Re: glibc-2.3.6 build error

[abhijeet tripathi <abhijeet dot tripathy at yahoo dot co dot in>]

Hi Carlos,

--- On Tue, 1/12/09, Carlos O'Donell <carlos@systemhalted.org> wrote:

> From: Carlos O'Donell <carlos@systemhalted.org>
> Subject: Re: glibc-2.3.6 build error
> To: "abhijeet tripathi" <abhijeet.tripathy@yahoo.co.in>
> Cc: "libc-help@sourceware.org" <libc-help@sourceware.org>, "abhijeet.tripathy@gmail.com" <abhijeet.tripathy@gmail.com>
> Date: Tuesday, 1 December, 2009, 9:30 PM
> On Tue, Dec 1, 2009 at 4:47 AM,
> abhijeet tripathi
> <abhijeet.tripathy@yahoo.co.in>
> wrote:
> > The issue what I am facing has been fixed for uclibc
> but I coulkd not see a similar fix for glibc. As I have
> selected ARM_EABI, so some of the syscalls are undefined but
> they are not sync with glibc.
> >
> > Fix for uclibc is at following link:
> > http://lists.uclibc.org/pipermail/uclibc/2006-October/037412.html
> >
> > Please correct me if I am wrong.
> 
> If you are building an ARM GNU/Linux EABI compatible
> toolchain (ARM
> EABI is generally used to talk about bare-metal firmware
> toolchains)
> then you must use a kernel newer than 2.6.16, and
> therefore
> __ASSUME_NEW_GETRLIMIT_SYSCALL should have been defined by
> the arm
> kernel-features.h, and the above code in your build failure
> should
> never get called, instead the new ugetrlimit syscall should
> have been
> used.
> 
> Given that explanation, you need to track down what is
> wrong, either
> the kernel is too old, or the arm kernel-features.h is
> wrong, or
> getrlimit.c is not using the new syscall.
> 
> Cheers,
> Carlos.
> 

Thank you for the explanation. I am using 2.6.28 kernel, so the problem is with glibc. I need to dig further to see what actually is going wrong.

--regards,
Abhijeet


      The INTERNET now has a personality. YOURS! See your Yahoo! Homepage. http://in.yahoo.com/