bas_gigaset: suspend support (v2)
Review
Quoted 12 lines · On Tue, 13 Nov 2007 18:30:30 +0100 (CET) Tilman Schmidt <[email protected]> wrote:
From: Tilman Schmidt <[email protected]>
This patch adds basic suspend/resume support to the
bas_gigasetISDN driver for the Siemens Gigaset SX255 series of ISDN DECT bases.Only the USB aspects are handled so far; the ISDN subsystem is not notified in any way, for lack of information about how to do that. The driver will refuse to suspend if a connection is active.
...
(quoted diff) +1
-0+ if (atomic_read(&cs->hw.bas->basstate) & BS_SUSPEND) {
that's pretty peculiar. We'd only expect to see atomics being used in conjunction with atomic_add/sub/inc/etc. Here the driver is using an atomic_t as a state variable. And here's the magic bit:
spin_lock_irqsave(&ucs->lock, flags); state = atomic_read(&ucs->basstate); atomic_set(&ucs->basstate, (state & ~clear) | set); spin_unlock_irqrestore(&ucs->lock, flags);
I'm suspecting that a plain old `int' would be more appropriate here.
Am 15.11.2007 23:50 schrieb Andrew Morton:
Quoted 8 lines
Review
Quoted 3 lines
...
(quoted diff) +1
-0+ if (atomic_read(&cs->hw.bas->basstate) & BS_SUSPEND) {that's pretty peculiar. We'd only expect to see atomics being used in conjunction with atomic_add/sub/inc/etc. Here the driver is using an
atomic_tas a state variable. And here's the magic bit:
Quoted 6 lines
spin_lock_irqsave(&ucs->lock, flags); state = atomic_read(&ucs->basstate); atomic_set(&ucs->basstate, (state & ~clear) | set); spin_unlock_irqrestore(&ucs->lock, flags);I'm suspecting that a plain old `int' would be more appropriate here.
You're right. That's a prehistoric leftover. That variable was originally accessed using atomic_set_mask() and atomic_clear_mask() which are unfortunately x86 platform specific.
I'll prepare a cleanup patch.
Thanks,
Tilman
Signature · 4 lines
Tilman Schmidt E-Mail: [email protected] Bonn, Germany Diese Nachricht besteht zu 100% aus wiederverwerteten Bits. Ungeöffnet mindestens haltbar bis: (siehe Rückseite)