**** BEGIN LOGGING AT Sun May 16 03:00:13 2021
May 16 03:15:26 <stintel> surprisingly little breakage with GCC10 and CONFIG_ALL=y
May 16 04:15:09 <mangix> stintel: :)
May 16 04:33:55 <owrt-2102-builds> Build [#47](https://buildbot.openwrt.org/openwrt-21.02/images/#builders/42/builds/47) of `sunxi/cortexa53` failed.
May 16 05:03:48 <silverwhitefish> does anyone have any documentation on this global option: Remove ipkg/opkg status data files in final images
May 16 05:04:09 <silverwhitefish> what exactly does it remove>
May 16 05:04:11 <silverwhitefish> ?
May 16 05:11:03 <silverwhitefish> is it files relating informational command 'status [pkg|globp]' 	-> Display all status for <pkg> ?
May 16 07:38:53 <silverwhitefish> guess i'll just have to put that one on a todo list
May 16 07:41:29 <silverwhitefish> wizards just told me to query code: https://github.com/openwrt/openwrt/search?q=clean_ipkg
May 16 08:56:28 <mangix> meh
May 16 09:09:33 <owrt-snap-builds> Build [#102](https://buildbot.openwrt.org/master/images/#builders/61/builds/102) of `arc770/generic` failed.
May 16 11:21:43 <Grommish> blogic: ping
May 16 12:15:03 <ldir> stintel: It's been suggested to me that adding "-fno-strict-aliasing" would be an interesting data point re: grub2 and legacy boot.  I'm not in a position to test.
May 16 12:38:59 <PaulFertser> I wonder if booting Linux directly as an EFI application was considered.
May 16 13:00:43 <stintel> ldir: you mean unfilter -O2 and add -fno-strict-aliasing ?
May 16 13:00:52 <stintel> PaulFertser: not relevant, our hw doesn't do EFI
May 16 13:01:19 <PaulFertser> stintel: :) must be old
May 16 13:01:28 <stintel> not *that* old
May 16 13:02:23 <ldir> stintel: yes
May 16 13:02:26 <stintel> actually not sure about coreboot
May 16 13:02:32 <stintel> APU2 BIOS is coreboot
May 16 13:02:38 <stintel> does that do EFI?
May 16 13:03:00 <stintel> ldir I'll give it a shot
May 16 13:03:19 <stintel> just arrived home after fleeing the rain in some event in some park :P
May 16 13:03:27 <ldir> I don't think so - I find the whole UEFI EFI thing confusing.
May 16 13:05:15 <ldir> 1st hit for me "Coreboot is an open source firmware, and used instead of EFI on supported platforms."
May 16 13:12:09 <PaulFertser> There's a UEFI implementation for coreboot, how else it would be booting windows :)
May 16 13:12:23 <PaulFertser> But yes, if it's coreboot I wouldn't bother with EFI of course.
May 16 13:12:36 <PaulFertser> btw, U-boot implements some essential parts of UEFI API too.
May 16 13:15:22 <ldir> it's beyond my brain cell today - and I've a headache
May 16 13:44:24 <ldir> stintel: Looks like I've managed to get virtualbox to do openwrt VMs - and the -fno-strict-aliasing doesn't fix the problem.
May 16 13:44:40 <stintel> oh
May 16 13:44:52 <stintel> interesting, virtualbox exposes the bug, qemu/kvm doesn't
May 16 13:44:55 <stintel> what the ..
May 16 13:47:34 <ldir> I could have done something wrong!
May 16 13:48:02 <stintel> :D
May 16 13:48:14 <stintel> ok I will try it
May 16 14:27:27 <stintel> ldir: same result
May 16 16:37:22 <wingsorc> I'm trying to add a new device (tl-wa1201v2) to the firmwarelayout in tplink-safeloader.c. Besides filling out the product_name and product_ver there is an entry called special_id. Where do I get that one from?
May 16 17:14:52 <PaulFertser> wingsorc: probably from vendor's firmware?
May 16 17:15:31 <Tapper> So people I am building for my r7800 all is working just fine. I used ./scripts/env to make a new env for my c7 and I cant get it to build. I cant get it to work no matter what I do.
May 16 17:15:32 <Tapper> Toolchane GCC final failed to build.
May 16 17:15:32 <Tapper> libtool: install: `libcc1plugin.la' is not a valid libtool archive
May 16 17:15:32 <Tapper>  libtool: install: Try `libtool --help --mode=install' for more information.
May 16 17:15:32 <Tapper>  full log: https://paste.ubuntu.com/p/2c8pfykt4y/
May 16 17:18:56 <PaulFertser> Tapper: is that with -j1?
May 16 17:18:56 <Tapper> No -j17
May 16 17:18:56 <PaulFertser> Tapper: when debugging build issues you need -j1
May 16 17:18:56 <Tapper> PaulFertser OK I will run a new build j1
May 16 17:18:59 <Tapper> It's building now.
