**** BEGIN LOGGING AT Sat Mar 28 02:59:59 2020
Mar 28 03:00:33 <pkgadd> e.g., I know that the WRT1900AC v1 is the odd one out, ARMADA XP based (instead of 370), non-NEON capable, avoid - but I wouldn't have the slightest idea that it's called mamba, if I hadn't grep'ed the kernel for the various DTS names
Mar 28 03:09:05 <mangix> swalker: faad2 version tracking is messed up. Upstream moved.
Mar 28 03:09:44 * mangix contemplates backporting fdk-aac/faad changes as master tracks BUILD_PATENTED properly.
Mar 28 07:03:40 <aparcar> pkgadd: sorry I list you messages. Disconnected from IRC :(
Mar 28 07:03:46 <aparcar> And the log server does not seem to work
Mar 28 07:10:57 <aparcar> nevermind, works again
Mar 28 07:11:13 <aparcar> I
Mar 28 07:11:26 <aparcar> ynezz: do you have an opinion on that?
Mar 28 07:11:30 <aparcar> or jow
Mar 28 09:58:40 <f00b4r0> nbd: ping?
Mar 28 10:33:29 <f00b4r0> Can someone remind me what is the make command to trigger the unpacking _and_ patching of the target kernel?
Mar 28 10:33:55 <f00b4r0> it seems I might have found why partial erase size is broken on 4.19: the patches don't seem to be applied
Mar 28 10:39:11 <ldir> iirc make target/linux/prepare V=s
Mar 28 10:40:18 <f00b4r0> thanks!
Mar 28 10:42:42 <Mint-IRC> Is there an archive of previously (successfully) compiled packages for a given architecture of snapshots?
Mar 28 10:45:46 <Mint-IRC> Or otherwise where should I report an absolutely colossal (build-breaking) failure of package compiling on the snapshot targets? Seriously, I can't even do the default "make image", a requisite package for it is missing.
Mar 28 10:58:07 <ldir> what pre-req ?
Mar 28 11:23:39 <ldir> what does PKG_BUILD_PARALLEL:=0 do vs PKG_BUILD_PARALLEL:=1 ?
Mar 28 11:41:34 <Redfoxmoon> ldir, I'd imagine make -j1 versus make -j<number of cores + 1>
Mar 28 11:43:37 <ldir> MAKE_J:=$(if $(MAKE_JOBSERVER),$(MAKE_JOBSERVER) $(if $(filter 3.% 4.0 4.1,$(MAKE_VERSION)),-j))
Mar 28 11:43:45 <ldir> PKG_JOBS?=$(if $(PKG_BUILD_PARALLEL),$(MAKE_J),-j1)
Mar 28 11:49:01 <f00b4r0> ok so the patches are applied, they just don't seem to work
Mar 28 11:49:51 * f00b4r0 spots http://git.openwrt.org/9261e7447 from Hauke, wonders if that could be a clue
Mar 28 11:57:50 <f00b4r0> Redfoxmoon: j+1 is a common myth, the problem is it usually brings no benefit and is more likely to hurt performance :)
Mar 28 11:59:14 <Borromini> f00b4r0: what's recommended for optimal parallel compilation then?
Mar 28 11:59:22 <f00b4r0> nproc
Mar 28 11:59:41 <Borromini> ?
Mar 28 12:00:09 <f00b4r0> ncpus
Mar 28 12:00:26 <Borromini> so just the amount of threads you have
Mar 28 12:00:32 <f00b4r0> exactly
Mar 28 12:01:03 <f00b4r0> HT gives a minor but noticeable boost. Going beyond that, performance start to suffer. Especially if the machine is memory starved
Mar 28 12:01:26 <ldir> this is all very interesting BUT none of it answers my question
Mar 28 12:02:41 <Borromini> ldir: we can't always be answering *your* questions :P
Mar 28 12:03:07 <Redfoxmoon> ldir, some packages are broken when built with parallelism
Mar 28 12:03:12 <Redfoxmoon> i.e. race conditions
Mar 28 12:03:21 <ldir> yes, I get that.
