| *** jjardon_ has joined #baserock | 00:21 | |
| *** jjardon_ has joined #baserock | 00:21 | |
| *** jjardon has quit IRC | 00:21 | |
| *** persia_ has joined #baserock | 00:29 | |
| *** persia_ has joined #baserock | 00:29 | |
| *** bwh_ has joined #baserock | 00:30 | |
| *** petefoth1ringham has joined #baserock | 00:31 | |
| *** richard_1aw has joined #baserock | 00:31 | |
| *** pdar_ has joined #baserock | 00:31 | |
| *** lachlanmackenzie has quit IRC | 00:35 | |
| *** bwh has quit IRC | 00:35 | |
| *** petefotheringham has quit IRC | 00:35 | |
| *** jjardon_ has quit IRC | 00:35 | |
| *** persia has quit IRC | 00:36 | |
| *** richard_maw has quit IRC | 00:36 | |
| *** pdar has quit IRC | 00:36 | |
| *** lachlanmackenzie has joined #baserock | 00:41 | |
| *** jjardon_ has joined #baserock | 00:42 | |
| *** jjardon_ has quit IRC | 00:42 | |
| *** jjardon_ has joined #baserock | 00:42 | |
| *** Walkerdine has joined #baserock | 01:01 | |
| *** zoli__ has joined #baserock | 01:45 | |
| *** zoli__ has quit IRC | 01:49 | |
| *** jjardon_ has quit IRC | 02:28 | |
| *** jjardon_ has joined #baserock | 02:33 | |
| *** petefoth has joined #baserock | 06:31 | |
| *** paulw has joined #baserock | 06:38 | |
| *** Walkerdine has quit IRC | 07:01 | |
| *** franred has joined #baserock | 07:11 | |
| *** petefoth has quit IRC | 07:19 | |
| *** jjardon_ has quit IRC | 07:19 | |
| *** toscalix__ has joined #baserock | 07:24 | |
| *** petefoth has joined #baserock | 07:24 | |
| *** jjardon_ has joined #baserock | 07:24 | |
| *** zoli__ has joined #baserock | 07:25 | |
| *** zoli___ has joined #baserock | 07:26 | |
| *** zoli__ has quit IRC | 07:26 | |
| *** bashrc has joined #baserock | 08:00 | |
| paulsherwood | hmmmm. any autotools/gcc whizzes able to guide me on http://paste.baserock.org/vupujoguke please? | 08:19 |
|---|---|---|
| paulsherwood | (moonshot failing to build gcc) | 08:19 |
| Kinnison | You're building aarch64 there, is that intended? | 08:24 |
| paulsherwood | i believe so... that's wht the moonshot is? | 08:25 |
| Kinnison | It is, yes. Okay. | 08:25 |
| Kinnison | configure: error: cannot compute suffix of object files: cannot compile | 08:26 |
| Kinnison | that's the error | 08:26 |
| Kinnison | If you still have the build tree then look for config.log in ..../gcc.build/o/aarch64*/libgcc/ | 08:26 |
| Kinnison | that may contain the actual error | 08:26 |
| Kinnison | my guess is that something has put either the stage2 libc or stage2 linkloader somewhere odd | 08:26 |
| Kinnison | Or possibly the stage 3 libc is squiffy | 08:27 |
| Kinnison | It's hard to guess | 08:27 |
| Kinnison | esp. given it has compiled somethings | 08:27 |
| *** mariaderidder has joined #baserock | 08:28 | |
| Kinnison | I'm now going to guess that the log it pastes is not the full log | 08:30 |
| Kinnison | because there's no way that's all the build log there is, not if it's using xgcc already | 08:31 |
| paulsherwood | it's 200 lines | 08:32 |
| paulsherwood | i'll put the full log up | 08:32 |
| paulsherwood | http://sprunge.us/AFZC | 08:33 |
| Kinnison | Okay, so yes, xgcc was built but nothing seems to be exercising it strongly | 08:34 |
| Kinnison | can you please look for the config.log I mentioned? | 08:34 |
| paulsherwood | http://sprunge.us/eaBi | 08:37 |
| * Kinnison looks | 08:40 | |
