https://buildd.debian.org/status/package.php?p=gcc-4.6&suite=experimental

Bootstrap comparison failure!
gcc/opts.o differs
make[4]: *** [compare] Error 1


https://buildd.debian.org/status/package.php?p=gcc-snapshot

fails to build with a link error.


--
To UNSUBSCRIBE, email to debian-gcc-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4D861A4B.4020506@debian.org">http://lists.debian.org/4D861A4B.4020506@debian.org

On 21/03/11 04:16, Matthias Klose wrote:

https://buildd.debian.org/status/package.php?p=gcc-snapshot

fails to build with a link error.


That's the GPREL16 relocation truncation error which is reputedly due to
a bug in the linker optimiser. Linking with -Wl,--no-relax resolves
those (presumably producing suboptimal, but at least working, code).


Cheers
Michael.


--
To UNSUBSCRIBE, email to debian-gcc-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4D87125C.1040705@orcon.net.nz">http://lists.debian.org/4D87125C.1040705@orcon.net.nz

On Sun, Mar 20, 2011 at 4:16 PM, Matthias Klose <doko@debian.org> wrote:
> https://buildd.debian.org/status/package.php?p=gcc-4.6&suite=experimental
>
> Bootstrap comparison failure!
> gcc/opts.o differs
> make[4]: *** [compare] Error 1

Hm, 4.6.0-prerelease works for me [1].

> https://buildd.debian.org/status/package.php?p=gcc-snapshot
>
> fails to build with a link error.

Also mainline, but ATM a patch from [2] is needed.

[1] http://gcc.gnu.org/ml/gcc-testresults/2011-03/msg02077.html
[2] http://gcc.gnu.org/ml/gcc-bugs/2011-03/msg02185.html

Uros.


--
To UNSUBSCRIBE, email to debian-gcc-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: AANLkTikiqHn8ud3Ma58szvPu2kqwgFVgaD6BUhbs8kzf@mail .gmail.com">http://lists.debian.org/AANLkTikiqHn8ud3Ma58szvPu2kqwgFVgaD6BUhbs8kzf@mail .gmail.com

On 21.03.2011 13:19, Uros Bizjak wrote:
> On Sun, Mar 20, 2011 at 4:16 PM, Matthias Klose <doko@debian.org> wrote:
>> https://buildd.debian.org/status/package.php?p=gcc-4.6&suite=experimental
>>
>> Bootstrap comparison failure!
>> gcc/opts.o differs
>> make[4]: *** [compare] Error 1
>
> Hm, 4.6.0-prerelease works for me [1].

maybe alpha-no-ev4-directive.diff targeting to ev4 makes a difference?

>> https://buildd.debian.org/status/package.php?p=gcc-snapshot
>>
>> fails to build with a link error.
>
> Also mainline, but ATM a patch from [2] is needed.
>
> [1] http://gcc.gnu.org/ml/gcc-testresults/2011-03/msg02077.html
> [2] http://gcc.gnu.org/ml/gcc-bugs/2011-03/msg02185.html


--
To UNSUBSCRIBE, email to debian-gcc-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4D87452E.6080109@debian.org">http://lists.debian.org/4D87452E.6080109@debian.org

On 22/03/11 01:31, Matthias Klose wrote:
> On 21.03.2011 13:19, Uros Bizjak wrote:
>> On Sun, Mar 20, 2011 at 4:16 PM, Matthias Klose <doko@debian.org> wrote:
>>> https://buildd.debian.org/status/package.php?p=gcc-4.6&suite=experimental
>>>
>>> Bootstrap comparison failure!
>>> gcc/opts.o differs
>>> make[4]: *** [compare] Error 1

I see gcc-4.6 still FTBFS on alpha with the boostrap comparison failure.

> maybe alpha-no-ev4-directive.diff targeting to ev4 makes a difference?

No, it's not that. Nor any of the other Debian patches. Interestingly
it compiles with upstream sources (well, it fails at a later stage with
a linker internal error but that's another story) if one just configures
with ./configure. But if one uses the Debian options used to configure
[1] then upstream source also has the bootstrap comparison failure.

Comparing gcc/opts.o from stage 3 against stage 2 by the following:

objdump -d ./stage2-gcc/opts.o > opts-2.txt
objdump -d ./stage2-gcc/opts.o > opts-2.txt
diff opts-2.txt opts-3.txt

I get from the diff:

90c90
< 14c: 01 00 84 20 lda t3,1(t3)
---
> 14c: 02 30 40 40 addl t1,0x1,t1
93c93
< 158: 02 30 40 40 addl t1,0x1,t1
---
> 158: 01 00 84 20 lda t3,1(t3)

showing that two instructions have been swapped in order. Actually the
order does not matter as we can see from opts-2.txt:

140: a5 05 eb 40 cmpeq t6,s2,t4
144: a3 07 03 41 cmpule t7,t2,t2
148: a5 15 a0 40 cmpeq t4,0,t4
14c: 01 00 84 20 lda t3,1(t3)
150: 03 00 a3 44 and t4,t2,t2
154: 1f 04 ff 5f fnop
158: 02 30 40 40 addl t1,0x1,t1
15c: 10 00 60 f4 bne t2,1a0 <wrap_help+0x1a0>

The instructions at 14c and 158 have no data dependencies with each
other or with instructions in between so can be safely swapped.

Cheers
Michael

[1] ../src/configure -v --with-pkgversion='Debian 4.6.0-8'
--with-bugurl='file:///usr/share/doc/gcc-4.6/README.Bugs'
--enable-languages=c,c++,fortran,objc,obj-c++ --prefix=/usr
--program-suffix=-4.6 --enable-shared --enable-multiarch
--with-multiarch-defaults=alpha-linux-gnu --enable-linker-build-id
--with-system-zlib --libexecdir=/usr/lib --without-included-gettext
--enable-threads=posix --with-gxx-include-dir=/usr/include/c++/4.6
--libdir=/usr/lib --enable-nls --enable-clocale=gnu
--enable-libstdcxx-debug --enable-libstdcxx-time=yes --disable-libssp
--enable-plugin --enable-objc-gc --with-long-double-128
--enable-checking=release --build=alpha-linux-gnu --host=alpha-linux-gnu
--target=alpha-linux-gnu


--
To UNSUBSCRIBE, email to debian-gcc-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4DE9B590.7070302@orcon.net.nz">http://lists.debian.org/4DE9B590.7070302@orcon.net.nz