linux-kernel - Re: [PATCH] kasan: allow sampling page_alloc allocations for HW

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Sat, 26 Nov 2022 20:13:31 +0100
From:   Andrey Konovalov <andreyknvl@...il.com>
To:     Marco Elver <elver@...gle.com>
Cc:     andrey.konovalov@...ux.dev,
        Alexander Potapenko <glider@...gle.com>,
        Dmitry Vyukov <dvyukov@...gle.com>,
        Andrey Ryabinin <ryabinin.a.a@...il.com>,
        kasan-dev@...glegroups.com, Peter Collingbourne <pcc@...gle.com>,
        Evgenii Stepanov <eugenis@...gle.com>,
        Florian Mayer <fmayer@...gle.com>,
        Andrew Morton <akpm@...ux-foundation.org>, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org,
        Andrey Konovalov <andreyknvl@...gle.com>
Subject: Re: [PATCH] kasan: allow sampling page_alloc allocations for HW_TAGS

On Sun, Oct 30, 2022 at 3:59 AM Marco Elver <elver@...gle.com> wrote:
>
> > +- ``kasan.page_alloc.sample=<sampling frequency>`` makes KASAN tag only
>
> Frequency is number of samples per frame (unit time, or if used
> non-temporally like here, population size).
>
> [1] https://en.wikipedia.org/wiki/Systematic_sampling
>
> You're using it as an interval, so I'd just replace uses of frequency
> with "interval" appropriately here and elsewhere.

Done in v2.

> > +static inline bool kasan_sample_page_alloc(void)
> > +{
> > +     unsigned long *count = this_cpu_ptr(&kasan_page_alloc_count);
>
> this_cpu_inc_return()
>
> without it, you need to ensure preemption is disabled around here.
>
> > +
> > +     return (*count)++ % kasan_page_alloc_sample == 0;
>
> Doing '%' is a potentially costly operation if called in a fast-path.
>
> We can generate better code with (rename 'count' -> 'skip'):
>
>         long skip_next = this_cpu_dec_return(kasan_page_alloc_skip);
>
>         if (skip_next < 0) {
>                 this_cpu_write(kasan_page_alloc_skip, kasan_page_alloc_sample - 1);
>                 return true;
>         }
>
>         return false;

Done in v2.

Thank you, Marco!