[PATCH] return to OF via trap, not exit

[PATCH] return to OF via trap, not exit

[Olaf Hering]

Do not call prom exit prom_panic. It clears the screen and the exit message is lost.
On some (or all?) pmacs it causes another crash when OF tries to print the
date and time in its banner.

Signed-off-by: Olaf Hering <olh@suse.de>

 arch/powerpc/kernel/prom_init.c |    4 +++-
 1 files changed, 3 insertions(+), 1 deletion(-)

Index: linux-2.6.16-rc5-olh/arch/powerpc/kernel/prom_init.c
===================================================================
--- linux-2.6.16-rc5-olh.orig/arch/powerpc/kernel/prom_init.c
+++ linux-2.6.16-rc5-olh/arch/powerpc/kernel/prom_init.c
@@ -398,7 +398,9 @@ static void __init __attribute__((noretu
 #endif
 	prom_print(reason);
 	/* ToDo: should put up an SRC here on p/iSeries */
-	call_prom("exit", 0, 0);
+	/* Do not call exit because it clears the screen on pmac
+	 * it also causes some sort of double-fault on early pmacs */
+	asm("trap\n");
 
 	for (;;)			/* should never get here */
 		;

Re: [PATCH] return to OF via trap, not exit

[Michael Ellerman]

[-- Attachment #1: Type: text/plain, Size: 2849 bytes --]

On Sun, 5 Mar 2006 06:10, Olaf Hering wrote:
> Do not call prom exit prom_panic. It clears the screen and the exit message
> is lost. On some (or all?) pmacs it causes another crash when OF tries to
> print the date and time in its banner.
>
> Signed-off-by: Olaf Hering <olh@suse.de>
>
>  arch/powerpc/kernel/prom_init.c |    4 +++-
>  1 files changed, 3 insertions(+), 1 deletion(-)
>
> Index: linux-2.6.16-rc5-olh/arch/powerpc/kernel/prom_init.c
> ===================================================================
> --- linux-2.6.16-rc5-olh.orig/arch/powerpc/kernel/prom_init.c
> +++ linux-2.6.16-rc5-olh/arch/powerpc/kernel/prom_init.c
> @@ -398,7 +398,9 @@ static void __init __attribute__((noretu
>  #endif
>  	prom_print(reason);
>  	/* ToDo: should put up an SRC here on p/iSeries */
> -	call_prom("exit", 0, 0);
> +	/* Do not call exit because it clears the screen on pmac
> +	 * it also causes some sort of double-fault on early pmacs */
> +	asm("trap\n");
>
>  	for (;;)			/* should never get here */
>  		;

I don't think I like it, on IBM firmware it takes us from this:

Elapsed time since release of system processors: 26005 mins 51 secs

zImage starting: loaded at 0x00400000 (sp: 0x01a1ffe0)
Allocating 0x7edc50 bytes for kernel ...
OF version = 'IBM,SF230_126'
gunzipping (0x1c00000 <- 0x407000:0x6a04e4)...done 0x766880 bytes
OF stdout device is: /vdevice/vty@30000000
Error: You can't boot a kdump kernel from OF!
EXIT called ok
0 >


To this:

Elapsed time since release of system processors: 25995 mins 56 secs

zImage starting: loaded at 0x00400000 (sp: 0x01a1ffe0)
Allocating 0x7edc50 bytes for kernel ...
OF version = 'IBM,SF230_126'
gunzipping (0x1c00000 <- 0x407000:0x6a044b)...done 0x766880 bytes
OF stdout device is: /vdevice/vty@30000000
Error: You can't boot a kdump kernel from OF!
DEFAULT CATCH!, exception-handler=fff00700
at   %SRR0: 00000000020eaf78   %SRR1: 8000000000023002
Open Firmware exception handler entered from non-OF code

