Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <cover.1610554432.git.andreyknvl@google.com> <77015767eb7cfe1cc112a564d31e749d68615a0f.1610554432.git.andreyknvl@google.com>
 <CANpmjNPX9yn5izxtYMq14Aas2y4NA1ijkcS9KN4QQ-7Hv8qZEQ@mail.gmail.com>
In-Reply-To: <CANpmjNPX9yn5izxtYMq14Aas2y4NA1ijkcS9KN4QQ-7Hv8qZEQ@mail.gmail.com>
From:   Andrey Konovalov <andreyknvl@google.com>
Date:   Thu, 14 Jan 2021 18:58:57 +0100
Message-ID: <CAAeHK+xberuUH0cn8U1N8X0n56Fy=uLCxZ=P3q+E2PPRsfuKNQ@mail.gmail.com>
Subject: Re: [PATCH v2 11/14] kasan: fix bug detection via ksize for HW_TAGS mode
To:     Marco Elver <elver@google.com>
Cc:     Catalin Marinas <catalin.marinas@arm.com>,
        Vincenzo Frascino <vincenzo.frascino@arm.com>,
        Dmitry Vyukov <dvyukov@google.com>,
        Alexander Potapenko <glider@google.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        Will Deacon <will.deacon@arm.com>,
        Andrey Ryabinin <aryabinin@virtuozzo.com>,
        Peter Collingbourne <pcc@google.com>,
        Evgenii Stepanov <eugenis@google.com>,
        Branislav Rankov <Branislav.Rankov@arm.com>,
        Kevin Brodsky <kevin.brodsky@arm.com>,
        kasan-dev <kasan-dev@googlegroups.com>,
        Linux ARM <linux-arm-kernel@lists.infradead.org>,
        Linux Memory Management List <linux-mm@kvack.org>,
        LKML <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Wed, Jan 13, 2021 at 5:54 PM Marco Elver <elver@google.com> wrote:
>
> > +bool __kasan_check_byte(const void *addr, unsigned long ip);
> > +static __always_inline bool kasan_check_byte(const void *addr, unsigned long ip)
> > +{
> > +       if (kasan_enabled())
> > +               return __kasan_check_byte(addr, ip);
> > +       return true;
> > +}
>
> Why was this not added to kasan-checks.h? I'd assume including all of
> kasan.h is also undesirable for tag-based modes if we just want to do
> a kasan_check_byte().
>
> Was requiring 'ip' intentional? Unlike the other
> kasan_check-functions, this takes an explicit 'ip'. In the case of
> ksize() usage, this is an advantage, so I'd probably keep it, but the
> rationale to introducing 'ip' vs. before wasn't mentioned.

Yes, to avoid having a ksize() frame in the report. However, I'll move
_RET_IP_ inside of kasan_check_byte() as it's inline.

> > +bool __kasan_check_byte(const void *address, unsigned long ip)
> > +{
> > +       if (!kasan_byte_accessible(address)) {
> > +               kasan_report((unsigned long)address, 1, false, ip);
> > +               return false;
> > +       }
> > +       return true;
> > +}
>
> Like the other __kasan_check*, should this have been EXPORT_SYMBOL()?
> Or was it intentional to not export as it's currently only used by
> non-modules?

We can add EXPORT_SYMBOL as soon as there's a need for it.

Thanks!