tcp: improve recv_skip_hint for tcp_zerocopy_receive
tcp_zerocopy_receive() rounds down the zc->length a multiple of
PAGE_SIZE. This results in two issues:
- tcp_zerocopy_receive sets recv_skip_hint to the length of the
receive queue if the zc->length input is smaller than the
PAGE_SIZE, even though the data in receive queue could be
zerocopied.
- tcp_zerocopy_receive would set recv_skip_hint of 0, in cases
where we have a little bit of data after the perfectly-sized
packets.
To fix these issues, do not store the rounded down value in
zc->length. Round down the length passed to zap_page_range(),
and return min(inq, zc->length) when the zap_range is 0.
Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index f98a188..9f41a76 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -1739,8 +1739,8 @@ static int tcp_zerocopy_receive(struct sock *sk,
struct tcp_zerocopy_receive *zc)
{
unsigned long address = (unsigned long)zc->address;
+ u32 length = 0, seq, offset, zap_len;
const skb_frag_t *frags = NULL;
- u32 length = 0, seq, offset;
struct vm_area_struct *vma;
struct sk_buff *skb = NULL;
struct tcp_sock *tp;
@@ -1767,12 +1767,12 @@ static int tcp_zerocopy_receive(struct sock *sk,
seq = tp->copied_seq;
inq = tcp_inq(sk);
zc->length = min_t(u32, zc->length, inq);
- zc->length &= ~(PAGE_SIZE - 1);
- if (zc->length) {
- zap_page_range(vma, address, zc->length);
+ zap_len = zc->length & ~(PAGE_SIZE - 1);
+ if (zap_len) {
+ zap_page_range(vma, address, zap_len);
zc->recv_skip_hint = 0;
} else {
- zc->recv_skip_hint = inq;
+ zc->recv_skip_hint = zc->length;
}
ret = 0;
while (length + PAGE_SIZE <= zc->length) {