bpf.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jakub Sitnicki <jakub@cloudflare.com>
To: bpf@vger.kernel.org
Cc: netdev@vger.kernel.org, kernel-team@cloudflare.com,
	Eric Dumazet <eric.dumazet@gmail.com>,
	John Fastabend <john.fastabend@gmail.com>
Subject: [PATCH bpf 0/2] Fixes for sock_hash_free
Date: Sun,  7 Jun 2020 22:52:27 +0200	[thread overview]
Message-ID: <20200607205229.2389672-1-jakub@cloudflare.com> (raw)

This series is an attempt to fix a race in sock_hash_free recently reported
by Eric [0]. The race, and a mem leak I found on the way, can be triggered
by the crude reproducer posted below.

[0] https://lore.kernel.org/bpf/6f8bb6d8-bb70-4533-f15b-310db595d334@gmail.com/

Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: John Fastabend <john.fastabend@gmail.com>

--8<--

enum { NUM_SOCKS = 1000 };

static void *close_map(void *map)
{
	close(*(int *)map);
	return NULL;
}

int main(void)
{
	int sock[NUM_SOCKS];
	pthread_t worker;
	int map;
	int i, err;

	map = bpf_create_map(BPF_MAP_TYPE_SOCKHASH, sizeof(int), sizeof(int), NUM_SOCKS, 0);
	if (map < 0)
		error(1, -map, "map create");

	for (i = 0; i < NUM_SOCKS; i++) {
		int fd = socket(AF_INET, SOCK_STREAM, 0);
		if (fd < 0)
			error(1, errno, "socket");

		err = listen(fd, SOMAXCONN);
		if (err)
			error(1, errno, "listen");

		sock[i] = fd;
		err = bpf_map_update_elem(map, &i, &fd, BPF_ANY);
		if (err)
			error(1, errno, "map update");
	}

	err = pthread_create(&worker, NULL, close_map, &map);
	if (err)
		error(1, err, "thread create");

	/* usleep(100); */

	for (int i = 0; i < NUM_SOCKS; i++)
		close(sock[i]);

	pthread_join(worker, NULL);
	return 0;
}
-->8--

Jakub Sitnicki (2):
  bpf, sockhash: Fix memory leak when unlinking sockets in
    sock_hash_free
  bpf, sockhash: Synchronize delete from bucket list on map free

 net/core/sock_map.c | 24 ++++++++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)

-- 
2.25.4


             reply	other threads:[~2020-06-07 20:52 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-07 20:52 Jakub Sitnicki [this message]
2020-06-07 20:52 ` [PATCH bpf 1/2] bpf, sockhash: Fix memory leak when unlinking sockets in sock_hash_free Jakub Sitnicki
2020-06-09 16:56   ` John Fastabend
2020-06-07 20:52 ` [PATCH bpf 2/2] bpf, sockhash: Synchronize delete from bucket list on map free Jakub Sitnicki
2020-06-09 17:09   ` John Fastabend
2020-06-09 18:04     ` Alexei Starovoitov
2020-06-09 20:29       ` John Fastabend

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=20200607205229.2389672-1-jakub@cloudflare.com \
    --to=jakub@cloudflare.com \
    --cc=bpf@vger.kernel.org \
    --cc=eric.dumazet@gmail.com \
    --cc=john.fastabend@gmail.com \
    --cc=kernel-team@cloudflare.com \
    --cc=netdev@vger.kernel.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).