Problems with syscall for SH

Thiago Galesi thiagogalesi at gmail.com
Thu Mar 6 18:22:45 UTC 2008


Hello

It is not a problem compiling an application that uses syscall, but
rather, building uclibc (using buildroot.uclibc as part of toolchain
build).

Yours

Thiago Galesi

On Thu, Mar 6, 2008 at 2:37 PM, Carmelo AMOROSO <carmelo.amoroso at st.com> wrote:
>
> Carmelo Amoroso wrote:
>  > Thiago Galesi wrote:
>  >
>  >> Hello
>  >>
>  >> I am trying this with svn uclibc, and even with (moderately) old
>  >> versions the result is the same.
>  >>
>  >> I saw this problem discussed in older messages (for other archs), but
>  >> it was not solved them
>  >>
>  >> There is a conflict with the definition of syscall (in .h) and the
>  >> implementation (for SH).
>  >>
>  >> this is the .h (include/unistd.h)
>  >>
>  >> extern long int syscall (long int __sysno, ...) __THROW;
>  >>
>  >> this is the .c (libc/sysdeps/linux/sh/syscall.c)
>  >>
>  >> long syscall(long sysnum,
>  >>                          long arg1, long arg2, long arg3,
>  >>                          long arg4, long arg5, long arg6)
>  >>
>  >> Problem is, if I take the .c prototype and put it in the .h, things
>  >> will not work (eg. there are problems with compiling pthread)
>  >>
>  >> What would be the correct way to deal with this? Most archs don't have
>  >> this problem, as they have a syscall.S file, ARM uses a .c, I'll try
>  >> to figure out from there.
>  >>
>  >> Thanks
>  >>
>  >
>  > Hello Thiago,
>  > I cannot do any test just now, hopefully tomorrow
>  > I will do a check on my sh4.
>  >
>  > Carmelo
>  >
>  Hello,
>  I did a simple test call syscall(100, 1, 2,3,4) and I did not get any
>  problem at compile
>  time.. where is the conflict you have ? can you post some error messages ?
>
>  Carmelo
>  > _______________________________________________
>  > uClibc mailing list
>  > uClibc at uclibc.org
>  > http://busybox.net/cgi-bin/mailman/listinfo/uclibc
>  >
>  >
>
>



-- 
-
Thiago Galesi



More information about the uClibc mailing list