* Re: [patch 2/2] Remove a few inlines in fs/namei.c
[not found] ` <20080729155527.799486507@free-electrons.com>
@ 2008-07-29 18:55 ` Al Viro
2008-07-29 19:33 ` Andrew Morton
0 siblings, 1 reply; 2+ messages in thread
From: Al Viro @ 2008-07-29 18:55 UTC (permalink / raw)
To: Thomas Petazzoni; +Cc: linux-kernel, linux-fsdevel, akpm
On Tue, Jul 29, 2008 at 05:52:31PM +0200, Thomas Petazzoni wrote:
> Uninline may_create() and do_follow_link() and use vfs_follow_link()
> instead of __vfs_follow_link() in __do_follow_link(). It allows to
> slightly reduce the size of kernel code:
>
> text data bss dec hex filename
> 1114299 118564 212992 1445855 160fdf vmlinux.old
> 1112971 118564 212992 1444527 160aaf vmlinux
> -1328 0 0 -1328 -530 +/-
>
> This patch is part of the Linux Tiny project and has been originally
> written by Matt Mackall <mpm@selenic.com>
Absolutely not. do_follow_link() et.al. are parts of mutual recursion
loop and we need to reduce the use of kernel *stack*.
^ permalink raw reply [flat|nested] 2+ messages in thread* Re: [patch 2/2] Remove a few inlines in fs/namei.c
2008-07-29 18:55 ` [patch 2/2] Remove a few inlines in fs/namei.c Al Viro
@ 2008-07-29 19:33 ` Andrew Morton
0 siblings, 0 replies; 2+ messages in thread
From: Andrew Morton @ 2008-07-29 19:33 UTC (permalink / raw)
To: Al Viro; +Cc: Thomas Petazzoni, linux-kernel, linux-fsdevel
On Tue, 29 Jul 2008 19:55:37 +0100 Al Viro <viro@ZenIV.linux.org.uk> wrote:
> On Tue, Jul 29, 2008 at 05:52:31PM +0200, Thomas Petazzoni wrote:
> > Uninline may_create() and do_follow_link() and use vfs_follow_link()
> > instead of __vfs_follow_link() in __do_follow_link(). It allows to
> > slightly reduce the size of kernel code:
> >
> > text data bss dec hex filename
> > 1114299 118564 212992 1445855 160fdf vmlinux.old
> > 1112971 118564 212992 1444527 160aaf vmlinux
> > -1328 0 0 -1328 -530 +/-
That's a really big difference, considering whereabouts in the
kernel this is.
> > This patch is part of the Linux Tiny project and has been originally
> > written by Matt Mackall <mpm@selenic.com>
>
> Absolutely not. do_follow_link() et.al. are parts of mutual recursion
> loop and we need to reduce the use of kernel *stack*.
Last time someone tried to "fix" this I suggested that we add some
comments explaining the design - why it is the way it is. And the
time before that.
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2008-07-29 19:33 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <20080729155229.489904417@free-electrons.com>
[not found] ` <20080729155527.799486507@free-electrons.com>
2008-07-29 18:55 ` [patch 2/2] Remove a few inlines in fs/namei.c Al Viro
2008-07-29 19:33 ` Andrew Morton
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).