Trac December 2012

trac@lists.einsteintoolkit.org

1 participants
134 discussions

[Einstein Toolkit] #963: Improve McLachlan accuracy
by Einstein Toolkit 15 Oct '13

15 Oct '13

#963: Improve McLachlan accuracy -----------------------------------+---------------------------------------- Reporter: eschnett | Owner: Type: enhancement | Status: new Priority: major | Milestone: Component: EinsteinToolkit thorn | Version: Keywords: | -----------------------------------+---------------------------------------- James van Meter provided me with an optimised version of McLachlan. He states: 1. You are not taking full advantage of the chi=exp(-2phi) variable. There are several terms you divide by chi or chi^2 in expressions with overall factors exp(-4phi). I rewrote the BSSN equations to make these cancellations before coding. So where you have an expression of the form chi^2(A+B/chi^2), I have chi^2A+B. This gives a slight but noticeable advantage in both accuracy and performance. 2. I added Hamiltonian-constraint-damping terms due to Duez et al. These terms don't seem to be well-known but they are effective. 3. I added a Gamma-constraint-damping term due to Yo et al. 4. I enforce det(g)=1. I have not tested this new version yet, but James suggests to include it into the Einstein Toolkit. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/963> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 14

[Einstein Toolkit] #605: Simfactory does not find the right 'path' for symlinked cactus directories
by Einstein Toolkit 13 Aug '13

13 Aug '13

#605: Simfactory does not find the right 'path' for symlinked cactus directories ------------------------+--------------------------------------------------- Reporter: knarf | Owner: eschnett Type: defect | Status: new Priority: minor | Milestone: ET_2011_10 Component: SimFactory | Version: development version Keywords: | ------------------------+--------------------------------------------------- Simfactory currently does not detect the 'right' path to a Cactus sourcetree if this is from within a symlink, e.g. Cactus -> /mnt/data/Cactus. This is because while `pwd` returns '/home/user/Cactus', simfactory uses a wrapper to the C-library getcwd(), which dereferences symlinks. Simfactory then gets '/mnt/data/Cactus' and tries to use this path on remote machines when syncing - which of course does not work. The attached patch fixes this by using os.environ.get("PWD") and, if this is not defined, using os.getcwd() as workaround. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/605> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 27

[Einstein Toolkit] #117: Error while building multiple configurations
by Einstein Toolkit 12 Aug '13

12 Aug '13

#117: Error while building multiple configurations ------------------------+--------------------------------------------------- Reporter: eschnett | Owner: mthomas Type: defect | Status: new Priority: major | Milestone: Component: SimFactory | Version: Keywords: | ------------------------+--------------------------------------------------- I wanted to build multiple configurations at once with the command ./simfactory/sim build sim-debug sim sim-debug is a debug configuration, sim is optimised. SimFactory built sim-debug, then aborted with the error: optionlist is: /Users/eschnett/EinsteinToolkit-hg/configs/sim/OptionList Error: pattern '^\s*DEBUG\s*=.*$' exists already It assumes that SimFactory doesn't properly distinguish between the different build options of the different configurations. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/117> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 2

[Einstein Toolkit] #874: Reduce warnings during an ET build
by Einstein Toolkit 12 Aug '13

12 Aug '13

#874: Reduce warnings during an ET build --------------------+------------------------------------------------------- Reporter: hinder | Owner: Type: defect | Status: new Priority: minor | Milestone: Component: Other | Version: Keywords: | --------------------+------------------------------------------------------- A new user of the ET might be disconcerted to see warnings during compilation such as: {{{ /home/ianhin/Cactus/EinsteinToolkit/arrangements/Carpet/CarpetLib/src/bboxset.hh(206): warning #1224: #warning directive: "TODO: this is incorrect" #warning "TODO: this is incorrect" ^ }}} I propose that such directives are replaced with a comment and a link to a TRAC ticket concerning the problem. Further, we should aim to eliminate as many warnings as possible from the build, similarly to avoid eroding confidence. Warnings can also make it hard to find the real error messages, when a build fails with errors. Other warnings I see are similar to: {{{ In file included from /home/ianhin/Cactus/GBSC12/arrangements/CactusUtils/Formaline/src/thornlist.cc:13: /home/ianhin/Cactus/GBSC12/configs/bbh/bindings/include/thornlist.h:61: warning: deprecated conversion from string constant to ‘char*’ }}} {{{ /home/ianhin/Cactus/GBSC12/arrangements/EinsteinInitialData/TwoPunctures/src/Newton.c: In function ‘TestRelax’: /home/ianhin/Cactus/GBSC12/arrangements/EinsteinInitialData/TwoPunctures/src/Newton.c:287: warning: passing argument 6 of ‘resid’ from incompatible pointer type /home/ianhin/Cactus/GBSC12/arrangements/EinsteinInitialData/TwoPunctures/src/Newton.c:75: note: expected ‘const int * const restrict* const restrict’ but argument is of type ‘int **’ }}} {{{ /home/ianhin/Cactus/GBSC12/configs/bbh/bindings/Variables/ADMBase.c: In function ‘CCTKi_BindingsFortranWrapperADMBase’: /home/ianhin/Cactus/GBSC12/configs/bbh/bindings/Variables/ADMBase.c:34: warning: passing argument 22 of ‘function’ from incompatible pointer type /home/ianhin/Cactus/GBSC12/configs/bbh/bindings/Variables/ADMBase.c:34: note: expected ‘const struct cGH * const*’ but argument is of type ‘struct cGH **’ }}} This last warning appears very many times during a build (though I believe it may be triggered only by GCC and not by the Intel compiler). {{{ /home/ianhin/Cactus/GBSC12/src/main/Parameters.c: In function ‘ParameterSetBoolean’: /home/ianhin/Cactus/GBSC12/src/main/Parameters.c:2394: warning: passing argument 4 of ‘Util_ExpressionEvaluate’ discards qualifiers from pointer target type /home/ianhin/Cactus/GBSC12/src/include/util_Expression.h:38: note: expected ‘void *’ but argument is of type ‘const int *’ }}} There are probably more as well. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/874> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

