[Devel] Re: [PATCH 1/3] lutimesat: simplify utime(2)

Alexey Dobriyan adobriyan at gmail.com
Sun Jan 28 07:24:59 PST 2007


On Fri, Jan 26, 2007 at 12:41:20PM -0800, Andrew Morton wrote:
> On Fri, 26 Jan 2007 14:21:42 +0300
> Alexey Dobriyan <adobriyan at openvz.org> wrote:
> 
> > Rewrite via do_utimes() like compat_sys_utime().
> 
> I'm somewhat surprised that this wasn't done earlier.

Because, the following patch didn't hit -mm. :)

	From adobriyan at gmail.com Wed Oct 25 20:32:24 2006
	Date: Wed, 25 Oct 2006 20:32:24 +0400
	From: Alexey Dobriyan <adobriyan at gmail.com>
	To: Andrew Morton <akpm at osdl.org>
	Cc: linux-kernel at vger.kernel.org
	Subject: [PATCH] Shrink sys_utime()
	Message-ID: <20061025163224.GA5356 at martell.zuzino.mipt.ru>
	Mime-Version: 1.0
	Content-Type: text/plain; charset=us-ascii
	Content-Disposition: inline
	User-Agent: Mutt/1.5.11
	Status: RO
	Content-Length: 3314
	Lines: 118

	All checks in sys_utime() and do_utimes() are duplicated as well as a
	comment. sys_utime() will now use do_utimes() after getting times from
	userspace and projecting them to struct timeval [2].

Nevermind.

> I wonder if there's some subtle reason why this won't work.

I don't know. Compat syscall -- I'm not touching it. Normal syscall --
time_t is long on all archs, suseconds_t is sometimes int, but we're
putting zero there.

> How well tested is this?

It passed utime tests in December's LTP.




More information about the Devel mailing list