 Gavin McCullagh
|
2072c228c9
[TCP]: use non-delayed ACK for congestion control RTT
|
il y a 17 ans |
 Satoru SATOH
|
488faa2ae3
[IPV4]: Make tcp_input_metrics() get minimum RTO via tcp_rto_min()
|
il y a 17 ans |
 Ilpo Jrvinen
|
52d3408150
[TCP]: Move prior_in_flight collect to more robust place
|
il y a 17 ans |
|
Ilpo Jrvinen
|
3e6f049e0c
[TCP] FRTO: Use of existing funcs make code more obvious & robust
|
il y a 17 ans |
|
Ilpo Jrvinen
|
e1cd8f78f8
[TCP] FRTO: Clear frto_highmark only after process_frto that uses it
|
il y a 17 ans |
|
Ilpo Jrvinen
|
96a2d41a3e
[TCP]: Make sure write_queue_from does not begin with NULL ptr
|
il y a 17 ans |
|
Ilpo Jrvinen
|
23aeeec365
[TCP] FRTO: Plug potential LOST-bit leak
|
il y a 17 ans |
|
Ilpo Jrvinen
|
746aa32d28
[TCP] FRTO: Limit snd_cwnd if TCP was application limited
|
il y a 17 ans |
|
Ilpo Jrvinen
|
fbd52eb2bd
[TCP]: Split SACK FRTO flag clearing (fixes FRTO corner case bug)
|
il y a 17 ans |
|
Ilpo Jrvinen
|
e49aa5d456
[TCP]: Add unlikely() to sacktag out-of-mem in fragment case
|
il y a 17 ans |
|
Ilpo Jrvinen
|
c7caf8d3ed
[TCP]: Fix reord detection due to snd_una covered holes
|
il y a 17 ans |
|
Ilpo Jrvinen
|
8dd71c5d28
[TCP]: Consider GSO while counting reord in sacktag
|
il y a 17 ans |
|
Ilpo Jrvinen
|
261ab365fa
[TCP]: Another TAGBITS -> SACKED_ACKED|LOST conversion
|
il y a 17 ans |
|
Ilpo Jrvinen
|
e56d6cd605
[TCP]: Process DSACKs that reside within a SACK block
|
il y a 17 ans |
 Ryousei Takano
|
94d3b1e586
[TCP]: fix D-SACK cwnd handling
|
il y a 17 ans |
 Adrian Bunk
|
0f79efdc23
[TCP]: Make tcp_match_skb_to_sack() static.
|
il y a 17 ans |
|
Ryousei Takano
|
564262c1f0
[TCP]: Fix inconsistency of terms.
|
il y a 17 ans |
 Chuck Lever
|
c2636b4d9e
[NET]: Treat the sign of the result of skb_headroom() consistently
|
il y a 17 ans |
|
Ilpo Järvinen
|
df2e014bfb
[TCP]: Remove lost_retrans zero seqno special cases
|
il y a 17 ans |
|
Ilpo Järvinen
|
f885c5b08e
[TCP]: high_seq parameter removed (all callers use tp->high_seq)
|
il y a 17 ans |
|
Ilpo Järvinen
|
b08d6cb22c
[TCP]: Limit processing lost_retrans loop to work-to-do cases
|
il y a 17 ans |
|
Ilpo Järvinen
|
f785a8e28b
[TCP]: Fix lost_retrans loop vs fastpath problems
|
il y a 17 ans |
|
Ilpo Järvinen
|
4cd829995b
[TCP]: No need to re-count fackets_out/sacked_out at RTO
|
il y a 17 ans |
|
Ilpo Järvinen
|
d193594299
[TCP]: Extract tcp_match_queue_to_sack from sacktag code
|
il y a 17 ans |
|
Ilpo Järvinen
|
f6fb128d27
[TCP]: Kill almost unused variable pcount from sacktag
|
il y a 17 ans |
|
Ilpo Järvinen
|
3eec0047d9
[TCP]: Fix mark_head_lost to ignore R-bit when trying to mark L
|
il y a 17 ans |
|
Ilpo Järvinen
|
16e906812f
[TCP]: Add bytes_acked (ABC) clearing to FRTO too
|
il y a 17 ans |
|
Ilpo Järvinen
|
1c1e87edb9
[TCP]: Separate lost_retrans loop into own function
|
il y a 17 ans |
 Stephen Hemminger
|
cfcabdcc2d
[NET]: sparse warning fixes
|
il y a 17 ans |
|
Ilpo Järvinen
|
de83c058af
[TCP]: "Annotate" another fackets_out state reset
|
il y a 17 ans |