[Einstein Toolkit] #780: missing small packages
by Einstein Toolkit 29 Jul '13

29 Jul '13

#780: missing small packages ----------------------------------------------+----------------------------- Reporter: knarf | Owner: dcastl2 Type: enhancement | Status: new Priority: major | Milestone: Component: Einstein Toolkit Virtual Machine | Version: Keywords: | ----------------------------------------------+----------------------------- Some tools to be installed: mmv screen tmux texlive (probably quite large, but also probably necessary) -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/780> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 5

[Einstein Toolkit] #1192: Speed up Hydro_InitExcision test cases
by Einstein Toolkit 08 Jul '13

08 Jul '13

#1192: Speed up Hydro_InitExcision test cases -----------------------------------+---------------------------------------- Reporter: eschnett | Owner: Type: enhancement | Status: new Priority: major | Milestone: Component: EinsteinToolkit thorn | Version: Keywords: | -----------------------------------+---------------------------------------- Hydro_InitExcision sets up an excision mask for hydrodynamics evolution. It is active at initial and poststep (and postregrid), and its logic depends only on parameters, not on the simulation state. Its test cases take a very long time to run: (1) Running this for 100 iterations with PUGH is overkill. 1 iteration suffices, testing both initial and poststep. (2) The excision mask is not even output! Instead, only the indirect results on the hydro variables are tested. (3) All test cases use PUGH, so Carpet's postregrid bin is not tested. (4) Each test exists three times, for different reconstruction methods. The reconstruction methods do not enter into this thorn's logic. I conclude: The existing tests confirm that this thorn has the right functionality. They are therefore important. However, they should not be in the Cactus test suite. We want regression tests in the Cactus test cases, and this could be achieved about 100x faster (fewer iterations, fewer test cases). Additionally, there should be Carpet tests, and the excision mask should be output. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/1192> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 1

[Einstein Toolkit] #456: Ensure functions are not defined twice
by Einstein Toolkit 08 Jul '13

08 Jul '13

#456: Ensure functions are not defined twice -------------------------+-------------------------------------------------- Reporter: hinder | Owner: Type: enhancement | Status: new Priority: major | Milestone: Component: Cactus | Version: Keywords: | -------------------------+-------------------------------------------------- When multiple source files define non-static functions with the same name, these are all compiled together into the resulting executable. Which one gets called is not predictable by the user. This can happen if a user copies a thorn, modifies it slightly, and compiles both thorns into the same configuration. This can lead to *very* difficult-to-find bugs, and it would be helpful if Cactus was able to prevent this, or at least to mitigate the problem. At the CST level, Cactus should be able to tell that there are two thorns which schedule functions with the same name. At the moment, this is not caught. I propose that this should be a fatal error, as there is no way to predict which function will be called eventually. This will solve the problem in some cases, but not in the case where there are instances of duplicate function names which are not scheduled. It should be possible to scan the object/library files using standard tools (nm etc) to determine if there are multiple globally visible symbols with the same name. There might even be standard tools for this purpose. Doing this in a portable way might not be straightforward, but having an implementation for Linux, for example, would catch the majority of cases. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/456> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 7

[Einstein Toolkit] #1195: Modify bisection algorithm in EOS_Omni's nuc_eos
by Einstein Toolkit 07 Jul '13

07 Jul '13

#1195: Modify bisection algorithm in EOS_Omni's nuc_eos -----------------------------------+---------------------------------------- Reporter: eschnett | Owner: Type: enhancement | Status: new Priority: major | Milestone: Component: EinsteinToolkit thorn | Version: Keywords: | -----------------------------------+---------------------------------------- EOS_Omni's nuc_eos routines use a bisection method to find the temperature for a given internal energy. This bisects on the temperature T. I suggest to bisect on log T instead, which is a more natural operation. For example, if the temperature range is [1...1000] in some units, then bisecting in T will examine the sequence 500, 250, 125, ..., and will thus perform many table lookups until (say) T=10 is reached. Bisecting in log T will find this result much faster. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/1195> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 5

[Einstein Toolkit] #447: simfactory 1.0 always uses -L 3
by Einstein Toolkit 06 Jul '13

06 Jul '13

#447: simfactory 1.0 always uses -L 3 ----------------------------------------------+----------------------------- Reporter: alexander.beck-ratzka@… | Owner: eschnett Type: defect | Status: new Priority: major | Milestone: Component: SimFactory | Version: Keywords: | ----------------------------------------------+----------------------------- The perl version of simfactory always uses the debugging loglevel for simulations. The loglevel of a cactus simulation can be specified by setting -L to a value between 0 (none) and 3 (debug). While the cactus default is 0, simfactory sets it to 3. The debug loglevel could lead to huge output files. I would suggest to set it to the cactus default, and allow a user to increase it. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/447> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 4

[Einstein Toolkit] #107: Update released einsteintoolkit.bib once we have a more final version
by Einstein Toolkit 06 Jul '13

06 Jul '13

#107: Update released einsteintoolkit.bib once we have a more final version -----------------------+---------------------------------------------------- Reporter: anonymous | Type: task Status: new | Priority: major Milestone: | Component: EinsteinToolkit website Version: | Keywords: -----------------------+---------------------------------------------------- -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/107> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

← Newer
1
...
5
6
7
8
9
10
11
...
14
Older →

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Trac December 2012