Search String: Display: Description: Sort:

Results:

References: [ +subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*Early\s+SPECWeb99\s+results\s+on\s+2\.5\.33\s+with\s+TSO\s+on\s+e1000\s*$/: 202 ]

Total 202 documents matching your query.

181. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: Alan Cox <alan@xxxxxxxxxxxxxxxxxxx>
Date: 12 Sep 2002 15:11:23 +0100
As far as I am aware it was original when Linux first did it (and we broke cisco pix, some boot proms, some sco in the process). Credit goes to Arnt Gulbrandsen probably better known nowdays for his
/archives/netdev/2002-09/msg00473.html (11,407 bytes)

182. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: todd-lkml@xxxxxxxxxxxxx
Date: Thu, 12 Sep 2002 08:41:25 -0600 (MDT)
alan, good to know. it's a nice piece of engineering. it's useful to note that linux has such a long and rich history of breaking de-facto standards in order to make things work better. t. -- todd un
/archives/netdev/2002-09/msg00474.html (11,794 bytes)

183. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: Nivedita Singhvi <niv@xxxxxxxxxx>
Date: Thu, 12 Sep 2002 10:18:37 -0700
Quoting Todd Underwood <todd@xxxxxxxxxxxxx>: Some of that may be a byproduct of the "all the worlds' a webserver" mindset - we are primarily focussed on the server side (aka money side ;)), and there
/archives/netdev/2002-09/msg00476.html (12,068 bytes)

184. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Thu, 12 Sep 2002 16:12:25 -0700 (PDT)
I disagree, at least for bulk receivers. We have no way currently to get rid of the data copy. We desperately need sys_receivefile() and appropriate ops all the way into the networking, then the nece
/archives/netdev/2002-09/msg00478.html (11,156 bytes)

185. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: todd-lkml@xxxxxxxxxxxxx
Date: Fri, 13 Sep 2002 15:59:15 -0600 (MDT)
dave, all, not sure i understand what you're proposing, but while we're at it, why not also make the api for apps to allocate a buffer in userland that (for nics that support it) the nic can dma dire
/archives/netdev/2002-09/msg00489.html (11,883 bytes)

186. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: Nivedita Singhvi <niv@xxxxxxxxxx>
Date: Fri, 13 Sep 2002 15:12:20 -0700
Quoting todd-lkml@xxxxxxxxxxxxx: I believe thats exactly what David was referring to - reverse direction sendfile() so to speak.. However, I dont think what youre saving are interrupts as much as the
/archives/netdev/2002-09/msg00490.html (11,157 bytes)

187. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Fri, 13 Sep 2002 15:04:39 -0700 (PDT)
Cards in the future at 10gbit and faster are going to provide facilities by which: 1) You register a IPV4 src_addr/dst_addr TCP src_port/dst_port cookie with the hardware when TCP connections are ope
/archives/netdev/2002-09/msg00491.html (12,848 bytes)

188. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: jamal <hadi@xxxxxxxxxx>
Date: Sun, 15 Sep 2002 16:16:13 -0400 (EDT)
10 gige becomes more of an interesting beast. Not sure if we would see servers with 10gige real soon now. Your proposal does make sense although compute power would still be a player. I think the key
/archives/netdev/2002-09/msg00500.html (11,298 bytes)

189. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Sun, 15 Sep 2002 21:23:21 -0700 (PDT)
Your proposal does make sense although compute power would still be a player. I think the key would be parallelization; Oh I forgot to mention that some of these cards also compute a cookie for you o
/archives/netdev/2002-09/msg00504.html (11,037 bytes)

190. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: todd-lkml@xxxxxxxxxxxxx
Date: Mon, 16 Sep 2002 08:16:47 -0600 (MDT)
david, comments/questions below... intriguing architecture. are there any standards in progress to support this. bascially, people doing high performance computing have been customizing non-commodity
/archives/netdev/2002-09/msg00506.html (14,653 bytes)

191. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Mon, 16 Sep 2002 12:52:11 -0700 (PDT)
are there any standards in progress to support this. Your question makes no sense, it is a hardware optimization of an existing standard. The chip merely is told what flows exist and it concatenates
/archives/netdev/2002-09/msg00509.html (11,200 bytes)

192. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: todd-lkml@xxxxxxxxxxxxx
Date: Mon, 16 Sep 2002 15:32:56 -0600 (MDT)
folx, perhaps i was insufficiently clear. hardware optimizations can be standardized. in fact, when they are, it is substantially easier to implement to them. my assumption (perhaps incorrect) is tha
/archives/netdev/2002-09/msg00516.html (12,799 bytes)

193. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Mon, 16 Sep 2002 14:29:31 -0700 (PDT)
The system calls would go into the VFS, sys_receivefile is not networking specific in any way shape or form. And to answer your question, if I had the time I'd work on it yes. Right now the answer to
/archives/netdev/2002-09/msg00517.html (11,138 bytes)

194. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: David Woodhouse <dwmw2@xxxxxxxxxxxxx>
Date: Mon, 16 Sep 2002 23:53:00 +0100
Er, surely the same goes for sys_sendfile? Why have a new system call rather than just swapping the 'in' and 'out' fds? -- dwmw2
/archives/netdev/2002-09/msg00518.html (11,113 bytes)

195. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Mon, 16 Sep 2002 15:46:40 -0700 (PDT)
There is an assumption that one is a linear stream of output (in this case a socket) and the other one is a page cache based file. It would be nice to extend sys_sendfile to work properly in both way
/archives/netdev/2002-09/msg00519.html (11,202 bytes)

196. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: David Woodhouse <dwmw2@xxxxxxxxxxxxx>
Date: Tue, 17 Sep 2002 00:03:19 +0100
That's an implementation detail and it's not clear we should be exposing it to the user. It's not entirely insane to contemplate socket->socket or too? File descriptors are file descriptors. Yeah --
/archives/netdev/2002-09/msg00520.html (12,000 bytes)

197. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: Jeff Garzik <jgarzik@xxxxxxxxxxxxxxxx>
Date: Mon, 16 Sep 2002 19:08:15 -0400
Er, surely the same goes for sys_sendfile? Why have a new system call rather than just swapping the 'in' and 'out' fds? There is an assumption that one is a linear stream of output (in this case a s
/archives/netdev/2002-09/msg00521.html (12,661 bytes)

198. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Mon, 16 Sep 2002 16:02:10 -0700 (PDT)
What change made this happen?
/archives/netdev/2002-09/msg00522.html (10,877 bytes)

199. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: Jeff Garzik <jgarzik@xxxxxxxxxxxxxxxx>
Date: Mon, 16 Sep 2002 19:48:37 -0400
David S. Miller wrote: From: Jeff Garzik <jgarzik@xxxxxxxxxxxxxxxx> Date: Mon, 16 Sep 2002 19:08:15 -0400 I was rather disappointed when file->file sendfile was [purposefully?] broken in 2.5.x... Wha
/archives/netdev/2002-09/msg00523.html (11,801 bytes)

200. Re: Early SPECWeb99 results on 2.5.33 with TSO on e1000 (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Mon, 16 Sep 2002 16:43:43 -0700 (PDT)
I dunno when it happened, but 2.5.x now returns EINVAL for all file->file cases. In 2.4.x, if sendpage is NULL, file_send_actor in mm/filemap.c faked a call to fops->write(). In 2.5.x, if sendpage is
/archives/netdev/2002-09/msg00524.html (11,318 bytes)


This search system is powered by Namazu