**** BEGIN LOGGING AT Fri Jan 01 03:00:13 2021
Jan 01 05:20:50 >KGB-0< https://tests.reproducible-builds.org/openwrt/openwrt_kirkwood.html has been updated. (100.0% images and 97.3% packages reproducible in our current test framework.)
Jan 01 09:39:21 <rsalvaterra> Hauke: The 74K is out-of-order, and there are lots of popular ath79 devices based on the 74K (WDR3600 and WDR4300, from the top of my head).
Jan 01 09:39:40 <rsalvaterra> Also, Happy New Year, everyone. :)
Jan 01 10:15:31 <Tapper> Happy New Year to everyone here.
Jan 01 12:04:56 <mangix> rsalvaterra: that's news to me
Jan 01 12:05:31 <rsalvaterra> mangix: The fact that 74K(Ec) is OoO?
Jan 01 12:06:50 <rsalvaterra> Here: https://s3-eu-west-1.amazonaws.com/downloads-mips/documents/MD00496-2B-74KC-DTS-01.07.pdf
Jan 01 12:07:19 <rsalvaterra> "The 74Kc core implements the MIPS32 Release 2 Architecture in a superscalar, out-of-order execution pipeline."
Jan 01 12:10:32 <mangix> wonder if compilation as 24kc makes a difference
Jan 01 12:12:47 <rsalvaterra> mangix: If that was the case, we'd just compile x86 code for Atom… :P
Jan 01 12:13:00 <rsalvaterra> (Bonnell)
Jan 01 12:14:25 <mangix> :D
Jan 01 12:20:17 <Borromini> wasn't there a patch for a short while that compiled 74k for certain SoCs on openwrt? and then abandoned because of the tradeoff (extra needed resources not worth the tiny performance benefit)
Jan 01 12:21:25 <Borromini> i know ldir keeps a 74kc patch in his tree for some ath79 targets
Jan 01 12:21:32 <Borromini> https://git.openwrt.org/?p=openwrt/staging/ldir.git;a=commit;h=49a360e97777d6069cdbd879ad6e76347171c606
Jan 01 12:22:45 <rsalvaterra> Borromini: I compile my images with -mtune=74kc for my 74Kc targets.
Jan 01 12:23:39 <rsalvaterra> https://gitlab.com/rsalvaterra/openwrt/-/commit/2a86e12297187527a9da3e806e9f906aad21bef3
Jan 01 12:23:43 <Borromini> ok
Jan 01 12:24:07 <Borromini> looks like the same patch ldir is carrying
Jan 01 12:25:07 <rsalvaterra> Yeah, pretty much.
Jan 01 12:28:10 <rsalvaterra> I also have this one I haven't sent yet… https://gitlab.com/rsalvaterra/openwrt/-/commit/082d3b702e3778535293d3cf92f56cf51633b6dd
Jan 01 12:29:30 <rsalvaterra> … but since it doesn't change anything in terms of code generation, I don't know if upstream is interested.
Jan 01 13:23:27 <Hauke> mangix: here is a comment from the former Linux mips maintainer: https://www.linux-mips.org/archives/linux-mips/2018-01/msg00101.html
Jan 01 13:39:25 <russell--> Borromini: i recall 74k existing for quite a while (years) before it got merged back to 24k to save on maintence effort
**** BEGIN LOGGING AT Fri Jan 01 17:57:33 2021
Jan 01 18:31:43 >KGB-0< https://tests.reproducible-builds.org/openwrt/openwrt_omap.html has been updated. (0% images and 97.0% packages reproducible in our current test framework.)
Jan 01 20:16:47 <aparcar[m]> jow: are you okay with moving usbutils to packages.git?