May 16 17:21:57 <Tapper> Here's the log https://paste.ubuntu.com/p/pvbYVqZQbx/
May 16 17:22:01 <Tapper> -j1
May 16 17:25:28 <PaulFertser> Tapper: I suggest you "rm /home/tapper/openwrt/build_dir/toolchain-mips_24kc_gcc-8.4.0_musl/gcc-8.4.0-final/libcc1/libcc1.la' and retry.
May 16 17:30:58 <Tapper> PaulFertser no mate still crashing. Do you want to see the new log?
May 16 17:32:15 <PaulFertser> Tapper: yes
May 16 17:32:15 <Tapper> https://paste.ubuntu.com/p/yM3bVsjJfW/
May 16 17:32:47 <Tapper> This is using GCC 8 I could switch to using gcc 9 or 10.
May 16 17:33:00 <PaulFertser> Tapper: ok, but that's different. I suggest this: find /home/tapper/openwrt/build_dir/toolchain-mips_24kc_gcc-8.4.0_musl/gcc-8.4.0-final/ -name \*.lo | xargs rm
May 16 17:33:08 <Tapper> My r7800 builds with gcc10
May 16 17:33:51 <PaulFertser> Tapper: or just remove the toolchain build dir and have it rebuilt from scratch automatically.
May 16 17:33:55 <Tapper> PaulFertser I don't know what you mene
May 16 17:34:26 <blogic> Grommish_: pong
May 16 17:34:28 <Tapper> How to remove the toolchane dir? just rm -r toolchane?
May 16 17:34:48 <PaulFertser> Tapper: I think you have some old artifacts, probably copied the build dir from another machine or something. Yes, rm -rf build_dir/toolchain*
May 16 17:35:14 <Tapper> The thing is I am using ./scripts/env to do my builds.
May 16 17:38:57 <ldir> If you're going to the level of removing and then rebuilding the toolchain, then you may as well do 'make dirclean'  - but before doing that, have you updated the env held config?  ie. have you done a 'make defconfig' and then use './scripts/env diff' to see if anything changed?
May 16 17:41:30 <Tapper> ldir Hi mate I did update the env config and I did not do a dir-clean because I don't want to lose my configs for my r7800 if I can help it
May 16 17:41:54 <Tapper> I am rebuilding now lets see
May 16 17:42:42 <ldir> 'make dirclean' doesn't destroy the env directory hence your configs are safe assuming that you've done ./scripts/env save  to keep them updated.
May 16 17:43:01 <Tapper> I am so glad I have a fast PC because rebuilding the toolchane takes a goddam age!
May 16 17:43:23 <Tapper> ldir yes mate I do env save.
May 16 17:43:33 <Tapper> I learned the hard way. lol
May 16 18:05:39 <Tapper> make dir-clean fails.
May 16 18:05:45 <Tapper> WTF?
May 16 18:06:18 <Tapper> even make clean fails
May 16 18:06:46 <Tapper> This is a new env I just set up.
May 16 18:07:38 <svanheule> blogic: did you see https://github.com/openwrt/openwrt/pull/4078 ?
May 16 18:08:28 <Tapper> PaulFertser ldir Any pointers?
May 16 18:09:23 <ldir> Tapper: make dirclean itself fails or the build after doing a make dirclean fails ?
May 16 18:09:33 <Tapper> WTF I just did make clean V=sc and it says that it cant rm files as it is a read onley filesystem
May 16 18:10:04 <Tapper> This could be a WSL bug.
May 16 18:10:24 <Tapper> I can still build for my r7800 tho.
May 16 18:10:42 <Tapper> how can that even be?
May 16 18:10:46 <ldir> I'm without clue!
May 16 18:10:54 <Tapper> lol me to mate
May 16 18:11:40 <Tapper> I think I will delete the hole openwrt dir and start from scratch
May 16 18:12:12 <Tapper> Can I save my .configs? So I don't have to use make menuconfig again for my r7800?
May 16 18:13:14 <ldir> Tapper: should be able to - the env directory is basically a git repository, each environment is a branch in that git repo
May 16 18:13:34 <Tapper> OK
May 16 18:33:59 <Hauke> ynezz: which is an alias on Fedora 34 and this check fails: https://git.openwrt.org/13069b1a1dcbef0bb9b42db98025fa9203283220
May 16 18:39:06 <swalker> updated openwrt/upstream, https://sdwalker.github.io/uscan/index.html
May 16 19:25:53 <karlp> Hauke: but the alias is still printed and grep selects which, so how does it fail?
May 16 19:54:03 <rsalvate_> Uhh… mangix closed his musl 1.2.2 pull request, WTF…?! o_O
May 16 20:17:57 <owrt-snap-builds> Build [#93](https://buildbot.openwrt.org/master/images/#builders/9/builds/93) of `lantiq/ase` failed.
