wifi: iwlwifi: rework firmware error handling

In order to later add the ability to do deeper resets of the
device when it crashes, first restructure the firmware error
handling. Instead of having just a single nic_error() method
that handles all, split it:
 - nic_error() just handles and prints the error itself,
 - dump_error() synchronously creates an error dump, and
 - sw_reset() will be called to request doing a SW reset.

This changes the architecture so that the transport is now
responsible for deciding how to do the reset, and therefore
the handling of reprobe if error occurs during reconfig
moves there, which necessitates adding a method there that
notifies the transport that the recovery was completed.

Actually introducing the model under which deeper resets can
be done will be in future patches.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20241227095718.6d4f741ae907.I96a9243e7877808ed6d1bff6967c15d6c24882f0@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

2 files changed, 58 insertions, 86 deletions

diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c b/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c
index b82a2ac50d5b..3e0f34dc6e1a 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c
@@ -1340,6 +1340,7 @@ static void iwl_mvm_restart_complete(struct iwl_mvm *mvm)
 	iwl_mvm_teardown_tdls_peers(mvm);
 
 	IWL_INFO(mvm, "restart completed\n");
+	iwl_trans_finish_sw_reset(mvm->trans);
 }
 
 void iwl_mvm_mac_reconfig_complete(struct ieee80211_hw *hw,
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/ops.c b/drivers/net/wireless/intel/iwlwifi/mvm/ops.c
index ef54ba818d1f..08c59df593b2 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/ops.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/ops.c
@@ -2008,27 +2008,62 @@ static void iwl_mvm_free_skb(struct iwl_op_mode *op_mode, struct sk_buff *skb)
 	ieee80211_free_txskb(mvm->hw, skb);
 }
 
-struct iwl_mvm_reprobe {
-	struct device *dev;
-	struct work_struct work;
-};
+static void iwl_mvm_nic_error(struct iwl_op_mode *op_mode,
+			      enum iwl_fw_error_type type)
+{
+	struct iwl_mvm *mvm = IWL_OP_MODE_GET_MVM(op_mode);
+
+	iwl_abort_notification_waits(&mvm->notif_wait);
+	iwl_dbg_tlv_del_timers(mvm->trans);
+
+	if (type == IWL_ERR_TYPE_CMD_QUEUE_FULL)
+		IWL_ERR(mvm, "Command queue full!\n");
+	else if (!test_bit(STATUS_TRANS_DEAD, &mvm->trans->status) &&
+		 !test_and_clear_bit(IWL_MVM_STATUS_SUPPRESS_ERROR_LOG_ONCE,
+				     &mvm->status))
+		iwl_mvm_dump_nic_error_log(mvm);
+
+	/*
+	 * This should be first thing before trying to collect any
+	 * data to avoid endless loops if any HW error happens while
+	 * collecting debug data.
+	 * It might not actually be true that we'll restart, but the
+	 * setting of the bit doesn't matter if we're going to be
+	 * unbound either.
+	 */
+	if (type != IWL_ERR_TYPE_RESET_HS_TIMEOUT)
+		set_bit(IWL_MVM_STATUS_HW_RESTART_REQUESTED, &mvm->status);
+}
 
-static void iwl_mvm_reprobe_wk(struct work_struct *wk)
+static void iwl_mvm_dump_error(struct iwl_op_mode *op_mode,
+			       struct iwl_fw_error_dump_mode *mode)
 {
-	struct iwl_mvm_reprobe *reprobe;
-
-	reprobe = container_of(wk, struct iwl_mvm_reprobe, work);
-	if (device_reprobe(reprobe->dev))
-		dev_err(reprobe->dev, "reprobe failed!\n");
-	put_device(reprobe->dev);
-	kfree(reprobe);
-	module_put(THIS_MODULE);
+	struct iwl_mvm *mvm = IWL_OP_MODE_GET_MVM(op_mode);
+
+	/* if we come in from opmode we have the mutex held */
+	if (mode->context == IWL_ERR_CONTEXT_FROM_OPMODE) {
+		lockdep_assert_held(&mvm->mutex);
+		iwl_fw_error_collect(&mvm->fwrt, true);
+	} else {
+		mutex_lock(&mvm->mutex);
+		if (mode->context != IWL_ERR_CONTEXT_ABORT)
+			iwl_fw_error_collect(&mvm->fwrt, true);
+		mutex_unlock(&mvm->mutex);
+	}
 }
 
-static void iwl_mvm_nic_restart(struct iwl_mvm *mvm)
+static bool iwl_mvm_sw_reset(struct iwl_op_mode *op_mode,
+			     enum iwl_fw_error_type type)
 {
-	iwl_abort_notification_waits(&mvm->notif_wait);
-	iwl_dbg_tlv_del_timers(mvm->trans);
+	struct iwl_mvm *mvm = IWL_OP_MODE_GET_MVM(op_mode);
+
+	/*
+	 * If the firmware crashes while we're already considering it
+	 * to be dead then don't ask for a restart, that cannot do
+	 * anything useful anyway.
+	 */
+	if (!test_bit(IWL_MVM_STATUS_FIRMWARE_RUNNING, &mvm->status))
+		return false;
 
