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
next prev 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 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).