lkml.org 
[lkml]   [2022]   [Jul]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [PATCH net-next v4 11/27] tcp: support externally provided ubufs
From
On 7/8/22 7:03 AM, Pavel Begunkov wrote:
>>> @@ -1356,9 +1362,11 @@ int tcp_sendmsg_locked(struct sock *sk, struct
>>> msghdr *msg, size_t size)
>>>                 copy = min_t(int, copy, pfrag->size - pfrag->offset);
>>>   -            if (tcp_downgrade_zcopy_pure(sk, skb))
>>> -                goto wait_for_space;
>>> -
>>> +            if (unlikely(skb_zcopy_pure(skb) ||
>>> skb_zcopy_managed(skb))) {
>>> +                if (tcp_downgrade_zcopy_pure(sk, skb))
>>> +                    goto wait_for_space;
>>> +                skb_zcopy_downgrade_managed(skb);
>>> +            }
>>>               copy = tcp_wmem_schedule(sk, copy);
>>>               if (!copy)
>>>                   goto wait_for_space;
>>
>> You dropped the msg->msg_ubuf checks on jump labels. Removing the one
>> you had at 'out_nopush' I agree with based on my tests (i.e, it is not
>> needed).
>
> It was an optimisation, which I dropped for simplicity. Will be sending it
> and couple more afterwards.
>
>
>> The one at 'out_err' seems like it is needed - but it has been a few
>> weeks since I debugged that case. I believe the error path I was hitting
>> was sk_stream_wait_memory with MSG_DONTWAIT flag set meaning timeout is
>> 0 and it jumps there with EPIPE.
>
> Currently, it's consistent with MSG_ZEROCOPY ubuf_info, we grab a ubuf_info
> reference at the beginning (msg_zerocopy_realloc() for MSG_ZEROCOPY and
> net_zcopy_get() for msg_ubuf), and then release it at the end
> with net_zcopy_put() or net_zcopy_put_abort().
>

my fault; I somehow dropped a line in the port to the 5.13 kernel.


\
 
 \ /
  Last update: 2022-07-14 01:39    [W:0.082 / U:0.044 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site