Hi,
x32 is an experimental setup. I remember x32 for dietlibc needs some directory
/ symlink naming workaround that was too dirty to just commit at the time I
played with it.
I’m not the biggest fan of the mess they call Perl - so patches welcome ;-) For
my mips builds it built last night after I updated and tweaked the .conf a bit.
René
On Jun 22, 2016, at 10:56, Barry Kauler <[email protected]> wrote:
> Rene,
> SVN 44898, generic minimal target, x86_64 host, target x86_64 with
> x32, cross-compile, still fails at dietlibc, for the same reasons.
>
> That was building on my Quirky/Puppy "Werewolf" which is built with Ubuntu
> DEBs.
>
> Trying the same on my Quirky "April" which is built with binary pkgs
> compiled in T2, this time dietlibc at stage1 succeeded, but failed at
> perl.
>
> I tried the older perl, from SVN 44859, but failed for same reason.
>
> The ERROR-LOG has a lot of "ignored" errors, like this one:
>
> makefile:671: recipe for target '_cleaner2' failed
> make[1]: [_cleaner2] Error 1 (ignored)
>
> Right at the end of ERROR-LOG is this:
>
> makefile:641: recipe for target '_mopup' failed
> make: [_mopup] Error 1 (ignored)
> rm -f perl.exp ext.libs pod/perltoc.pod extra.pods
> pod/perl5240delta.pod pod/perlapi.pod pod/perlintern.pod
> pod/perlmodlib.pod pod/perluniprops.pod uni.data opmini.o perlmini.o
> pod/roffitall
> rm -f perl.export perl.dll perl.libexp perl.map perl.def
> rm -f *perl.xok
> rm -f cygwin.c libperl*.def libperl*.dll cygperl*.dll *.exe.stackdump
> rm -f perl miniperl libperl.a libperl.* microperl
> rm -f config.arch config.over
> Note that 'make realclean' does not delete config.sh or Policy.sh
> make: *** No rule to make target 'config.h'. Stop.
> Due to previous errors, no 1-perl.log file!
> (Try enabling xtrace in the config to track an error inside the build system.)
> --- BUILD ERROR ---
> Creating file list and doing final adaptions ...
> Searching for orphaned files ...
> Found 5 files for this package.
> Found 0 orphaned files for this package.
> Clear (old) md5sums ...
> Creating md5sum files ... done.
> Creating package description ...
> Making post-install adaptions.
>
> Regards,
> Barry
>
>
>
>
> On 6/22/16, René Rebe <[email protected]> wrote:
>> Hey,
>>
>> I spent some extra nighttime to tweak some more stuff to cross build again.
>> Like perl and texinfo.
>>
>> A minimal target now (mostly) cross builds again - tested to mips and
>> mips64. The later sort-of test booted on a Sgi Octane - yay!
>> (I cheated a bit by not rebuilding the kernel though -
>> https://www.instagram.com/p/BG4qjXtgx99/)
>> (and I ignored some missing dietlibc/udev symbol, … and also grub2 on mips,
>> … looked a bit Loongson-only)
>>
>> Cheers,
>> René
>>
>> On Jun 21, 2016, at 18:50, René Rebe <[email protected]> wrote:
>>
>>> Hey,
>>>
>>> sorry for the late replay, busy lately.
>>>
>>> Yes, compiler and packages and such should not read host files (except
>>> when they build natively run helper tools obviously).
>>>
>>> However, there is no fancy sandbox or so that stops packagers to
>>> accidentally doing so.
>>> This is why cross compiled phase requires on proper guidance with
>>> configure and Makefile switchers, patches etc. to guide packages to cross
>>> compile proper.
>>>
>>> I do not remember seeing such error, and recently cross compiled trunk on
>>> a relative fresh trunk installation on an amd64 to mips64 and aside minor
>>> hiccups which I patched did not encounter major problems, ...
