| 00:14.02 | *** join/#oe rcw (~rcw@65-110-208-159.cpe.pppoe.ca) |
| 00:19.15 | *** join/#oe infobot (ibot@208.53.50.136) |
| 00:19.15 | *** topic/#oe is OpenEmbedded Developer Lounge | Web: http://openembedded.org | Repositories: http://git.openembedded.org/ | Primary Repo Mirrors: https://github.com/openembedded | This is not a distro or machine support channel |
| 00:19.38 | khem | from the changes it does not look a musl bug to me. Rather a gcc bug that is exposed by this change |
| 02:33.31 | *** join/#oe florian_kc (~florian_k@Maemo/community/contributor/florian) |
| 04:00.59 | *** join/#oe behanw (uid110099@gateway/web/irccloud.com/x-mclvmxizyzammqqx) |
| 04:08.49 | *** join/#oe kanavin_home (~ak@46.189.28.201) |
| 04:30.43 | *** join/#oe armpit (~armpit@2601:202:4000:1184:89c3:4e4b:df9b:b7de) |
| 04:32.30 | *** join/#oe kanavin_home (~ak@46.189.28.187) |
| 04:45.24 | *** join/#oe aloril_ (~aloril@80.246.146.6) |
| 05:14.18 | *** join/#oe nerdboy (~sarnold@gatekeeper.gentoogeek.org) |
| 05:14.18 | *** join/#oe nerdboy (~sarnold@gentoo/developer/nerdboy) |
| 06:35.30 | *** join/#oe cristiano (~cris@pdpc/supporter/active/cristiano) |
| 07:04.09 | *** join/#oe dqx_ (~dqx@unaffiliated/dqx) |
| 07:39.21 | *** join/#oe ao2 (~ao2@host102-86-dynamic.13-79-r.retail.telecomitalia.it) |
| 07:43.19 | *** join/#oe Jybz (~jibz@ip-37-201-6-238.hsi13.unitymediagroup.de) |
| 08:26.05 | *** join/#oe nerdboy (~sarnold@gatekeeper.gentoogeek.org) |
| 08:26.05 | *** join/#oe nerdboy (~sarnold@gentoo/developer/nerdboy) |
| 08:29.10 | *** join/#oe kanavin_home (~ak@46.189.28.198) |
| 08:48.14 | *** join/#oe nerdboy (~sarnold@gatekeeper.gentoogeek.org) |
| 08:48.14 | *** join/#oe nerdboy (~sarnold@gentoo/developer/nerdboy) |
| 09:00.39 | ant_home | khem, I'll retry the same checkout but with gcc7.3 now |
| 09:10.05 | ant_home | glibc first |
| 09:11.46 | ant_home | (I expect both to produce flawed binaries, as it was last week) |
| 09:51.05 | *** join/#oe _Chris_ (~Chris@p4FCA09E8.dip0.t-ipconnect.de) |
| 09:52.15 | _Chris_ | Anyone using here Gentoo as some host for building OE? |
| 10:59.12 | *** join/#oe nerdboy (~sarnold@gatekeeper.gentoogeek.org) |
| 12:14.54 | ant_home | khem, glibc image is ok with gcc7.3 |
| 12:15.20 | ant_home | glibc 2.27 |
| 12:16.31 | *** join/#oe Noor (~quassel@110.93.212.98) |
| 12:16.38 | *** join/#oe georgem_home (uid210681@gateway/web/irccloud.com/x-jhniqemysjqwauks) |
| 12:27.26 | *** join/#oe florian (~florian_k@Maemo/community/contributor/florian) |
| 12:32.15 | *** join/#oe koen (~koen@185.47.135.167) |
| 12:32.15 | *** join/#oe koen (~koen@linaro/koen) |
| 14:27.42 | *** join/#oe ntl (~nathanl@65-36-80-8.dyn.grandenetworks.net) |
| 14:34.07 | ant_home | khem, while musl image spits align traps... |
| 14:35.01 | ant_home | so with both gcc 7.3 and 8.1 |
| 14:46.37 | *** join/#oe gbritton (~Adium@209.6.40.118) |
| 14:55.35 | *** join/#oe sr105 (~sr105@72.187.19.10) |
| 15:05.59 | *** join/#oe koen (~koen@185.47.135.167) |
| 15:05.59 | *** join/#oe koen (~koen@linaro/koen) |
| 16:51.21 | *** join/#oe blight (~greg@80-109-10-222.cable.dynamic.surfer.at) |
| 16:51.21 | *** join/#oe blight (~greg@reactos/developer/blight) |
| 17:02.36 | khem | ant_home: hey |
| 17:05.34 | *** join/#oe ao2 (~ao2@host102-86-dynamic.13-79-r.retail.telecomitalia.it) |
| 17:18.24 | ant_home | hi khem |
| 17:22.53 | ant_home | khem, samegcc7.3 with musl c4bc0b1a64e1 (just before that commit) is ok |
| 17:24.09 | ant_home | so musl starting with 32482f61da76 fails with both gcc's |
| 17:24.42 | ant_home | same binutils |
| 17:26.32 | ant_home | I don't see compile warnings |
| 17:47.20 | *** join/#oe _Chris_ (~Chris@p4FCA09E8.dip0.t-ipconnect.de) |
| 17:51.54 | _Chris_ | When I want to build sumo under Gentoo, I get some segfault for m4 related to ld-2.27 (glibc) in the syslog and it fails to build autofonc. I also have this problem with rocko and pyro, but NOT with morty. It works here. Bitbake version is unrelated to this it seems. Anyone has some clue? It's really bad. |
| 17:52.22 | _Chris_ | *autoconf |
| 17:53.28 | _Chris_ | The buildlog don't show anything special then. I just shows that I lack m4, which is simply because of the segfault. |
