# Re: \[PATCH v3 01/10\] mm: add Kernel Electric-Fence infrastructure

- From: Marco Elver \<elver@google.com\>
- Date: 2020-09-29 13:11:47 
- List: linux-arm-kernel
- Page: https://ratatoskr.run/linux-arm-kernel/2020/09/6124340
- Thread: https://ratatoskr.run/linux-doc/2020/09/6113322/t.md

```text
On Tue, Sep 29, 2020 at 02:42PM +0200, Andrey Konovalov wrote:
[...]
> > +        */
> > +       index = (addr - (unsigned long)__kfence_pool) / (PAGE_SIZE * 2) - 1;
> 
> Why do we subtract 1 here? We do have the metadata entry reserved for something?

Above the declaration of __kfence_pool it says:

	* We allocate an even number of pages, as it simplifies calculations to map
	* address to metadata indices; effectively, the very first page serves as an
	* extended guard page, but otherwise has no special purpose.

Hopefully that clarifies the `- 1` here.

[...]
> > +       /* Allocation and free stack information. */
> > +       int num_alloc_stack;
> > +       int num_free_stack;
> > +       unsigned long alloc_stack[KFENCE_STACK_DEPTH];
> > +       unsigned long free_stack[KFENCE_STACK_DEPTH];
> 
> It was a concious decision to not use stackdepot, right? Perhaps it
> makes sense to document the reason somewhere.

Yes; we want to avoid the dynamic allocations that stackdepot does.

[...]

Thanks,
-- Marco

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
```