Client's Fix Pt Regs:
 00 00000000020eaf78 0000000001a1fe90 00000000023681d0 0000000000000002
 04 0000000044000024 0000000000000000 0000000000000000 0000000002063e1e
 08 0000000000000000 0000000001a1fc9c 0000000000003002 0000000000003002
 0c 2000000000000000 0000000000000000 0000000000000000 0000000000000000
 10 0000000000000000 0000000000000000 0000000000000000 0000000000000000
 14 0000000000c00000 0000000000000008 0000000000000000 0000000000000000
 18 0000000000000000 0000000000000000 0000000000000000 0000000000c39a48
 1c c000000002453cc8 0000000002108938 0000000000000000 fffffffffffffffd
Special Regs:
    %IV: 00000700     %CR: 44000022    %XER: 00000000  %DSISR: 00000000
  %SRR0: 00000000020eaf78   %SRR1: 8000000000023002
    %LR: 00000000020eaf78    %CTR: 0000000000000000
   %DAR: 0000000000000000
Virtual PID = 0
PFW: Unable to send error log!
 ok
0 >

cheers

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [PATCH] return to OF via trap, not exit

[Olaf Hering]

 On Mon, Mar 06, Michael Ellerman wrote:


> I don't think I like it, on IBM firmware it takes us from this:
> 
> Elapsed time since release of system processors: 26005 mins 51 secs
> 
> zImage starting: loaded at 0x00400000 (sp: 0x01a1ffe0)
> Allocating 0x7edc50 bytes for kernel ...
> OF version = 'IBM,SF230_126'
> gunzipping (0x1c00000 <- 0x407000:0x6a04e4)...done 0x766880 bytes
> OF stdout device is: /vdevice/vty@30000000
> Error: You can't boot a kdump kernel from OF!
> EXIT called ok
> 0 >
> 
> 
> To this:

Which is likely ok because it saves us from typing .registers manually.

Re: [PATCH] return to OF via trap, not exit

[Segher Boessenkool]

> Which is likely ok because it saves us from typing .registers  
> manually.

I can't say that's a great argument.  How about doing the trap thing
on "known-bad" platforms only?


Segher

Re: [PATCH] return to OF via trap, not exit

[Olaf Hering]

 On Mon, Mar 06, Segher Boessenkool wrote:

> >Which is likely ok because it saves us from typing .registers  
> >manually.
> 
> I can't say that's a great argument.  How about doing the trap thing
> on "known-bad" platforms only?

Can you do anything at that time anyway? I mean, reset-all pending
either way.

Re: [PATCH] return to OF via trap, not exit

[Segher Boessenkool]

>>> Which is likely ok because it saves us from typing .registers
>>> manually.
>>
>> I can't say that's a great argument.  How about doing the trap thing
>> on "known-bad" platforms only?
>
> Can you do anything at that time anyway? I mean, reset-all pending
> either way.

You can boot a different kernel, or try to work out what went wrong.


Segher

Re: [PATCH] return to OF via trap, not exit

[Olaf Hering]

 On Mon, Mar 06, Segher Boessenkool wrote:

> >>> Which is likely ok because it saves us from typing .registers
> >>> manually.
> >>
> >> I can't say that's a great argument.  How about doing the trap thing
> >> on "known-bad" platforms only?
> >
> > Can you do anything at that time anyway? I mean, reset-all pending
> > either way.
> 
> You can boot a different kernel, or try to work out what went wrong.

No, I cant, all resources are possible busy. Either way, maybe checking
what type stdout is and trap or exit depending on that.
And I dont see whats bad about 'trap handler called', other than being
verbose.

Re: [PATCH] return to OF via trap, not exit

[Segher Boessenkool]

>>> Can you do anything at that time anyway? I mean, reset-all pending
>>> either way.
>>
>> You can boot a different kernel, or try to work out what went wrong.
>
> No, I cant, all resources are possible busy.

That's a separate bug.  What resources do you see busy?  Allocated
memory, I/O devices?

> Either way, maybe checking
> what type stdout is and trap or exit depending on that.

That's better than always calling trap, sure.  Is there any reason
you can't just do it on Macs though?  Because the problem you're trying
to work around only happens there.

> And I dont see whats bad about 'trap handler called', other than being
> verbose.

Too verbose -- it prints useless information, and the actual error
message might scroll away.  No biggie of course, but why do the
workaround when not needed?


Segher

Re: [PATCH] return to OF via trap, not exit

[Olaf Hering]

 On Mon, Mar 06, Segher Boessenkool wrote:

> That's better than always calling trap, sure.  Is there any reason
> you can't just do it on Macs though?  Because the problem you're trying
> to work around only happens there.

