more precisely, the following (the original was putting the new global variable in the wrong place)
704a705 > static Obj ProtectFname; 727a729,731 > /* For some reason itanium GC seems unable to spot fname */ > ProtectFname = fname; > 730a735,736 > ProtectFname = (Obj)0L; > 1071a1078 > 1078a1086 > InitGlobalBag(&ProtectFname, "Protected Filename for SaveWorkspace"); On Jan 11, 9:53 pm, Dima Pasechnik <dimp...@gmail.com> wrote: > OK, this patch appears to fix the problem, indeed. > Dima > > 2010/1/11 Stephen Linton <s...@cs.st-andrews.ac.uk>: > > > > > Try this patch for saveload.c: > > > 708a709 > >> static Obj ProtectFname; > > 727a729,731 > >> /* For some reason itanium GC seems unable to spot fname */ > >> ProtectFname = fname; > > > 730a735,736 > >> ProtectFname = (Obj)0L; > > > 1070a1077,1078 > >> InitGlobalBag(&ProtectFname, "Protected Filename for SaveWorkspace"); > > > It should bullet-proof the filename through the garbage collection. > > > Steve > > > On 10 Jan 2010, at 21:26, William Stein wrote: > > >> On Sun, Jan 10, 2010 at 1:06 PM, Stephen Linton <s...@cs.st-andrews.ac.uk> > >> wrote: > >>> As far as I can tell, all the relevant code is identical in 4.4.10 + > >>> patch and 4.4.12. Of course, I may (must, perhaps) have too narrow a > >>> definition of relevant, but I wonder if its actually a matter of 4.4.10 > >>> being lucky in some way, rather than a bug having been introduced later. > > >>> If we think the problem is just SaveWorkspace, I can hack around it, but > >>> I for one, would certainly be holding my breath waiting for the next > >>> similar bug to show up. > > >>> Steve > > >> It's likely better to hack around it if it is limited to > >> SaveWorkspace, then to simply stick with 4.4.10, since presumably > >> there are many bugs that got fixed in 4.4.12 that are in 4.4.10. > >> Moreover, whatever is the real bug causing this is probably not way > >> worse in 4.4.12 than in 4.4.10. > > >> Do you have still know how to access the Itanium account(s) I got you? > >> Do I need to setup anything for you? > > >> -- William > > >>> On 10 Jan 2010, at 19:11, William Stein wrote: > > >>>> On Sun, Jan 10, 2010 at 10:17 AM, Dima Pasechnik <dimp...@gmail.com> > >>>> wrote: > > >>>>> On Jan 10, 11:56 pm, William Stein <wst...@gmail.com> wrote: > >>>>>> On Sat, Jan 9, 2010 at 11:10 PM, Dima Pasechnik <dimp...@gmail.com> > >>>>>> wrote: > >>>>>>> William, > >>>>>>> I think I can reproduce one of your Itanium GAP bug; the workspace > >>>>>>> filename gets mangled by SaveWorkspace. > >>>>>>> (so the workspace gets saved, but with a horribly wrong name...) > >>>>>>> Should be next to trivial to fix... > >>>>>>> Dima > > >>>>>> Sweet. If I remember correctly, workspace bugs were the *only* bugs > >>>>>> we had. > > >>>>> turned out to be a Itanium-specific GC bug, so fixing might take a > >>>>> little while. > >>>>> Oh well. > > >>>>> Dmitrii > > >>>> Since the bug appears not to be in 4.10, can you look what was changed > >>>> between 4.10 and 4.12 that is related to Itanium specific garbage > >>>> collection? (I may have already tried that and found that it wasn't > >>>> so easy...) > > >>>>>> William > > >>>>> -- > >>>>> To post to this group, send email to sage-support@googlegroups.com > >>>>> To unsubscribe from this group, send email to > >>>>> sage-support+unsubscr...@googlegroups.com > >>>>> For more options, visit this group > >>>>> athttp://groups.google.com/group/sage-support > >>>>> URL:http://www.sagemath.org > > >>>> -- > >>>> William Stein > >>>> Associate Professor of Mathematics > >>>> University of Washington > >>>>http://wstein.org > > >> -- > >> William Stein > >> Associate Professor of Mathematics > >> University of Washington > >>http://wstein.org > > -- > Dmitrii Pasechnik > ----- > DISCLAIMER: Any text following this sentence does not constitute a > part of this message, and was added automatically during transmission.
-- To post to this group, send email to sage-support@googlegroups.com To unsubscribe from this group, send email to sage-support+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-support URL: http://www.sagemath.org
