Hi Eric,
On Fri, 1 Mar 2024 10:54:19 +0100 Eric Dumazet <edumazet@xxxxxxxxxx> wrote:
>
> OK, I think the issue is caused by a hole at the start of
> tcp_sock_write_rx group
>
> I will send this patch for review, thanks !
>
> diff --git a/include/linux/tcp.h b/include/linux/tcp.h
> index 988a30ef6bfe956fa573f1f18c8284aa382dc1cc..55399ee2a57e736b55ed067fc06ea620bbe62fd3
> 100644
> --- a/include/linux/tcp.h
> +++ b/include/linux/tcp.h
> @@ -304,7 +304,7 @@ struct tcp_sock {
> __cacheline_group_end(tcp_sock_write_txrx);
>
> /* RX read-write hotpath cache lines */
> - __cacheline_group_begin(tcp_sock_write_rx);
> + __cacheline_group_begin(tcp_sock_write_rx) __aligned(8);
> u64 bytes_received;
> /* RFC4898 tcpEStatsAppHCThruOctetsReceived
> * sum(delta(rcv_nxt)), or how many bytes
> diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
> index c82dc42f57c65df112f79080ff407cd98d11ce68..7e1b848398d04f2da2a91c3af97b1e2e3895b8ee
> 100644
> --- a/net/ipv4/tcp.c
> +++ b/net/ipv4/tcp.c
> @@ -4651,7 +4651,7 @@ static void __init tcp_struct_check(void)
> CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock,
> tcp_sock_write_tx, tsorted_sent_queue);
> CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock,
> tcp_sock_write_tx, highest_sack);
> CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock,
> tcp_sock_write_tx, ecn_flags);
> - CACHELINE_ASSERT_GROUP_SIZE(struct tcp_sock, tcp_sock_write_tx, 113);
> + CACHELINE_ASSERT_GROUP_SIZE(struct tcp_sock, tcp_sock_write_tx, 105);
>
> /* TXRX read-write hotpath cache lines */
> CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock,
> tcp_sock_write_txrx, pred_flags);
I will apply this patch to them merge of the net-next tree until it is
applied there.
--
Cheers,
Stephen Rothwell
Attachment:
pgpBZiObh35_F.pgp
Description: OpenPGP digital signature
