Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1707737

[PATCH 4.9 36/93] tcp_bbr: cut pacing rate only if filled pipe

Show all headers | View raw

4.9-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Neal Cardwell <ncardwell@google.com>


[ Upstream commit 4aea287e90dd61a48268ff2994b56f9799441b62 ]

In bbr_set_pacing_rate(), which decides whether to cut the pacing
rate, there was some code that considered exiting STARTUP to be
equivalent to the notion of filling the pipe (i.e.,
bbr_full_bw_reached()). Specifically, as the code was structured,
exiting STARTUP and going into PROBE_RTT could cause us to cut the
pacing rate down to something silly and low, based on whatever
bandwidth samples we've had so far, when it's possible that all of
them have been small app-limited bandwidth samples that are not
representative of the bandwidth available in the path. (The code was
correct at the time it was written, but the state machine changed
without this spot being adjusted correspondingly.)

Fixes: 0f8782ea1497 ("tcp_bbr: add BBR congestion control")
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 net/ipv4/tcp_bbr.c |    3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

--- a/net/ipv4/tcp_bbr.c
+++ b/net/ipv4/tcp_bbr.c
@@ -191,12 +191,11 @@ static u64 bbr_rate_bytes_per_sec(struct
  */
 static void bbr_set_pacing_rate(struct sock *sk, u32 bw, int gain)
 {
-	struct bbr *bbr = inet_csk_ca(sk);
 	u64 rate = bw;
 
 	rate = bbr_rate_bytes_per_sec(sk, rate, gain);
 	rate = min_t(u64, rate, sk->sk_max_pacing_rate);
-	if (bbr->mode != BBR_STARTUP || rate > sk->sk_pacing_rate)
+	if (bbr_full_bw_reached(sk) || rate > sk->sk_pacing_rate)
 		sk->sk_pacing_rate = rate;
 }
 

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH 4.9 00/93] 4.9.42-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 23/93] gpiolib: skip unwanted events, dont convert them to opposite edge Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 19/93] timers: Fix overflow in get_next_timer_interrupt Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 26/93] ARM: dts: armada-38x: Fix irq type for pca955 Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 06/93] iwlwifi: dvm: prevent an out of bounds access Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 41/93] ipv4: ipv6: initialize treq->txhash in cookie_v[46]_check() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 37/93] tcp_bbr: introduce bbr_bw_to_pacing_rate() helper Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 10/93] device property: Make dev_fwnode() public Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 40/93] tcp_bbr: init pacing rate on first RTT sample Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 11/93] mmc: core: Fix access to HS400-ES devices Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 21/93] powerpc/64: Fix __check_irq_replay missing decrementer interrupt Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 36/93] tcp_bbr: cut pacing rate only if filled pipe Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 12/93] mm, mprotect: flush TLB if potentially racing with a parallel reclaim leaving stale TLB entries Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 24/93] ext4: fix SEEK_HOLE/SEEK_DATA for blocksize < pagesize Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 20/93] powerpc/tm: Fix saving of TM SPRs in core dump Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 33/93] f2fs: sanity check checkpoint segno and blkoff Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 29/93] iscsi-target: Fix initial login PDU asynchronous socket close OOPs Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 35/93] [media] saa7164: fix double fetch PCIe access condition Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 32/93] media: lirc: LIRC_GET_REC_RESOLUTION should return microseconds Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 01/93] parisc: Handle vmas whose context is not current in flush_cache_range Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 44/93] net: dsa: b53: Add missing ARL entries for BCM53125 Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 08/93] NFSv4: Fix EXCHANGE_ID corrupt verifier issue Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  [PATCH 4.9 43/93] ipv6: avoid overflow of offset in ip6_find_1stfragopt Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-08-09 20:40 +0200
  Re: [PATCH 4.9 00/93] 4.9.42-stable review Shuah Khan <shuahkh@osg.samsung.com> - 2017-08-10 02:20 +0200
  Re: [PATCH 4.9 00/93] 4.9.42-stable review Guenter Roeck <linux@roeck-us.net> - 2017-08-10 02:50 +0200
  Re: [PATCH 4.9 00/93] 4.9.42-stable review Sumit Semwal <sumit.semwal@linaro.org> - 2017-08-10 05:10 +0200

csiph-web