the subject line is what is recieved after attempting to run inkscape 0.41 from the command line. using the gui interface, it quitely dies, no error messages. this is on gentoo. i've filed bug report #1176668, this happens with the CVS version and the emerge version. there where no other errors reported with or during the emerge.
any other troubleshooting that i can do before dumping this version and going back to 0.37, which works fine ? thanks for the suggestions jerry
__________________________________ Do you Yahoo!? Yahoo! Personals - Better first dates. More second dates. http://personals.yahoo.com
Nearly all of the problems of this class on Gentoo have been due to an upgraded gcc, so that's what I'm going to bet this is, as well.
Basically, the C++ linking ABI between gcc 3.3 and 3.4 changed, and if you mix C++ libs compiled with one version, with code from the other, you get all sorts of oddball problems - 'Absurd stack bottom value' sure sounds oddball. ;-)
Look at gcc --version. If it's >3.3, then that's probably the issue.
If this is indeed the cause, then the solution is straightforward: Re-emerge gtkmm, glibmm, and sigc++, then re-compile inkscape.
On the off chance that this *isn't* the problem, you could run gdb and get a backtrace (see the Report Bug page on the Inkscape website.) Attach that to the bug report. Again, though, my bet is that it's the well-known ABI bug, in which case gdb will be of no help; try re-emerging the libs so they match your compiler version.
Bryce
On Mon, Apr 04, 2005 at 03:46:38PM -0700, jerry ondola wrote:
the subject line is what is recieved after attempting to run inkscape 0.41 from the command line. using the gui interface, it quitely dies, no error messages. this is on gentoo. i've filed bug report #1176668, this happens with the CVS version and the emerge version. there where no other errors reported with or during the emerge.
any other troubleshooting that i can do before dumping this version and going back to 0.37, which works fine ? thanks for the suggestions jerry
__________________________________ Do you Yahoo!? Yahoo! Personals - Better first dates. More second dates. http://personals.yahoo.com
SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click _______________________________________________ Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel
On Mon, Apr 04, 2005 at 03:46:38PM -0700, jerry ondola wrote:
i've filed bug report #1176668,
Please direct any responses to that bug report.
News so far: the message comes from libgc. Submitter is using x86; was compiling with -fomit-frame-pointer -march=686 -O2, gcc 3.3. I suggested that -fomit-frame-pointer could be the culprit given that it's complaining about the stack.
pjrm.
--- Peter Moulder <Peter.Moulder@...38...> wrote:
On Mon, Apr 04, 2005 at 03:46:38PM -0700, jerry ondola wrote:
i've filed bug report #1176668,
Please direct any responses to that bug report.
News so far: the message comes from libgc. Submitter is using x86; was compiling with -fomit-frame-pointer -march=686 -O2, gcc 3.3. I suggested that -fomit-frame-pointer could be the culprit given that it's complaining about the stack.
pjrm.
So are you suggesting I recompile without -fomit-frame-pointer ? jerry
__________________________________ Do you Yahoo!? Yahoo! Personals - Better first dates. More second dates. http://personals.yahoo.com
On Mon, Apr 04, 2005 at 09:07:22PM -0700, jerry ondola wrote:
So are you suggesting I recompile without -fomit-frame-pointer ?
Yes; please see the bug report for bug #1176668 on sourceforge (https://sourceforge.net/tracker/index.php?func=detail&aid=1176668&gr...) for more detail.
pjrm.
Quoting jerry ondola <bluestrkchain@...36...>:
the subject line is what is recieved after attempting to run inkscape 0.41 from the command line. using the gui interface, it quitely dies, no error messages. this is on gentoo. i've filed bug report #1176668, this happens with the CVS version and the emerge version. there where no other errors reported with or during the emerge.
any other troubleshooting that i can do before dumping this version and going back to 0.37, which works fine ?
We need to refer this to the libgc list, I think. The normal means for libgc to find the top and bottom of the stack apparently aren't working in your case.
Are you using any special kernel build options?
-mental
--- mental@...3... wrote:
We need to refer this to the libgc list, I think. The normal means for libgc to find the top and bottom of the stack apparently aren't working in your case.
Are you using any special kernel build options?
-mental
nothing special, just grsec
__________________________________ Yahoo! Messenger Show us what our next emoticon should look like. Join the fun. http://www.advision.webevents.yahoo.com/emoticontest
On Tue, 2005-04-05 at 14:16 -0700, jerry ondola wrote:
--- mental@...3... wrote:
Are you using any special kernel build options?
-mental
nothing special, just grsec
So I guess you're joking? "Nothing special," but grsecurity? There's a reason that grsecurity doesn't end up in the mainline kernel. Hint: It's not that mainline kernel developers don't care about security. Of course, this might still mean that gc actually uses something that lies on the limits of what is guaranteed by POSIX, it's just that grsecurity is likely breaking something that many believe is an accepted standard.
Truely blatant breaking of POSIX rules often gets hammered by Exec- shield and pals in Fedora; Inkscape seems fine here as far as I can see, so grsecurity seems like a likely problem candidate.
Jerry, any chance you can try with a more mainstream kernel? If not a vanilla 2.6.11.<latest> then perhaps whatever your distro feels like distributing?
/Per
I haven't heard of any problems with libgc with grsec, but I've never tried myself. I don't think it diddles with the address space much, but it might affect the behavior of some of the nodes in /proc/self/ which libgc uses.
Again, the libgc mailing list is probably the right place to pursue this.
--- Per Bjornsson <perbj@...604...> wrote:
nothing special, just grsec
So I guess you're joking? "Nothing special," but grsecurity? There's a reason that grsecurity doesn't end up in the mainline kernel. Hint: It's not that mainline kernel developers don't care about security. Of course, this might still mean that gc actually uses something that lies on the limits of what is guaranteed by POSIX, it's just that grsecurity is likely breaking something that many believe is an accepted standard.
Truely blatant breaking of POSIX rules often gets hammered by Exec- shield and pals in Fedora; Inkscape seems fine here as far as I can see, so grsecurity seems like a likely problem candidate.
Jerry, any chance you can try with a more mainstream kernel? If not a vanilla 2.6.11.<latest> then perhaps whatever your distro feels like distributing?
/Per
-- Per Bjornsson <perbj@...604...>
Per Bjornsson and all those who offered advice
Linux screamer 2.6.11-gentoo-r5 #2 Tue Apr 5 16:51:11 AKDT 2005 i686 AMD Duron(tm) Processor AuthenticAMD GNU/Linux
inkscape 0.41 runs fine now learn something everyday, now to go back and check the 2.4 kernel, without the grsec checked.
Thanks for all the assistance in getting MY problem solved. I look forward to learning and using this great program.
Thank you jerry
__________________________________ Do you Yahoo!? Yahoo! Personals - Better first dates. More second dates. http://personals.yahoo.com
participants (5)
-
unknown＠example.com -
Bryce Harrington -
jerry ondola -
Per Bjornsson -
Peter Moulder