[Clfs-support] glibc-2.19 -m32 build error clfs-multilib

[William Harrington]

On Thu, January 29, 2015 23:44, lux-integ wrote:
> On Thursday 29 January 2015 22:51:35 William Harrington wrote:
>>  don't see much help. Have you attempted to build it without scripts.
> yes I have copied and pasted with the same result
>
> (And just to make sure I went just now did a run of copying and pasting
> from
> the   CLFS book  and I have   atttached  the result
>
>
> Similar build logs have been    repeatd   on several runs on different
> hosts
> via scripting and on copy and paste..  I have tried these:-
> --3 different machines with amd64 cpus
> --running clfs(pure64) , fedora and slackware
> And all gave the same result.
>
>
>
> AND As I said in my last email   the only report of this  was  in 2010
> here
> https://sourceware.org/bugzilla/show_bug.cgi?id=11245
> ( using a similar method to build glibc )
> It was tracked down to the need for kernel headers before linux-2.6.32.  I
> should hope that   for glibc2.19 linux-2.6.31 headers are not needed.
>
> ---
>
>
>
> ( If you have built clfsmultilib as per the recipe  given   here
> http://clfs.org/view/git/x86_64/
> could you let me have your build logs  please ?
>
>  AND   I will have to wonder if all the machines I have tried  have been
> hackd/compromised/disabled    somehow??)



Hello,

I could go through a quick multilib build with jhalfs. The last modified
CLFS build I did using our future toolchain upgrades worked fine with
Glibc 2.20 and linux 3.18 kernel headers, so not sure what is going on
over there. I don't know if it has something to do with how your build
system is laid out, or not.

I haven't ran across that error, as far as I know, since I started working
with CLFS in 2007.

You can compare your build logs with what I did last summer which uses the
same versions as you are using:

http://clfs.org/testsuite-logs/git/x86_64/Core-i7-2700k/

Also, reply to all since your messages aren't going to the clfs mailing list.

Sincerely,

William Harrington