linux_dsm_epyc7002/include
Ilpo Järvinen 46d0de4ed9 [TCP] FRTO: Entry is allowed only during (New)Reno like recovery
This interpretation comes from RFC4138:
    "If the sender implements some loss recovery algorithm other
     than Reno or NewReno [FHG04], the F-RTO algorithm SHOULD
     NOT be entered when earlier fast recovery is underway."

I think the RFC means to say (especially in the light of
Appendix B) that ...recovery is underway (not just fast recovery)
or was underway when it was interrupted by an earlier (F-)RTO
that hasn't yet been resolved (snd_una has not advanced enough).
Thus, my interpretation is that whenever TCP has ever
retransmitted other than head, basic version cannot be used
because then the order assumptions which are used as FRTO basis
do not hold.

NewReno has only the head segment retransmitted at a time.
Therefore, walk up to the segment that has not been SACKed, if
that segment is not retransmitted nor anything before it, we know
for sure, that nothing after the non-SACKed segment should be
either. This assumption is valid because TCPCB_EVER_RETRANS does
not leave holes but each non-SACKed segment is rexmitted
in-order.

Check for retrans_out > 1 avoids more expensive walk through the
skb list, as we can know the result beforehand: F-RTO will not be
allowed.

SACKed skb can turn into non-SACked only in the extremely rare
case of SACK reneging, in this case we might fail to detect
retransmissions if there were them for any other than head. To
get rid of that feature, whole rexmit queue would have to be
walked (always) or FRTO should be prevented when SACK reneging
happens. Of course RTO should still trigger after reneging which
makes this issue even less likely to show up. And as long as the
response is as conservative as it's now, nothing bad happens even
then.

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-25 22:23:12 -07:00
..
acpi Pull bugzilla-8171 into release branch 2007-03-20 11:06:00 -04:00
asm-alpha alpha: build fixes - force architecture 2007-04-17 16:36:27 -07:00
asm-arm [ARM] 4298/1: fix memory barriers for DMA coherent and SMP platforms 2007-04-01 22:38:36 +01:00
asm-arm26 Storage class should be before const qualifier 2007-02-17 20:11:19 +01:00
asm-avr32 [PATCH] gpio_direction_output() needs an initial value 2007-03-16 19:25:04 -07:00
asm-cris [PATCH] consolidate line discipline number definitions 2007-02-11 10:51:26 -08:00
asm-frv [PATCH] FRV: Add some missng lazy MMU hooks for NOMMU mode 2007-03-01 14:53:36 -08:00
asm-generic [PATCH] Proper fix for highmem kmap_atomic functions for VMI for 2.6.21 2007-04-08 19:47:55 -07:00
asm-h8300 [PATCH] consolidate line discipline number definitions 2007-02-11 10:51:26 -08:00
asm-i386 [PATCH] Proper fix for highmem kmap_atomic functions for VMI for 2.6.21 2007-04-08 19:47:55 -07:00
asm-ia64 [IA64] SGI Altix : fix pcibr_dmamap_ate32() bug 2007-04-06 15:38:12 -07:00
asm-m32r [PATCH] m32r dma-mapping.h should simply include generic/dma-mapping-broken.h 2007-03-14 15:27:49 -07:00
asm-m68k [PATCH] m68k dma-mapping: gfp_t annotations 2007-03-14 15:27:51 -07:00
asm-m68knommu [PATCH] m68knommu: GPIO line defines for the ColdFire 5282 2007-03-06 18:08:38 -08:00
asm-mips [MIPS] Fix wrong checksum for split TCP packets on 64-bit MIPS 2007-04-20 14:58:37 +01:00
asm-parisc Merge master.kernel.org:/pub/scm/linux/kernel/git/kyle/parisc-2.6 2007-02-26 12:48:06 -08:00
asm-powerpc allow vmsplice to work in 32-bit mode on ppc64 2007-04-17 16:36:26 -07:00
asm-ppc [POWERPC] ppc: Add support for AMCC Taishan 440GX eval board 2007-02-13 15:35:52 +11:00
asm-s390 [S390] Fix TCP/UDP pseudo header checksum computation. 2007-03-26 20:43:46 +02:00
asm-sh sh: Trivial fix for hp6xx build. 2007-03-28 19:45:59 +09:00
asm-sh64 [PATCH] consolidate line discipline number definitions 2007-02-11 10:51:26 -08:00
asm-sparc [SPARC]: Add unsigned to unused bit field in a.out.h 2007-04-02 14:26:21 -07:00
asm-sparc64 [SPARC]: Add unsigned to unused bit field in a.out.h 2007-04-02 14:26:21 -07:00
asm-um [PATCH] uml: fix unreasonably long udelay 2007-04-02 10:06:08 -07:00
asm-v850 [PATCH] consolidate line discipline number definitions 2007-02-11 10:51:26 -08:00
asm-x86_64 [PATCH] x86_64 irq: Fix comments after changing IRQ0_VECTOR from 0x20 to 0x30 2007-03-29 08:16:23 -07:00
asm-xtensa [PATCH] consolidate line discipline number definitions 2007-02-11 10:51:26 -08:00
crypto [CRYPTO] api: Allow multiple frontends per backend 2007-02-07 09:21:01 +11:00
keys
linux Merge master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2007-04-24 18:20:32 -07:00
math-emu
media V4L/DVB (5441): Saa7146: Fix allocation of clipping memory 2007-03-27 08:45:56 -03:00
mtd [MTD] remove unused ecctype,eccsize fields from struct mtd_info 2007-02-09 15:27:12 +00:00
net [TCP] FRTO: Entry is allowed only during (New)Reno like recovery 2007-04-25 22:23:12 -07:00
pcmcia serial: Add PCMCIA IDs for Quatech DSP-100 dual RS232 adapter. 2007-02-16 15:19:16 -08:00
rdma RDMA/cma: Add multicast communication support 2007-02-16 14:29:07 -08:00
rxrpc
scsi Merge master.kernel.org:/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6 2007-02-19 13:32:28 -08:00
sound [ALSA] version 1.0.14rc3 2007-03-14 08:25:52 +01:00
video [PATCH] Video: fb, add true ref_count atomicity 2007-02-12 09:48:42 -08:00
Kbuild