>>>
>>> René
>>>
>>> On Jun 6, 2016, at 15:37, Barry Kauler <[email protected]> wrote:
>>>
>>>>
>>>> It is reading the header file in my host system, should it be doing
>>>> that in stage1?
>>>>
>>>> Anyway, I inserted this into /usr/include/x86_64-linux-gnu/sys/cdefs.h:
>>>>
>>>> --------------------------
>>>> /* Convenience macros to test the version of gcc. */
>>>> #undef __GNUC_PREREQ
>>>> #if defined __GNUC__ && defined __GNUC_MINOR__
>>>> # define __GNUC_PREREQ(maj, min) \
>>>> ((__GNUC__ << 16) + __GNUC_MINOR__ >= ((maj) << 16) + (min))
>>>> #else
>>>> # define __GNUC_PREREQ(maj, min) 0
>>>> #endif
>>>> ---------------------------
>>>>
>>>> Which seems to have fixed it, but then it has stopped on another error
>>>> in another header file in my host system.
>>>>
>>>>
>>>> On 6/6/16, Barry Kauler <[email protected]> wrote:
>>>>> Continuing the "generic" x32 build, dietlibc has failed at stage1.
>>>>>
>>>>> ---------------------
>>>>> x86_64-t2-linux-gnux32-cc -I. -isystem include -Os -fstrict-aliasing
>>>>> -momit-leaf-frame-pointer -mfancy-math-387 -W -Wall -Wextra
>>>>> -Wchar-subscripts -Wmissing-prototypes -Wmissing-declarations
>>>>> -Wno-switch -Wno-unused -Wredundant-decls -c dyn_start.c -o
>>>>> bin-x86_64/dyn_start.o -D__dietlibc__
>>>>> In file included from dyn_start.c:4:0:
>>>>> /usr/include/x86_64-linux-gnu/sys/cdefs.h:41:20: error: missing binary
>>>>> operator before token "("
>>>>> # if __GNUC_PREREQ (4, 6) && !defined _LIBC
>>>>> ---------------------
>>>>>
>>>>> Rene,
>>>>> does that error mean anything to you? It is that "(" it is objecting
>>>>> to.
>>>>>
>>>>> Regards,
>>>>> Barry
>>>>>
>>>>
>>>>
>>>>
>>>> -----------------------------------------------------------
>>>> If you wish to unsubscribe from this mailing, send mail to
>>>> [email protected] with a subject of: unsubscribe t2
>>>
>>> --
>>> ExactCODE GmbH, Lietzenburger Str. 42, DE-10789 Berlin
>>> http://exactcode.com | http://exactscan.com | http://ocrkit.com |
>>> http://t2-project.org | http://rene.rebe.de
>>>
>>> -----------------------------------------------------------
>>> If you wish to unsubscribe from this mailing, send mail to
>>> [email protected] with a subject of: unsubscribe t2
>>
>> --
>> ExactCODE GmbH, Lietzenburger Str. 42, DE-10789 Berlin
>> DE Legal: Amtsgericht Berlin (Charlottenburg) HRB 105123B, Tax-ID#:
>> DE251602478
>> Managing Director: René Rebe
>> http://exactcode.com | http://exactscan.com | http://ocrkit.com |
>> http://t2-project.org | http://rene.rebe.de
>>
>>
>
> -----------------------------------------------------------
> If you wish to unsubscribe from this mailing, send mail to
> [email protected] with a subject of: unsubscribe t2
--
ExactCODE GmbH, Lietzenburger Str. 42, DE-10789 Berlin
DE Legal: Amtsgericht Berlin (Charlottenburg) HRB 105123B, Tax-ID#: DE251602478
Managing Director: René Rebe
http://exactcode.com | http://exactscan.com | http://ocrkit.com |
http://t2-project.org | http://rene.rebe.de
-----------------------------------------------------------
If you wish to unsubscribe from this mailing, send mail to
[email protected] with a subject of: unsubscribe t2
