linux-kernel - Re: [PATCH v3 1/1] dcache: Translating dentry into pathname without taking rename

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] [thread-next>] [day] [month] [year] [list]

Message-ID: <CA+55aFw8UuEmKsSdGZPUmcekZheNimEt4U81i7Bz0LipGzsFmg@mail.gmail.com>
Date:	Fri, 6 Sep 2013 13:52:49 -0700
From:	Linus Torvalds <torvalds@...ux-foundation.org>
To:	Waiman Long <Waiman.Long@...com>
Cc:	Alexander Viro <viro@...iv.linux.org.uk>,
	linux-fsdevel <linux-fsdevel@...r.kernel.org>,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
	"Chandramouleeswaran, Aswin" <aswin@...com>,
	"Norton, Scott J" <scott.norton@...com>,
	George Spelvin <linux@...izon.com>,
	John Stoffel <john@...ffel.org>
Subject: Re: [PATCH v3 1/1] dcache: Translating dentry into pathname without
 taking rename_lock

On Fri, Sep 6, 2013 at 9:08 AM, Waiman Long <Waiman.Long@...com> wrote:
>
> This patch will replace the writer's write_seqlock/write_sequnlock
> sequence of the rename_lock of the callers of the prepend_path() and
> __dentry_path() functions with the reader's read_seqbegin/read_seqretry
> sequence within these 2 functions.

Ok, this actually looks really good.

I do have one comment, from just reading the patch:

I would really like the stuff inside the

   restart:
      bptr = *buffer;
      blen = *buflen;
      if (retry_cnt) {
            seq = read_seqbegin(&rename_lock);
            rcu_read_lock();
      } else
            write_seqlock(&rename_lock);

      ... guts of path generation ...

      if (retry_cnt) {
            retry_cnt--;
            rcu_read_unlock();
            if (read_seqretry(&rename_lock, seq))
                  goto restart;
      } else
            write_sequnlock(&rename_lock);

could possible be done as a separate function?

Alternatively (or perhaps additionally), maybe the locking could be
done as an inline function too (taking the retry count as an argument)
to make things a bit more easy to understand.

Right now there is a lot of fairly subtle things going on in that
__dentry_path() function. It's not a huge function, but I think that
"while()" loop inside that locking could be done as its own function
and make it even more readable.

But I could already apply this as-is, so it's not a big deal.

Al - do you have comments? Do you want to take this through your tree,
or are you working on other things? I can take this directly too..

                 Linus
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/