Mar 28 12:03:35 <f00b4r0> the second part of the line seems to look for specific make versions
Mar 28 12:03:46 <ldir> so - PKG_BUILD_PARALLEL:=0
Mar 28 12:04:19 <f00b4r0> the final "-j" scares me a bit out of context
Mar 28 12:04:24 <f00b4r0> it could turn into a fork bomb
Mar 28 12:04:44 * ldir hands f00b4r0 a spoon 
Mar 28 12:04:49 <f00b4r0> ;)
Mar 28 12:05:56 <Borromini> :P
Mar 28 12:06:11 <ldir> so what does the '-j' do then, other than fork bomb
Mar 28 12:06:40 <f00b4r0> hugh
Mar 28 12:06:52 <f00b4r0> ok I'm now convince partial erase is completely broken on 4.19
Mar 28 12:06:56 * ldir reads man page
Mar 28 12:07:01 <f00b4r0> convinced*
Mar 28 12:07:19 <f00b4r0> the changes introduced by Hauke in the above patch indeed don't work, as he suspected
Mar 28 12:07:53 <f00b4r0> they completely break the operating logic of the code
Mar 28 12:08:49 <tmn505> f00b4r0: that's the outcome: https://bugs.openwrt.org/index.php?do=details&task_id=2428
Mar 28 12:10:52 <tmn505> ldir: PKG_BUILD_PARALLEL:=0 forces package to be built with -j1, even if You specified -j2 or more.
Mar 28 12:11:28 <ldir> and PKG_BUILD_PARALLEL:=1 ?
Mar 28 12:11:53 <tmn505> allows pcakage to be built in parallel
Mar 28 12:12:10 <f00b4r0> tmn505: I replied to the FS. It also breaks mikrotik devices on ath79 (which uses 4.19) in a way that could also lead to a brick
Mar 28 12:12:27 <ynezz> tmn505: ldir: I think, that it doesnt check value
Mar 28 12:12:38 <f00b4r0> tmn505: this code is a bit above my head, I'd very much like for nbd to take a look; it also seems that 4K_SECTOR_LIMIT also no longer works
Mar 28 12:12:53 <ynezz> PKG_JOBS?=$(if $(PKG_BUILD_PARALLEL),$(MAKE_J),-j1)
Mar 28 12:13:30 <ynezz> is PKG_BUILD_PARALLEL is set(0,1,y,whatever), then build in parallel or -j1
Mar 28 12:14:04 <ynezz> ah there is `ifeq ($(strip $(PKG_BUILD_PARALLEL)),0)` above
Mar 28 12:14:25 <ynezz> so 0=-j1, any other value -jY
Mar 28 12:14:36 <ynezz> sorry for the noise :)
Mar 28 12:15:28 <ldir> so you think PKG_BUILD_PARALLEL:=2 equals -j2 ?
Mar 28 12:15:35 <ynezz> no
Mar 28 12:15:53 * ldir is confused...again
Mar 28 12:15:56 <ynezz> it equals to MAKE_J
Mar 28 12:16:12 <ynezz> $(MAKE_J) = to whatever you've set with -jY
Mar 28 12:16:29 <tmn505> f00b4r0: also this: http://lists.infradead.org/pipermail/openwrt-devel/2020-March/022159.html
Mar 28 12:17:04 <ldir> Ahh, so 0 = disable parallel, !0 = build parallel with the passed -jY
Mar 28 12:17:14 <ynezz> yep
Mar 28 12:17:14 <tmn505> yep
Mar 28 12:17:44 <f00b4r0> tmn505: this seems to be a different problem
Mar 28 12:18:20 <ldir> ok, that leads to stupid question dept part 2 - why is 'PKG_BUILD_PARALLEL != 0' a thing?
Mar 28 12:18:49 <tmn505> f00this is above my head unfortunately, so im open for any code to test.
Mar 28 12:18:51 <ldir> if you just wanted to disable parallel building you set PKG_BUILD_PARALLEL=0.
