All of lore.kernel.org
 help / color / mirror / Atom feed
From: Maya Erez <merez@codeaurora.org>
To: Kalle Valo <kvalo@codeaurora.org>
Cc: Ahmad Masri <amasri@codeaurora.org>,
	linux-wireless@vger.kernel.org, wil6210@qti.qualcomm.com,
	Maya Erez <merez@codeaurora.org>
Subject: [PATCH 16/21] wil6210: fix report of rx packet checksum in edma mode
Date: Fri, 22 Feb 2019 16:21:14 +0200	[thread overview]
Message-ID: <1550845279-16103-17-git-send-email-merez@codeaurora.org> (raw)
In-Reply-To: <1550845279-16103-1-git-send-email-merez@codeaurora.org>

From: Ahmad Masri <amasri@codeaurora.org>

Update the rx packet checksum of received packet according to edma
HW spec:

No need to calculate checksum in the following cases:
L4_status=0 and L3_status=0 - No L3 and no L4 known protocols found
L4_status=0 and L3_status=1 - L3 was found, and checksum check passed.
No known L4 protocol was found.
L4_status=1 - L4 was found, and checksum check passed.

Recalculate checksum in the following cases:
L4_status=3 and L3_status=1 - It means that L3 protocol was found,
and checksum passed, but L4 checksum failed.
L4_status=3 and L3_status=2	- Both L3 and L4 checksum check failed.

Signed-off-by: Ahmad Masri <amasri@codeaurora.org>
Signed-off-by: Maya Erez <merez@codeaurora.org>
---
 drivers/net/wireless/ath/wil6210/txrx_edma.c | 21 +-------------
 drivers/net/wireless/ath/wil6210/txrx_edma.h | 41 +++++++++++++++++++++++++++-
 2 files changed, 41 insertions(+), 21 deletions(-)

diff --git a/drivers/net/wireless/ath/wil6210/txrx_edma.c b/drivers/net/wireless/ath/wil6210/txrx_edma.c
index 48098ae..492f7bd 100644
--- a/drivers/net/wireless/ath/wil6210/txrx_edma.c
+++ b/drivers/net/wireless/ath/wil6210/txrx_edma.c
@@ -807,18 +807,9 @@ static int wil_rx_error_check_edma(struct wil6210_priv *wil,
 				   struct sk_buff *skb,
 				   struct wil_net_stats *stats)
 {
-	int error;
 	int l2_rx_status;
-	int l3_rx_status;
-	int l4_rx_status;
 	void *msg = wil_skb_rxstatus(skb);
 
-	error = wil_rx_status_get_error(msg);
-	if (!error) {
-		skb->ip_summed = CHECKSUM_UNNECESSARY;
-		return 0;
-	}
-
 	l2_rx_status = wil_rx_status_get_l2_rx_status(msg);
 	if (l2_rx_status != 0) {
 		wil_dbg_txrx(wil, "L2 RX error, l2_rx_status=0x%x\n",
@@ -847,17 +838,7 @@ static int wil_rx_error_check_edma(struct wil6210_priv *wil,
 		return -EFAULT;
 	}
 
-	l3_rx_status = wil_rx_status_get_l3_rx_status(msg);
-	l4_rx_status = wil_rx_status_get_l4_rx_status(msg);
-	if (!l3_rx_status && !l4_rx_status)
-		skb->ip_summed = CHECKSUM_UNNECESSARY;
-	/* If HW reports bad checksum, let IP stack re-check it
-	 * For example, HW don't understand Microsoft IP stack that
-	 * mis-calculates TCP checksum - if it should be 0x0,
-	 * it writes 0xffff in violation of RFC 1624
-	 */
-	else
-		stats->rx_csum_err++;
+	skb->ip_summed = wil_rx_status_get_checksum(msg, stats);
 
 	return 0;
 }
diff --git a/drivers/net/wireless/ath/wil6210/txrx_edma.h b/drivers/net/wireless/ath/wil6210/txrx_edma.h
index 343516a..9cf9ecf 100644
--- a/drivers/net/wireless/ath/wil6210/txrx_edma.h
+++ b/drivers/net/wireless/ath/wil6210/txrx_edma.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2012-2016,2018, The Linux Foundation. All rights reserved.
+ * Copyright (c) 2012-2016,2018-2019, The Linux Foundation. All rights reserved.
  *
  * Permission to use, copy, modify, and/or distribute this software for any
  * purpose with or without fee is hereby granted, provided that the above
@@ -511,6 +511,45 @@ static inline int wil_rx_status_get_l4_rx_status(void *msg)
 			    5, 6);
 }
 
