From: Alexander Lobakin <alexandr.lobakin@intel•com>
To: Gerhard Engleder <gerhard@engleder-embedded•com>
Cc: <davem@davemloft•net>, <kuba@kernel•org>, <edumazet@google•com>,
<pabeni@redhat•com>, <netdev@vger•kernel.org>
Subject: Re: [PATCH net-next v3 1/9] tsnep: Use spin_lock_bh for TX
Date: Thu, 5 Jan 2023 13:40:17 +0100 [thread overview]
Message-ID: <79d2e66a-3633-ff3a-bdfd-656c735c85e0@intel.com> (raw)
In-Reply-To: <20230104194132.24637-2-gerhard@engleder-embedded.com>
From: Gerhard Engleder <gerhard@engleder-embedded•com>
Date: Wed Jan 04 2023 20:41:24 GMT+0100
> TX processing is done only within process or BH context. Therefore,
> _irqsafe variant is not necessary.
NAPI and .ndo_{start,xdp}_xmit are BH-only, where can the process
context happen here?
>
> Signed-off-by: Gerhard Engleder <gerhard@engleder-embedded•com>
> ---
> drivers/net/ethernet/engleder/tsnep_main.c | 19 ++++++++-----------
> 1 file changed, 8 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/net/ethernet/engleder/tsnep_main.c b/drivers/net/ethernet/engleder/tsnep_main.c
> index bf0190e1d2ea..7cc5e2407809 100644
> --- a/drivers/net/ethernet/engleder/tsnep_main.c
> +++ b/drivers/net/ethernet/engleder/tsnep_main.c
> @@ -434,7 +434,6 @@ static int tsnep_tx_unmap(struct tsnep_tx *tx, int index, int count)
> static netdev_tx_t tsnep_xmit_frame_ring(struct sk_buff *skb,
> struct tsnep_tx *tx)
> {
> - unsigned long flags;
> int count = 1;
> struct tsnep_tx_entry *entry;
> int length;
> @@ -444,7 +443,7 @@ static netdev_tx_t tsnep_xmit_frame_ring(struct sk_buff *skb,
> if (skb_shinfo(skb)->nr_frags > 0)
> count += skb_shinfo(skb)->nr_frags;
>
> - spin_lock_irqsave(&tx->lock, flags);
> + spin_lock_bh(&tx->lock);
>
> if (tsnep_tx_desc_available(tx) < count) {
> /* ring full, shall not happen because queue is stopped if full
> @@ -452,7 +451,7 @@ static netdev_tx_t tsnep_xmit_frame_ring(struct sk_buff *skb,
> */
> netif_stop_queue(tx->adapter->netdev);
>
> - spin_unlock_irqrestore(&tx->lock, flags);
> + spin_unlock_bh(&tx->lock);
>
> return NETDEV_TX_BUSY;
> }
> @@ -468,7 +467,7 @@ static netdev_tx_t tsnep_xmit_frame_ring(struct sk_buff *skb,
>
> tx->dropped++;
>
> - spin_unlock_irqrestore(&tx->lock, flags);
> + spin_unlock_bh(&tx->lock);
>
> netdev_err(tx->adapter->netdev, "TX DMA map failed\n");
>
> @@ -496,20 +495,19 @@ static netdev_tx_t tsnep_xmit_frame_ring(struct sk_buff *skb,
> netif_stop_queue(tx->adapter->netdev);
> }
>
> - spin_unlock_irqrestore(&tx->lock, flags);
> + spin_unlock_bh(&tx->lock);
>
> return NETDEV_TX_OK;
> }
>
> static bool tsnep_tx_poll(struct tsnep_tx *tx, int napi_budget)
> {
> - unsigned long flags;
> int budget = 128;
> struct tsnep_tx_entry *entry;
> int count;
> int length;
>
> - spin_lock_irqsave(&tx->lock, flags);
> + spin_lock_bh(&tx->lock);
>
> do {
