Received-SPF: pass (google.com: domain of linux-wireless-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
To:     "linux-wireless@vger.kernel.org" <linux-wireless@vger.kernel.org>
From:   Ben Greear <greearb@candelatech.com>
Subject: iwlwifi: tx-fail reported with zero retry count.
Organization: Candela Technologies
Message-ID: <6b920700-e9a2-bbeb-98ce-7a2d09e76975@candelatech.com>
Date:   Tue, 6 Jun 2023 06:36:11 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.10.1
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-MW
Content-Transfer-Encoding: 7bit
Precedence: bulk

It seems that when tx fails (due to TX_STATUS_FAIL_LONG_LIMIT in this case), the
retry-count is sometimes reported as zero.  I would expect that it actually must have retried
many times (15 it seems?).  At least for fail-long-limit failures, is there a known constant amount of
times the firmware retransmits the frame before giving up?  If so, I can just hack that
value into the retry counter in this case.

Debug patch:

[greearb@ben-dt5 iwlwifi]$ git diff
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/tx.c b/drivers/net/wireless/intel/iwlwifi/mvm/tx.c
index 1cddc65dd51e..6544dabb092b 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/tx.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/tx.c
@@ -1929,8 +1929,14 @@ static void iwl_mvm_rx_tx_cmd_single(struct iwl_mvm *mvm,
                 mvm->ethtool_stats.tx_mpdu_attempts += info->status.rates[0].count;
                 mvm->ethtool_stats.tx_mpdu_retry += tx_resp->failure_frame;
                 if (cb.flags & IWL_TX_CB_TXO_USED) {
+                       u32 idx = status & TX_STATUS_MSK;
+
                         mvm->ethtool_stats.txo_tx_mpdu_attempts += info->status.rates[0].count;
                         mvm->ethtool_stats.txo_tx_mpdu_retry += tx_resp->failure_frame;
+                       if (idx != TX_STATUS_SUCCESS) {
+                               pr_info("txo tx status failed: %d  tx_resp->failure_frame: %d\n",
+                                       idx, tx_resp->failure_frame);
+                       }
                 }

                 iwl_mvm_hwrate_to_tx_status(mvm, mvm->fw,


 From dmesg:

Jun 06 06:29:40 ct523c-0b29 kernel: txo tx status failed: 131  tx_resp->failure_frame: 15
Jun 06 06:29:43 ct523c-0b29 kernel: txo tx status failed: 131  tx_resp->failure_frame: 15
Jun 06 06:29:48 ct523c-0b29 kernel: txo tx status failed: 131  tx_resp->failure_frame: 15
Jun 06 06:29:49 ct523c-0b29 kernel: txo tx status failed: 131  tx_resp->failure_frame: 15
Jun 06 06:29:52 ct523c-0b29 kernel: txo tx status failed: 131  tx_resp->failure_frame: 15
Jun 06 06:29:57 ct523c-0b29 kernel: txo tx status failed: 131  tx_resp->failure_frame: 0
Jun 06 06:29:57 ct523c-0b29 kernel: txo tx status failed: 131  tx_resp->failure_frame: 0
Jun 06 06:29:58 ct523c-0b29 kernel: txo tx status failed: 131  tx_resp->failure_frame: 0
Jun 06 06:30:01 ct523c-0b29 kernel: txo tx status failed: 131  tx_resp->failure_frame: 15
Jun 06 06:30:02 ct523c-0b29 kernel: txo tx status failed: 131  tx_resp->failure_frame: 15
Jun 06 06:30:03 ct523c-0b29 kernel: txo tx status failed: 131  tx_resp->failure_frame: 15
....


Thanks,
Ben

-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com