Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1695664
| From | Li Kun <hw.likun@huawei.com> |
|---|---|
| Newsgroups | linux.kernel |
| Subject | Re: [kernel-hardening] [PATCH v8 3/3] x86/refcount: Implement fast refcount overflow protection |
| Date | 2017-07-25 14:10 +0200 |
| Message-ID | <u76P0-2BR-15@gated-at.bofh.it> (permalink) |
| References | <u6QqR-dR-3@gated-at.bofh.it> <u6QqR-dR-11@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
Hi Kees,
on 2017/7/25 2:35, Kees Cook wrote:
> +static __always_inline __must_check
> +int __refcount_add_unless(refcount_t *r, int a, int u)
> +{
> + int c, new;
> +
> + c = atomic_read(&(r->refs));
> + do {
> + if (unlikely(c == u))
> + break;
> +
> + asm volatile("addl %2,%0\n\t"
> + REFCOUNT_CHECK_LT_ZERO
> + : "=r" (new)
> + : "0" (c), "ir" (a),
> + [counter] "m" (r->refs.counter)
> + : "cc", "cx");
here when the result LT_ZERO, you will saturate the r->refs.counter and
make the
atomic_try_cmpxchg(&(r->refs), &c, new) bound to fail first time.
maybe we can just saturate the value of variable "new" ?
> +
> + } while (!atomic_try_cmpxchg(&(r->refs), &c, new));
> +
> + return c;
> +}
> +
--
Best Regards
Li Kun
Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
[PATCH v8 0/3] x86: Implement fast refcount overflow protection Kees Cook <keescook@chromium.org> - 2017-07-24 20:40 +0200
[PATCH v8 3/3] x86/refcount: Implement fast refcount overflow protection Kees Cook <keescook@chromium.org> - 2017-07-24 20:40 +0200
Re: [kernel-hardening] [PATCH v8 3/3] x86/refcount: Implement fast refcount overflow protection Li Kun <hw.likun@huawei.com> - 2017-07-25 14:10 +0200
Re: [kernel-hardening] [PATCH v8 3/3] x86/refcount: Implement fast refcount overflow protection Kees Cook <keescook@chromium.org> - 2017-07-25 21:20 +0200
csiph-web