| Kinnison | /gcc.build/o/./gcc/cc1: error while loading shared libraries: libz.so.1: cannot open shared object file: No such file or directory | 08:41 |
| Kinnison | That's odd | 08:41 |
| Kinnison | that implies that the proper zlib isn't present in the right way | 08:41 |
| Kinnison | but it was built according to your log | 08:41 |
| Kinnison | Could you pick apart the artifact for zlib.417fa73fa8876c0f59ac9723e408b06ca97de1e578ec4277a4fd3a09073bc15d and pastebin a file listing? | 08:42 |
| paulsherwood | http://paste.baserock.org/acuhigixiv | 08:43 |
| Kinnison | okay that looks not unreasonable | 08:44 |
| Kinnison | are those files present in your assembly for building gcc? | 08:44 |
| paulsherwood | http://paste.baserock.org/mecujihude | 08:46 |
| paulsherwood | i think so | 08:46 |
| Kinnison | and can you please check what usr/lib/libz.so.1 is | 08:47 |
| paulsherwood | odd thgin is this built previously. this was my second run | 08:47 |
| paulsherwood | 0 lrwxrwxrwx 1 root root 13 Sep 16 05:59 usr/lib/libz.so.1 -> libz.so.1.2.8 | 08:47 |
| paulsherwood | root@ciat-large-0:/src/tmp/tmpiV02co# ls -ls usr/lib/libz.so.1.2.8 | 08:47 |
| paulsherwood | 104 -rwxr-xr-x 3 root root 104280 Nov 11 2011 usr/lib/libz.so.1.2.8 | 08:47 |
| Kinnison | okay so there's no reason that wouldn't work | 08:48 |
| * Kinnison hrms | 08:48 | |
| Kinnison | What's interesting is that it managed to link the cc1 against that library | 08:49 |
| Kinnison | and yet it can't run it | 08:49 |
| Kinnison | okay, final question | 08:49 |
| Kinnison | in your build tree, is there an /etc/ld.so.conf (and if so what's in it) and is there an /etc/ld.so.cache (and if so, does strings /etc/ld.so.cache | grep libz list the library?) | 08:50 |
| paulsherwood | root@ciat-large-0:/src/tmp/tmpiV02co# cat etc/ld.so.conf | 08:51 |
| paulsherwood | /lib | 08:51 |
| paulsherwood | /usr/lib | 08:51 |
| paulsherwood | /usr/local/lib | 08:51 |
| paulsherwood | include /etc/ld.so.conf.d/*.conf | 08:51 |
| radiofree | i had this exact same problem paulsherwood | 08:51 |
| paulsherwood | really? | 08:51 |
| radiofree | yep | 08:51 |
| paulsherwood | root@ciat-large-0:/src/tmp/tmpiV02co# strings etc/ld.so.cache | grep libz | 08:51 |
| paulsherwood | libz.so.1 | 08:51 |
| paulsherwood | /usr/lib/libz.so.1 | 08:51 |
| paulsherwood | libz.so | 08:51 |
| paulsherwood | /usr/lib/libz.so | 08:51 |
| paulsherwood | radiofree: did you find out the cause? | 08:52 |
| radiofree | https://irclogs.baserock.org/%23baserock.2015-06-09.log.html#t2015-06-09T15:40:56 | 08:52 |
| radiofree | paulsherwood: it ended up just working, i suspect if you start the build again it will work | 08:52 |
| Kinnison | That's not very heartening | 08:52 |
| paulsherwood | radiofree: i'm sure it'll work. so i'm not re-running it yet... | 08:53 |
| Kinnison | I have no more ideas I can query you for information for | 08:53 |
| Kinnison | that's very odd | 08:53 |
| Kinnison | everything looks like it should have worked | 08:53 |
| Kinnison | radiofree: were you building aarch64 at the time? | 08:53 |
| paulsherwood | given this has never been seen on x86 (and i must've run the build a few hundred times so far) i'm thinking it's something specific to this box/chip | 08:54 |
| paulsherwood | Kinnison: he was, same soc | 08:54 |
