**** BEGIN LOGGING AT Mon Nov 07 02:59:56 2022
Nov 07 08:17:29 <mckoan> good morning
Nov 07 14:14:52 <zhmylove> RP: hi! So you're suggesting to remove all the mtime stuff + change fstab only in case it was touched, right?
Nov 07 14:27:43 <RP> zhmylove: yes, I see a v3 of the patch
Nov 07 14:29:45 <zhmylove> RP: actually I do really sure that there should be a way to deal with mtimes... We have REPRODUCIBLE_TIMESTAMP_ROOTFS variable though
Nov 07 14:33:37 <RP> zhmylove: I still suspect upstream might take a patch which sets all the times to mtime if you set the option
Nov 07 14:53:53 <RP> zhmylove: in fact now I think about it, why doesn't the extfs tooling set ctime/crtime to SDE itself?
Nov 07 14:54:02 <RP> I suspect they might take that patch
Nov 07 14:54:59 <zhmylove> RP: it's due to we fill rootfs using "mkfs.ext4 -d ${IMAGE_ROOTFS}" which is not aware (and has no way to specify) about SDE-related stuff
Nov 07 15:01:03 <RP> zhmylove: If SOURCE_DATE_EPOCH is set in the environment, it could act on that though?
Nov 07 15:07:48 <zhmylove> RP: maybe. It'd be great to discuss this w/ extfs toolchain maintainers
Nov 07 15:23:26 <JPEW> Does it just pull the mtimes from source files?
Nov 07 15:26:21 <JPEW> Ah, there is some precendence there. mkfs.fat supports SOURCE_DATE_EPOCH
Nov 07 15:39:20 <JPEW> Reading through the code, there is a variable called `E2FSPROGS_FAKE_TIME` that might help; it only sets the time for some of the FS metadata stuff (journal & such), not the actual files being copied in though. I'm not sure if that is enough or not
**** ENDING LOGGING AT Tue Nov 08 02:59:56 2022