On Mon, 7 Mar 2005 17:46:41 -0800
Nishanth Aravamudan <[EMAIL PROTECTED]> wrote:
> On Tue, Mar 08, 2005 at 02:27:20AM +0300, Evgeniy Polyakov wrote:
> > On Mon, 7 Mar 2005 14:51:21 -0800
> > Nish Aravamudan <[EMAIL PROTECTED]> wrote:
> >
> > > On Tue, 8 Mar 2005 02:14:31 +0300, Evgeniy Polyakov <[EMAIL PROTECTED]>
> > > wrote:
> > > > On Mon, 7 Mar 2005 14:40:52 -0800
> > > > Nish Aravamudan <[EMAIL PROTECTED]> wrote:
> > > >
> > > > > On Mon, 7 Mar 2005 23:37:34 +0300, Evgeniy Polyakov <[EMAIL
> > > > > PROTECTED]> wrote:
> > > > > > --- /tmp/empty/crypto_dev.c 1970-01-01 03:00:00.000000000 +0300
> > > > > > +++ ./acrypto/crypto_dev.c 2005-03-07 20:35:36.000000000 +0300
> > > > > > @@ -0,0 +1,421 @@
> > > > > > +/*
> > > > > > + * crypto_dev.c
> > > > >
> > > > > <snip>
> > > > >
> > > > > > + while (atomic_read(&__dev->refcnt)) {
> > >
> > > <snip>
> > >
> > > > > > +
> > > > > > set_current_state(TASK_UNINTERRUPTIBLE);
> > > > > > + schedule_timeout(HZ);
> > > > >
> > > > > I don't see any wait-queues in the immediate area of this code. Can
> > > > > this be an ssleep(1)?
> > > >
> > > > Yes, you are right, this loop just spins until all pending sessions
> > > > are removed from given crypto device, so it can just ssleep(1) here.
> > >
> > > Would you like me to send an incremental patch or will you be changing
> > > it yourself?
> >
> > That would be nice to see your changes in the acrypto.
> > If it will be commited...
>
> Well, here is an incremental patch, then:
>
> Description: Use ssleep() instead of schedule_timeout() to guarantee the
> task delays as expected.
>
> Signed-off-by: Nishanth Aravamudan <[EMAIL PROTECTED]>
>
Thank you, I've applied it to my tree.
> --- 2.6.11-v/acrypto/crypto_dev.c 2005-03-07 17:41:31.000000000 -0800
> +++ 2.6.11/acrypto/crypto_dev.c 2005-03-07 17:41:57.000000000 -0800
> @@ -28,6 +28,7 @@
> #include <linux/interrupt.h>
> #include <linux/spinlock.h>
> #include <linux/device.h>
> +#include <linux/delay.h>
>
> #include "acrypto.h"
>
> @@ -399,8 +400,7 @@ void crypto_device_remove(struct crypto_
> */
>
> __dev->data_ready(__dev);
> - set_current_state(TASK_UNINTERRUPTIBLE);
> - schedule_timeout(HZ);
> + ssleep(1);
> }
>
> dprintk(KERN_ERR "Crypto device %s was unregistered.\n",
Evgeniy Polyakov
Only failure makes us experts. -- Theo de Raadt
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
