Re: [PATCH 2/2] kdump shutdown hook support

[Michael Ellerman <michael@ellerman.id.au>]

[-- Attachment #1: Type: text/plain, Size: 4958 bytes --]

On Wed, 2007-12-12 at 16:45 +1100, Michael Neuling wrote:
> This adds hooks into the default_machine_crash_shutdown so drivers can
> register a function to be run in the first kernel before we hand off
> to the second kernel.  This should only be used in exceptional
> circumstances, like where the device can't be reset in the second
> kernel alone (as is the case with eHEA).  To emphasize this, the
> number of handles allowed to be registered is currently #def to 1.
> 
> This uses the setjmp/longjmp code to call out to the registered hooks,
> so any bogus exceptions we encounter will hopefully be recoverable.  
> 
> I've tested with bogus data and instruction exceptions.
> 
> Signed-off-by: Michael Neuling <mikey@neuling.org>
> ---
> 
>  arch/powerpc/kernel/crash.c |   90 +++++++++++++++++++++++++++++++++++++++++---
>  include/asm-powerpc/kexec.h |    3 +
>  2 files changed, 88 insertions(+), 5 deletions(-)
> 
> Index: linux-2.6-ozlabs/arch/powerpc/kernel/crash.c
> ===================================================================
> --- linux-2.6-ozlabs.orig/arch/powerpc/kernel/crash.c
> +++ linux-2.6-ozlabs/arch/powerpc/kernel/crash.c
> @@ -32,6 +32,8 @@
>  #include <asm/lmb.h>
>  #include <asm/firmware.h>
>  #include <asm/smp.h>
> +#include <asm/system.h>
> +#include <asm/setjmp.h>
>  
>  #ifdef DEBUG
>  #include <asm/udbg.h>
> @@ -45,6 +47,11 @@ int crashing_cpu = -1;
>  static cpumask_t cpus_in_crash = CPU_MASK_NONE;
>  cpumask_t cpus_in_sr = CPU_MASK_NONE;
>  
> +#define CRASH_SHUTDOWN_HANDLES_NUM 1

CRASH_HANDLER_MAX ?

> +/* NULL terminated list of shutdown handles */
> +static crash_shutdown_t crash_shutdown_handles[CRASH_SHUTDOWN_HANDLES_NUM+1];
> +static DEFINE_SPINLOCK(crash_handles_lock);
> +
>  #ifdef CONFIG_SMP
>  static atomic_t enter_on_soft_reset = ATOMIC_INIT(0);
>  
> @@ -285,9 +292,69 @@ static inline void crash_kexec_stop_spus
>  }
>  #endif /* CONFIG_SPU_BASE */
>  
> +/* 
> + * Register a function to be called on shutdown.  Only use this if you
> + * can't reset your device in the second kernel.
> + */
> +int crash_shutdown_register(crash_shutdown_t handler)
> +{
> +	unsigned int i, rc;
> +
> +	spin_lock(&crash_handles_lock);

> +	for(i = 0 ; i <= CRASH_SHUTDOWN_HANDLES_NUM; i++) {
> +		if (!crash_shutdown_handles[i]) {
> +			/* Insert handle at end */
> +			crash_shutdown_handles[i] = handler;
> +			rc = 0;
> +			break;
> +		}
> +	}
> 
> +	if (i == CRASH_SHUTDOWN_HANDLES_NUM){
> +		printk(KERN_ERR "Crash shutdown handles full, "
> +		       "not registered.\n");
> +		rc = 1;
> +	}
> +
> +	spin_unlock(&crash_handles_lock);

Perhaps?

> +	return rc;
> +}
> +EXPORT_SYMBOL(crash_shutdown_register);
> +
> +int crash_shutdown_unregister(crash_shutdown_t handler)
> +{
> +	unsigned int i;
> +
> +	spin_lock(&crash_handles_lock);
> +	for(i = 0 ; i <= CRASH_SHUTDOWN_HANDLES_NUM; i++)
> +		if (crash_shutdown_handles[i] == handler)
> +			break;
> +
> +	if (i == CRASH_SHUTDOWN_HANDLES_NUM){
> +		printk(KERN_ERR "Crash shutdown handle not found\n");
> +		spin_unlock(&crash_handles_lock);
> +		return 1;
> +	}
> +
> +	/* Shift handles down */
> +	while(crash_shutdown_handles[i]) {
> +		crash_shutdown_handles[i] = crash_shutdown_handles[i+1];
> +		i++;
> +	}
> +	spin_unlock(&crash_handles_lock);
> +	return 0;
> +}
> +EXPORT_SYMBOL(crash_shutdown_unregister);
> +
> +static long crash_shutdown_buf[SETJMP_BUF_LEN];

unsigned long?

> +
> +static int handle_fault(struct pt_regs *regs)
> +{
> +	longjmp(crash_shutdown_buf, 1);
> +	return 0;
> +}
> +
>  void default_machine_crash_shutdown(struct pt_regs *regs)
>  {
> -	unsigned int irq;
> +	unsigned int i;
>  
>  	/*
>  	 * This function is only called after the system
> @@ -301,14 +368,27 @@ void default_machine_crash_shutdown(stru
>  	 */
>  	hard_irq_disable();
>  
> -	for_each_irq(irq) {
> -		struct irq_desc *desc = irq_desc + irq;
> +	for_each_irq(i) {
> +		struct irq_desc *desc = irq_desc + i;
>  
>  		if (desc->status & IRQ_INPROGRESS)
> -			desc->chip->eoi(irq);
> +			desc->chip->eoi(i);
>  
>  		if (!(desc->status & IRQ_DISABLED))
> -			desc->chip->disable(irq);
> +			desc->chip->disable(i);
> +	}
> +
> +	/* Call registered shutdown routines */
> +	__debugger_fault_handler = handle_fault;
> +	i = 0;
> +	while(crash_shutdown_handles[i]){
> +		if (setjmp(crash_shutdown_buf) == 0) {
> +			asm volatile("sync; isync");
> +			crash_shutdown_handles[i]();
> +			asm volatile("sync; isync");
> +			__delay(200);
> +		}
> +		i++;
>  	}

You should probably reset __debugger_fault_handler, just to be safe.

cheers

-- 
Michael Ellerman
OzLabs, IBM Australia Development Lab

wwweb: http://michael.ellerman.id.au
phone: +61 2 6212 1183 (tie line 70 21183)

We do not inherit the earth from our ancestors,
we borrow it from our children. - S.M.A.R.T Person

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]