Trac

trac@lists.einsteintoolkit.org

8708 discussions

[Einstein Toolkit] #155: SimFactory tutorial does not say how to download GetComponents
by Einstein Toolkit 08 Mar '11

08 Mar '11

#155: SimFactory tutorial does not say how to download GetComponents ------------------------+--------------------------------------------------- Reporter: hinder | Owner: mthomas Type: defect | Status: new Priority: major | Milestone: Component: SimFactory | Version: Keywords: | ------------------------+--------------------------------------------------- The page http://docs.einsteintoolkit.org/et-docs/Pysim_Tutorial_for_New_Users assumes that you have GetComponents already. I suggest to copy the relevant text from the cactuscode.org/download page. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/155> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 1

[Einstein Toolkit] #330: Allow the user to choose whether to receive job notification emails
by Einstein Toolkit 08 Mar '11

08 Mar '11

#330: Allow the user to choose whether to receive job notification emails -------------------------+-------------------------------------------------- Reporter: hinder | Owner: mthomas Type: enhancement | Status: new Priority: minor | Milestone: Component: SimFactory | Version: Keywords: | -------------------------+-------------------------------------------------- Most HPC batch systems can be configured to send users an email when their job starts and stops. This is usually achieved by adding a specific line to the job submission script including the user's email address. At present, SimFactory includes this line in all job submission scripts. This means that the user cannot choose whether to receive these emails or not. I propose a new machine-specific option which allows the user to enable or disable all job-related emails. This could be an entry in the machine database which was by default unset, could be specified in the [default] machine, or customised on a per-machine basis. For example, I probably don't want job notification emails for my laptop, but for Kraken where jobs can take a long time to start I probably do. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/330> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 0

[Einstein Toolkit] #263: run-script cannot find output-0000-active
by Einstein Toolkit 04 Mar '11

04 Mar '11

#263: run-script cannot find output-0000-active ------------------------+--------------------------------------------------- Reporter: knarf | Owner: mthomas Type: defect | Status: new Priority: critical | Milestone: Component: SimFactory | Version: Keywords: | ------------------------+--------------------------------------------------- While trying to run on philip: /work/knarf/philip/simulations/leak001/output-0000/SIMFACTORY/RunScript: line 7: cd: /work/knarf/phi lip/simulations/leak001/output-0000-active: No such file or directory And indeed, the directory is not there (I did build, create and submit before). -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/263> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 4

[Einstein Toolkit] #319: more OpenMP WORKSHARE sections in GRHydro Reconstruct
by Einstein Toolkit 02 Mar '11

02 Mar '11

#319: more OpenMP WORKSHARE sections in GRHydro Reconstruct -------------------------+-------------------------------------------------- Reporter: rhaas | Owner: Type: enhancement | Status: new Priority: minor | Milestone: Component: Other | Version: Keywords: | -------------------------+-------------------------------------------------- the attached patch compiles and should at least do no harm. It adds some new OMP WORKSHARE sections around array operations and adds PARALLEL to some existing ones. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/319> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

[Einstein Toolkit] #324: Unknown MDB keys should lead to syntax error
by Einstein Toolkit 02 Mar '11

02 Mar '11

#324: Unknown MDB keys should lead to syntax error ------------------------+--------------------------------------------------- Reporter: eschnett | Owner: mthomas Type: defect | Status: new Priority: minor | Milestone: Component: SimFactory | Version: Keywords: | ------------------------+--------------------------------------------------- If SimFactory encounters and MDB key that it does not know, then this is most likely a typo and should be reported as syntax error. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/324> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 0

[Einstein Toolkit] #310: segfault with kerr initial data using IDAnalyticBH
by Einstein Toolkit 28 Feb '11

28 Feb '11

#310: segfault with kerr initial data using IDAnalyticBH --------------------------+------------------------------------------------- Reporter: bmundim | Owner: bmundim Type: defect | Status: new Priority: major | Milestone: Component: Other | Version: Keywords: IDAnalyticBH | --------------------------+------------------------------------------------- Yosef's bug report and patch: Hi, If initial_data=kerr and metric_type=physical, IDAnalyticBH writes to psi[x-z] and psi[x-z][x-z] (it tries to set these GFs to zero) even if no storage is allocated. The problem is with the code IDAnalyticBH_zero_CCTK_REAL_array(npoints, psix); etc.., which is only executed when the metric is physical. I was able to use the Kerr data with the following patch, but just deleting the offending lines should be sufficient. --- Kerr.c 2011-02-10 09:08:58.000000000 -0500 +++ /tmp/Kerr.c 2011-02-23 18:15:38.000000000 -0500 @@ -103,6 +103,10 @@ conformal_storage); /*NOTREACHED*/ } } + else + { + *conformal_state = 0; + } /* printf("npoints: %i\n",npoints); */ @@ -311,15 +315,22 @@ } IDAnalyticBH_zero_CCTK_REAL_array(npoints, gxz); IDAnalyticBH_zero_CCTK_REAL_array(npoints, gyz); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psix); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiy); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiz); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixx); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixy); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixz); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiyy); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiyz); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psizz); + + if (psix && *conformal_state < 2) + { + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psix); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiy); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiz); + } + if (psixx && *conformal_state < 3) + { + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixx); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixy); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixz); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiyy); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiyz); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psizz); + } } return; -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/310> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 1