 	/*
 	 * This is a bit racy, but worst case we tell mac80211 about
@@ -2043,47 +2078,11 @@ static void iwl_mvm_nic_restart(struct iwl_mvm *mvm)
 	iwl_mvm_report_scan_aborted(mvm);
 
 	/*
-	 * If we're restarting already, don't cycle restarts.
 	 * If INIT fw asserted, it will likely fail again.
 	 * If WoWLAN fw asserted, don't restart either, mac80211
 	 * can't recover this since we're already half suspended.
 	 */
-	if (test_bit(IWL_MVM_STATUS_IN_HW_RESTART, &mvm->status)) {
-		struct iwl_mvm_reprobe *reprobe;
-
-		IWL_ERR(mvm,
-			"Firmware error during reconfiguration - reprobe!\n");
-
-		/*
-		 * get a module reference to avoid doing this while unloading
-		 * anyway and to avoid scheduling a work with code that's
-		 * being removed.
-		 */
-		if (!try_module_get(THIS_MODULE)) {
-			IWL_ERR(mvm, "Module is being unloaded - abort\n");
-			return;
-		}
-
-		reprobe = kzalloc(sizeof(*reprobe), GFP_ATOMIC);
-		if (!reprobe) {
-			module_put(THIS_MODULE);
-			return;
-		}
-		reprobe->dev = get_device(mvm->trans->dev);
-		INIT_WORK(&reprobe->work, iwl_mvm_reprobe_wk);
-		schedule_work(&reprobe->work);
-	} else if (test_bit(IWL_MVM_STATUS_HW_RESTART_REQUESTED,
-			    &mvm->status)) {
-		IWL_ERR(mvm, "HW restart already requested, but not started\n");
-	} else if (mvm->fwrt.cur_fw_img == IWL_UCODE_REGULAR &&
-		   mvm->hw_registered &&
-		   !test_bit(STATUS_TRANS_DEAD, &mvm->trans->status)) {
-		/* This should be first thing before trying to collect any
-		 * data to avoid endless loops if any HW error happens while
-		 * collecting debug data.
-		 */
-		set_bit(IWL_MVM_STATUS_HW_RESTART_REQUESTED, &mvm->status);
-
+	if (mvm->fwrt.cur_fw_img == IWL_UCODE_REGULAR && mvm->hw_registered) {
 		if (mvm->fw->ucode_capa.error_log_size) {
 			u32 src_size = mvm->fw->ucode_capa.error_log_size;
 			u32 src_addr = mvm->fw->ucode_capa.error_log_addr;
@@ -2098,48 +2097,18 @@ static void iwl_mvm_nic_restart(struct iwl_mvm *mvm)
 			}
 		}
 
-		iwl_fw_error_collect(&mvm->fwrt, false);
-
-		if (!iwlwifi_mod_params.fw_restart)
-			return;
-
 		if (mvm->fwrt.trans->dbg.restart_required) {
 			IWL_DEBUG_INFO(mvm, "FW restart requested after debug collection\n");
 			mvm->fwrt.trans->dbg.restart_required = false;
 			ieee80211_restart_hw(mvm->hw);
+			return true;
 		} else if (mvm->trans->trans_cfg->device_family <= IWL_DEVICE_FAMILY_8000) {
 			ieee80211_restart_hw(mvm->hw);
+			return true;
 		}
 	}
-}
-
-static void iwl_mvm_nic_error(struct iwl_op_mode *op_mode,
-			      enum iwl_fw_error_type type)
-{
-	struct iwl_mvm *mvm = IWL_OP_MODE_GET_MVM(op_mode);
-
-	if (type == IWL_ERR_TYPE_CMD_QUEUE_FULL)
-		IWL_ERR(mvm, "Command queue full!\n");
-	else if (!test_bit(STATUS_TRANS_DEAD, &mvm->trans->status) &&
-		 !test_and_clear_bit(IWL_MVM_STATUS_SUPPRESS_ERROR_LOG_ONCE,
-				     &mvm->status))
-		iwl_mvm_dump_nic_error_log(mvm);
-
-	/* reset HS timeout is during shutdown, so collect right now */
-	if (type == IWL_ERR_TYPE_RESET_HS_TIMEOUT) {
-		iwl_fw_error_collect(&mvm->fwrt, true);
-		return;
-	}
-
-	/*
-	 * If the firmware crashes while we're already considering it
-	 * to be dead then don't ask for a restart, that cannot do
-	 * anything useful anyway.
-	 */
-	if (!test_bit(IWL_MVM_STATUS_FIRMWARE_RUNNING, &mvm->status))
-		return;
 
-	iwl_mvm_nic_restart(mvm);
+	return false;
 }
 
 static void iwl_op_mode_mvm_time_point(struct iwl_op_mode *op_mode,
@@ -2175,6 +2144,8 @@ static void iwl_op_mode_mvm_device_powered_off(struct iwl_op_mode *op_mode)
 	.hw_rf_kill = iwl_mvm_set_hw_rfkill_state,		\
 	.free_skb = iwl_mvm_free_skb,				\
 	.nic_error = iwl_mvm_nic_error,				\
+	.dump_error = iwl_mvm_dump_error,			\
+	.sw_reset = iwl_mvm_sw_reset,				\
 	.nic_config = iwl_mvm_nic_config,			\
 	/* as we only register one, these MUST be common! */	\
 	.start = iwl_op_mode_mvm_start,				\