Jan 01 20:23:47 <owrt-snap-builds> build #709 of ramips/rt3883 is complete: Failure [failed pkgbuild]  Build details are at http://buildbot.openwrt.org/master/images/builders/ramips%2Frt3883/builds/709  blamelist: Rosen Penev <rosenp@gmail.com>, Rui Salvaterra <rsalvaterra@gmail.com>, David Bauer <mail@david-bauer.net>, Adrian Schmutzler <freifunk@adrianschmutzler.de>, Russell Senior
Jan 01 20:23:47 <owrt-snap-builds> <russell@personaltelco.net>, Stijn Segers <foss@volatilesystems.org>, Eneas U de Queiroz <cotequeiroz@gmail.com>, Hauke Mehrtens <hauke@hauke-m.de>
Jan 01 20:25:20 <rsalvaterra> Yay, I'm on the blame list! \o/
Jan 01 20:31:42 <Borromini> me too :P
Jan 01 20:35:01 <rsalvaterra> Indeed you are, but I had to look it up, since I didn't know how you sign off. :)
Jan 01 20:36:47 <Borromini> yeah nick is different from real name and from mail i use for openwrt
Jan 01 20:38:56 <Hauke> it is a problem in wolfssl update
Jan 01 20:38:59 <Hauke> it is broken again
Jan 01 20:39:03 <Borromini> lovely
Jan 01 20:39:15 <aparcar[m]> what's the status of mbedtls + wpa3?
Jan 01 20:40:14 <rsalvaterra> aparcar[m]: Dear God, not another combinatorial explosion in the wireless APD menu…
Jan 01 20:40:28 <aparcar[m]> :D
Jan 01 20:40:42 <aparcar[m]> I thought mbedtls is slick and great
Jan 01 20:45:49 <mangix> wolfssl devs contributed the code upstream I believe
Jan 01 20:46:09 <mangix> mbedtls ones have not :)
Jan 01 20:47:15 <aparcar[m]> should we kindly ask them?
Jan 01 20:51:07 <mangix> wouldn't hurt I guess
Jan 01 20:51:28 <mangix> I think wolfssl contributed the code for business reasons
Jan 01 20:52:09 <mangix> thry also contributed code to OpenSSH but that was rejected over licensing
Jan 01 21:06:21 <Hauke> I creaetd a fix for the wolfssl build, LTO removed a function used by hostapd
Jan 01 21:06:37 <Hauke> in 4.5 this was marked as exported ABI
Jan 01 21:22:37 <Hauke> the problem found by build bot should be fixed now
Jan 01 21:48:28 <owrt-snap-builds> build #686 of mediatek/mt7622 is complete: Failure [failed pkgbuild]  Build details are at http://buildbot.openwrt.org/master/images/builders/mediatek%2Fmt7622/builds/686  blamelist: Alexandru Ardelean <ardeleanalex@gmail.com>, Rosen Penev <rosenp@gmail.com>, Rui Salvaterra <rsalvaterra@gmail.com>, David Bauer <mail@david-bauer.net>, Adrian Schmutzler
Jan 01 21:48:28 <owrt-snap-builds> <freifunk@adrianschmutzler.de>, Russell Senior <russell@personaltelco.net>, Sven Wegener <sven.wegener@stealer.net>, Old?ich Jedli?ka <oldium.pro@gmail.com>, Stijn Segers <foss@volatilesystems.org>, Eneas U de Queiroz <cotequeiroz@gmail.com>, Hauke Mehrtens <hauke@hauke-m.de>
Jan 01 23:18:56 <aparcar[m]> Hauke: do you have an opinion on the qemu patch?
Jan 01 23:19:00 <aparcar[m]> aka qemu removal?
Jan 02 00:26:04 <mangix> cool. another patch from https://github.com/openwrt/openwrt/pull/3004 merged
Jan 02 00:30:23 <aparcar[m]> mangix: do you have an idea how to implement the prereq thing in a better way?
Jan 02 00:31:53 <mangix> Applying: build/prereq: add qemu-img as optional host dep
Jan 02 00:31:54 <mangix> error: patch failed: include/prereq-build.mk:62
Jan 02 00:31:54 <mangix> error: include/prereq-build.mk: patch does not apply
Jan 02 00:31:56 <mangix> hmmmm
Jan 02 00:32:43 <aparcar[m]> w00t
Jan 02 00:32:51 <aparcar[m]> I'll rebase it
Jan 02 00:34:14 <mangix> I think my branch was out of date actually
Jan 02 00:34:36 <mangix> hmmm nvm
Jan 02 00:36:00 <aparcar[m]> yea it rebases withouth problems...
