2024-03-08 11:57:46

by Geethasowjanya Akula

Subject: [net PATCH] octeontx2-pf: Do not use HW TSO when gso_size < 16bytes

Hardware doesn't support packet segmentation when segment size
is < 16 bytes. Hence add an additional check and use SW
segmentation in such case.

Fixes: 86d7476078b8 ("octeontx2-pf: TCP segmentation offload support").
Signed-off-by: Geetha sowjanya <[email protected]>
---
drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c | 7 +++++++
1 file changed, 7 insertions(+)

diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
index f828d32737af..2ac56abb3a0e 100644
--- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
+++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
@@ -967,6 +967,13 @@ static bool is_hw_tso_supported(struct otx2_nic *pfvf,
{
int payload_len, last_seg_size;

+ /* Due to hw issue segment size less than 16 bytes
+ * are not supported. Hence do not offload such TSO
+ * segments.
+ */
+ if (skb_shinfo(skb)->gso_size < 16)
+ return false;
+
if (test_bit(HW_TSO, &pfvf->hw.cap_flag))
return true;

--
2.25.1

2024-03-08 13:59:03

by Eric Dumazet

Subject: Re: [net PATCH] octeontx2-pf: Do not use HW TSO when gso_size < 16bytes

On Fri, Mar 8, 2024 at 12:57 PM Geetha sowjanya <[email protected]> wrote:
>
> Hardware doesn't support packet segmentation when segment size
> is < 16 bytes. Hence add an additional check and use SW
> segmentation in such case.
>
> Fixes: 86d7476078b8 ("octeontx2-pf: TCP segmentation offload support").
> Signed-off-by: Geetha sowjanya <[email protected]>
> ---
> drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
> index f828d32737af..2ac56abb3a0e 100644
> --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
> +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
> @@ -967,6 +967,13 @@ static bool is_hw_tso_supported(struct otx2_nic *pfvf,
> {
> int payload_len, last_seg_size;
>
> + /* Due to hw issue segment size less than 16 bytes
> + * are not supported. Hence do not offload such TSO
> + * segments.
> + */
> + if (skb_shinfo(skb)->gso_size < 16)
> + return false;
> +
> if (test_bit(HW_TSO, &pfvf->hw.cap_flag))
> return true;

How is this driver doing SW segmentation at this stage ?

You might perform this check in ndo_features_check() instead ?

otx2_sq_append_skb() is also forcing a linearization if
skb_shinfo(skb)->nr_frags >= OTX2_MAX_FRAGS_IN_SQE

This is quite bad, this one definitely should use ndo_features_check()
to mask NETIF_F_GSO_MASK for GSO packets.

Look at typhoon_features_check() for an example.

2024-03-11 06:41:42

by Geethasowjanya Akula

Subject: RE: [EXTERNAL] Re: [net PATCH] octeontx2-pf: Do not use HW TSO when gso_size < 16bytes

> -----Original Message-----
> From: Eric Dumazet <[email protected]>
> Sent: Friday, March 8, 2024 7:29 PM
> To: Geethasowjanya Akula <[email protected]>
> Cc: [email protected]; [email protected]; [email protected];
> [email protected]; [email protected]; Sunil Kovvuri Goutham
> <[email protected]>; Subbaraya Sundeep Bhatta
> <[email protected]>; Hariprasad Kelam <[email protected]>
> Subject: [EXTERNAL] Re: [net PATCH] octeontx2-pf: Do not use HW TSO when
> gso_size < 16bytes
> On Fri, Mar 8, 2024 at 12:57 PM Geetha sowjanya <[email protected]>
> wrote:
> >
> > Hardware doesn't support packet segmentation when segment size is < 16
> > bytes. Hence add an additional check and use SW segmentation in such
> > case.
> >
> > Fixes: 86d7476078b8 ("octeontx2-pf: TCP segmentation offload support").
> > Signed-off-by: Geetha sowjanya <[email protected]>
> > ---
> > drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c | 7 +++++++
> > 1 file changed, 7 insertions(+)
> >
> > diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
> > b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
> > index f828d32737af..2ac56abb3a0e 100644
> > --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
> > +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
> > @@ -967,6 +967,13 @@ static bool is_hw_tso_supported(struct otx2_nic
> > *pfvf, {
> > int payload_len, last_seg_size;
> >
> > + /* Due to hw issue segment size less than 16 bytes
> > + * are not supported. Hence do not offload such TSO
> > + * segments.
> > + */
> > + if (skb_shinfo(skb)->gso_size < 16)
> > + return false;
> > +
> > if (test_bit(HW_TSO, &pfvf->hw.cap_flag))
> > return true;
>
> How is this driver doing SW segmentation at this stage ?
>
> You might perform this check in ndo_features_check() instead ?
>
> otx2_sq_append_skb() is also forcing a linearization if skb_shinfo(skb)-
> >nr_frags >= OTX2_MAX_FRAGS_IN_SQE
>
> This is quite bad, this one definitely should use ndo_features_check() to mask
> NETIF_F_GSO_MASK for GSO packets.
>
> Look at typhoon_features_check() for an example.

Thanks for the suggestions. Will submit v2 with suggested changes.