Re: [PATCH] vmacache: change vmacache_find() to always check ->vm_mm

["Srivatsa S. Bhat" <srivatsa.bhat@linux.vnet.ibm.com>]

On 04/29/2014 06:22 PM, Oleg Nesterov wrote:
> On 04/29, Srivatsa S. Bhat wrote:
>>
>> I guess I'll hold off on testing this fix until I get to reproduce
>> the bug more reliably..
> 
> perhaps the patch below can help a bit?
> 
> -------------------------------------------------------------------------------
> Subject: [PATCH] vmacache: change vmacache_find() to always check ->vm_mm
> 
> If ->vmacache was corrupted it would be better to detect and report
> the problem asap, check vma->vm_mm before vm_start/vm_end.
> 
> Signed-off-by: Oleg Nesterov <oleg@redhat.com>
> ---
>  mm/vmacache.c |    5 +++--
>  1 files changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/mm/vmacache.c b/mm/vmacache.c
> index d4224b3..952a324 100644
> --- a/mm/vmacache.c
> +++ b/mm/vmacache.c
> @@ -81,9 +81,10 @@ struct vm_area_struct *vmacache_find(struct mm_struct *mm, unsigned long addr)
>  	for (i = 0; i < VMACACHE_SIZE; i++) {
>  		struct vm_area_struct *vma = current->vmacache[i];
> 
> -		if (vma && vma->vm_start <= addr && vma->vm_end > addr) {
> +		if (vma) {
>  			BUG_ON(vma->vm_mm != mm);
> -			return vma;
> +			if (vma->vm_start <= addr && vma->vm_end > addr)
> +				return vma;
>  		}
>  	}
> 

IIUC, this is similar to commit 50f5aa8a9b2 (mm: don't pointlessly use
BUG_ON() for sanity check). But even with that commit included I was
not able to reproduce the bug again, as reported here:

https://lkml.org/lkml/2014/4/29/187

Regards,
Srivatsa S. Bhat

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>