Jan 02 00:36:11 <Hauke> aparcar[m]: it is nice that OpenWrt directly creates images for VMware and Virtual box, I used them in the past, but nowadays I am using qemu
Jan 02 00:36:38 <aparcar[m]> I don't want to stop that
Jan 02 00:36:40 <mangix> Hauke: I use GNOME Boxes, which uses QEMU AFAIK
Jan 02 00:36:44 <aparcar[m]> the patch still allows the usage of it
Jan 02 00:36:53 <aparcar[m]> it just moves it to the building host
Jan 02 00:36:59 <aparcar[m]> I think our buildbots can handle that
Jan 02 00:37:24 <mangix> aparcar[m]: https://github.com/openwrt/openwrt/commit/cc9d5b5a488d3a43068f621a1ef184f06e43bede
Jan 02 00:37:46 <aparcar[m]> mangix: rebased and pushed
Jan 02 00:37:54 <Hauke> aparcar[m]: I am ok if this only works when I have the qemu tools installed locally: https://github.com/openwrt/openwrt/pull/3739
Jan 02 00:38:21 <Hauke> mangix: I would like to use musl 1.1.X for the next release, but switch to 1.2.X for the release after the next one
Jan 02 00:38:50 <mangix> aparcar[m]: works now, ty
Jan 02 00:39:12 <mangix> Hauke: right. I'm talking about the other commits
Jan 02 00:39:16 <Hauke> I assume that there will be multiple thinks breaking wehn we switch to musl 1.2.X
Jan 02 00:39:26 <aparcar[m]> Hauke: do you have an idea how to retrigger the prereq check?
Jan 02 00:39:28 <mangix> Hauke: nah. I've fixed most of them.
Jan 02 00:39:42 <mangix> the ones in that PR are the remaining ones.
Jan 02 00:40:04 <mangix> note that glibc already is doing 64-bit time_t on 32-ib
Jan 02 00:40:06 <mangix> *bit
Jan 02 00:40:09 <Hauke> mangix: the ppp maintainers at github would like to merge the patches all the linux distributions are currently shipping
Jan 02 00:40:15 <Hauke> you could try to sedn it upstream
Jan 02 00:40:36 <Hauke> aparcar[m]: I do not know how to retrigger rethis
Jan 02 00:41:23 <mangix> Hauke: the ppp patch is not applicable upstream. they fixed it differently: https://github.com/paulusmack/ppp/commit/75870d7b55e36af526a0786fff94912989c73fd1
Jan 02 00:41:33 <mangix> It's not wise to backport something like that
Jan 02 00:42:37 <mangix> btw the umdns commit also fixes https://downloads.openwrt.org/snapshots/faillogs/arc_archs/base/umdns/compile.txt
Jan 02 00:42:47 <mangix> the problem is the same between glibc and musl 1.2.x
Jan 02 00:44:15 <mangix> aparcar[m]: add my ACK if you need it. I don't see qemu depends when building for mips.
Jan 02 00:46:47 <Hauke> mangix: I will have a look at the umdns changes later
Jan 02 00:47:42 <mangix> Hauke: note the GCC10 one was rejected by ynezz. The relevant one is: https://patchwork.ozlabs.org/project/openwrt/patch/20201015061027.7614-1-rosenp@gmail.com/
Jan 02 00:55:37 <aparcar[m]> dango already mentioned that sizeof(time_t) == 8 is only always true for musl.
Jan 02 01:11:20 <mangix> aparcar[m]: ARC is a little...special :)
Jan 02 01:11:47 <mangix> that is to say, every new architecture ported to glibc is automatically using 64-bit time_t
Jan 02 01:13:02 <mangix> ARC support was merged to glibc last year I believe
Jan 02 01:20:31 <aparcar[m]> Okay
**** ENDING LOGGING AT Sat Jan 02 02:59:56 2021