linux-wireless.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: greearb@candelatech.com
To: linux-wireless@vger.kernel.org
Cc: ath10k@lists.infradead.org, Ben Greear <greearb@candelatech.com>
Subject: [PATCH 2/3] ath10k:  Ensure peer_map references are cleaned up.
Date: Thu, 31 Mar 2016 13:59:45 -0700	[thread overview]
Message-ID: <1459457986-29222-2-git-send-email-greearb@candelatech.com> (raw)
In-Reply-To: <1459457986-29222-1-git-send-email-greearb@candelatech.com>

From: Ben Greear <greearb@candelatech.com>

While debugging OS crashes due to firmware crashes, I enabled
kasan, and it noticed that peer objects were being used-after-freed.

Looks like there are two places we could be leaving stale references
in the peer-map, so clean that up.

Signed-off-by: Ben Greear <greearb@candelatech.com>
---
 drivers/net/wireless/ath/ath10k/mac.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/drivers/net/wireless/ath/ath10k/mac.c b/drivers/net/wireless/ath/ath10k/mac.c
index 0a81ca2..07b155d 100644
--- a/drivers/net/wireless/ath/ath10k/mac.c
+++ b/drivers/net/wireless/ath/ath10k/mac.c
@@ -794,6 +794,7 @@ static void ath10k_peer_cleanup(struct ath10k *ar, u32 vdev_id)
 {
 	struct ath10k_peer *peer, *tmp;
 	int peer_id;
+	int i;
 
 	lockdep_assert_held(&ar->conf_mutex);
 
@@ -812,6 +813,15 @@ static void ath10k_peer_cleanup(struct ath10k *ar, u32 vdev_id)
 			ar->peer_map[peer_id] = NULL;
 		}
 
+		/* Double check that peer is properly un-referenced from the peer_map */
+		for (i = 0; i < ARRAY_SIZE(ar->peer_map); i++) {
+			if (ar->peer_map[i] == peer) {
+				ath10k_warn(ar, "ERROR:  Faile to properly clean up peer: %p %pM, idx: %d, will fix.\n",
+					    peer, peer->addr, i);
+				ar->peer_map[i] = NULL;
+			}
+		}
+
 		list_del(&peer->list);
 		kfree(peer);
 		ar->num_peers--;
@@ -840,6 +850,7 @@ void ath10k_dump_peer_info(struct ath10k *ar)
 static void ath10k_peer_cleanup_all(struct ath10k *ar)
 {
 	struct ath10k_peer *peer, *tmp;
+	int i;
 
 	lockdep_assert_held(&ar->conf_mutex);
 
@@ -850,6 +861,11 @@ static void ath10k_peer_cleanup_all(struct ath10k *ar)
 		list_del(&peer->list);
 		kfree(peer);
 	}
+
+	/* Clean up peer-map */
+	for (i = 0; i < ARRAY_SIZE(ar->peer_map); i++)
+		ar->peer_map[i] = NULL;
+
 	spin_unlock_bh(&ar->data_lock);
 
 	ar->num_peers = 0;
-- 
2.4.3


  reply	other threads:[~2016-03-31 20:59 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-03-31 20:59 [PATCH 1/3] ath10k: Ensure txrx-compl-task is stopped when cleaning htt-tx greearb
2016-03-31 20:59 ` greearb [this message]
2016-04-01  6:18   ` [PATCH 2/3] ath10k: Ensure peer_map references are cleaned up Michal Kazior
2016-03-31 20:59 ` [PATCH 3/3] ath10k: Add BUG_ON if we over-write peer-map pointer greearb
2016-04-01  6:09   ` Michal Kazior
2016-04-01 14:04     ` Ben Greear
2016-04-01  6:12 ` [PATCH 1/3] ath10k: Ensure txrx-compl-task is stopped when cleaning htt-tx Michal Kazior
2016-04-01 20:36   ` Ben Greear

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=1459457986-29222-2-git-send-email-greearb@candelatech.com \
    --to=greearb@candelatech.com \
    --cc=ath10k@lists.infradead.org \
    --cc=linux-wireless@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).