Re: 2.6.19-rc1 genirq causes either boot hang or "do_IRQ: cannot handle IRQ -1"

[Andrew Vasquez]

On Fri, 06 Oct 2006, Andrew Morton wrote:

> On Fri, 6 Oct 2006 12:00:39 -0700
> Andrew Vasquez <andrew.vasquez@xxxxxxxxxx> wrote:
> 
> > [   27.510539] Booting processor 1/2 APIC 0x1
> > [   27.514684] Unable to handle kernel NULL pointer dereference at 0000000000000088 RIP: 
> > [   27.520204]  [<ffffffff80225fb0>] profile_tick+0x40/0x90
> > [   27.528118] PGD 0 
> > [   27.530222] Oops: 0000 [1] SMP 
> > [   27.533505] CPU 0 
> > [   27.535610] Modules linked in:
> > [   27.538755] Pid: 1, comm: swapper Not tainted 2.6.19-rc1 #5
> > [   27.544367] RIP: 0010:[<ffffffff80225fb0>]  [<ffffffff80225fb0>] profile_tick+0x40/0x90
> > [   27.552483] RSP: 0000:ffffffff8059ff78  EFLAGS: 00010046
> > [   27.557842] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000001
> > [   27.565024] RDX: ffff810081a77f40 RSI: 0000000000000046 RDI: 0000000000000001
> > [   27.572203] RBP: 0000000000000001 R08: 0000000000000000 R09: 0000000000000007
> 
> 
> hm, we seem to have broken x86_64 completely.
> 
> smp_apic_timer_interrupt() needs to do
> 
> 	struct pt_regs *old_regs = set_irq_regs(regs);
> 
> on entry and
> 
> 	set_irq_regs(old_regs);
> 
> on exit.
> 
> But it doesn't get passed the pt_regs*
> 
> >From my reading of `macro apicinterrupt' in arch/x86_64/kernel/entry.S,
> smp_apic_timer_interrupt() actually _does_ get passed the pt_reg*, only it
> doesn't declare it.  I think - Andi would need to confirm.
> 
> If I'm right...
> 
> 
> diff -puN arch/x86_64/kernel/apic.c~x86_64-irq_regs-fix arch/x86_64/kernel/apic.c
> --- a/arch/x86_64/kernel/apic.c~x86_64-irq_regs-fix
> +++ a/arch/x86_64/kernel/apic.c
> @@ -913,8 +913,10 @@ void smp_local_timer_interrupt(void)
>   * [ if a single-CPU system runs an SMP kernel then we call the local
>   *   interrupt as well. Thus we cannot inline the local irq ... ]
>   */
> -void smp_apic_timer_interrupt(void)
> +void smp_apic_timer_interrupt(struct pt_regs *regs)
>  {
> +	struct pt_regs *old_regs = set_irq_regs(regs);
> +
>  	/*
>  	 * the NMI deadlock-detector uses this.
>  	 */
> @@ -934,6 +936,7 @@ void smp_apic_timer_interrupt(void)
>  	irq_enter();
>  	smp_local_timer_interrupt();
>  	irq_exit();
> +	set_irq_regs(old_regs);
>  }
>  
>  /*

Patch appears to work.

At least I can now boot my x86_64 box.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/