**** BEGIN LOGGING AT Mon Oct 23 03:00:01 2017 Oct 23 14:44:00 hm, is there some nice article about which type of FS to use for a RW layer on a readonly RFS setup? I'm currently using ext4 and if I do either e2fsck or mkfs.ext4 during boot on a damaged ext4 FS I run into a kernel panic and that's the thing I'm trying to avoid with a RO RFS... Suggestions for a readup? Oct 23 14:45:11 noc0lour: a kernel panic is very bad, sounds like your kernel is broken Oct 23 14:45:22 ext4 should be solid Oct 23 14:47:21 can anyone who was in the OEDEM remember who was talking about cve-report-tool? Oct 23 14:50:48 hm, should be, for testing I'm overwriting the first 4KB on the partition which is probably too hard to recover from? Oct 23 14:50:58 and does not reflect reality? Oct 23 14:51:50 that's a pretty tragic corruption but there are copies of the superblock so it should be able to recover? Oct 23 14:53:57 if I run e2fsck -y it repairs the FS but kernel panics half way through, and after a power cycle it comes up clean, I'd like to solve this without the need for a user to power cycle the device, and just run mkfs.ext4 if e2fsck returns non-zero Oct 23 14:54:24 -n will always run into a kernel panic Oct 23 14:58:23 unfortunately I'm stuck with 3.14 right now, so maybe ext4 is sucky there :/ Oct 23 15:00:15 let's see if ext3 yields the same, thanks rburton for the quick response! (: Oct 23 15:59:16 If I manually run mkfs.ext3/4 not during init everything is fine, no kernel panic nothing, but during init it panics ? Oct 23 16:30:31 noc0lour: can you pastebin the panic output somewhere? Oct 23 18:34:49 Did some ever consider building a native (as in pc build version) of the qt SDK from the OE sources, so they are in sync... Or is qemu typically being used? Oct 23 19:53:15 otavio: did you ever do that, build an qt sdk for a pc based on OE? Oct 23 20:03:10 ntl: Sure, tomorrow, for the short term I probably have to reside with tmpfs and enough swap, even if that's shitty Oct 23 22:01:50 khem, seen the new gcc7 patch about ldrd/strd? [ARM] PR 82445 Oct 23 22:02:56 ...incorrect generation of ldrd/strd with unaligned accesses that could Oct 23 22:02:56 previously have occurred on ARMv5e ... Oct 23 22:03:23 s/could/reportedly/ Oct 23 22:04:38 I am a bit lost here because even with -fno-store-merging I did see strd instructions Oct 23 22:05:00 I am surely misunderstanding Oct 23 23:00:53 khem: straight link https://github.com/gcc-mirror/gcc/commit/f59996b56aaa1c1d62a16cbb4010775b624cbde0 Oct 23 23:04:47 gn **** ENDING LOGGING AT Tue Oct 24 03:00:00 2017