All of lore.kernel.org
 help / color / mirror / Atom feed
From: Vincent Donnefort <vdonnefort@google.com>
To: Steven Rostedt <rostedt@goodmis.org>
Cc: Linux Trace Devel <linux-trace-devel@vger.kernel.org>
Subject: Re: [PATCH] libtracefs: Add ring buffer memory mapping APIs
Date: Mon, 8 Jan 2024 14:25:03 +0000	[thread overview]
Message-ID: <ZZwFvyGvm0C38eBh@google.com> (raw)
In-Reply-To: <20240105152906.743d7e03@gandalf.local.home>

[...]

> +/**
> + * trace_mmap - try to mmap the ring buffer
> + * @fd: The file descriptor to the trace_pipe_raw file
> + * @kbuf: The kbuffer to load the subbuffer to
> + *
> + * Will try to mmap the ring buffer if it is supported, and
> + * if not, will return NULL, otherwise it returns a descriptor
> + * to handle the mapping.
> + */
> +__hidden void *trace_mmap(int fd, struct kbuffer *kbuf)
> +{
> +	struct trace_mmap *tmap;
> +	int page_size;
> +	void *meta;
> +	void *data;
> +
> +	page_size = getpagesize();
> +	meta = mmap(NULL, page_size, PROT_READ, MAP_SHARED, fd, 0);
> +	if (meta == MAP_FAILED)
> +		return NULL;
> +
> +	tmap = calloc(1, sizeof(*tmap));
> +	if (!tmap) {
> +		munmap(meta, page_size);
> +		return NULL;
> +	}
> +
> +	tmap->kbuf = kbuffer_dup(kbuf);
> +	if (!tmap->kbuf) {
> +		munmap(meta, page_size);
> +		free(tmap);
> +	}
> +
> +	tmap->fd = fd;
> +
> +	tmap->map = meta;
> +	tmap->meta_len = tmap->map->meta_page_size;
> +
> +	if (tmap->meta_len > page_size) {
> +		munmap(meta, page_size);
> +		meta = mmap(NULL, tmap->meta_len, PROT_READ, MAP_SHARED, fd, 0);
> +		if (meta == MAP_FAILED) {
> +			kbuffer_free(tmap->kbuf);
> +			free(tmap);
> +			return NULL;
> +		}
> +		tmap->map = meta;
> +	}
> +
> +	tmap->data_pages = meta + tmap->meta_len;
> +
> +	tmap->data_len = tmap->map->subbuf_size * tmap->map->nr_subbufs;
> +
> +	tmap->data = mmap(NULL, tmap->data_len, PROT_READ, MAP_SHARED,
> +			  fd, tmap->meta_len);
> +	if (tmap->data == MAP_FAILED) {
> +		munmap(meta, tmap->meta_len);
> +		kbuffer_free(tmap->kbuf);
> +		free(tmap);
> +		return NULL;
> +	}
> +
> +	tmap->last_idx = tmap->map->reader.id;
> +
> +	data = tmap->data + tmap->map->subbuf_size * tmap->last_idx;
> +	kbuffer_load_subbuffer(kbuf, data);
> +
> +	/*
> +	 * The page could have left over data on it that was already
> +	 * consumed. Move the "read" forward in that case.
> +	 */
> +	if (tmap->map->reader.read) {
> +		int size = kbuffer_start_of_data(kbuf) + tmap->map->reader.read;
> +		char tmpbuf[size];
> +		kbuffer_read_buffer(kbuf, tmpbuf, size);

It does not seem to update the kbuf timestamp. To observe the problem I did:

### Create few events on the page

$ echo 0 > /sys/kernel/tracing/trace
$ <read ring-buffer>
$ cat /proc/uptime | awk '{print $1}' > /sys/kernel/debug/tracing/trace_marker
<...>-2305 279515.453542096     print: tracing_mark_write: 279515.33
<...>-2307 279522.090413680     print: tracing_mark_write: 279521.97
<...>-2309 279522.960932976     print: tracing_mark_write: 279522.85
$ <unmap and close ring-buffer>


### Re-map again the ring-buffer to trigger the fast-forward

$ <read ring-buffer>
  before fast-forward kbuf->timestamp=279515453542096
  after fast-forward kbuf->timestamp=279515453542096
$ cat /proc/uptime | awk '{print $1}' > /sys/kernel/debug/tracing/trace_marker
<...>-2312 279549.725524688     print: tracing_mark_write: 279557.12

The timestamp above is a few seconds off, which I believe might be due to an
outdated kbuf->timestamp.

> +	}
> +
> +	return tmap;
> +}
> +
> +__hidden void trace_unmap(void *mapping)
> +{
> +	struct trace_mmap *tmap = mapping;
> +
> +	munmap(tmap->data, tmap->data_len);
> +	munmap(tmap->map, tmap->meta_len);
> +	kbuffer_free(tmap->kbuf);
> +	free(tmap);
> +}
> +

  parent reply	other threads:[~2024-01-08 14:25 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-01-05 20:29 [PATCH] libtracefs: Add ring buffer memory mapping APIs Steven Rostedt
2024-01-06 13:08 ` [PATCH v2] " Steven Rostedt
2024-01-08 14:25 ` Vincent Donnefort [this message]
2024-01-08 17:16   ` [PATCH] " Steven Rostedt
2024-01-08 17:34     ` Vincent Donnefort
2024-01-23  9:52 ` Vincent Donnefort
2024-01-23 15:15   ` Steven Rostedt
  -- strict thread matches above, loose matches on Subject: below --
2023-12-29  1:11 Steven Rostedt
2024-01-05  9:17 ` Vincent Donnefort
2024-01-05 13:41   ` Steven Rostedt
2024-01-05 14:25     ` Vincent Donnefort
2024-01-05 17:31       ` Steven Rostedt
2024-01-05 18:23         ` Vincent Donnefort
2024-01-05 19:00           ` Steven Rostedt
2024-01-05 20:11         ` Steven Rostedt
2024-01-05 20:22     ` Steven Rostedt

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=ZZwFvyGvm0C38eBh@google.com \
    --to=vdonnefort@google.com \
    --cc=linux-trace-devel@vger.kernel.org \
    --cc=rostedt@goodmis.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.