2007-10-22 11:57:35

by Alexey Dobriyan

Subject: [PATCH] Bogus KERN_ALERT on oops

printing eip: f881b9f3 *pdpt = 0000000000003001 <1>*pde = 000000000480a067 *pte = 0000000000000000
^^^

Signed-off-by: Alexey Dobriyan <[email protected]>
---

arch/x86/mm/fault_32.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

--- a/arch/x86/mm/fault_32.c
+++ b/arch/x86/mm/fault_32.c
@@ -550,7 +550,7 @@ no_context:
page &= PAGE_MASK;
page = ((__typeof__(page) *) __va(page))[(address >> PMD_SHIFT)
& (PTRS_PER_PMD - 1)];
- printk(KERN_ALERT "*pde = %016Lx ", page);
+ printk("*pde = %016Lx ", page);
page &= ~_PAGE_NX;
}
#else

2007-10-22 13:06:31

by Pekka Enberg

Subject: Re: [PATCH] Bogus KERN_ALERT on oops

Hi Alexey,

On 10/22/07, Alexey Dobriyan <[email protected]> wrote:
> printing eip: f881b9f3 *pdpt = 0000000000003001 <1>*pde = 000000000480a067 *pte = 0000000000000000

> --- a/arch/x86/mm/fault_32.c
> +++ b/arch/x86/mm/fault_32.c
> @@ -550,7 +550,7 @@ no_context:
> page &= PAGE_MASK;
> page = ((__typeof__(page) *) __va(page))[(address >> PMD_SHIFT)
> & (PTRS_PER_PMD - 1)];
> - printk(KERN_ALERT "*pde = %016Lx ", page);
> + printk("*pde = %016Lx ", page);

Use the new KERN_CONT annotation here?

2007-10-24 10:33:15

by Ingo Molnar

Subject: Re: [PATCH] Bogus KERN_ALERT on oops

* Alexey Dobriyan <[email protected]> wrote:

> printing eip: f881b9f3 *pdpt = 0000000000003001 <1>*pde = 000000000480a067 *pte = 0000000000000000
> ^^^

thanks, added this to the x86 queue.

Ingo

2007-10-24 10:33:46

by Ingo Molnar

Subject: Re: [PATCH] Bogus KERN_ALERT on oops

* Pekka Enberg <[email protected]> wrote:

> > - printk(KERN_ALERT "*pde = %016Lx ", page);
> > + printk("*pde = %016Lx ", page);
>
> Use the new KERN_CONT annotation here?

indeed - i changed the patch to do that.

Ingo

2007-10-24 11:40:06

by Alexey Dobriyan

Subject: Re: [PATCH] Bogus KERN_ALERT on oops

On Wed, Oct 24, 2007 at 12:33:18PM +0200, Ingo Molnar wrote:
>
> * Pekka Enberg <[email protected]> wrote:
>
> > > - printk(KERN_ALERT "*pde = %016Lx ", page);
> > > + printk("*pde = %016Lx ", page);
> >
> > Use the new KERN_CONT annotation here?
>
> indeed - i changed the patch to do that.

Might as well change comment around KERN_CONT -- for starters it lied
about "early bootup" phase since day one.

Proposed text:

/*
* Annotation for a "continued" line of log printout (only done
* after a line that had no enclosing \n).
*
* Introduced because checkpatch.pl couldn't be arsed to learn C
* and distinguish continued printk() from the one that starts
* new line.
*
* Caveat #1: Empty string-literal, so compiler can't check for
* KERN_CONT misuse.
* Caveat #2: Empty string-literal, so it can't be used in
* printk(var); situations.
* Caveat #3: takes characters on the screen, so code is harder
* to read.
* Caveat #4: checkpatch.pl doesn't know C, so it can't check
* for KERN_CONT misuse, anyway.
*/