lkml.org 
[lkml]   [2013]   [Aug]   [1]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [RFC][PATCH 4/4] tracing/uprobes: Fail to unregister if probe event files are open
On 07/03, Steven Rostedt wrote:
>
> From: "Steven Rostedt (Red Hat)" <rostedt@goodmis.org>

Acked-by: Oleg Nesterov <oleg@redhat.com>


Just a couple of nits in the case you are going to redo this change,

> Modules do with with the module owner set (automated
> from the VFS layer).

This logic is dead, I think.

> The ftrace buffer instances have a ref count added
> to the trace_array when the enabled file is opened

This is too.

> -static void cleanup_all_probes(void)
> +static int cleanup_all_probes(void)
> {
> struct trace_uprobe *tu;
> + int ret = 0;
>
> mutex_lock(&uprobe_lock);
> while (!list_empty(&uprobe_list)) {
> tu = list_entry(uprobe_list.next, struct trace_uprobe, list);
> - unregister_trace_uprobe(tu);
> + ret = unregister_trace_uprobe(tu);
> + if (ret)
> + break;
> }
> mutex_unlock(&uprobe_lock);
> + return ret;
> }

Again, it is not clear what exactly we should do and I won't argue
either way. But note that (with or without this patch) this doesn't
match kprobe's release_all_trace_probes() which checks (tries to,
actually) trace_probe_is_enabled() for every probe first. Perhaps
we should cleanup this later.

> static int probes_open(struct inode *inode, struct file *file)
> {
> + int ret = 0;
> +
> if ((file->f_mode & FMODE_WRITE) && (file->f_flags & O_TRUNC))
> - cleanup_all_probes();
> + ret = cleanup_all_probes();
> + if (ret)
> + return ret;

Cosmetic, but perhaps it would be a bit more clean to move this check
(with "int ret") under if (WRITE && TRUNC) block.

Oleg.



\
 
 \ /
  Last update: 2013-08-01 16:41    [W:0.292 / U:0.172 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site