| Kinnison | hmm | 08:54 |
| * Kinnison goes to ask some linaro/arm peeps | 08:54 | |
| radiofree | Kinnison: yep | 08:54 |
| radiofree | are you sure libz is in your build dir? | 08:55 |
| radiofree | ah, yes it is | 08:55 |
| Kinnison | Steve McIntyre doesn't think they've seen anything like that before | 08:56 |
| Kinnison | It's interesting that it only seems to happen on aarch64 | 08:56 |
| paulsherwood | Kinnison: i expect that if i re-run the commmand, it will work. i could make a copy of the tmp dir for forensics, maybe? | 08:57 |
| Kinnison | please hive it of (just mv it out of the way) | 08:57 |
| Kinnison | I may get some time to take a look or I may tell you to abandon | 08:58 |
| * Kinnison has a meeting now, sadly | 08:58 | |
| *** ssam2 has joined #baserock | 09:01 | |
| *** ChanServ sets mode: +v ssam2 | 09:01 | |
| *** tiagogomes_ has quit IRC | 09:10 | |
| ratmice___ | paulsherwood: if you can try with LD_DEBUG="files" or "all", that might show something, or if you have the output of 'readelf -d' on the binary/lib could be relevant | 09:29 |
| *** tiagogomes_ has joined #baserock | 09:32 | |
| richard_1aw | paulsherwood: you could get that ENOENT of not being able to find the library, even when the file is in the search path and being loaded, if the library itself depends on something that it can't find, i.e. are you sure the libz you have is linked against a libc that you can find in your search path? | 11:04 |
| *** richard_1aw is now known as richard_maw | 11:04 | |
| richard_maw | if it's an intermittent failure, my guesses are 1. dentry cache corruption, I heard that was happening in some of the early ARM64 kernel versions, 2. dodgy page tables from similar bugs, 3. you could have corrupted files in your staging area/assembly. | 11:08 |
| richard_maw | 3. can happen from bit-rot, running builds without full sandboxing in the staging area, or accidentally making changes to a staging area when debugging failed builds | 11:09 |
| paulsherwood | i'm currently discounting 3 - afaik this has only ever been seen on aarch64 | 11:28 |
| *** lachlanmackenzie has quit IRC | 12:45 | |
| *** lachlanmackenzie has joined #baserock | 12:46 | |
| *** persia_ has quit IRC | 12:48 | |
| *** persia has joined #baserock | 12:50 | |
| *** persia_ has joined #baserock | 12:50 | |
| *** persia_ has joined #baserock | 12:50 | |
| *** franred has quit IRC | 12:54 | |
| *** petefoth_ has joined #baserock | 13:28 | |
| *** jjardon_ has quit IRC | 13:29 | |
| *** petefoth has quit IRC | 13:29 | |
| *** petefoth_ is now known as petefoth | 13:29 | |
| *** jjardon__ has joined #baserock | 13:30 | |
| *** zoli___ has quit IRC | 13:36 | |
| *** franred has joined #baserock | 14:02 | |
| *** paulw has quit IRC | 15:01 | |
| *** jjardon__ is now known as jjardon | 15:10 | |
| *** toscalix__ has quit IRC | 15:16 | |
| *** mariaderidder has quit IRC | 16:43 | |
| *** ssam2 has quit IRC | 16:58 | |
| *** bashrc has quit IRC | 17:08 | |
| *** franred has quit IRC | 17:12 | |
| *** zoli__ has joined #baserock | 17:15 | |
| *** franred has joined #baserock | 17:17 | |
| *** lachlanmackenzie has quit IRC | 18:31 | |
| *** zoli__ has quit IRC | 18:32 | |
| *** franred has quit IRC | 18:57 | |
| *** lachlanmackenzie has joined #baserock | 19:01 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!