**** BEGIN LOGGING AT Fri Jun 22 03:00:01 2018
Jun 22 08:08:23 <greguu> ant_work: hi, how did you go with the alignment issue ?
Jun 22 08:08:59 <greguu> ant_work: I like to try iwmmxt as march, mcpu and mtune, but I think it uses version 4 ABI and not EABI ?
Jun 22 08:12:10 <ant_work> hi greguu
Jun 22 08:12:32 <ant_work> I could bisect up to a musl upgrade
Jun 22 08:13:40 <ant_work> but I had align traps in the logs with glibc as well so I can not trust completely th etoolchain
Jun 22 08:14:11 <ant_work> thgis is the musl upgrade
Jun 22 08:14:13 <ant_work> -SRCREV = "628cf979b249fa76a80962e2eefe05073216a4db"
Jun 22 08:14:13 <ant_work> +SRCREV = "55df09bfccbfe21fc9dd7d8f94550c0ff25ace04"
Jun 22 08:14:24 <ant_work> first good, second bad
Jun 22 08:15:31 <greguu> hm, maybe still OE issue.. I just rebuild void based on their latest git, using 7.3.0
Jun 22 08:16:02 <greguu> alignment issues may point to abi mismatch ?
Jun 22 08:17:20 <ant_work> is it musl?
Jun 22 08:17:42 <ant_work> I think it is this single change
Jun 22 08:17:43 <ant_work> https://git.musl-libc.org/cgit/musl/commit/?id=1bc10ffeaa7c7ce44b3e214e02e302642511c7c7
Jun 22 08:18:34 <ant_work> there is somewhere a mismatch with off_t size
Jun 22 08:18:55 <ant_work> so I guess it is +#define _DIRENT_HAVE_D_OFF
Jun 22 08:19:13 <ant_work> wrong offset from some header..I'll test latert
Jun 22 08:21:35 <ant_work> greguu, about mcpu=xscale and the failure using thexscale instead of arm56te
Jun 22 08:21:48 <ant_work> there are fixes for gcc7 and gcc8 like
Jun 22 08:21:49 <ant_work> https://github.com/gcc-mirror/gcc/commit/f18d79595db4ae88dbd40d73bb2d1ea191a756a1#diff-092e0ffb260f6293b5cd7bfb660278d6
Jun 22 08:22:00 <ant_work> very recent
Jun 22 08:22:42 <ant_work> argh I have desk full and can not type today
Jun 22 08:23:41 <ant_work> greguu, I doubt your toolchain already has this...
Jun 22 09:59:47 <greguu> ant_work: let me check tomorrow
Jun 22 10:00:33 <greguu> I doubt its musl, but...
Jun 22 10:01:56 <ant_work> I verified it
Jun 22 10:02:16 <ant_work> but it happens with glibc as well so I am dubious
Jun 22 10:02:55 <ant_work> I'll bisect glibc this night
Jun 22 10:03:34 <ant_work> for musl, there are very few commits between god and bad
**** ENDING LOGGING AT Sat Jun 23 03:00:01 2018