+/* L4	L3	Expected result
+ * 0	0	Ok. No L3 and no L4 known protocols found.
+ *		Treated as L2 packet. (no offloads on this packet)
+ * 0	1	Ok. It means that L3 was found, and checksum check passed.
+ *		No known L4 protocol was found.
+ * 0	2	It means that L3 protocol was found, and checksum check failed.
+ *		No L4 known protocol was found.
+ * 1	any	Ok. It means that L4 was found, and checksum check passed.
+ * 3	0	Not a possible scenario.
+ * 3	1	Recalculate. It means that L3 protocol was found, and checksum
+ *		passed. But L4 checksum failed. Need to see if really failed,
+ *		or due to fragmentation.
+ * 3	2	Both L3 and L4 checksum check failed.
+ */
+static inline int wil_rx_status_get_checksum(void *msg,
+					     struct wil_net_stats *stats)
+{
+	int l3_rx_status = wil_rx_status_get_l3_rx_status(msg);
+	int l4_rx_status = wil_rx_status_get_l4_rx_status(msg);
+
+	if (l4_rx_status == 1)
+		return CHECKSUM_UNNECESSARY;
+
+	if (l4_rx_status == 0 && l3_rx_status == 1)
+		return CHECKSUM_UNNECESSARY;
+
+	if (l3_rx_status == 0 && l4_rx_status == 0)
+		/* L2 packet */
+		return CHECKSUM_NONE;
+
+	/* If HW reports bad checksum, let IP stack re-check it
+	 * For example, HW doesn't understand Microsoft IP stack that
+	 * mis-calculates TCP checksum - if it should be 0x0,
+	 * it writes 0xffff in violation of RFC 1624
+	 */
+	stats->rx_csum_err++;
+	return CHECKSUM_NONE;
+}
+
 static inline int wil_rx_status_get_security(void *msg)
 {
 	return WIL_GET_BITS(((struct wil_rx_status_compressed *)msg)->d0,
-- 
1.9.1


  parent reply	other threads:[~2019-02-22 14:21 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-02-22 14:20 [PATCH 00/21] wil6210 patches Maya Erez
2019-02-22 14:20 ` [PATCH 01/21] wil6210: remove rtap_include_phy_info module param Maya Erez
2019-02-28  9:25   ` Kalle Valo
2019-02-22 14:21 ` [PATCH 02/21] wil6210: add option to drop Tx packets when Tx ring is full Maya Erez
2019-02-22 14:21 ` [PATCH 03/21] wil6210: support up to 20 stations in AP mode Maya Erez
2019-02-28  9:17   ` Kalle Valo
2019-02-28 10:36     ` merez
2019-02-22 14:21 ` [PATCH 04/21] wil6210: accessing 802.3 addresses via utility functions Maya Erez
2019-02-22 14:21 ` [PATCH 05/21] wil6210: fix invalid sta statistics update Maya Erez
2019-02-22 14:21 ` [PATCH 06/21] wil6210: ignore HALP ICR if already handled Maya Erez
2019-02-22 14:21 ` [PATCH 07/21] wil6210: check null pointer in _wil_cfg80211_merge_extra_ies Maya Erez
2019-02-22 14:21 ` [PATCH 08/21] wil6210: align to latest auto generated wmi.h Maya Erez
2019-04-03 12:43   ` Kalle Valo
2019-02-22 14:21 ` [PATCH 09/21] wil6210: prevent device memory access while in reset or suspend Maya Erez
2019-02-22 14:21 ` [PATCH 10/21] wil6210: increase PCP stop command timeout Maya Erez
2019-02-22 14:21 ` [PATCH 11/21] wil6210: do not set BIT_USER_SUPPORT_T_POWER_ON_0 in Talyn-MB Maya Erez
2019-02-22 14:21 ` [PATCH 12/21] wil6210: update WIL_MCS_MAX to 15 Maya Erez
2019-02-22 14:21 ` [PATCH 13/21] wil6210: check mid is valid Maya Erez
2019-02-22 14:21 ` [PATCH 14/21] wil6210: use OEM MAC address from OTP Maya Erez
2019-02-22 14:21 ` [PATCH 15/21] wil6210: free edma_rx_swtail upon reset Maya Erez
2019-02-22 14:21 ` Maya Erez [this message]
2019-04-03 12:46   ` [PATCH 16/21] wil6210: fix report of rx packet checksum in edma mode Kalle Valo
2019-02-22 14:21 ` [PATCH 17/21] wil6210: fix return code of wmi_mgmt_tx and wmi_mgmt_tx_ext Maya Erez
2019-02-22 14:21 ` [PATCH 18/21] wil6210: prevent access to RGF_CAF_ICR in Talyn Maya Erez
2019-02-22 14:21 ` [PATCH 19/21] wil6210: add support for ucode tracing Maya Erez
2019-02-22 14:21 ` [PATCH 20/21] wil6210: reset buff id in status message after completion Maya Erez
2019-02-22 14:21 ` [PATCH 21/21] wil6210: print error in FW and board files load failures Maya Erez
2019-02-28  9:05 ` [PATCH 00/21] wil6210 patches Kalle Valo

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=1550845279-16103-17-git-send-email-merez@codeaurora.org \
    --to=merez@codeaurora.org \
    --cc=amasri@codeaurora.org \
    --cc=kvalo@codeaurora.org \
    --cc=linux-wireless@vger.kernel.org \
    --cc=wil6210@qti.qualcomm.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.