> if (tx->read == tx->write)
> @@ -568,18 +566,17 @@ static bool tsnep_tx_poll(struct tsnep_tx *tx, int napi_budget)
> netif_wake_queue(tx->adapter->netdev);
> }
>
> - spin_unlock_irqrestore(&tx->lock, flags);
> + spin_unlock_bh(&tx->lock);
>
> return (budget != 0);
> }
>
> static bool tsnep_tx_pending(struct tsnep_tx *tx)
> {
> - unsigned long flags;
> struct tsnep_tx_entry *entry;
> bool pending = false;
>
> - spin_lock_irqsave(&tx->lock, flags);
> + spin_lock_bh(&tx->lock);
>
> if (tx->read != tx->write) {
> entry = &tx->entry[tx->read];
> @@ -589,7 +586,7 @@ static bool tsnep_tx_pending(struct tsnep_tx *tx)
> pending = true;
> }
>
> - spin_unlock_irqrestore(&tx->lock, flags);
> + spin_unlock_bh(&tx->lock);
>
> return pending;
> }
Thanks,
Olek
next prev parent reply other threads:[~2023-01-05 12:40 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-01-04 19:41 [PATCH net-next v3 0/9] tsnep: XDP support Gerhard Engleder
2023-01-04 19:41 ` [PATCH net-next v3 1/9] tsnep: Use spin_lock_bh for TX Gerhard Engleder
2023-01-05 12:40 ` Alexander Lobakin [this message]
2023-01-05 19:48 ` Gerhard Engleder
2023-01-04 19:41 ` [PATCH net-next v3 2/9] tsnep: Do not print DMA mapping error Gerhard Engleder
2023-01-04 19:41 ` [PATCH net-next v3 3/9] tsnep: Add adapter down state Gerhard Engleder
2023-01-05 12:44 ` Alexander Lobakin
2023-01-05 19:54 ` Gerhard Engleder
2023-01-04 19:41 ` [PATCH net-next v3 4/9] tsnep: Add XDP TX support Gerhard Engleder
2023-01-05 13:01 ` Alexander Lobakin
2023-01-05 21:13 ` Gerhard Engleder
2023-01-06 22:13 ` Jakub Kicinski
2023-01-06 23:34 ` Gerhard Engleder
2023-01-04 19:41 ` [PATCH net-next v3 5/9] tsnep: Substract TSNEP_RX_INLINE_METADATA_SIZE once Gerhard Engleder
2023-01-04 19:41 ` [PATCH net-next v3 6/9] tsnep: Support XDP BPF program setup Gerhard Engleder
2023-01-05 17:24 ` Alexander Lobakin
2023-01-05 22:09 ` Gerhard Engleder
2023-01-04 19:41 ` [PATCH net-next v3 7/9] tsnep: Prepare RX buffer for XDP support Gerhard Engleder
2023-01-04 19:41 ` [PATCH net-next v3 8/9] tsnep: Add RX queue info " Gerhard Engleder
2023-01-05 17:35 ` Alexander Lobakin
2023-01-05 22:28 ` Gerhard Engleder
2023-01-04 19:41 ` [PATCH net-next v3 9/9] tsnep: Add XDP RX support Gerhard Engleder
2023-01-05 17:52 ` Alexander Lobakin
2023-01-05 23:22 ` Gerhard Engleder
2023-01-07 8:46 ` Dan Carpenter
2023-01-07 20:12 ` Gerhard Engleder
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=79d2e66a-3633-ff3a-bdfd-656c735c85e0@intel.com \
--to=alexandr.lobakin@intel$(echo .)com \
--cc=davem@davemloft$(echo .)net \
--cc=edumazet@google$(echo .)com \
--cc=gerhard@engleder-embedded$(echo .)com \
--cc=kuba@kernel$(echo .)org \
--cc=netdev@vger$(echo .)kernel.org \
--cc=pabeni@redhat$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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