Switching TCP to GSO mode, relying on core networking layers to perform eventual adaptation for dumb devices was overdue.
1) Most TCP developments are done with TSO in mind. 2) Less high-resolution timers needs to be armed for TCP-pacing 3) GSO can benefit of xmit_more hint 4) Receiver GRO is more effective (as if TSO was used for real on sender) -> less ACK packets and overhead. 5) Write queues have less overhead (one skb holds about 64KB of payload) 6) SACK coalescing just works. (no payload in skb->head) 7) rtx rb-tree contains less packets, SACK is cheaper. 8) Removal of legacy code. Less maintenance hassles. Note that I have left the sendpage/zerocopy paths, but they probably can benefit from the same strategy. Thanks to Oleksandr Natalenko for reporting a performance issue for BBR/fq_codel, which was the main reason I worked on this patch series. Eric Dumazet (6): tcp: switch to GSO being always on tcp: remove sk_can_gso() use tcp: remove sk_check_csum_caps() tcp: tcp_sendmsg() only deals with CHECKSUM_PARTIAL tcp: remove dead code from tcp_set_skb_tso_segs() tcp: remove dead code after CHECKSUM_PARTIAL adoption include/net/sock.h | 10 +------- net/core/sock.c | 2 +- net/ipv4/tcp.c | 57 ++++++++++++------------------------------- net/ipv4/tcp_input.c | 3 --- net/ipv4/tcp_ipv4.c | 13 +++------- net/ipv4/tcp_output.c | 40 +++++------------------------- 6 files changed, 26 insertions(+), 99 deletions(-) -- 2.16.1.291.g4437f3f132-goog