Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20221104231235.348958-1-cmllamas@google.com>
In-Reply-To: <20221104231235.348958-1-cmllamas@google.com>
From:   Todd Kjos <tkjos@google.com>
Date:   Fri, 4 Nov 2022 16:18:55 -0700
Message-ID: <CAHRSSEy4tfkFtnxSXrgX5PhUdj2xf+yrfC7Zujf-mvt2ekJe+Q@mail.gmail.com>
Subject: Re: [PATCH] binder: validate alloc->mm in ->mmap() handler
To:     Carlos Llamas <cmllamas@google.com>
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        =?UTF-8?B?QXJ2ZSBIasO4bm5ldsOlZw==?= <arve@android.com>,
        Todd Kjos <tkjos@android.com>,
        Martijn Coenen <maco@android.com>,
        Joel Fernandes <joel@joelfernandes.org>,
        Christian Brauner <brauner@kernel.org>,
        Suren Baghdasaryan <surenb@google.com>,
        "Liam R. Howlett" <Liam.Howlett@oracle.com>,
        linux-kernel@vger.kernel.org, kernel-team@android.com,
        Jann Horn <jannh@google.com>, stable@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Fri, Nov 4, 2022 at 4:12 PM Carlos Llamas <cmllamas@google.com> wrote:
>
> Since commit 1da52815d5f1 ("binder: fix alloc->vma_vm_mm null-ptr
> dereference") binder caches a pointer to the current->mm during open().
> This fixes a null-ptr dereference reported by syzkaller. Unfortunately,
> it also opens the door for a process to update its mm after the open(),
> (e.g. via execve) making the cached alloc->mm pointer invalid.
>
> Things get worse when the process continues to mmap() a vma. From this
> point forward, binder will attempt to find this vma using an obsolete
> alloc->mm reference. Such as in binder_update_page_range(), where the
> wrong vma is obtained via vma_lookup(), yet binder proceeds to happily
> insert new pages into it.
>
> To avoid this issue fail the ->mmap() callback if we detect a mismatch
> between the vma->vm_mm and the original alloc->mm pointer. This prevents
> alloc->vm_addr from getting set, so that any subsequent vma_lookup()
> calls fail as expected.
>
> Fixes: 1da52815d5f1 ("binder: fix alloc->vma_vm_mm null-ptr dereference")
> Reported-by: Jann Horn <jannh@google.com>
> Cc: <stable@vger.kernel.org> # 5.15+
> Signed-off-by: Carlos Llamas <cmllamas@google.com>

Acked-by: Todd Kjos <tkjos@google.com>

> ---
>  drivers/android/binder_alloc.c | 7 +++++++
>  1 file changed, 7 insertions(+)
>
> diff --git a/drivers/android/binder_alloc.c b/drivers/android/binder_alloc.c
> index 1c39cfce32fa..4ad42b0f75cd 100644
> --- a/drivers/android/binder_alloc.c
> +++ b/drivers/android/binder_alloc.c
> @@ -739,6 +739,12 @@ int binder_alloc_mmap_handler(struct binder_alloc *alloc,
>         const char *failure_string;
>         struct binder_buffer *buffer;
>
> +       if (unlikely(vma->vm_mm != alloc->mm)) {
> +               ret = -EINVAL;
> +               failure_string = "invalid vma->vm_mm";
> +               goto err_invalid_mm;
> +       }
> +
>         mutex_lock(&binder_alloc_mmap_lock);
>         if (alloc->buffer_size) {
>                 ret = -EBUSY;
> @@ -785,6 +791,7 @@ int binder_alloc_mmap_handler(struct binder_alloc *alloc,
>         alloc->buffer_size = 0;
>  err_already_mapped:
>         mutex_unlock(&binder_alloc_mmap_lock);
> +err_invalid_mm:
>         binder_alloc_debug(BINDER_DEBUG_USER_ERROR,
>                            "%s: %d %lx-%lx %s failed %d\n", __func__,
>                            alloc->pid, vma->vm_start, vma->vm_end,
> --
> 2.38.1.431.g37b22c650d-goog
>