This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: gij 3.3.3 core dump
- From: Mark Wielaard <mark at klomp dot org>
- To: Steven Augart <augart at watson dot ibm dot com>
- Cc: gcc at gcc dot gnu dot org, java at gcc dot gnu dot org
- Date: Wed, 31 Mar 2004 09:43:33 +0200
- Subject: Re: gij 3.3.3 core dump
- References: <200403260429.i2Q4TPl25447@bilbo.watson.ibm.com>
Hi Steven,
On Thu, Mar 25, 2004 at 11:29:25PM -0500, Steven Augart wrote:
> I'm reporting a segmentation fault when I invoke gij version 3.3.3. I
> built gij on x86/Linux (Ibm Thinkpad T-23, Pentium III, Red Hat kernel
> 2.4.20-18.9).
Discussion about gcj/gij issues are mostly done on the java@gcc.gnu.org
list (CCed) or reported directly into http://gcc.gnu.org/bugzilla/.
> To reproduce it:
Sorry didn't have time.
> Here is the stack trace:
> [...]
> at ../../../gcc-3.3.3/libjava/java/util/ResourceBundle.java:244
> #7 0x403fcef5 in java.util.Calendar.getInstance(java.util.TimeZone, java.util.L
> ocale) (zone=0x8110b40, locale=0x809dc60)
> at ../../../gcc-3.3.3/libjava/java/util/Calendar.java:433
> #8 0x403fcd55 in java.util.Calendar.getInstance() ()
> at ../../../gcc-3.3.3/libjava/java/util/Calendar.java:401
> #9 0x404bf1f5 in java.util.zip.ZipEntry.getCalendar() ()
> at ../../../gcc-3.3.3/libjava/java/util/zip/ZipEntry.java:218
> #10 0x404beecb in java.util.zip.ZipEntry.setTime(long) (this=0x80f5e70,
> time=1079375321)
> at ../../../gcc-3.3.3/libjava/java/util/zip/ZipEntry.java:167
> #11 0x404bf692 in java.util.zip.ZipEntry.setExtra(byte[]) (this=0x80f5e70,
> extra=0x0)@../../../gcc-3.3.3/libjava/java/util/zip/ZipEntry.java:344
> #12 0x404c0586 in java.util.zip.ZipFile.readEntries() (this=0x809eec0)
> [...]
Ugh. I know this one. It is a nasty bootstrap loader issue when the
bootstrap loader is reading a zip file that has extra bytes set for
which it then tries to instantiate a Calendar which needs to get a
ResourceBundle which it loads through the zipfiles known to the
bootstrap loader which...
See also ithis thread http://gcc.gnu.org/ml/java/2003-03/msg00267.html
which should contain a workaround.
This might be fixed in CVS (I haven't checked).
If you have time please try CVS and/or look through bugzilla to see if
it is already in there.
Thanks,
Mark