**** BEGIN LOGGING AT Sun Jan 31 02:59:58 2016 Jan 31 05:39:12 build #201 of x86.64 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/x86.64/builds/201 Jan 31 05:49:32 build #194 of ppc40x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc40x/builds/194 Jan 31 08:05:20 Grrrrr.... this whole linux-cryptodev vs ocf-cryptodev thing is a PITA. On the other hand, I have kirkwood updated to 4.4 and running on real hardware with CESA doing its thing. Jan 31 08:47:09 12 Jan 31 09:20:56 build #184 of brcm2708 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/brcm2708/builds/184 Jan 31 10:27:45 build #186 of x86.xen_domu is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/x86.xen_domu/builds/186 Jan 31 10:40:47 build #172 of arm64 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/arm64/builds/172 Jan 31 11:40:52 build #152 of uml is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/152 **** BEGIN LOGGING AT Sun Jan 31 12:28:53 2016 Jan 31 12:29:50 nbd r48578 trunk/target/linux/ar71xx/files/drivers/net/ethernet/atheros/ag71xx/ag71xx_main.c * ar71xx: use page fragment API in the ethernet driver Jan 31 12:35:11 nbd r48579 trunk/target/linux/cns3xxx/files/drivers/net/ethernet/cavium/cns3xxx_eth.c * cns3xxx: fix ethernet DMA ring allocation issues Jan 31 12:35:26 nbd r48580 trunk/target/linux/cns3xxx/files/drivers/net/ethernet/cavium/cns3xxx_eth.c * cns3xxx: improve ethernet performance by using the page fragment allocation API Jan 31 12:35:36 nbd r48581 trunk/target/linux/cns3xxx/config-4.4 * cns3xxx: enable CONFIG_PROC_STRIPPED to improve network performance Jan 31 12:38:47 nbd how would you classify the code quality of cavium? :) Jan 31 12:41:13 no idea, i haven't read much code written by cavium Jan 31 12:41:24 that ethernet driver is for cavium hw, but it wasn't written by cavium Jan 31 12:41:35 ah, that explains it Jan 31 12:41:51 oh, wait, i did see a cavium sdk a few years ago Jan 31 12:42:01 the code was complete garbage Jan 31 12:42:12 hahaha, so I figure.. Jan 31 12:42:19 but that is expected in a vendor sdk Jan 31 12:42:20 ;) Jan 31 12:42:39 probably one of the reasons they don't release their "hw offload" code Jan 31 12:42:51 I'm joking, but.. Jan 31 12:42:51 what hardware offload are you referring to? Jan 31 12:43:07 the ones used in ubiquiti hardware Jan 31 12:43:42 ah, so for a completely different soc Jan 31 12:43:56 not sure if offload is even necessary there Jan 31 12:44:22 yeah, I guess ERLs doesn't use cns3xxx Jan 31 12:45:04 mine for example, if I don't enable vlan offload, it can't push more than 500Mbit/s between vlans Jan 31 12:45:52 what version did you test? Jan 31 12:46:18 that was with their own software, edgeos in this case Jan 31 12:46:45 maybe you should try openwrt Jan 31 12:47:00 users report that it has a lot better no-offload performance Jan 31 12:47:51 I might try it, stintel has one as well, I ask him how is the performance with openwrt Jan 31 12:48:47 i only have the er pro, which is probably too fast to run into that kind of performance limit Jan 31 12:49:46 iirc ERLs have dual-core 500MHz Jan 31 12:50:27 I just wish one day that fastpath "hack" can go into the kernel somehow Jan 31 12:51:43 er pro has dual-core 1 GHz Jan 31 12:52:16 yes, way faster hehe Jan 31 12:53:19 i'm more interested in real network stack performance than fastpath hackery Jan 31 12:53:46 it sounds like a broken concept anyhow Jan 31 12:53:57 "fastpath" implies that features or actions are bypassed Jan 31 12:54:15 build #165 of avr32 is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/165 Jan 31 12:54:18 build #169 of xburst is complete: Failure [failed svn shell_15 compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/xburst/builds/169 Jan 31 12:54:20 and the linux network stack guys are getting more interested in embedded system performance these days Jan 31 12:54:38 I've seen a discussion in netdev about a some kind of fastpath for virt solutions Jan 31 12:55:07 some discussions that i'm involved in are not on netdev@ Jan 31 12:55:23 and i'm going to meet a lot of the people involved on the upcoming netdev conference Jan 31 12:55:34 great! Jan 31 14:48:42 build #162 of ixp4xx is complete: Failure [failed svn shell_15 compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/ixp4xx/builds/162 Jan 31 14:48:47 build #166 of kirkwood is complete: Failure [failed svn shell_15 compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/kirkwood/builds/166 Jan 31 14:54:39 build #172 of x86 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/x86/builds/172 Jan 31 15:07:09 build #209 of lantiq is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/lantiq/builds/209 Jan 31 16:14:14 nbd r48582 trunk/ package/kernel/mac80211/patches/004-backports-add-skb_free_frag.patch package/kernel/mac80211/patches/005-backports-add-napi_alloc_frag.patch * mac80211: backport skb_free_frag and napi_alloc_frag Jan 31 17:20:31 build #206 of ramips.mt7628 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/ramips.mt7628/builds/206 Jan 31 17:23:27 build #146 of mpc52xx is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/mpc52xx/builds/146 Jan 31 21:00:50 build #151 of mvebu is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/mvebu/builds/151 Jan 31 21:01:06 build #153 of ath25 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/ath25/builds/153 Jan 31 23:29:02 nbd r48583 trunk/scripts/portable_date.sh * add scripts/portable_date: wrapper around date for bsd/linux Jan 31 23:29:09 nbd r48584 trunk/ include/toplevel.mk scripts/get_source_date_epoch.sh * build: introduce SOURCE_DATE_EPOCH variable Jan 31 23:29:14 nbd r48585 trunk/package/Makefile * package/Makefile: override opkg installation time when SOURCE_DATE_EPOCH set Jan 31 23:29:18 nbd r48586 trunk/ include/image.mk include/kernel-build.mk scripts/sysupgrade-nand.sh * build system: have tar use $SOURCE_DATE_EPOCH for --mtime Jan 31 23:29:25 nbd r48587 trunk/include/image.mk * include/image: don't save gzip name/timestamp metadata Jan 31 23:29:34 nbd r48588 trunk/include/image.mk * include/image: use SOURCE_DATE_EPOCH if set for squashfs + ext4 images Jan 31 23:29:44 nbd r48589 trunk/target/linux/ar71xx/image/Makefile * ar71xx/image: use SOURCE_DATE_EPOCH for mksquashfs-lzma Feb 01 00:38:04 build #206 of cobalt is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/206 Feb 01 00:42:51 build #206 of cns21xx is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/cns21xx/builds/206 Feb 01 00:47:06 build #206 of orion is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/206 Feb 01 01:06:41 florian r48590 trunk/config/Config-kernel.in * buildroot: add options to build the kernel for NFS boot Feb 01 01:06:58 florian r48591 trunk/include/kernel-defaults.mk * kernel: set root on NFS when enabled Feb 01 01:35:50 build #209 of bcm53xx is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/bcm53xx/builds/209 **** ENDING LOGGING AT Mon Feb 01 02:59:58 2016