin, set 2d masks -------------------+-------------------------------------------------------- Reporter: rhaas | Type: enhancement Status: new | Priority: minor Milestone: | Component: Other Version: | Keywords: CarpetHDF5 -------------------+-------------------------------------------------------- Hello all, the attached patches (when all of them are applied to VisIt's CarpetHDF5 plugin) speed up opening (large) HDF5 output files in VisIt by about a factor of 8. The two main speedups are replacing a linear search when translating from the Cactus iteration cctk_iteration to timestep (index) number and (surprisingly enough) the parsing of the Cactus variable name out of the dataset name rather than the "name" attribute (but it falls back to reading the attribute if the parsing fails). The third speed improvement is only visible when more than one file are opened and plotted. VisIt seems to consider reading metadata a cheap operation and creates and destroys the metadata object (avtCarpet...) very often. The patch caches metadata when VisIt destroys the object. This speeds up plotting several frames (or using the timestep slider) considerably, but has the unfortunate side effect that one cannot fully close files anymore (re-loading still works though). I attach timing information to show the gains. Data files and a python script to use with visit -cli -s openfile.py are provided at http://www.numrel.org/~rhaas3/CarpetHDF5/ . The data files are about 20MB when compressed and about 2-4GB when decompressed (they are actual data from one of my simulations with all values set to zero). There are three data sets provided. "flat" is the direct output from the Cactus simulation. "grouped" contains datasets from each timestep in a group of their own (grouping is mostly useful when mergin hdf5 files which is unbearably slow otherwise). "unchunked" used a python script to merge the individual Carpet components into the CarpetRegrid2 boxes (or equivalent), which results in much faster load times and smaller files (but is itself a slow operation). Code and scripts to group and unchunk hdf5 data is not yet public. If there is interested and I can post those as well but they are not nicely coded at all. The last patch (actually the first two by number) changes the way CarpetHDF5 sets up the nesting of components for 2D data so that eg. contour plots work properly (no more duplicate lines from the coarse points under fine points). -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/479> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 2

[Einstein Toolkit] #459: Autodetect sched_getaffinity
by Einstein Toolkit 25 Sep '11

25 Sep '11

#459: Autodetect sched_getaffinity -------------------------+-------------------------------------------------- Reporter: eschnett | Owner: Type: enhancement | Status: new Priority: minor | Milestone: Component: Cactus | Version: Keywords: | -------------------------+-------------------------------------------------- The attached patch autodetects whether sched_getaffinity is available. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/459> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 4

[Einstein Toolkit] #551: Output metadata from TwoPunctures
by Einstein Toolkit 25 Sep '11

25 Sep '11

#551: Output metadata from TwoPunctures -----------------------------------+---------------------------------------- Reporter: hinder | Owner: Type: enhancement | Status: new Priority: minor | Milestone: Component: EinsteinToolkit thorn | Version: Keywords: TwoPunctures | -----------------------------------+---------------------------------------- It is desirable to be able to determine the properties of !TwoPunctures initial data from the simulation output. Currently, some quantities (initial momenta etc.) can be determined by parsing the parameter file and param.ccl, and others by parsing the standard output of the run (ADM mass). Neither of these are optimal. The attached patch causes !TwoPunctures to output a metadata file in the form described in [http://arxiv.org/abs/0709.0093 Data formats for numerical relativity waves]. The keys used include current draft proposals used in the NR-AR project. To encourage the use of this, I have made the output happen unconditionally. The output filename is !TwoPunctures.bbh and it is written into IO::out_dir. The metadata file looks like this: {{{ # ================================== # Numerical Relativity Metadata file # ================================== # # This file contains information about the simulation provided by the # TwoPunctures thorn. The format is described in the NR Data Format Document # http://arxiv.org/abs/0709.0093 [draft SVN r707]. [metadata] initial-ADM-energy = 1.190594340667362472 initial-ADM-angular-momentumx = 0.1100000000000000006 initial-ADM-angular-momentumy = 0.2200000000000000011 initial-ADM-angular-momentumz = 1.195856641772157003 initial-separation = 7 initial-data-type = Bowen-York initial-data-bibtex-keys = Bowen:1980yu Brandt:1997tf Ansorg:2004ds initial-bh-position1x = 3.5 initial-bh-position1y = 0 initial-bh-position1z = 0 initial-bh-position2x = -3.5 initial-bh-position2y = 0 initial-bh-position2z = 0 initial-bh-momentum1x = -0.003358311330506950042 initial-bh-momentum1y = 0.1236938059674509982 initial-bh-momentum1z = 0 initial-bh-momentum2x = 0.003358311330506950042 initial-bh-momentum2y = -0.1236938059674509982 initial-bh-momentum2z = 0 initial-bh-spin1x = 0.1000000000000000056 initial-bh-spin1y = 0.2000000000000000111 initial-bh-spin1z = 0.2999999999999999889 initial-bh-spin2x = 0.01000000000000000021 initial-bh-spin2y = 0.02000000000000000042 initial-bh-spin2z = 0.02999999999999999889 initial-bh-puncture-adm-mass1 = 0.7117197159189734279 initial-bh-puncture-adm-mass2 = 0.5107035566228820267 }}} -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/551> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 6

← Newer
1
...
12
13
14
15
16
17
18
19
Older →

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Trac September 2011