May 16 20:34:49 <Grommish_> blogic: I've got this ER10x which uses both the MT7621 and the RTL8367RB.  I've got a 4.14 kernel from Ubuiquti and I was working to port the drivers.  Did the ./drivers/net/ethernet/mediatek get migrated to ralink?  Am I understanding that correctly?
May 16 20:35:23 <Grommish_> Tapper: Do wat?
May 16 20:36:19 <Tapper> Grommish_ I don't know mate
May 16 20:36:30 <Tapper> I don't know what I put.
May 16 20:37:07 <Grommish_> Tapper: If you really want to flush the sysmte:  rm -rf bin/ build_dir/ staging_dir/ tmp/ .ccache/
May 16 20:37:10 <Tapper> lol was chilling and eating some verry good pixa
May 16 20:37:48 <Grommish_> Tapper: I'm still dicking with this driver.. I did setup the studio again though
May 16 20:37:53 <Grommish_> We'll have to try it later
May 16 20:38:01 <Tapper> OK thanks mate, but WSL was telling me that it cant remove files because I have a read onley filesystem.
May 16 20:38:10 <Tapper> fucking messed up
May 16 20:38:16 <Grommish> Tapper: I'm using WSL without issues.. What did you do ;p
May 16 20:38:33 <Grommish> Tapper: Hell, I even rebuilt a 5.10 kernel for WSL2 Ubuntu hehe
May 16 20:38:46 <Grommish> You busted it on a level I'm impressed by
May 16 20:38:52 <Tapper> lol I don't know what I broke mate
May 16 20:39:07 <Tapper> hahah that is how I get down
May 16 20:39:41 <Grommish> Tapper: I did Find out something for when you get it fixed.. You can \\wsl$\ to access your WSL Filesystem from Windows
May 16 20:39:50 <Tapper> I will jump on discord a bit later k
May 16 20:40:00 <Tapper> Cool
May 16 20:40:04 <Grommish> from file explorer/etc
May 16 20:42:42 <PaulFertser> Grommish: I'm fairly sure it did not. ethernet/mediatek are current upstream drivers, and before they were available ethernet/ralink OpenWrt code was added.
May 16 20:44:01 <Grommish> PaulFertser: Well, I know blogic is listed as the Module author, so I figured I could go to the source ;)  Maybe roll a single tear and see if I can get the master to help hehe although, I did find this.. https://github.com/openwrt/openwrt/pull/4178
May 16 20:47:25 <Grommish> PaulFertser: I think one of the biggest differences i'm seeing i the DSA driver for the mt7621 which wasn't a thing in the 4.14 kernel I have
May 16 21:01:19 <blogic> svanheule: apparently not
May 16 21:41:42 <karlp> Hauke: well, confirmed it now after actually trying to build at least :)
May 16 21:42:51 <karlp> chekcing for two versions of g++ seems odd too? https://paste.jvnv.net/view/uUL6k
May 16 21:50:39 <ldir> karlp: ha - the text is misleading - the 1st line is 'I'm trying to find the compiler', the 2nd line is 'I used the found compiler but it didn't work'
May 16 21:50:55 <karlp> yeah, I'm just reconfiguing git send email to send a patch :)
May 16 21:51:07 <owrt-2102-builds> Build [#48](https://buildbot.openwrt.org/openwrt-21.02/images/#builders/42/builds/48) of `sunxi/cortexa53` completed successfully.
May 16 21:51:10 <ldir> the version numbering for the 2nd line should be altered.
May 16 21:51:18 <karlp> yeah, already fixed it locally :)
May 16 21:57:46 <karlp> now sure how to fix the "which" thing though
May 16 21:58:24 <karlp> heh, says, "use FORCE=1 to override"
May 16 21:58:35 <karlp> do that... still just says no,
May 16 21:58:57 <karlp> make menuconfig FORCE=1, exported, no difference.
**** BEGIN LOGGING AT Mon May 17 01:17:10 2021
May 17 02:31:32 <owrt-snap-builds> Build [#97](https://buildbot.openwrt.org/master/images/#builders/64/builds/97) of `realtek/generic` failed.
May 17 02:31:51 <owrt-snap-builds> Build [#103](https://buildbot.openwrt.org/master/images/#builders/61/builds/103) of `arc770/generic` completed successfully.
May 17 02:48:57 <m4t> noltari: for https://github.com/openwrt/packages/commit/e9a126c14509a2d83c2b679a3a471099c5fd9eb3 , what if a kmod is needed before the hw rng enumerates (e.g. ttyACM0)? those seem to  come in a bit later. should it move to S11 after boot (kmodloader) at S10 ?
May 17 02:54:21 <m4t> i moved it to START=11 and it actually starts now...
**** ENDING LOGGING AT Mon May 17 02:59:57 2021