Re: stage1 gcc bootstrap fails looking for bits/libc-header-start.h

[Dennis Clarke <dclarke@blastwave.org>]

On 5/10/22 03:00, Jonathan Wakely wrote:
> On Tue, 10 May 2022 at 01:54, Dennis Clarke <dclarke@blastwave.org> wrote:
>>
>> On 5/9/22 14:51, Jonathan Wakely wrote:
>>> On Mon, 9 May 2022, 18:16 Dennis Clarke via Gcc-help, <gcc-help@gcc.gnu.org>
>>> wrote:
>>>
>>>>
>>>> Not sure what is causing this on a Linux armv7 32-bit machine but I have
>>>> libc6-dev installed and that provides :
>>>>
>>>> # ls -lapb /usr/include/arm-linux-gnueabihf/bits/libc-header-start.h
>>>>
>>>
>>>
>>> But this is not a standard header location, so GCC won't look there.
>>>
>>> You need to use --enable-multiarch to use the Debian MultiArch directory
>>> layout.
>>>
>>
>> Sadly things have not worked out any better.   :(
>>
>> Configure still looks nice :
>>
>> io$
>> io$ uname -a
>> Linux io 4.4.194 #52 SMP Tue Oct 26 08:35:58 UTC 2021 armv7l GNU/Linux
>> io$ gcc --version
>> gcc (Debian 8.3.0-6) 8.3.0
>> Copyright (C) 2018 Free Software Foundation, Inc.
>> This is free software; see the source for copying conditions.  There is NO
>> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
>>
>> io$
>>
>> All the usual baseline tools are around. Also gmp/mpfr/mpc are extracted
>> in the source tree as per normal.
>>
>> io$ CC='gcc -march=armv7-a -mtune=cortex-a17 -mfpu=vfpv4-d16 -marm' \
>>   > CXX='g++ -march=armv7-a -mtune=cortex-a17 -mfpu=vfpv4-d16 -marm' \
>>   > LD_RUN_PATH='/opt/gcc/imed/gcc12/lib:/opt/bw/lib' \
>>   > ../gcc-12.1.0/configure --prefix=/opt/gcc/imed/gcc12 \
>>   > --enable-multiarch \
>>   > --disable-nls --enable-threads=posix --enable-shared \
>>   > --with-gmp=/opt/bw --with-mpfr=/opt/bw --with-mpc=/opt/bw \
> 
> Wait, what? You said "gmp/mpfr/mpc are extracted in the source tree as
> per normal" so why are you telling GCC to get them from /opt/bw?
> 

I have gone both directions and used the sources for those in tree as
well as included and it seems to have nothing at all to do with the real
problem.

> Xi Ruoyao is sending replies that you're apparently not receiving, see
> https://gcc.gnu.org/pipermail/gcc-help/2022-May/141536.html

I think my mail server has gone into full rogue mode with spamassassin
  and that is the issue.

In any case I see :

io$
io$  /opt/bw/build/gcc-12.1.0_debian_10_armv7l_imed.004/./gcc/xgcc 
-print-multiarch
arm-linux-gnueabi
io$





-- 
Dennis Clarke
RISC-V/SPARC/PPC/ARM/CISC
UNIX and Linux spoken
GreyBeard and suspenders optional