This is the mail archive of the
libc-ports@sources.redhat.com
mailing list for the libc-ports project.
Re: [PATCH] Fix some system calls with long long arguments
- From: Atsushi Nemoto <anemo at mba dot ocn dot ne dot jp>
- To: linux-mips at linux-mips dot org
- Cc: ralf at linux-mips dot org, kraj at mvista dot com, libc-ports at sourceware dot org
- Date: Thu, 15 Mar 2007 10:35:11 +0900 (JST)
- Subject: Re: [PATCH] Fix some system calls with long long arguments
- References: <20070307.003931.25235381.anemo@mba.ocn.ne.jp> <20070307.231410.15268922.anemo@mba.ocn.ne.jp> <20070309.003749.39154822.anemo@mba.ocn.ne.jp>
On Fri, 09 Mar 2007 00:37:49 +0900 (JST), Atsushi Nemoto <anemo@mba.ocn.ne.jp> wrote:
> > fadvise64(), readahead(), sync_file_range() have long long argument(s)
> > but glibc passes it by hi/lo pair without padding, on both O32 and
> > N32.
> >
> > Also wire up fadvise64_64() and fixup confusion of it with
> > fadvise64().
>
> If best performance was preferred, the O32 readahead and
> sync_file_range should not changed and libc should provide MIPS
> specific syscall wrappers, like pread64. The N32 can also use
> standard sys_readahead(), etc. and libc should provide wrappers, too.
>
> Anyway fadvice64() needs to be fixed.
>
> Any comments from libc side? Original patch is here:
> http://www.linux-mips.org/archives/linux-mips/2007-03/msg00092.html
Any comments?
I think this patch has less maintainance cost but a little bit slow.
These syscalls can be a little bit faster, but needs more works on
glibc (and uClibc, etc.) side.
Anyway we should take some action while current implementation is
broken (except N64).
Which is a way to go?
---
Atsushi Nemoto