From: Vladimir Oltean <vladimir.oltean@nxp.com>
To: Wei Fang <wei.fang@nxp.com>
Cc: claudiu.manoil@nxp.com, xiaoning.wang@nxp.com,
andrew+netdev@lunn.ch, davem@davemloft.net, edumazet@google.com,
kuba@kernel.org, pabeni@redhat.com, ioana.ciornei@nxp.com,
yangbo.lu@nxp.com, michal.swiatkowski@linux.intel.com,
netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
imx@lists.linux.dev, stable@vger.kernel.org
Subject: Re: [PATCH v2 net 2/9] net: enetc: correct the tx_swbd statistics
Date: Thu, 20 Feb 2025 18:01:23 +0200 [thread overview]
Message-ID: <20250220160123.5evmuxlbuzo7djgr@skbuf> (raw)
In-Reply-To: <20250219054247.733243-3-wei.fang@nxp.com>
On Wed, Feb 19, 2025 at 01:42:40PM +0800, Wei Fang wrote:
> When creating a TSO header, if the skb is VLAN tagged, the extended BD
> will be used and the 'count' should be increased by 2 instead of 1.
> Otherwise, when an error occurs, less tx_swbd will be freed than the
> actual number.
>
> Fixes: fb8629e2cbfc ("net: enetc: add support for software TSO")
> Cc: stable@vger.kernel.org
> Signed-off-by: Wei Fang <wei.fang@nxp.com>
> Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
> ---
I'm not sure "correct the statistics" is the best way to describe this
change. Maybe "keep track of correct TXBD count in enetc_map_tx_tso_buffs()"?
The bug is that not all TX buffers are freed on error, not that some
statistics are wrong.
> drivers/net/ethernet/freescale/enetc/enetc.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/net/ethernet/freescale/enetc/enetc.c b/drivers/net/ethernet/freescale/enetc/enetc.c
> index 01c09fd26f9f..0658c06a23c1 100644
> --- a/drivers/net/ethernet/freescale/enetc/enetc.c
> +++ b/drivers/net/ethernet/freescale/enetc/enetc.c
> @@ -759,6 +759,7 @@ static int enetc_lso_hw_offload(struct enetc_bdr *tx_ring, struct sk_buff *skb)
> static int enetc_map_tx_tso_buffs(struct enetc_bdr *tx_ring, struct sk_buff *skb)
> {
> struct enetc_ndev_priv *priv = netdev_priv(tx_ring->ndev);
> + bool ext_bd = skb_vlan_tag_present(skb);
> int hdr_len, total_len, data_len;
> struct enetc_tx_swbd *tx_swbd;
> union enetc_tx_bd *txbd;
> @@ -792,7 +793,7 @@ static int enetc_map_tx_tso_buffs(struct enetc_bdr *tx_ring, struct sk_buff *skb
> csum = enetc_tso_hdr_csum(&tso, skb, hdr, hdr_len, &pos);
> enetc_map_tx_tso_hdr(tx_ring, skb, tx_swbd, txbd, &i, hdr_len, data_len);
> bd_data_num = 0;
> - count++;
> + count += ext_bd ? 2 : 1;
>
> while (data_len > 0) {
> int size;
> --
> 2.34.1
>
stylistic nitpick: I think this implementation choice obscures the fact,
to an unfamiliar reader, that the requirement for an extended TXBD comes
from enetc_map_tx_tso_hdr(). This is because you repeat the condition
for skb_vlan_tag_present(), but it's not obvious it's correlated to the
other one. Something like the change below is more expressive in this
regard, in my opinion:
diff --git a/drivers/net/ethernet/freescale/enetc/enetc.c b/drivers/net/ethernet/freescale/enetc/enetc.c
index fe3967268a19..6178157611db 100644
--- a/drivers/net/ethernet/freescale/enetc/enetc.c
+++ b/drivers/net/ethernet/freescale/enetc/enetc.c
@@ -410,14 +410,15 @@ static int enetc_map_tx_buffs(struct enetc_bdr *tx_ring, struct sk_buff *skb)
return 0;
}
-static void enetc_map_tx_tso_hdr(struct enetc_bdr *tx_ring, struct sk_buff *skb,
- struct enetc_tx_swbd *tx_swbd,
- union enetc_tx_bd *txbd, int *i, int hdr_len,
- int data_len)
+static int enetc_map_tx_tso_hdr(struct enetc_bdr *tx_ring, struct sk_buff *skb,
+ struct enetc_tx_swbd *tx_swbd,
+ union enetc_tx_bd *txbd, int *i, int hdr_len,
+ int data_len)
{
union enetc_tx_bd txbd_tmp;
u8 flags = 0, e_flags = 0;
dma_addr_t addr;
+ int count = 1;
enetc_clear_tx_bd(&txbd_tmp);
addr = tx_ring->tso_headers_dma + *i * TSO_HEADER_SIZE;
