From: Jason Gunthorpe <jgg@nvidia.com>
To: Jann Horn <jannh@google.com>
Cc: Andrew Morton <akpm@linux-foundation.org>, <linux-mm@kvack.org>,
<linux-kernel@vger.kernel.org>,
"Eric W . Biederman" <ebiederm@xmission.com>,
Michel Lespinasse <walken@google.com>,
Mauro Carvalho Chehab <mchehab@kernel.org>,
Sakari Ailus <sakari.ailus@linux.intel.com>,
Jeff Dike <jdike@addtoit.com>,
Richard Weinberger <richard@nod.at>,
Anton Ivanov <anton.ivanov@cambridgegreys.com>,
<linux-um@lists.infradead.org>,
"John Hubbard" <jhubbard@nvidia.com>,
Johannes Berg <johannes@sipsolutions.net>
Subject: Re: [PATCH resend v3 2/2] exec: Broadly lock nascent mm until setup_arg_pages()
Date: Tue, 20 Oct 2020 16:15:40 -0300 [thread overview]
Message-ID: <20201020191540.GM6219@nvidia.com> (raw)
In-Reply-To: <20201016225713.1971256-3-jannh@google.com>
On Sat, Oct 17, 2020 at 12:57:13AM +0200, Jann Horn wrote:
> @@ -374,17 +366,12 @@ static int bprm_mm_init(struct linux_binprm *bprm)
> task_unlock(current->group_leader);
>
> err = __bprm_mm_init(bprm);
> - if (err)
> - goto err;
> -
> - return 0;
> -
> -err:
> - if (mm) {
> - bprm->mm = NULL;
> - mmdrop(mm);
> - }
> + if (!err)
> + return 0;
>
> + bprm->mm = NULL;
> + mmap_write_unlock(mm);
> + mmdrop(mm);
> return err;
nit, but prefer 'success-oriented-flow' eg invert the 'if (!err)' and
put the error unwind in the {}
> @@ -1545,6 +1532,18 @@ void setup_new_exec(struct linux_binprm * bprm)
> me->mm->task_size = TASK_SIZE;
> mutex_unlock(&me->signal->exec_update_mutex);
> mutex_unlock(&me->signal->cred_guard_mutex);
> +
> + if (!IS_ENABLED(CONFIG_MMU)) {
> + /*
> + * On MMU, setup_arg_pages() wants to access bprm->vma after
> + * this point, so we can't drop the mmap lock yet.
> + * On !MMU, we have neither setup_arg_pages() nor bprm->vma,
> + * so we should drop the lock here.
> + */
> + mmap_write_unlock(bprm->mm);
> + mmput(bprm->mm);
> + bprm->mm = NULL;
> + }
The only thing I dislike about this is how tricky the lock lifetime
is, it all looks correct, but expecting the setup_arg_pages() or
setup_new_exec() to unlock (depending!) is quite tricky.
It feels like it would be clearer to have an explicit function to do
this, like 'release_brp_mm()' indicating that current->mm is now the
only way to get the mm and it must be locked.
Or, more practically, the load_binary functionc can now call
vm_mmap().
Anyhow, it took a bit to study all the parts but I think it looks
right as is.
Jason
next prev parent reply other threads:[~2020-10-20 19:15 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-10-16 22:57 [PATCH resend v3 0/2] Broad write-locking of nascent mm in execve Jann Horn
2020-10-16 22:57 ` [PATCH resend v3 1/2] mmap locking API: Order lock of nascent mm outside lock of live mm Jann Horn
2020-10-16 22:57 ` [PATCH resend v3 2/2] exec: Broadly lock nascent mm until setup_arg_pages() Jann Horn
2020-10-20 19:15 ` Jason Gunthorpe [this message]
2020-11-03 3:53 ` Jann Horn
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20201020191540.GM6219@nvidia.com \
--to=jgg@nvidia.com \
--cc=akpm@linux-foundation.org \
--cc=anton.ivanov@cambridgegreys.com \
--cc=ebiederm@xmission.com \
--cc=jannh@google.com \
--cc=jdike@addtoit.com \
--cc=jhubbard@nvidia.com \
--cc=johannes@sipsolutions.net \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-um@lists.infradead.org \
--cc=mchehab@kernel.org \
--cc=richard@nod.at \
--cc=sakari.ailus@linux.intel.com \
--cc=walken@google.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).