Re: [PATCH] mm: page_ext: validate section in for_each_page_ext iterator

[David Hildenbrand (Arm)]

On 6/19/26 15:38, Matthew Wilcox wrote:
> On Fri, Jun 19, 2026 at 10:00:59AM +0200, David Hildenbrand (Arm) wrote:
>> Likely we should just handle that in the iterator, looking up page-ext for something
>> that doesn't exist is weird.
> 
> Yes, but I prefer to pass 'max' to the _next step explicitly, rather
> than put it in the iterator struct.  Untested patch attached.

Yeah, no strong opinion, same fix idea.

> 
> The other way to do it is s/index/remaining/ and count it down rather
> than upwards.  But that assumes none of the iterator users look at
> iter->index, and I don't have the patience to check that.
> 
> I find it weird that somebody botheresd to add kernel-doc for
> page_ext_iter_begin/next.  They're internal implementation and really
> shouldn't have public documentation.

Yeah, Luiz was just very thorough :)


-- 
Cheers,

David