regressions.lists.linux.dev archive mirror
 help / color / mirror / Atom feed
* [PATCH] wifi: iwlwifi: mvm: spin_lock_bh() to fix lockdep regression
@ 2023-06-09 21:29 Hugh Dickins
  0 siblings, 0 replies; only message in thread
From: Hugh Dickins @ 2023-06-09 21:29 UTC (permalink / raw)
  To: Gregory Greenman
  Cc: Ariel Malamud, Johannes Berg, linux-wireless, linux-kernel, regressions

Lockdep on 6.4-rc on ThinkPad X1 Carbon 5th says
=====================================================
WARNING: SOFTIRQ-safe -> SOFTIRQ-unsafe lock order detected
6.4.0-rc5 #1 Not tainted
-----------------------------------------------------
kworker/3:1/49 [HC0[0]:SC0[4]:HE1:SE0] is trying to acquire:
ffff8881066fa368 (&mvm_sta->deflink.lq_sta.rs_drv.pers.lock){+.+.}-{2:2}, at: rs_drv_get_rate+0x46/0xe7

and this task is already holding:
ffff8881066f80a8 (&sta->rate_ctrl_lock){+.-.}-{2:2}, at: rate_control_get_rate+0xbd/0x126
which would create a new lock dependency:
 (&sta->rate_ctrl_lock){+.-.}-{2:2} -> (&mvm_sta->deflink.lq_sta.rs_drv.pers.lock){+.+.}-{2:2}

but this new dependency connects a SOFTIRQ-irq-safe lock:
 (&sta->rate_ctrl_lock){+.-.}-{2:2}
etc. etc. etc.

Changing the spin_lock() in rs_drv_get_rate() to spin_lock_bh() was not
enough to pacify lockdep, but changing them all on pers.lock has worked.

Fixes: a8938bc881d2 ("wifi: iwlwifi: mvm: Add locking to the rate read flow")
Signed-off-by: Hugh Dickins <hughd@google.com>
---

 drivers/net/wireless/intel/iwlwifi/mvm/rs.c |   12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

--- 6.4-rc5/drivers/net/wireless/intel/iwlwifi/mvm/rs.c
+++ linux/drivers/net/wireless/intel/iwlwifi/mvm/rs.c
@@ -2692,7 +2692,7 @@ static void rs_drv_get_rate(void *mvm_r,
 
 	lq_sta = mvm_sta;
 
-	spin_lock(&lq_sta->pers.lock);
+	spin_lock_bh(&lq_sta->pers.lock);
 	iwl_mvm_hwrate_to_tx_rate_v1(lq_sta->last_rate_n_flags,
 				     info->band, &info->control.rates[0]);
 	info->control.rates[0].count = 1;
@@ -2707,7 +2707,7 @@ static void rs_drv_get_rate(void *mvm_r,
 		iwl_mvm_hwrate_to_tx_rate_v1(last_ucode_rate, info->band,
 					     &txrc->reported_rate);
 	}
-	spin_unlock(&lq_sta->pers.lock);
+	spin_unlock_bh(&lq_sta->pers.lock);
 }
 
 static void *rs_drv_alloc_sta(void *mvm_rate, struct ieee80211_sta *sta,
@@ -3264,11 +3264,11 @@ void iwl_mvm_rs_tx_status(struct iwl_mvm
 	/* If it's locked we are in middle of init flow
 	 * just wait for next tx status to update the lq_sta data
 	 */
-	if (!spin_trylock(&mvmsta->deflink.lq_sta.rs_drv.pers.lock))
+	if (!spin_trylock_bh(&mvmsta->deflink.lq_sta.rs_drv.pers.lock))
 		return;
 
 	__iwl_mvm_rs_tx_status(mvm, sta, tid, info, ndp);
-	spin_unlock(&mvmsta->deflink.lq_sta.rs_drv.pers.lock);
+	spin_unlock_bh(&mvmsta->deflink.lq_sta.rs_drv.pers.lock);
 }
 
 #ifdef CONFIG_MAC80211_DEBUGFS
@@ -4117,9 +4117,9 @@ void iwl_mvm_rs_rate_init(struct iwl_mvm
 	} else {
 		struct iwl_mvm_sta *mvmsta = iwl_mvm_sta_from_mac80211(sta);
 
-		spin_lock(&mvmsta->deflink.lq_sta.rs_drv.pers.lock);
+		spin_lock_bh(&mvmsta->deflink.lq_sta.rs_drv.pers.lock);
 		rs_drv_rate_init(mvm, sta, band);
-		spin_unlock(&mvmsta->deflink.lq_sta.rs_drv.pers.lock);
+		spin_unlock_bh(&mvmsta->deflink.lq_sta.rs_drv.pers.lock);
 	}
 }
 

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2023-06-09 21:29 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-06-09 21:29 [PATCH] wifi: iwlwifi: mvm: spin_lock_bh() to fix lockdep regression Hugh Dickins

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).