public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Oleg Nesterov <oleg@redhat.com>
To: Steven Rostedt <rostedt@goodmis.org>,
	Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Alexander Z Lam <azl@google.com>,
	Arnaldo Carvalho de Melo <acme@ghostprotocols.net>,
	David Sharp <dhsharp@google.com>,
	Frederic Weisbecker <fweisbec@gmail.com>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Ingo Molnar <mingo@redhat.com>,
	Peter Zijlstra <a.p.zijlstra@chello.nl>,
	Srikar Dronamraju <srikar@linux.vnet.ibm.com>,
	Vaibhav Nagarnaik <vnagarnaik@google.com>,
	"zhangwei(Jovi)" <jovi.zhangwei@huawei.com>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 0/6] tracing: open/delete fixes
Date: Sun, 28 Jul 2013 20:34:07 +0200	[thread overview]
Message-ID: <20130728183407.GA16723@redhat.com> (raw)
In-Reply-To: <20130726172513.GA3586@redhat.com>

On 07/26, Oleg Nesterov wrote:
>
> Testing: seems to work, but assumes that debugs was fixed too. Hopefully
> "debugfs: debugfs_remove_recursive() must not rely on list_empty(d_subdirs)"
> is fine.
>
> (We should probably make debugfs_remove_recursive() to return the error
>  and trace_events.c should warn if it fails)
>
> However: I am still testing this patches, so this is still mostly for
> review. I'll report if I find anything,

I tried to play more with these patches, and didn't find anything wrong.

However, after I re-read 6/6 I think it should be updated:

	- remove_event_file_dir() needs to check file->dir != NULL

	  it can be NULL if event_create_dir()->debugfs_create_dir()
	  fails, or if event_create_dir() fails before.

	- I've also added spin_lock(d_lock) and d_inode != NULL check.

	  I do not think this is needed. Afaics __create_file() can't
	  leave the !debugfs_positive() file on ->d_subdirs list if
	  debugfs_get_inode() fails, dput()->d_kill() should remove it.

	  But I do not understand this code enough, and debugfs checks
	  d_inode != NULL all the time.

I am sending "PATCH v3 6/6" in reply to "PATCH v2 6/6". See the diff
below.

So I am waiting for your review. If you and Masami agree with these
changes I'll resend other fixes (1 from me and 2 from Steven) on top
of this series.

Oleg.

--- x/kernel/trace/trace_events.c
+++ x/kernel/trace/trace_events.c
@@ -431,12 +431,18 @@ static void remove_event_file_dir(struct
 	struct dentry *dir = file->dir;
 	struct dentry *child;
 
-	/* ->i_mutex is not needed, nobody can create/remove a file */
-	list_for_each_entry(child, &dir->d_subdirs, d_u.d_child)
-		child->d_inode->i_private = NULL;
+	if (dir) {
+		spin_lock(&dir->d_lock);	/* probably unneeded */
+		list_for_each_entry(child, &dir->d_subdirs, d_u.d_child) {
+			if (child->d_inode)	/* probably unneeded */
+				child->d_inode->i_private = NULL;
+		}
+		spin_unlock(&dir->d_lock);
+
+		debugfs_remove_recursive(dir);
+	}
 
 	list_del(&file->list);
-	debugfs_remove_recursive(dir);
 	remove_subsystem(file->system);
 	kmem_cache_free(file_cachep, file);
 }


  parent reply	other threads:[~2013-07-28 18:39 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-07-26 17:25 [PATCH v2 0/6] tracing: open/delete fixes Oleg Nesterov
2013-07-26 17:25 ` [PATCH v2 1/6] tracing: Turn event/id->i_private into call->event.type Oleg Nesterov
2013-07-30  1:29   ` Masami Hiramatsu
2013-07-26 17:25 ` [PATCH v2 2/6] tracing: Change event_enable/disable_read() to verify i_private != NULL Oleg Nesterov
2013-07-30  1:31   ` Masami Hiramatsu
2013-07-30  1:42     ` Steven Rostedt
2013-07-26 17:25 ` [PATCH v2 3/6] tracing: Change event_filter_read/write " Oleg Nesterov
2013-07-30  1:34   ` Masami Hiramatsu
2013-07-26 17:25 ` [PATCH v2 4/6] tracing: Change f_start() to take event_mutex and " Oleg Nesterov
2013-07-30  1:33   ` Masami Hiramatsu
2013-07-26 17:25 ` [PATCH v2 5/6] tracing: Introduce remove_event_file_dir() Oleg Nesterov
2013-07-30  1:33   ` Masami Hiramatsu
2013-07-26 17:25 ` [PATCH v2 6/6] tracing: Change remove_event_file_dir() to clear "d_subdirs"->i_private Oleg Nesterov
2013-07-28 18:35   ` [PATCH v3 " Oleg Nesterov
2013-07-30  1:36     ` Masami Hiramatsu
2013-07-30  2:01       ` Steven Rostedt
2013-07-29  9:43   ` [PATCH v2 " Masami Hiramatsu
2013-07-29 14:21     ` Oleg Nesterov
2013-07-30  1:28       ` Masami Hiramatsu
2013-07-30  1:59         ` Steven Rostedt
2013-07-28 18:34 ` Oleg Nesterov [this message]
2013-07-29 11:36 ` [PATCH v2 0/6] tracing: open/delete fixes Masami Hiramatsu
2013-07-29 14:43   ` Oleg Nesterov

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=20130728183407.GA16723@redhat.com \
    --to=oleg@redhat.com \
    --cc=a.p.zijlstra@chello.nl \
    --cc=acme@ghostprotocols.net \
    --cc=azl@google.com \
    --cc=dhsharp@google.com \
    --cc=fweisbec@gmail.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=jovi.zhangwei@huawei.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=masami.hiramatsu.pt@hitachi.com \
    --cc=mingo@redhat.com \
    --cc=rostedt@goodmis.org \
    --cc=srikar@linux.vnet.ibm.com \
    --cc=vnagarnaik@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