Linux Trace Kernel
 help / color / mirror / Atom feed
From: Crystal Wood <crwood@redhat.com>
To: Steven Rostedt <rostedt@goodmis.org>
Cc: Tomas Glozar <tglozar@redhat.com>, John Kacur <jkacur@redhat.com>,
	linux-trace-kernel@vger.kernel.org,
	Crystal Wood <crwood@redhat.com>
Subject: [PATCH 4/4] tracing/osnoise: Take trace_types_lock in timerlat_fd_open
Date: Mon, 24 Aug 2026 16:15:44 -0500	[thread overview]
Message-ID: <20260824211544.3984835-5-crwood@redhat.com> (raw)
In-Reply-To: <20260824211544.3984835-1-crwood@redhat.com>

This prevents a user fd from being opened when osnoise_options_write()
has already stopped threads but not yet set OSN_WORKLOAD, and matches what
all callers of start_kthread() do.

The original intent was probably for interface_lock to cover things like
this, but it never covered the full race window, and currently can't
due to lock order issues stemming from using interface_lock in
osnoise_sleep().  There does not appear to be any remaining reason
to take interface_lock in timerlat_fd_open(), so remove it.

Signed-off-by: Crystal Wood <crwood@redhat.com>
---
Do we really need that interface_lock in osnoise_sleep()?  Technically
we should probably be grabbing it every time we access osnoise_data, but
we don't seem to do so anywhere else.  Perhaps this was due to concerns
about the atomicity of "period - runtime", but that's just synchronization
theater given that userspace has no way of writing to both atomically.

In any case, we shouldn't use the same lock for protecting osnoise_data
members that we use for major reconfiguration procedures.
---
 kernel/trace/trace_osnoise.c | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/kernel/trace/trace_osnoise.c b/kernel/trace/trace_osnoise.c
index 0f27af6ade3e..3d12ce8a591e 100644
--- a/kernel/trace/trace_osnoise.c
+++ b/kernel/trace/trace_osnoise.c
@@ -2466,16 +2466,14 @@ static int timerlat_fd_open(struct inode *inode, struct file *file)
 	long cpu = (long) inode->i_cdev;
 	int ret = 0;
 
-	mutex_lock(&interface_lock);
+	guard(mutex)(&trace_types_lock);
 
 	/*
 	 * This file is accessible only if timerlat is enabled, and
 	 * NO_OSNOISE_WORKLOAD is set.
 	 */
-	if (!timerlat_enabled() || test_bit(OSN_WORKLOAD, &osnoise_options)) {
-		mutex_unlock(&interface_lock);
+	if (!timerlat_enabled() || test_bit(OSN_WORKLOAD, &osnoise_options))
 		return -EINVAL;
-	}
 
 	migrate_disable();
 
@@ -2524,7 +2522,6 @@ static int timerlat_fd_open(struct inode *inode, struct file *file)
 	 * Setup is done.
 	 */
 	mutex_unlock(&osn_var->lock);
-	mutex_unlock(&interface_lock);
 
 	tlat = this_cpu_tmr_var();
 	tlat->count = 0;
@@ -2536,7 +2533,6 @@ static int timerlat_fd_open(struct inode *inode, struct file *file)
 
 err:
 	mutex_unlock(&osn_var->lock);
-	mutex_unlock(&interface_lock);
 	migrate_enable();
 	return ret;
 };
-- 
2.54.0


  parent reply	other threads:[~2026-08-24 21:16 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-24 21:15 [PATCH 0/4] tracing/osnoise: Synchronization fixes Crystal Wood
2026-08-24 21:15 ` [PATCH 1/4] tracing/osnoise: Per-cpu mutex and fd detachment Crystal Wood
2026-08-24 21:30   ` sashiko-bot
2026-08-24 21:15 ` [PATCH 2/4] tracing/osnoise: timerlat_main: Disable migration before per-cpu access Crystal Wood
2026-08-24 21:29   ` sashiko-bot
2026-08-24 21:15 ` [PATCH 3/4] tracing/osnoise: start_kthread: Always check OSN_WORKLOAD Crystal Wood
2026-08-24 21:15 ` Crystal Wood [this message]
2026-08-24 21:24   ` [PATCH 4/4] tracing/osnoise: Take trace_types_lock in timerlat_fd_open sashiko-bot

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=20260824211544.3984835-5-crwood@redhat.com \
    --to=crwood@redhat.com \
    --cc=jkacur@redhat.com \
    --cc=linux-trace-kernel@vger.kernel.org \
    --cc=rostedt@goodmis.org \
    --cc=tglozar@redhat.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