Re: [PATCH mm 4/4] kasan: simplify kasan_complete_mode_report_info for tag-based modes

[Andrey Konovalov <andreyknvl@gmail.com>]

On Thu, Dec 21, 2023 at 9:14 PM Marco Elver <elver@google.com> wrote:
>
> On Thu, 21 Dec 2023 at 19:35, <andrey.konovalov@linux.dev> wrote:
> >
> > From: Andrey Konovalov <andreyknvl@google.com>
> >
> > memcpy the alloc/free tracks when collecting the information about a bad
> > access instead of copying fields one by one.
> >
> > Fixes: 5d4c6ac94694 ("kasan: record and report more information")
> > Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
>
> Reviewed-by: Marco Elver <elver@google.com>
>
> > ---
> >  mm/kasan/report_tags.c | 23 ++++-------------------
> >  1 file changed, 4 insertions(+), 19 deletions(-)
> >
> > diff --git a/mm/kasan/report_tags.c b/mm/kasan/report_tags.c
> > index 688b9d70b04a..d15f8f580e2c 100644
> > --- a/mm/kasan/report_tags.c
> > +++ b/mm/kasan/report_tags.c
> > @@ -27,15 +27,6 @@ static const char *get_common_bug_type(struct kasan_report_info *info)
> >         return "invalid-access";
> >  }
> >
> > -#ifdef CONFIG_KASAN_EXTRA_INFO
> > -static void kasan_complete_extra_report_info(struct kasan_track *track,
> > -                                        struct kasan_stack_ring_entry *entry)
> > -{
> > -       track->cpu = entry->track.cpu;
> > -       track->timestamp = entry->track.timestamp;
> > -}
> > -#endif /* CONFIG_KASAN_EXTRA_INFO */
> > -
> >  void kasan_complete_mode_report_info(struct kasan_report_info *info)
> >  {
> >         unsigned long flags;
> > @@ -80,11 +71,8 @@ void kasan_complete_mode_report_info(struct kasan_report_info *info)
> >                         if (free_found)
> >                                 break;
> >
> > -                       info->free_track.pid = entry->track.pid;
> > -                       info->free_track.stack = entry->track.stack;
> > -#ifdef CONFIG_KASAN_EXTRA_INFO
> > -                       kasan_complete_extra_report_info(&info->free_track, entry);
> > -#endif /* CONFIG_KASAN_EXTRA_INFO */
> > +                       memcpy(&info->free_track, &entry->track,
> > +                              sizeof(info->free_track));
>
> Not sure why the line break is necessary.

Ah, just the old desire to use 80-column line limit :)

Let's keep this as this for now, but I'll fix it if I end up sending
v2 of this series.

Thanks!