Maybe something like this? Only compile tested


Do not call prom exit prom_panic. It clears the screen and the exit message is lost.
On some (or all?) pmacs it causes another crash when OF tries to print the
date and time in its banner.

Set of_platform earlier to catch more prom_panic() calls.

Signed-off-by: Olaf Hering <olh@suse.de>

 arch/powerpc/kernel/prom_init.c |   23 ++++++++++++++---------
 1 file changed, 14 insertions(+), 9 deletions(-)

Index: linux-2.6.16-olh/arch/powerpc/kernel/prom_init.c
===================================================================
--- linux-2.6.16-olh.orig/arch/powerpc/kernel/prom_init.c
+++ linux-2.6.16-olh/arch/powerpc/kernel/prom_init.c
@@ -397,6 +397,11 @@ static void __init __attribute__((noretu
 	reason = PTRRELOC(reason);
 #endif
 	prom_print(reason);
+	/* Do not call exit because it clears the screen on pmac
+	 * it also causes some sort of double-fault on early pmacs */
+	if (RELOC(of_platform) == PLATFORM_POWERMAC)
+		asm("trap\n");
+
 	/* ToDo: should put up an SRC here on p/iSeries */
 	call_prom("exit", 0, 0);
 
@@ -2060,15 +2065,6 @@ unsigned long __init prom_init(unsigned 
 	 */
 	prom_init_stdout();
 
-	/* Bail if this is a kdump kernel. */
-	if (PHYSICAL_START > 0)
-		prom_panic("Error: You can't boot a kdump kernel from OF!\n");
-
-	/*
-	 * Check for an initrd
-	 */
-	prom_check_initrd(r3, r4);
-
 	/*
 	 * Get default machine type. At this point, we do not differentiate
 	 * between pSeries SMP and pSeries LPAR
@@ -2078,6 +2074,15 @@ unsigned long __init prom_init(unsigned 
 	prom_setprop(_prom->chosen, "/chosen", "linux,platform",
 		     &getprop_rval, sizeof(getprop_rval));
 
+	/* Bail if this is a kdump kernel. */
+	if (PHYSICAL_START > 0)
+		prom_panic("Error: You can't boot a kdump kernel from OF!\n");
+
+	/*
+	 * Check for an initrd
+	 */
+	prom_check_initrd(r3, r4);
+
 #ifdef CONFIG_PPC_PSERIES
 	/*
 	 * On pSeries, inform the firmware about our capabilities

Re: [PATCH] return to OF via trap, not exit

[Michael Ellerman]

On Thu, 2006-03-23 at 22:03 +0100, Olaf Hering wrote:
>  On Mon, Mar 06, Segher Boessenkool wrote:
> 
> > That's better than always calling trap, sure.  Is there any reason
> > you can't just do it on Macs though?  Because the problem you're trying
> > to work around only happens there.
> 
> Maybe something like this? Only compile tested
> 
> 
> Do not call prom exit prom_panic. It clears the screen and the exit message is lost.
> On some (or all?) pmacs it causes another crash when OF tries to print the
> date and time in its banner.
> 
> Set of_platform earlier to catch more prom_panic() calls.

Looks reasonable, assuming it works :)

cheers

-- 
Michael Ellerman
IBM OzLabs

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

Re: [PATCH] return to OF via trap, not exit

[Segher Boessenkool]

On 23-mrt-2006, at 22:03, Olaf Hering wrote:
>  On Mon, Mar 06, Segher Boessenkool wrote:
>> That's better than always calling trap, sure.  Is there any reason
>> you can't just do it on Macs though?  Because the problem you're  
>> trying
>> to work around only happens there.
>
> Maybe something like this? Only compile tested

That looks fine yes (if it runs ;-) ), thanks!

> Do not call prom exit prom_panic. It clears the screen and the exit  
> message is lost.
> On some (or all?) pmacs it causes another crash when OF tries to  
> print the
> date and time in its banner.
>
> Set of_platform earlier to catch more prom_panic() calls.
>
> Signed-off-by: Olaf Hering <olh@suse.de>

Acked-by: Segher Boessenkool <segher@kernel.crashing.org>