Hi, Alle domenica 27 dicembre 2009, Thomas Schwinge ha scritto: > On Fri, Dec 25, 2009 at 03:26:05PM +0200, Niko Tyni wrote: > > Looks rather like a problem with the Storable module, but it's hard to > > say more without a test case. > > That's what I thought, too, but I now bisected this down to perlmagick > being installed vs. not being installed. Involving Pino Toscano who > recently did some imagemagick patching > (cf. <http://bugs.debian.org/551017>) -- Pino, please note that I'm not > saying that you're responsible for this perlmagick breakage which we're > discussing in <http://bugs.debian.org/557771>, but perhaps you have an > idea / know how to run some testing / etc. The only change needed was patching GetExecutionPath() (defined in magick/utility.c); note that the released version is slightly different from the one I sent. Knowing whether that function is the actual issue should be matter of taking the current implementation in the "#if defined(__GNU__) ... #endif" block out of there, put only a "return (MagickFalse);" in its place, recompile and test. If it still doesn't work with the above change, another option would be adding 1 to the calculation of "extent" (always in the __GNU__ block said above). A further option would be asking upstream to provide more unit test for that function, in case the current unit tests (which all pass on Hurd) do not covert it yet/fully. -- Pino Toscano
Attachment:
signature.asc
Description: This is a digitally signed message part.