| 17:54.05 | khem | _Chris_: you might have debug through it, seems host related |
| 17:55.19 | _Chris_ | "[23972.181338] m4[13146]: segfault at 4 ip b7ee5538 sp bfe9c100 error 4 in ld-2.27.so[b7ed9000+25000]" |
| 17:55.50 | _Chris_ | RAM is okay and I am building anything other. I don't have any clue. It simply makes no sense. |
| 17:57.57 | ant_home | khem, I can compile with debug enabled if it helps |
| 17:58.35 | _Chris_ | How can I do that? |
| 17:59.47 | ant_home | khem, btw, one of last musl commits is about avoiding r7 for arm |
| 17:59.57 | ant_home | in fact here we have a syscall |
| 18:00.14 | ant_home | which happens to use r7 |
| 18:02.10 | ant_home | but that sould happen in thumb mode and we did test disabling thumb was not enough, the point was dsp |
| 18:15.11 | _Chris_ | Now it segfaults with make. |
| 18:22.13 | khem | ant_home: gcc is failing to align ldrd |
| 18:22.23 | khem | thats really the problem |
| 18:30.49 | khem | that change in musl is just highlighting it |
| 18:31.24 | khem | so I would be happy if you can reduce the given musl function into a stand alone file with -E -P |
| 18:31.38 | khem | that way we can narrow down the case and submit it upstream |
| 18:32.07 | khem | _Chris_: JaMa uses gentoo so he might have some insights |
| 18:32.31 | khem | my suggestion would be to use one of supported distros for best results |
| 18:33.22 | khem | I personally use arch but I also run into the obscure issues that no one else sees, guess thats the price I pay for using what I want :) |
| 18:42.59 | _Chris_ | khem: I can't use anything different than Gentoo, because.... I only have Gentoo. Since more than 10 years.... Bad argument. |
| 18:43.31 | _Chris_ | JaMa: Do you have any clue what could be the problem? |
| 18:46.26 | *** part/#oe christner_home (~dchristne@70-91-95-105-washington.dc.hfc.comcastbusiness.net) |
| 18:47.53 | _Chris_ | When will be kernel 4.16 or 4.17 included? Most recent version I saw was 4.15. |
| 18:59.29 | _Chris_ | Is it possible to start with some precompiled toolchain for building OE? |
| 19:02.40 | _Chris_ | Or that I can use some precompiled autoconf package? |
| 19:07.04 | *** join/#oe nerdboy (~sarnold@gatekeeper.gentoogeek.org) |
| 19:07.04 | *** join/#oe nerdboy (~sarnold@gentoo/developer/nerdboy) |
| 19:09.35 | *** join/#oe _Chris_ (~Chris@p4FCA09E8.dip0.t-ipconnect.de) |
| 19:17.29 | *** join/#oe georgem_home (uid210681@gateway/web/irccloud.com/x-yfcaptdfgpxynwhm) |
| 19:41.22 | *** join/#oe behanw (uid110099@gateway/web/irccloud.com/x-pxbrkvfytzpmqlcf) |
| 19:43.26 | JaMa | _Chris_: I'm using Gentoo but for last half year or so I'm using docker images with Ubuntu for OE builds (to match what the rest of the company is using) |
| 19:45.47 | *** join/#oe bluelightning (~paul@pdpc/supporter/professional/bluelightning) |
| 19:47.28 | _Chris_ | JaMa: It seems to be a host specific problem. To be honest, I have to say, that I experienced this segfaults also with different binutils versions. |
| 19:48.38 | bluelightning | generally segfaults during the build mean either a host distro issue or a hardware issue |
| 19:52.09 | _Chris_ | I experienced some kind of segfaults on all my x86 hosts since some time. But until now they were only existing with specific binutils versions. 2.30 and 2.29.1 not, for example. |
| 19:52.33 | _Chris_ | Only while building. Memtest verified that the hardware seems to be okay. |
| 19:52.52 | _Chris_ | Sadly I don't have some amd64 machine here. |
| 20:02.22 | *** join/#oe nerdboy (~sarnold@gatekeeper.gentoogeek.org) |
| 20:02.23 | *** join/#oe nerdboy (~sarnold@gentoo/developer/nerdboy) |
| 20:15.05 | *** join/#oe Marex (~Marex@195.140.253.167) |
| 20:48.33 | *** join/#oe Chris_ (~Chris@p4FCA083A.dip0.t-ipconnect.de) |
| 22:13.17 | *** join/#oe lpotter (~quassel@2001:8003:7c13:e000:ba27:ebff:febb:59b) |
| 22:32.59 | *** join/#oe florian (~florian_k@Maemo/community/contributor/florian) |
| 22:38.03 | *** join/#oe Chris_ (~Chris@p4FCA083A.dip0.t-ipconnect.de) |
| 22:47.35 | *** join/#oe Chris_ (~Chris@p4FCA083A.dip0.t-ipconnect.de) |
| 22:57.15 | *** join/#oe Chris_ (~Chris@p4FCA083A.dip0.t-ipconnect.de) |
| 23:17.54 | *** join/#oe Chris_ (~Chris@p4FCA083A.dip0.t-ipconnect.de) |