netdev - Re: [PATCH net-next v2] net: skb: prevent the split of kfree_skb

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]

Message-ID: <CADxym3aD-7Yek75RzAPuwgib+6UOKukcVqGz4bDCY8HPfop_mQ@mail.gmail.com>
Date:   Mon, 15 Aug 2022 14:20:14 +0800
From:   Menglong Dong <menglong8.dong@...il.com>
To:     Miguel Ojeda <miguel.ojeda.sandonis@...il.com>
Cc:     kuba@...nel.org, ojeda@...nel.org, ndesaulniers@...gle.com,
        davem@...emloft.net, edumazet@...gle.com, pabeni@...hat.com,
        asml.silence@...il.com, imagedong@...cent.com,
        luiz.von.dentz@...el.com, vasily.averin@...ux.dev,
        jk@...econstruct.com.au, linux-kernel@...r.kernel.org,
        netdev@...r.kernel.org
Subject: Re: [PATCH net-next v2] net: skb: prevent the split of
 kfree_skb_reason() by gcc

Hello,

On Fri, Aug 12, 2022 at 4:50 PM Miguel Ojeda
<miguel.ojeda.sandonis@...il.com> wrote:
>
> On Fri, Aug 12, 2022 at 4:50 AM <menglong8.dong@...il.com> wrote:
> >
> >  #define __noreturn                      __attribute__((__noreturn__))
> >
> > +#define __nofnsplit                     __attribute__((__optimize__("O1")))
>
> This is still in the wrong place...
>
> Also, from what the bot says, Clang does not support it. I took a
> look, and that seems to be the case. ICC doesn't, either. Thus you
> would need to guard it and also add the docs as needed, like the other
> attributes.
>
> (Not saying that solving the issue with the attribute is a good idea,
> but if you really wanted to add one, it should be done properly)
>

I have dug it deeper, and found that this function-split optimization
is only used by GCC. Therefore, I think I need only to consider it
for GCC.

I'll send a V3, thanks~

Menglong Dong

> Cheers,
> Miguel