Hi,

> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On Behalf Of Kees
> Cook
> Sent: Friday, July 29, 2016 4:35 AM
> To: 岩松信洋 / IWAMATSU,NOBUHIRO
> Cc: Anton Vorontsov; Colin Cross; Tony Luck; LKML; Hiraku Toyooka; Mark
> Salyzyn; 阿口誠司 / AGUCHI,SEIJI
> Subject: Re: [PATCH v2 1/5] ramoops: use persistent_ram_free() instead of
> kfree() for freeing prz
> 
> On Sun, Jul 24, 2016 at 8:56 PM, Nobuhiro Iwamatsu
> <[email protected]> wrote:
> > From: Hiraku Toyooka <[email protected]>
> >
> > persistent_ram_zone(=prz) structures are allocated by
> > persistent_ram_new(), which includes vmap() or ioremap(). But they are
> > currently freed by kfree(). This uses persistent_ram_free() for correct
> this asymmetry usage.
> >
> > Signed-off-by: Hiraku Toyooka <[email protected]>
> > Signed-off-by: Nobuhiro Iwamatsu <[email protected]>
> > Cc: Mark Salyzyn <[email protected]>
> > Cc: Seiji Aguchi <[email protected]>
> 
> Oh, yes, oops. I may extract this patch and get it into v4.8, since this
> is an explicit bug fix. Thanks!
> 
> -Kees

Thanks. Please pickup this commit as bug fix.
And if you have a time, please review other patches in this patch series.

Best regards,
  Nobuhiro 

> 
> > ---
> >  fs/pstore/ram.c | 6 +++---
> >  1 file changed, 3 insertions(+), 3 deletions(-)
> >
> > diff --git a/fs/pstore/ram.c b/fs/pstore/ram.c index bd9812e..22416c0
> > 100644
> > --- a/fs/pstore/ram.c
> > +++ b/fs/pstore/ram.c
> > @@ -567,11 +567,11 @@ fail_buf:
> >         kfree(cxt->pstore.buf);
> >  fail_clear:
> >         cxt->pstore.bufsize = 0;
> > -       kfree(cxt->mprz);
> > +       persistent_ram_free(cxt->mprz);
> >  fail_init_mprz:
> > -       kfree(cxt->fprz);
> > +       persistent_ram_free(cxt->fprz);
> >  fail_init_fprz:
> > -       kfree(cxt->cprz);
> > +       persistent_ram_free(cxt->cprz);
> >  fail_init_cprz:
> >         ramoops_free_przs(cxt);
> >  fail_out:
> > --
> > 2.8.1
> >
> >
> 
> 
> 
> --
> Kees Cook
> Chrome OS & Brillo Security

Reply via email to