lkml.org 
[lkml]   [2002]   [Jun]   [25]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [PATCH][RFC] IDE locking #2
From
Date
Zwane Mwaikambo <zwane@linux.realnet.co.sz> writes:

> Hi Martin, Jens
>
> This patch gets a little further and is independent of the first one.
> However it is still not there yet.
>
> This patch tries to address the following;
>
> CURRENT
> o remove racy spin_unlock_irq() ... foo() .. spin_lock_irq() scenarios
> o moves locking from interrupt handling helpers/primitives
> (ide-disk.c:*_intr) to the caller.
> o removes ide__sti() from some areas, personally i think this is an evil
> macro for many reasons, is it supposed to be disable_irq(ide_irq);
> __sti(); ?
>
> o fixes some 'FIXME' entries... adds a couple more ;)
>
> TODO
> o fix breakage in ide-floppy/tape etc...
> o fix ide device setup/tear down locking
> o remove ide_wait commands from interrupt paths
>
> Index: linux-2.5.24/drivers/ide//ide-cd.c
> ===================================================================
> RCS file: /build/cvsroot/linux-2.5.24/drivers/ide/ide-cd.c,v
> retrieving revision 1.1.1.1
> diff -u -r1.1.1.1 ide-cd.c
> --- linux-2.5.24/drivers/ide//ide-cd.c 23 Jun 2002 13:32:22 -0000 1.1.1.1
> +++ linux-2.5.24/drivers/ide//ide-cd.c 23 Jun 2002 20:45:52 -0000
> @@ -319,6 +319,8 @@
>
> /****************************************************************************
> * Generic packet command support and error handling routines.
> + * Note. these are called with the channel lock held and irq disabled in most
> + * cases.
> */
>
> /* Mark that we've seen a media change, and invalidate our internal
> @@ -728,13 +730,13 @@
> int xferlen,
> ata_handler_t handler)
> {
> - unsigned long flags;
> struct ata_channel *ch = drive->channel;
> ide_startstop_t startstop;
> struct cdrom_info *info = drive->driver_data;
> int ret;
>
> - spin_lock_irqsave(ch->lock, flags);
> + BUG_ON(!spin_is_locked(ch->lock));

spin_is_locked() returns 0 always on the UP system.
Instead, the following macro may be useful.

#ifdef CONFIG_SMP
#define assert_spin_is_locked(lock) BUG_ON(!spin_is_locked(lock))
#else
#define assert_spin_is_locked(lock) do {} while(0)
#endif
--
OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:22    [W:0.065 / U:1.080 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site