Received-SPF: softfail (google.com: domain of transitioning linux-kernel-owner@vger.kernel.org does not designate 23.128.96.19 as permitted sender) client-ip=23.128.96.19;
MIME-Version: 1.0
References: <20220421072212.608884-1-song@kernel.org> <CAHk-=wi3eu8mdKmXOCSPeTxABVbstbDg1q5Fkak+A9kVwF+fVw@mail.gmail.com>
 <CAADnVQKyDwXUMCfmdabbVE0vSGxdpqmWAwHRBqbPLW=LdCnHBQ@mail.gmail.com>
 <CAHk-=whFeBezdSrPy31iYv-UZNnNavymrhqrwCptE4uW8aeaHw@mail.gmail.com>
 <CAPhsuW7M6exGD3C1cPBGjhU0Y5efxtJ3=0BWNnbuH87TgQMzdg@mail.gmail.com>
 <CAHk-=wh1mO5HdrOMTq68WHM51-=jdmQS=KipVYxS+5u3uRc5rg@mail.gmail.com>
 <1A4FF473-0988-48BE-9993-0F5E9F0AAC95@fb.com> <CAHk-=wi62LDc5B3DOr5pyVtOUOuLkLzHvmZQApH9q=raqaGkUg@mail.gmail.com>
 <8F788446-899C-4BA3-8236-612A94D98582@fb.com>
In-Reply-To: <8F788446-899C-4BA3-8236-612A94D98582@fb.com>
From:   Linus Torvalds <torvalds@linux-foundation.org>
Date:   Thu, 21 Apr 2022 16:10:07 -0700
Message-ID: <CAHk-=wgW2vxREeH0Bgr8hGxVavfRsNAX3cyaS9eCcg9A77zhLw@mail.gmail.com>
Subject: Re: [PATCH bpf] bpf: invalidate unused part of bpf_prog_pack
To:     Song Liu <songliubraving@fb.com>
Cc:     Peter Zijlstra <peterz@infradead.org>, Song Liu <song@kernel.org>,
        Alexei Starovoitov <alexei.starovoitov@gmail.com>,
        bpf <bpf@vger.kernel.org>, Linux-MM <linux-mm@kvack.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Alexei Starovoitov <ast@kernel.org>,
        Daniel Borkmann <daniel@iogearbox.net>,
        Kernel Team <Kernel-team@fb.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        "Edgecombe, Rick P" <rick.p.edgecombe@intel.com>,
        Christoph Hellwig <hch@infradead.org>,
        Andrii Nakryiko <andrii@kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Thu, Apr 21, 2022 at 3:52 PM Song Liu <songliubraving@fb.com> wrote:
>
> I think this won=E2=80=99t work, as set_memory_ro makes all the aliases o=
f
> these pages read only.

Argh. I thought we only did that for the whole memory type thing
(history: nasty machine checks possible on some hardware if you mix
memory types for the same physical page with virtual mappings), but if
we do it for RO too, then yeah.

It's sad to use that horrid machinery for basically non-live code, but
whatever.

                  Linus