Re: [Xen-devel] [PATCH v2 2/2] xmalloc: add a Kconfig option to poison free pool memory

[Paul Durrant <Paul.Durrant@xxxxxxxxxx>]

> -----Original Message-----
> From: Jan Beulich <JBeulich@xxxxxxxx>
> Sent: 05 July 2019 14:41
> To: Paul Durrant <Paul.Durrant@xxxxxxxxxx>; xen-devel@xxxxxxxxxxxxxxxxxxxx
> Cc: Julien Grall <julien.grall@xxxxxxx>; Andrew Cooper 
> <Andrew.Cooper3@xxxxxxxxxx>; George Dunlap
> <George.Dunlap@xxxxxxxxxx>; Ian Jackson <Ian.Jackson@xxxxxxxxxx>; Stefano 
> Stabellini
> <sstabellini@xxxxxxxxxx>; Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>; Tim 
> (Xen.org) <tim@xxxxxxx>;
> Wei Liu <wl@xxxxxxx>
> Subject: Re: [PATCH v2 2/2] xmalloc: add a Kconfig option to poison free pool 
> memory
> 
> On 05.07.2019 11:02, Paul Durrant wrote:
> > --- a/xen/Kconfig.debug
> > +++ b/xen/Kconfig.debug
> > @@ -105,6 +105,13 @@ config DEBUG_TRACE
> >       either directly to the console or are printed to console in case of
> >       a system crash.
> >
> > +config XMEM_POOL_POISON
> > +       bool "Poison free xenpool blocks"
> > +       default DEBUG
> > +       ---help---
> > +    Poison free blocks with 0xAA bytes and verify them when a block is
> > +    allocated in order to spot use-after-free issues.
> 
> There looks to be a spaces vs tabs problem here: Only the help
> text has tabs for initial indentation.

Oh yes. I'll fix that.

> 
> > --- a/xen/common/xmalloc_tlsf.c
> > +++ b/xen/common/xmalloc_tlsf.c
> > @@ -238,6 +238,11 @@ static inline void EXTRACT_BLOCK(struct bhdr *b, 
> > struct xmem_pool *p, int fl,
> >           }
> >       }
> >       b->ptr.free_ptr = (struct free_ptr) {NULL, NULL};
> > +#ifdef CONFIG_XMEM_POOL_POISON
> > +    if ( (b->size & BLOCK_SIZE_MASK) > MIN_BLOCK_SIZE )
> > +        ASSERT(!memchr_inv(b->ptr.buffer + MIN_BLOCK_SIZE, 0xAA,
> > +                           (b->size & BLOCK_SIZE_MASK) - MIN_BLOCK_SIZE));
> > +#endif /* CONFIG_XMEM_POOL_POISON */
> >   }
> >
> >   /**
> > @@ -245,6 +250,11 @@ static inline void EXTRACT_BLOCK(struct bhdr *b, 
> > struct xmem_pool *p, int fl,
> >    */
> >   static inline void INSERT_BLOCK(struct bhdr *b, struct xmem_pool *p, int 
> > fl, int sl)
> >   {
> > +#ifdef CONFIG_XMEM_POOL_POISON
> > +    if ( (b->size & BLOCK_SIZE_MASK) > MIN_BLOCK_SIZE )
> > +        memset(b->ptr.buffer + MIN_BLOCK_SIZE, 0xAA,
> > +               (b->size & BLOCK_SIZE_MASK) - MIN_BLOCK_SIZE);
> > +#endif /* CONFIG_XMEM_POOL_POISON */
> 
> Can you tie together the two instances of 0xAA via a #define,
> please?
> 

Ok, sure.

> It would also be nice if both #ifdef blocks were separated from
> their neighboring code by a blank line, despite the file using
> blank lines rather sparingly so far.
> 
> With the adjustments here I think it would be best if you could
> also resend patch 1 with the slightly adjusted commit message.

Alright, will do.

  Paul

> 
> Jan
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel