Re: [PATCH 5.10 1/2] iwlwifi: Fix softirq/hardirq disabling in iwl_pcie_gen2_enqueue_hcmd()

[Jari Ruusu]

On Sunday, May 2, 2021 2:03 PM, Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> wrote:
> If you could provide backported patches to those kernels you think this
> is needed to, I can take them directly. Otherwise running sed isn't
> always the easiest thing to do on my end :)

iwlwifi: Fix softirq/hardirq disabling in iwl_pcie_enqueue_hcmd()
upstream commit 2800aadc18a64c96b051bcb7da8a7df7d505db3f,
backport for linux-5.4.y and linux-4.19.y (booted and ping tested)
Signed-off-by: Jari Ruusu <jariruusu@xxxxxxxxxxxxxx>

--- a/drivers/net/wireless/intel/iwlwifi/pcie/tx.c
+++ b/drivers/net/wireless/intel/iwlwifi/pcie/tx.c
@@ -1544,6 +1544,7 @@ static int iwl_pcie_enqueue_hcmd(struct iwl_trans *trans,
 	u32 cmd_pos;
 	const u8 *cmddata[IWL_MAX_CMD_TBS_PER_TFD];
 	u16 cmdlen[IWL_MAX_CMD_TBS_PER_TFD];
+	unsigned long flags2;

 	if (WARN(!trans->wide_cmd_header &&
 		 group_id > IWL_ALWAYS_LONG_GROUP,
@@ -1627,10 +1628,10 @@ static int iwl_pcie_enqueue_hcmd(struct iwl_trans *trans,
 		goto free_dup_buf;
 	}

-	spin_lock_bh(&txq->lock);
+	spin_lock_irqsave(&txq->lock, flags2);

 	if (iwl_queue_space(trans, txq) < ((cmd->flags & CMD_ASYNC) ? 2 : 1)) {
-		spin_unlock_bh(&txq->lock);
+		spin_unlock_irqrestore(&txq->lock, flags2);

 		IWL_ERR(trans, "No space in command queue\n");
 		iwl_op_mode_cmd_queue_full(trans->op_mode);
@@ -1791,7 +1792,7 @@ static int iwl_pcie_enqueue_hcmd(struct iwl_trans *trans,
 	spin_unlock_irqrestore(&trans_pcie->reg_lock, flags);

  out:
-	spin_unlock_bh(&txq->lock);
+	spin_unlock_irqrestore(&txq->lock, flags2);
  free_dup_buf:
 	if (idx < 0)
 		kfree(dup_buf);

--
Jari Ruusu  4096R/8132F189 12D6 4C3A DCDA 0AA4 27BD  ACDF F073 3C80 8132 F189