Mar 28 12:19:01 <tmn505> f00b4r0: this is above my head unfortunately, so im open for any code to test.
Mar 28 12:19:51 <f00b4r0> tmn505: ack. nbd is the mtd guru here: he authored this code. I'll try to see if I can sort it out
Mar 28 12:19:55 <ldir> I'm assuming we build in parallel by default
Mar 28 12:20:07 <f00b4r0> but first I need to offload some of my patches to the mailing list, I have too much code in flight already ;P
Mar 28 12:23:04 <tmn505> ldir: if You specify jobs with -j then yes.
Mar 28 12:23:44 <ldir> tmn505: right, so we only need a disable option, we don't need an enable option
Mar 28 12:24:25 <ldir> I'm looking at an existing package definition and trying NOT to cargo cult all that I see :-)
Mar 28 12:25:36 <ldir> There's an awful lot of "PKG_BUILD_PARALLEL:=1" in the base repo alone that now just doesn't make sense
Mar 28 12:26:12 <ldir> Only 5 packages say PKG_BUILD_PARALLEL:=0
Mar 28 12:26:54 <tmn505> that might be a remnat of the past, I think jow would elaborate more on it.
Mar 28 12:28:21 <tmn505> and from fast glance I don't see any option in menuconfig to adjust that.
Mar 28 12:37:44 * ldir fires off macbot
Mar 28 12:50:18 <ynezz> tmn505: https://patchwork.ozlabs.org/patch/1257691/ (mvebu: cortexa9: correct cpu subtype)
Mar 28 12:50:27 <ynezz> BOARDNAME:=Marvell Armada 37x/38x/XP
Mar 28 12:50:36 <ynezz> "Armada 370  processors have only 16 double-precision registers."
Mar 28 12:51:04 <ynezz> what about 38x and XP? (dont have any mvebu so maybe stupid question)
Mar 28 13:18:29 <tmn505> ynezz: 38x definitely has 32 dp, and XP: https://www.marvell.com/content/dam/marvell/en/public-collateral/Fembedded-processors/marvell-embedded-processors-armada-xp-mv78230-hardware-specifications-2014-07.pdf, page 7 at the bottom.
Mar 28 13:19:14 <tmn505> wrong link: https://www.marvell.com/content/dam/marvell/en/public-collateral/embedded-processors/marvell-embedded-processors-armada-xp-mv78230-hardware-specifications-2014-07.pdf
Mar 28 13:20:49 <tmn505> ynezz: but it seems like XP is also dp32, since no one reported breackage on wrt1900 v1.
Mar 28 13:25:58 <ynezz> tmn505: so arent going the 38x/XP users complain now?
Mar 28 13:30:10 <ynezz> ah, now I get it, it was previously implicit to vfpv3-d16, but now its to vfpv3-d32, so it has to be set to vfpv3-d16 explicitly, right?
Mar 28 13:30:26 <tmn505> yes
Mar 28 13:53:53 <owrt-snap-builds> build #255 of ath79/tiny is complete: Failure [failed pkgbuild]  Build details are at http://buildbot.openwrt.org/master/images/builders/ath79%2Ftiny/builds/255  blamelist: John Crispin <john@phrozen.org>, Leon M. George <leon@georgemail.eu>, Petr ?tetiar <ynezz@true.cz>, Nick Bowler <nbowler@draconx.ca>, Thibaut VAR?NE <hacks@slashdirt.org>, Paul
Mar 28 13:53:53 <owrt-snap-builds> Spooren <mail@aparcar.org>, Alberto Bursi <bobafetthotmail@gmail.com>, Eneas U de Queiroz <cotequeiroz@gmail.com>, Baptiste Jonglez <git@bitsofnetworks.org>, Aleksander Jan Bajkowski <A.Bajkowski@stud.elka.pw.edu.pl>
Mar 28 14:21:03 * f00b4r0 dumps a few KB of code to the m-l and runs ;D
Mar 28 14:26:46 <swalker> mangix: thanks, the watch files have been updated
Mar 28 17:33:41 <owrt-snap-builds> build #348 of ath25/generic is complete: Failure [failed pkgbuild]  Build details are at http://buildbot.openwrt.org/master/images/builders/ath25%2Fgeneric/builds/348  blamelist: Leon M. George <leon@georgemail.eu>, Petr ?tetiar <ynezz@true.cz>, Nick Bowler <nbowler@draconx.ca>, Paul Spooren <mail@aparcar.org>, Baptiste Jonglez <git@bitsofnetworks.org>,
Mar 28 17:33:41 <owrt-snap-builds> Aleksander Jan Bajkowski <A.Bajkowski@stud.elka.pw.edu.pl>, Eneas U de Queiroz <cotequeiroz@gmail.com>
Mar 28 18:28:19 <Tapper> Which make clean command will clean up the most and let me keep my ./scripts/env files?