[Einstein Toolkit] #309: segfault with kerr initial data using IDAnalyticBH
by Einstein Toolkit 28 Feb '11

28 Feb '11

#309: segfault with kerr initial data using IDAnalyticBH --------------------------+------------------------------------------------- Reporter: bmundim | Owner: bmundim Type: defect | Status: new Priority: major | Milestone: Component: Other | Version: Keywords: IDAnalyticBH | --------------------------+------------------------------------------------- Yosef's bug report and patch: Hi, If initial_data=kerr and metric_type=physical, IDAnalyticBH writes to psi[x-z] and psi[x-z][x-z] (it tries to set these GFs to zero) even if no storage is allocated. The problem is with the code IDAnalyticBH_zero_CCTK_REAL_array(npoints, psix); etc.., which is only executed when the metric is physical. I was able to use the Kerr data with the following patch, but just deleting the offending lines should be sufficient. --- Kerr.c 2011-02-10 09:08:58.000000000 -0500 +++ /tmp/Kerr.c 2011-02-23 18:15:38.000000000 -0500 @@ -103,6 +103,10 @@ conformal_storage); /*NOTREACHED*/ } } + else + { + *conformal_state = 0; + } /* printf("npoints: %i\n",npoints); */ @@ -311,15 +315,22 @@ } IDAnalyticBH_zero_CCTK_REAL_array(npoints, gxz); IDAnalyticBH_zero_CCTK_REAL_array(npoints, gyz); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psix); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiy); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiz); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixx); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixy); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixz); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiyy); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiyz); - IDAnalyticBH_zero_CCTK_REAL_array(npoints, psizz); + + if (psix && *conformal_state < 2) + { + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psix); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiy); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiz); + } + if (psixx && *conformal_state < 3) + { + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixx); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixy); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psixz); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiyy); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psiyz); + IDAnalyticBH_zero_CCTK_REAL_array(npoints, psizz); + } } return; -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/309> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 1

[Einstein Toolkit] #312: typo in eosgeneral con2prim causes assignment to full array instead of single element
by Einstein Toolkit 26 Feb '11

26 Feb '11

#312: typo in eosgeneral con2prim causes assignment to full array instead of single element ---------------------+------------------------------------------------------ Reporter: rhaas | Owner: Type: defect | Status: new Priority: minor | Milestone: Component: Cactus | Version: ET_2010_11 Keywords: GRHydro | ---------------------+------------------------------------------------------ in Con2PrimGeneral inside a "do k .. do j .. do i" loop an assignment to the array "f" is made where only f(i,j,k) is used later on (all of "f" had been used earlier). In Con2PrimGeneral_Polytype this does no happen. This causes no actual error, just slower execution. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/312> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 2

[Einstein Toolkit] #303: expand $1 and $2 in GIT_REPO, DARCS_REPO, HG_REPO
by Einstein Toolkit 26 Feb '11

26 Feb '11

#303: expand $1 and $2 in GIT_REPO,DARCS_REPO,HG_REPO -------------------------+-------------------------------------------------- Reporter: rhaas | Owner: Type: enhancement | Status: new Priority: minor | Milestone: Component: Other | Version: Keywords: | -------------------------+-------------------------------------------------- I propose to apply the patch below to allow $1 and $2 to be expanded in GIT_REPO etc. this allows for things like: {{{ !TARGET = $ARR !TYPE = git !URL = $USER@numrel.org:/home/numrel/arrangements/$1/$2 !REPO_PATH = ../$2 !CHECKOUT = Scotch/SoundSpeed Scotch/TotalRestMass }}} to check out both SoundSpeed an TotalRestMass, which are both separate git repositories. Without the patch one has to repeat the whole $TARGET stanza for each and every thorn (since each thorn is a repository) only changing $1/$2 to Scotch/ThornName. I tested the patch with einsteintoolkit.th and the resulting checkout (using --parallel) is identical with and without the patch. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/303> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 3

[Einstein Toolkit] #287: NVCC option should be renamed CUCC
by Einstein Toolkit 25 Feb '11

25 Feb '11

#287: NVCC option should be renamed CUCC ----------------------+----------------------------------------------------- Reporter: eschnett | Owner: Type: defect | Status: new Priority: minor | Milestone: Component: Cactus | Version: Keywords: | ----------------------+----------------------------------------------------- The configuration option to specify the CUDA C compiler should be called CUCC, not NVCC. nvcc is only NVidia's CUDA C compiler. -- Ticket URL: <https://trac.einsteintoolkit.org/ticket/287> Einstein Toolkit <http://einsteintoolkit.org> The Einstein Toolkit

1 6

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Trac