@@ -460,7 +461,10 @@ static void enetc_map_tx_tso_hdr(struct enetc_bdr *tx_ring, struct sk_buff *skb,
/* Write the BD */
txbd_tmp.ext.e_flags = e_flags;
*txbd = txbd_tmp;
+ count++;
}
+
+ return count;
}
static int enetc_map_tx_tso_data(struct enetc_bdr *tx_ring, struct sk_buff *skb,
@@ -786,7 +790,6 @@ static int enetc_lso_hw_offload(struct enetc_bdr *tx_ring, struct sk_buff *skb)
static int enetc_map_tx_tso_buffs(struct enetc_bdr *tx_ring, struct sk_buff *skb)
{
struct enetc_ndev_priv *priv = netdev_priv(tx_ring->ndev);
- bool ext_bd = skb_vlan_tag_present(skb);
int hdr_len, total_len, data_len;
struct enetc_tx_swbd *tx_swbd;
union enetc_tx_bd *txbd;
@@ -818,9 +821,9 @@ static int enetc_map_tx_tso_buffs(struct enetc_bdr *tx_ring, struct sk_buff *skb
/* compute the csum over the L4 header */
csum = enetc_tso_hdr_csum(&tso, skb, hdr, hdr_len, &pos);
- enetc_map_tx_tso_hdr(tx_ring, skb, tx_swbd, txbd, &i, hdr_len, data_len);
+ count += enetc_map_tx_tso_hdr(tx_ring, skb, tx_swbd, txbd, &i,
+ hdr_len, data_len);
bd_data_num = 0;
- count += ext_bd ? 2 : 1;
while (data_len > 0) {
int size;
next prev parent reply other threads:[~2025-02-20 16:01 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-02-19 5:42 [PATCH v2 net 0/9] net: enetc: fix some known issues Wei Fang
2025-02-19 5:42 ` [PATCH v2 net 1/9] net: enetc: fix the off-by-one issue in enetc_map_tx_buffs() Wei Fang
2025-02-19 10:48 ` Claudiu Manoil
2025-02-20 13:01 ` Vladimir Oltean
2025-02-21 1:26 ` Wei Fang
2025-02-19 5:42 ` [PATCH v2 net 2/9] net: enetc: correct the tx_swbd statistics Wei Fang
2025-02-20 8:31 ` Ioana Ciornei
2025-02-20 16:01 ` Vladimir Oltean [this message]
2025-02-21 1:42 ` Wei Fang
2025-02-21 8:03 ` Claudiu Manoil
2025-02-21 8:34 ` Wei Fang
2025-02-21 9:22 ` Claudiu Manoil
2025-02-21 9:18 ` Vladimir Oltean
2025-02-24 3:27 ` Wei Fang
2025-02-19 5:42 ` [PATCH v2 net 3/9] net: enetc: correct the xdp_tx statistics Wei Fang
2025-02-20 8:37 ` Ioana Ciornei
2025-02-20 16:58 ` Vladimir Oltean
2025-02-19 5:42 ` [PATCH v2 net 4/9] net: enetc: VFs do not support HWTSTAMP_TX_ONESTEP_SYNC Wei Fang
2025-02-20 16:52 ` Vladimir Oltean
2025-02-19 5:42 ` [PATCH v2 net 5/9] net: enetc: update UDP checksum when updating originTimestamp field Wei Fang
2025-02-20 15:50 ` Vladimir Oltean
2025-02-19 5:42 ` [PATCH v2 net 6/9] net: enetc: add missing enetc4_link_deinit() Wei Fang
2025-02-20 16:50 ` Vladimir Oltean
2025-02-19 5:42 ` [PATCH v2 net 7/9] net: enetc: remove the mm_lock from the ENETC v4 driver Wei Fang
2025-02-20 16:03 ` Vladimir Oltean
2025-02-19 5:42 ` [PATCH v2 net 8/9] net: enetc: correct the EMDIO base offset for ENETC v4 Wei Fang
2025-02-19 20:42 ` Andrew Lunn
2025-02-20 8:05 ` Wei Fang
2025-02-20 15:58 ` Vladimir Oltean
2025-02-19 5:42 ` [PATCH v2 net 9/9] net: enetc: fix the off-by-one issue in enetc_map_tx_tso_buffs() Wei Fang
2025-02-19 11:04 ` Michal Swiatkowski
2025-02-19 17:46 ` Claudiu Manoil
2025-02-20 5:06 ` Wei Fang
2025-02-20 8:39 ` Claudiu Manoil
2025-02-20 16:32 ` Vladimir Oltean
2025-02-20 16:46 ` Vladimir Oltean
2025-02-21 2:56 ` Wei Fang
2025-02-21 9:30 ` Vladimir Oltean
2025-02-24 4:42 ` Wei Fang
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=20250220160123.5evmuxlbuzo7djgr@skbuf \
--to=vladimir.oltean@nxp.com \
--cc=andrew+netdev@lunn.ch \
--cc=claudiu.manoil@nxp.com \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=imx@lists.linux.dev \
--cc=ioana.ciornei@nxp.com \
--cc=kuba@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=michal.swiatkowski@linux.intel.com \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=stable@vger.kernel.org \
--cc=wei.fang@nxp.com \
--cc=xiaoning.wang@nxp.com \
--cc=yangbo.lu@nxp.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).