public inbox for gentoo-user@lists.gentoo.org
 help / color / mirror / Atom feed
* [gentoo-user] Re: glibc-2.17 fails and warning: setlocale: LC_ALL error
  2014-07-16 17:21                       ` Dale
@ 2014-07-17  0:39                         ` walt
  2014-07-17  1:01                           ` Dale
  0 siblings, 1 reply; 3+ messages in thread
From: walt @ 2014-07-17  0:39 UTC (permalink / raw
  To: gentoo-user

On 07/16/2014 10:21 AM, Dale wrote:
> I'm beginning to wonder about
> those stage3 tarballs for x86.  Do they test those from time to time to
> make sure they work or do they just autobuild them and upload them?  I'm
> following the install guide so I'm pretty sure I'm not doing anything
> wrong but at the same time, I don't want to file a bug when it is just me. 

Over the years I remember maybe three times I've had to untar a stage3 on
top of a "working" system, and IIRC the catastrophic problem was glibc each
time :/

Anyway, the first time I tried it I learned that the stage3 overwrites the
existing /etc directory. (That was so painful I remembered to move /etc out
of the way the next two times it did it :)

I mention all this because /etc/locale.conf may not be the file you expect,
based on how your compile environment and chroot stuff is configured, etc.

I been there/done that :)





^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [gentoo-user] Re: glibc-2.17 fails and warning: setlocale: LC_ALL error
  2014-07-17  0:39                         ` [gentoo-user] " walt
@ 2014-07-17  1:01                           ` Dale
  0 siblings, 0 replies; 3+ messages in thread
From: Dale @ 2014-07-17  1:01 UTC (permalink / raw
  To: gentoo-user

walt wrote:
> On 07/16/2014 10:21 AM, Dale wrote:
>> I'm beginning to wonder about
>> those stage3 tarballs for x86.  Do they test those from time to time to
>> make sure they work or do they just autobuild them and upload them?  I'm
>> following the install guide so I'm pretty sure I'm not doing anything
>> wrong but at the same time, I don't want to file a bug when it is just me. 
> Over the years I remember maybe three times I've had to untar a stage3 on
> top of a "working" system, and IIRC the catastrophic problem was glibc each
> time :/
>
> Anyway, the first time I tried it I learned that the stage3 overwrites the
> existing /etc directory. (That was so painful I remembered to move /etc out
> of the way the next two times it did it :)
>
> I mention all this because /etc/locale.conf may not be the file you expect,
> based on how your compile environment and chroot stuff is configured, etc.
>
> I been there/done that :)
>

I haven't done this on top of a working system tho.  The first tarball
was a fresh install.  The Gentoo that was there hadn't been updated in
like 3 years or so.  I wasn't about to jump in and try to update all
that.  The 2nd time, I just unpacked the tarball in /mnt/gentoo just
like I would a fresh install just that it was in the first install. 

The locale.gen file was untouched from the tarball so it was the wrong
file type when I downloaded it.  No clue on that yet. 

Dale

:-)  :-) 


^ permalink raw reply	[flat|nested] 3+ messages in thread

* [gentoo-user] Re: glibc-2.17 fails and warning: setlocale: LC_ALL error
@ 2014-08-05 23:48 ((RayDar))
  0 siblings, 0 replies; 3+ messages in thread
From: ((RayDar)) @ 2014-08-05 23:48 UTC (permalink / raw
  To: gentoo-user

Hello,

I think I can offer some some more clues to the problem in thread started by 
Dale, as archived here:
https://www.mail-archive.com/gentoo-user@lists.gentoo.org/msg147315.html

I too get the "Your old kernel is broken.  You need to update it to a newer 
version as syscall(<bignum>) will break." message whenever I try to emerge 
glibc-2.17. This happens on all three of my machines, which are kept up to 
date and have slightly different configurations. It is likely that many people 
are in a broken state, but they don't know it because they haven't tried re-
emerging glibc.

Previously in this thread, I see that Hinnerk van Bruinehsen asked:

> Can you run;

 > printf "#include <unistd.h>\n#include <sys/syscall.h>\nint main(){return \
 > syscall(1000)!=-1;}\n" > syscall.c && make syscall && ./syscall && echo $? 

When I try this on each of my systems, I get the same output:

cc     syscall.c   -o syscall

Notice that the line with the 0 is missing.

I found a work around: move to an earlier kernel. This problem happens for me 
on kernel-3.14.14 and for Dale on kernel-3.15.5. But, it DOESN'T happen for me 
on 3.12.21-r1.

When I boot into the 3.12.21-r1 kernel, the printf() above outputs:

cc     syscall.c   -o syscall
0

as it should.

Could the emerge glibc fail be due to a required kernel configuration option 
that didn't make it into 3.14.14 or above?



^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2014-08-05 23:46 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-08-05 23:48 [gentoo-user] Re: glibc-2.17 fails and warning: setlocale: LC_ALL error ((RayDar))
  -- strict thread matches above, loose matches on Subject: below --
2014-07-15  1:50 [gentoo-user] " Dale
2014-07-15  5:33 ` Dale
2014-07-15  7:04   ` Hinnerk van Bruinehsen
2014-07-15  7:40     ` Dale
2014-07-15  8:15       ` Hinnerk van Bruinehsen
2014-07-15  8:40         ` Hinnerk van Bruinehsen
2014-07-15  9:25           ` Dale
2014-07-16  7:32             ` Dale
2014-07-16  7:46               ` Helmut Jarausch
2014-07-16  8:43                 ` Dale
2014-07-16  9:19                   ` Dale
2014-07-16 10:33                     ` J. Roeleveld
2014-07-16 17:21                       ` Dale
2014-07-17  0:39                         ` [gentoo-user] " walt
2014-07-17  1:01                           ` Dale

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox