Re: [Pkg-ia32-libs-maintainers] ia32-libs-core_20100610_ia64.changes ACCEPTED
dann frazier <dannf@dannf.org> writes:
> On Wed, Jun 23, 2010 at 06:52:35PM +0200, Goswin von Brederlow wrote:
>> dann frazier <dannf@dannf.org> writes:
>>
>> > On Wed, Jun 23, 2010 at 04:37:52PM +0200, Goswin von Brederlow wrote:
>> >> Archive Administrator <installer@ftp-master.debian.org> writes:
>> >>
>> >> > Accepted:
>> >> > ia32-libs-core_20100610.dsc
>> >> > to main/i/ia32-libs-core/ia32-libs-core_20100610.dsc
>> >> > ia32-libs-core_20100610.tar.gz
>> >> > to main/i/ia32-libs-core/ia32-libs-core_20100610.tar.gz
>> >> > ia32-libs-core_20100610_ia64.deb
>> >> > to main/i/ia32-libs-core/ia32-libs-core_20100610_ia64.deb
>> >> >
>> >> >
>> >> > Override entries for your package:
>> >> > ia32-libs-core_20100610.dsc - extra libs
>> >> > ia32-libs-core_20100610_ia64.deb - extra libs
>> >> >
>> >> > Announcing to debian-devel-changes@lists.debian.org
>> >>
>> >> Can someone install this on ia64 and test the failed java build please?
>> >
>> > Which failed java build?
>> >
>>
>> The one from
>>
>> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=563402
>
> dannf@biglpk:~/sun-java6-6.20-dlj$ fakeroot ./debian/rules build 2>&1 | tee ../sun-java6.
> Build information
> ==========================
> TOP = usr/lib/jvm
> VENDOR = sun
> RELEASE = 6
> PRODUCT = java6
> basename = sun-java6
> version = 6
> releng_ver = 20
> unpackdir = jdk1.6.0_20
> srcdir = ia64-jdk
> all_archs = amd64 i586
> arch = ia64
> archdir = ia64
>
> rm -rf jdk1.6.0_20
> binsize=$(wc -c jdk-6u20-dlj-linux-i586.bin | awk '{print $1}'); \
> zipstart=$(unzip -ql jdk-6u20-dlj-linux-i586.bin 2>&1 >/dev/null | sed -n -e 's/.* \([0-9][0-9]*\) extra bytes.*/\1/p'); \
> tail -c $(expr $binsize - $zipstart) jdk-6u20-dlj-linux-i586.bin > tmp-jdk.zip
> unzip -q tmp-jdk.zip
> mv jdk1.6.0_20 i586-jdk
> i586
> diff -u i586-jdk/LICENSE i586-jdk/jre/LICENSE
> diff -u i586-jdk/THIRDPARTYLICENSEREADME.txt \
> i586-jdk/jre/THIRDPARTYLICENSEREADME.txt
> set -e; \
> for i in $(find i586-jdk -name '*.pack'); do \
> i586-jdk/bin/unpack200 $i \
> $(dirname $i)/$(basename $i .pack).jar; \
> rm -f $i; \
> done
> /bin/bash: line 1: 32364 Segmentation fault i586-jdk/bin/unpack200 $i $(dirname $i)/$(basename $i .pack).jar
> make: *** [unpack-i586-stamp] Error 139
> dannf@biglpk:~/sun-java6-6.20-dlj$ dpkg -l | grep ia32-libs
> ii ia32-libs-core 20100610 core ia32 shared libraries for use on ia64 s
>
>
> How can I help further?
That at least is something new.
Only thing I can think of is to test if the i586-jdk/bin/unpack200 with
the same versions of libs segfaults on native i386 too or if the problem
is in the 32bit emulation on ia64.
What could be is that the amd64 libc6-i386, which is used in
ia32-libs-core, is compiled with cpu features that ia64 simply does not
have. I would expect a SIGILL then though.
MfG
Goswin
Reply to: