Skip to contentSkip to conversation

Patch "ring-buffer: Force absolute timestamp on discard of event" has been added to the 6.1-stable tree

DISCUSSION Steven Rostedt 2 participants 2 messages Dec 30, 2023 – Dec 31, 2023 1 message not in the archive Stable branch

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

#1 lore raw

Sasha,

Was this automated or did you do this manually?

I'm asking because I was walking through my INBOX to see what FAILED backports I could clean up, and I started on this one:

https://lore.kernel.org/all/2023120938-unclamped-fleshy-688e@gregkh/

I did the cherry pick, fixed up the conflict, but when I tried to commit it, it failed because there was nothing to commit.

This confused me for a bit, and then when I did a git blame, I saw that you had done the fix already.

When you fix a FAILED patch, can you do a reply to the FAILED message that Greg sends out, so that I don't waste my time on trying to fix something that was already fixed?

Thanks!

-- Steve

Quoted 80 lines · On Sun, 10 Dec 2023 14:40:35 -0500 Sasha Levin <[email protected]> wrote:

This is a note to let you know that I've just added the patch titled

ring-buffer: Force absolute timestamp on discard of event

to the 6.1-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
ring-buffer-force-absolute-timestamp-on-discard-of-e.patch
and it can be found in the queue-6.1 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree, please let <[email protected]> know about it.

commit 1249c67fa9a9b3ae207b53fcbefa8dac3acbc308

Author: Steven Rostedt (Google) <[email protected]>
Date:   Wed Dec 6 10:02:44 2023 -0500

ring-buffer: Force absolute timestamp on discard of event

[ Upstream commit b2dd797543cfa6580eac8408dd67fa02164d9e56 ]

There's a race where if an event is discarded from the ring buffer and an interrupt were to happen at that time and insert an event, the time stamp is still used from the discarded event as an offset. This can screw up the timings.

If the event is going to be discarded, set the "before_stamp" to zero. When a new event comes in, it compares the "before_stamp" with the "write_stamp" and if they are not equal, it will insert an absolute timestamp. This will prevent the timings from getting out of sync due to the discarded event.

Cc Masami Hiramatsu <[email protected]>
Cc Mark Rutland <[email protected]>
Cc Mathieu Desnoyers <[email protected]>
Fixes 6f6be606e763f ("ring-buffer: Force before_stamp and write_stamp to be different on discard")
kernel/trace/ring_buffer.c +8 -11
index f3c4bb54a0485..c02a4cb879913 100644
@@ -3025,22 +3025,19 @@ rb_try_to_discard(struct ring_buffer_per_cpu *cpu_buffer,
 			local_read(&bpage->write) & ~RB_WRITE_MASK;
 		unsigned long event_length = rb_event_length(event);
 
+		/*
+		 * For the before_stamp to be different than the write_stamp
+		 * to make sure that the next event adds an absolute
+		 * value and does not rely on the saved write stamp, which
+		 * is now going to be bogus.
+		 */
+		rb_time_set(&cpu_buffer->before_stamp, 0);
+
 		/* Something came in, can't discard */
 		if (!rb_time_cmpxchg(&cpu_buffer->write_stamp,
 				       write_stamp, write_stamp - delta))
 			return 0;
 
-		/*
-		 * It's possible that the event time delta is zero
-		 * (has the same time stamp as the previous event)
-		 * in which case write_stamp and before_stamp could
-		 * be the same. In such a case, force before_stamp
-		 * to be different than write_stamp. It doesn't
-		 * matter what it is, as long as its different.
-		 */
-		if (!delta)
-			rb_time_set(&cpu_buffer->before_stamp, 0);
-
 		/*
 		 * If an event were to come in now, it would see that the
 		 * write_stamp and the before_stamp are different, and assume
maintainer
#2 lore raw

Hey Steve,

Quoted 4 lines · Steven Rostedt

Sasha,

Was this automated or did you do this manually?

This is mostly automated.

Quoted 8 lines

I'm asking because I was walking through my INBOX to see what FAILED backports I could clean up, and I started on this one:

https://lore.kernel.org/all/2023120938-unclamped-fleshy-688e@gregkh/

I did the cherry pick, fixed up the conflict, but when I tried to commit it, it failed because there was nothing to commit.

Quoted 6 lines

This confused me for a bit, and then when I did a git blame, I saw that you had done the fix already.

When you fix a FAILED patch, can you do a reply to the FAILED message that Greg sends out, so that I don't waste my time on trying to fix something that was already fixed?

The process I have is that I go over the stable tagged commits with a week or two delay, and attempt to apply the commits that aren't in the tree yet using the dependency bot.

I can look into how to tie this into Greg's FAILED emails; I used to do the above manually and reply to those mails, but now the process is slightly different and there's no direct connection between the two.

Signature · 2 lines
Thanks,
Sasha