**** BEGIN LOGGING AT Thu Jan 14 02:59:57 2021 Jan 14 04:22:38 on an Intel i210 using the igb driver, why would "ethtool --config-ntuple eth0 flow-type udp4 dst-ip 224.0.0.0 m 255.255.255.0 dst-port 443 action 1" give me "rmgr: Cannot insert RX class rule: Invalid argument" How does one troubleshoot this, last mention I found for it is from 2015? - Jan 14 04:22:38 https://patchwork.ozlabs.org/project/intel-wired-lan/patch/1451456399-13353-1-git-send-email-gangfeng.huang@ni.com/ Jan 14 04:24:51 ethtool -n eth0 shows: 4 RX rings available Total 0 rules Jan 14 08:54:41 build #45 of realtek/generic is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/realtek%2Fgeneric/builds/45 Jan 14 10:14:25 blogic: ping Jan 14 10:16:07 mangix: where are they able to conflict? URL_FILENAME is only used for download, the file ending up in dl/ is zstd-v1.4.4.tar.gz Jan 14 10:21:33 just tested, the sources fallback is also working with this.or is the conflict expected to be somewhere else? Jan 14 10:22:09 oh I see. In any case, the downloaded tarball is generated by github's codeload service, which has been known to change hashes Jan 14 10:22:27 best to use the release tarballs as in the packages feed Jan 14 10:22:50 thoseare stable Jan 14 10:25:07 I think I fixed the download URL in a version update but that one got rejected IIRC. Jan 14 10:25:45 ah right, there was a right or wrong on tarballs from github Jan 14 10:26:00 however, the hash matches the one which was there previously Jan 14 10:28:24 note that the packages feed uses a different tarball. azstd one. Jan 14 10:28:56 not for 1.4.4 Jan 14 10:29:58 1.4.4 on the sources mirror came from codeload whose hash still matches with the one served from github today Jan 14 10:30:51 So switching to the release tarball would end up breaking the sources mirror Jan 14 10:31:10 so we are currently stuck with what we have here, aren't we? Jan 14 10:31:29 at least until someone updates to a more recent release Jan 14 10:31:37 right. Jan 14 10:34:31 i'll try to whip one up tomorrow. Jan 14 10:39:19 mangix: 5.0.4.0 is the latest i can see from late 2018 Jan 14 10:39:47 aparcar[m]: I've had a workaround in my tree since the bump which at least compiles - I don't think fakeroot actually works under Mac OS 11 aka Big Sur but I don't need it to for a build https://git.openwrt.org/?p=openwrt/staging/ldir.git;a=commit;h=ce976f1e164622f03b504f5c5de68875f5b1bc31 Jan 14 11:04:07 blogic: ok Jan 14 11:20:58 build #807 of ipq806x/generic is complete: Failure [failed tools] Build details are at http://buildbot.openwrt.org/master/images/builders/ipq806x%2Fgeneric/builds/807 blamelist: Etan Kissling , Daniel Golle , Andy Walsh , Leon M. George Jan 14 12:07:01 mangix: do you still want that ? Jan 14 13:39:25 https://github.com/SELinuxProject/selinux/blob/master/secilc/docs/README.md Jan 14 13:39:30 wrong chan Jan 14 13:40:47 >KGB-1< https://tests.reproducible-builds.org/openwrt/openwrt_mediatek.html has been updated. (100.0% images and 98.3% packages reproducible in our current test framework.) Jan 14 14:47:16 jow: ping Jan 14 15:05:37 WTF?! Who thought was a good idea to unconditionally enable KERNEL_IPV6_SEG6_LWTUNNEL if !SMALL_FLASH? Jan 14 15:06:35 It should be a prompt, not everyone will want a 25 kiB bloat for a niche feature. Jan 14 15:07:05 I know *I* don't. Jan 14 15:24:18 build #808 of ipq806x/generic is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/ipq806x%2Fgeneric/builds/808 Jan 14 16:52:12 hello everybody, i wonder why mtune is 24kc instead of 1004kc with xiomi mi 3 pro? Jan 14 16:54:13 gcc treats 1004kc as an alias for 24kc Jan 14 16:54:34 ah, thats why. makes sense. thank you nbd Jan 14 17:08:54 nbd: that reminds me, the 34kc is also aliased internally to 24kc. :) Jan 14 17:10:33 But the 308-mips32r2_tune.patch sets -mtune=34kc. ;) Jan 14 17:28:18 wasn't the difference that 34k has the dsp extension that is not used regardless ? Jan 14 17:28:46 and kernel level code will alwasy use the mips32r2 opcode set which is a subset of all the other flvours ? Jan 14 17:49:43 blogic: I think it's the 74Kc which has the DSPv2 extensions. Don't know about the 34Kc. Anyway, every time I tried to build the userspace with DSPv2 instructions, my 74Kc machines would just hang. :P Jan 14 18:18:29 ldir: i think it used to work for a bit. I'll post an issue upstream but worst case I should revert it... Jan 14 18:19:01 * aparcar[m] speaking of testing, can I have access to your Mac mini again for ci testing? Jan 14 22:06:51 free on openwrt give me: Jan 14 22:07:00 total used free shared buff/cache available Jan 14 22:07:07 123216 54400 24996 38256 43820 4192 Jan 14 22:07:35 total = used + free + buff/cache which make sense Jan 14 22:08:15 but, isn't avaiable ~= free + buff/cache ? what is the difference between free vs available? Jan 14 22:09:28 shared is for tmpfs/shmem mostly Jan 14 22:21:11 figured it out: shared took 38256 out of 43820 and can not be reclaimed freely, so avaiable is only 4192, now all the memory I can use will be 'free+avaiable' as the shared memory can not be freed by the OS Jan 14 22:22:12 in short, buff/cache is not just kernel internal buffers, it also includes shared which is shm, well, also kernel buffers actually Jan 14 22:48:46 aparcar[m]: https://www.reddit.com/r/hackintosh Jan 14 22:49:16 mangix: not wants Jan 14 22:51:16 could also set up a QEMU i stance Jan 14 22:51:26 *instance Jan 14 23:04:45 mhh will think about it thank you Jan 15 00:00:32 local patches went down from 50 to 48 Jan 15 00:00:33 cool Jan 15 00:05:55 aparcar[m]: please look into https://patchwork.ozlabs.org/project/openwrt/list/?series=222569 when you have the time. Jan 15 00:10:54 what are these non readable errors? https://downloads.openwrt.org/snapshots/faillogs/mips_mips32/packages/lxc/compile.txt Jan 15 01:29:40 mangix: it's in the mix https://buildmaster.aparcar.org/master/phase1/ Jan 15 01:43:33 rsalvaterra: ping Jan 15 01:45:33 anyone here familiar with hotplugging? Jan 15 01:45:43 I'm trying to port udev rules to hotplug.d Jan 15 02:01:20 oh no. this looks like I need to patch usbmode to handle this Jan 15 02:05:27 build #789 of arc770/generic is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/arc770%2Fgeneric/builds/789 **** ENDING LOGGING AT Fri Jan 15 02:59:57 2021