Bug#612542: libhunspell-1.2-0: Iceweasel is crashing inside libhunspell when producing a suggestion list
tag 612542 + upstream
tag 612542 fixed-upstream
forwarded 612542 Caolan McNamara <caolanm@redhat.com>
thanks
On Tue, Feb 08, 2011 at 04:53:27PM -0800, David Starner wrote:
> I've been have crashes in Iceweasel when right-clicking on a word. I
> ran it under valgrind, and produced the message below, showing that
> mkallcap was trying to read invalid memory. Rolling it back to the
> version in testing, 1.2.11-1, made it work again. Iceweasel is version
> 3.5.16-4.
>
> ==16715== Invalid read of size 1
> ==16715== at 0x9920EAF: mkallcap(char*, cs_info const*) (in
> /usr/lib/libhunspell-1.2.so.0.0.0)
> ==16715== by 0x992F513: SuggestMgr::ngsuggest(char**, char*, int,
> HashMgr**, int) (in /usr/lib/libhunspell-1.2.so.0.0.0)
> ==16715== by 0x992A1E0: Hunspell::suggest(char***, char const*) (in
> /usr/lib/libhunspell-1.2.so.0.0.0)
> ==16715== by 0x6F25E4D: ??? (in /usr/lib/xulrunner-1.9.1/libxul.so)
> ==16715== by 0x6F1DD98: ??? (in /usr/lib/xulrunner-1.9.1/libxul.so)
> ==16715== by 0x6E29617: ??? (in /usr/lib/xulrunner-1.9.1/libxul.so)
> ==16715== by 0x6F6517D: NS_InvokeByIndex_P (in
> /usr/lib/xulrunner-1.9.1/libxul.so)
> ==16715== by 0x67A7BC7: ??? (in /usr/lib/xulrunner-1.9.1/libxul.so)
> ==16715== by 0x67AF492: ??? (in /usr/lib/xulrunner-1.9.1/libxul.so)
> ==16715== by 0x7E88456: js_Invoke (in /usr/lib/libmozjs.so.2d)
> ==16715== by 0x7E887F9: ??? (in /usr/lib/libmozjs.so.2d)
> ==16715== by 0x7E507E4: JS_CallFunctionValue (in /usr/lib/libmozjs.so.2d)
> ==16715== Address 0x137 is not stack'd, malloc'd or (recently) free'd
Just talked with upstream on IRC about this:
11:32 <@_rene_> caolan: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=612542
[...]
11:36 <@caolan> _rene_: I think that one is fixed in 1.2.15, if its the "assume
dict is not utf-8, oops it is" problem
11:36 <@caolan> hard to just grab the fix/log seeing as (for me anyway)
sourceforge cvs is still busted
11:37 <@_rene_> mmh, ok
11:37 <@_rene_> so I either choose between this or the potential ABI break. I
guess I'll choose the latter ;)
[...]
11:41 <@caolan> _rene_: you could unpack 1.2.14 and 1.2.15 and the bits of the
diff of -mkallcap +if (utf8) { do mkallcap_utf8 } else { do
mkallcap } are the fix
[...]
11:50 <@_rene_> caolan: thanks
Grüße/Regards,
René
--
.''`. René Engelhard -- Debian GNU/Linux Developer
: :' : http://www.debian.org | http://people.debian.org/~rene/
`. `' rene@debian.org | GnuPG-Key ID: D03E3E70
`- Fingerprint: E12D EA46 7506 70CF A960 801D 0AA0 4571 D03E 3E70
Reply to: