[vtk-developers] TestSetGet timeouts on RogueResearch machines

[Bill Lorensen]

Who needs CSI for excitement? Not us...

On Wed, Feb 20, 2013 at 7:32 PM, David Cole <dlrdave at aol.com> wrote:

> Run "cmake -E environment" (or even just "env | sort") in your login
> prompt and the cron env and then we can analyze the env diffs and see
> what might cause it. Maybe something in that report will pop out at
> somebody if this doesn't already trigger a thought about what it could be.
>
>
>
>
>
>  -----Original Message-----
> From: Sean McBride <sean at rogue-research.com>
> To: David E DeMarle <dave.demarle at kitware.com>
> Cc: David Cole <dlrdave at aol.com>; bill.lorensen <bill.lorensen at gmail.com>;
> vtk-developers <vtk-developers at vtk.org>
> Sent: Wed, Feb 20, 2013 7:28 pm
> Subject: Re: [vtk-developers] TestSetGet timeouts on RogueResearch machines
>
>  On Wed, 20 Feb 2013 16:45:38 -0500, David E DeMarle said:
>
> >Could be an environment difference between login and cron shell.
> >Put that command in a crontab that runs a few minutes from now, and
> >make sure output is redirected somewhere (ctest <what_dave_c_said> foo
> >2>&1 outfile.txt).
>
> Ah ha!  So I ran the test again via launchctl (OS X's cron) and it repros now.
> I attached with gdb and paused the process every once in a while, the common
> part of the backtrace is always:
>
> #1  0x0000000102b5eec8 in Tk_MeasureChars ()
> #2  0x0000000102b30927 in TkpTesttextCmd ()
> #3  0x0000000102b302b9 in TkpTesttextCmd ()
> #4  0x0000000102b2ff38 in TkpTesttextCmd ()
> #5  0x0000000102b2cd3e in TkpTesttextCmd ()
> #6  0x0000000102b2e5df in TkpTesttextCmd ()
> #7  0x0000000102b2dc83 in TkpTesttextCmd ()
> #8  0x0000000102b2e7de in TkpTesttextCmd ()
> #9  0x0000000102b220e7 in TkpTesttextCmd ()
> #10 0x0000000102bfe9cf in Tcl_ListMathFuncs ()
> #11 0x0000000102c3b283 in Tcl_ExprObj ()
> #12 0x0000000102c76a49 in TclObjInterpProcCore ()
> #13 0x0000000102bfe9cf in Tcl_ListMathFuncs ()
> #14 0x0000000102bff2a5 in Tcl_EvalObjv ()
> #15 0x0000000102c00518 in TclEvalObjEx ()
> #16 0x0000000102ac1728 in Tk_CreateConsoleWindow ()
> #17 0x0000000102c4f822 in Tcl_ClearChannelHandlers ()
> #18 0x0000000102c54f8e in Tcl_GetChannelError ()
> #19 0x0000000102c4f479 in Tcl_SpliceChannel ()
> #20 0x0000000102c504ae in Tcl_WriteObj ()
> #21 0x0000000102c55329 in Tcl_GetChannelError ()
> #22 0x0000000102bfe9cf in Tcl_ListMathFuncs ()
> #23 0x0000000102c3b283 in Tcl_ExprObj ()
> #24 0x0000000102c76a49 in TclObjInterpProcCore ()
> #25 0x0000000102bfe9cf in Tcl_ListMathFuncs ()
> #26 0x0000000102c3b283 in Tcl_ExprObj ()
> #27 0x0000000102c76a49 in TclObjInterpProcCore ()
> #28 0x0000000102bfe9cf in Tcl_ListMathFuncs ()
> #29 0x0000000102bffa7a in Tcl_EvalEx ()
> #30 0x0000000102bff444 in Tcl_EvalEx ()
> #31 0x0000000102c5be85 in Tcl_FSEvalFileEx ()
> #32 0x0000000102acea9a in Tk_MainEx ()
> #33 0x0000000102919bad in main (argc=4, argv=0x7fff5d2e8bd8) at
> /Users/builder/external/VTK/Wrapping/Tcl/vtkTkAppInit.cxx:123
>
> After Tk_MeasureChars() it's always different, but generally the function names
> look like font/glyph stuff.
>
> Does this ring any bells to anyone?
>
> If I let it continue, it seems to still pass, slowly, ex 900 seconds.
>
> I also tried bypassing ctest and running the process directly, and that changes
> nothing (still runs super slowly via launchctl, and not otherwise).
>
> All my buildbots do parallel testing, but Rogue11 will run one test at a time
> tonight, as a test...
>
> Cheers,
>
> --
> ____________________________________________________________
> Sean McBride, B. Eng                 sean at rogue-research.com
> Rogue Research                        www.rogue-research.com
> Mac Software Developer              Montréal, Québec, Canada
>
>
>
>


-- 
Unpaid intern in BillsBasement at noware dot com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/vtk-developers/attachments/20130220/a83acaff/attachment.html>