Re: Any hope for ide-scsi (error handling)?

[dan carpenter]

On Saturday 15 March 2003 10:01 pm, Willem Riede wrote:
> It may not be elegant to schedule(1) with the lock taken, but it
> does work.
>
> However, my latest patch doesn't seem to be applied, since in my
> version I have a set_current_state(TASK_INTERRUPTIBLE); before
> the schedule.

I don't see how it works.

spin_lock_irqsave() increments  preempt_count()
in_atomic checks is defined as:
# define in_atomic()    ((preempt_count() & ~PREEMPT_ACTIVE) != kernel_locked())
kernel_locked() is defined as:
#define kernel_locked()         (current->lock_depth >= 0)

If you call schedule while in_atomic() then it prints out the error 
"bad: scheduling while atomic!\n".  

As far as I can see set_current_state() doesn't affect
preempt_count() or current->lock_depth.  I must be 
missing something...

regards,
dan carpenter