Alan wrote:

>On Sat, 2003-03-01 at 21:05, Dr. David Alan Gilbert wrote:
>> Hi,
>>   2.5.63 had a good go at trying to boot for me; the only error during
>> boot was 'Debug: sleeping function called from illegal context at
>> mm/slab.c:1617' during the IDE startup.
>
>Known problem. Its a bug in the request_irq code on x86. IDE just
>happens to be a victim of it.
>  
>
I would call it an IDE bug: request_irq is not atomic, there are dozends 
of kmalloc(GFP_KERNEL) calls in the implementation.
And on some archs it might be impossible to implement it without 
sleeping - e.g. on a NUMA arch an IPI to the right node could be necessary.
What about fixing ide? If ide can't handle early interrupts, then it can use

    disable_irq();
    request_irq();
    spin_lock_irq(&ide_lock);
    do_setup();
    spin_unlock_irq(&ide_lock);
    enable_irq();

I've attached a proposal - what do you think?
--
    Manfred