Menu
▾
▴
Re: [uml-devel] Patch v3
|
From: Anton I. <ant...@ko...> - 2017-10-05 13:50:23
|
Hi List, hi Richard, Slight bump - we are now at 5.6Gbit. Same host hardware - 3.5GHz Athlon. This is an incremental on top of the last patchset from this week which tidies up further the SG and TSO/GSO code and enables it both way to the extent possible. Richard, I would like, if possible, to start working on cleaning this up. IMHO it will be better to work off the last submitted patch version and add TSO/GSO in a next revision after (and of course if) we have it merged. TSO/GSO will drag a huge blob of ethtool (or mconsole) related code to control it, it will be better if it comes as an incremental and not with the core driver code. Example: aivanov@MonstrousNightmare:~$ iperf -c 192.168.125.118 ------------------------------------------------------------ Client connecting to 192.168.125.118, TCP port 5001 TCP window size: 85.0 KByte (default) ------------------------------------------------------------ [ 3] local 192.168.125.1 port 34304 connected with 192.168.125.118 port 5001 [ ID] Interval Transfer Bandwidth [ 3] 0.0-10.0 sec 5.70 GBytes 4.89 Gbits/sec aivanov@MonstrousNightmare:~$ iperf -s ------------------------------------------------------------ Server listening on TCP port 5001 TCP window size: 85.3 KByte (default) ------------------------------------------------------------ [ 4] local 192.168.3.80 port 5001 connected with 192.168.125.118 port 52306 [ ID] Interval Transfer Bandwidth [ 4] 0.0-10.0 sec 6.12 GBytes 5.26 Gbits/sec A. On 10/04/17 08:29, Anton Ivanov wrote: > I got the TSO and some offloads working for some (not all) cases of my > patchsets. > > Preliminary speeds and feeds on iperf: > > [ 4] local 192.168.98.1 port 5001 connected with 192.168.98.132 port > 54838 > [ 4] 0.0-10.0 sec 4.78 GBytes 4.11 Gbits/sec > > As a comparison - kvm/qemu on same machine does about a Gbit and vEth > to container does 20Gbit - so somewhere in between. > > I am having some trouble with TSO and raw sockets. Once I figure out > what is going on (and hopefully fix it) there will be an updated > version which should take us into 4x what kvm/QEMU does on one CPU > networkwise. > > A. > > > On 09/26/17 11:26, ant...@ko... wrote: >> Hi List, hi Richard, >> >> I found a logical error in the SG code. This revision fixes that. No >> other changes - the rest appears to work as it should in my testing. >> >> Brgds, >> >> A. >> >> >> ------------------------------------------------------------------------------ >> >> Check out the vibrant tech community on one of the world's most >> engaging tech sites, Slashdot.org! http://sdm.link/slashdot >> _______________________________________________ >> User-mode-linux-devel mailing list >> Use...@li... >> https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel >> > |