Mar 28 18:44:38 <olmari> Tapper: I'd say "make dirclean"
Mar 28 18:45:06 <reiffert> Hi. I'm probably just too dumb to google. My / is full and I can't find a way what to do next. Help. Thanks.
Mar 28 19:04:05 <Tapper> Thanks olmari I will give it a go now
Mar 28 19:34:35 <ldir> tmp/.config-package.in:135008:error: recursive dependency detected!
Mar 28 19:34:36 <ldir> For a resolution refer to Documentation/kbuild/kconfig-language.txt
Mar 28 19:34:36 <ldir> subsection "Kconfig recursive dependency limitations"
Mar 28 19:34:36 <ldir> tmp/.config-package.in:135008:	symbol PACKAGE_adblock is selected by PACKAGE_luci-app-adblock
Mar 28 19:34:36 <ldir> For a resolution refer to Documentation/kbuild/kconfig-language.txt
Mar 28 19:34:37 <ldir> subsection "Kconfig recursive dependency limitations"
Mar 28 19:34:37 <ldir> tmp/.config-package.in:73342:	symbol PACKAGE_luci-app-adblock depends on PACKAGE_wget
Mar 28 19:34:37 <ldir> For a resolution refer to Documentation/kbuild/kconfig-language.txt
Mar 28 19:34:37 <ldir> subsection "Kconfig recursive dependency limitations"
Mar 28 19:34:38 <ldir> tmp/.config-package.in:109545:	symbol PACKAGE_wget is selected by PACKAGE_acme
Mar 28 19:34:39 <ldir> For a resolution refer to Documentation/kbuild/kconfig-language.txt
Mar 28 19:34:41 <ldir> subsection "Kconfig recursive dependency limitations"
Mar 28 19:34:43 <ldir> tmp/.config-package.in:134978:	symbol PACKAGE_acme is selected by PACKAGE_luci-app-acme
Mar 28 19:34:45 <ldir> For a resolution refer to Documentation/kbuild/kconfig-language.txt
Mar 28 19:34:47 <ldir> subsection "Kconfig recursive dependency limitations"
Mar 28 19:34:49 <ldir> tmp/.config-package.in:73326:	symbol PACKAGE_luci-app-acme depends on PACKAGE_lua
Mar 28 19:34:51 <ldir> For a resolution refer to Documentation/kbuild/kconfig-language.txt
Mar 28 19:34:53 <ldir> subsection "Kconfig recursive dependency limitations"
Mar 28 19:34:55 <ldir> tmp/.config-package.in:21901:	symbol PACKAGE_lua is selected by PACKAGE_lua-cjson
Mar 28 19:34:57 <ldir> For a resolution refer to Documentation/kbuild/kconfig-language.txt
Mar 28 19:34:59 <ldir> subsection "Kconfig recursive dependency limitations"
Mar 28 19:35:01 <ldir> tmp/.config-package.in:21947:	symbol PACKAGE_lua-cjson is selected by PACKAGE_prometheus-node-exporter-lua-bmx7
Mar 28 19:35:04 <ldir> For a resolution refer to Documentation/kbuild/kconfig-language.txt
Mar 28 19:35:06 <ldir> subsection "Kconfig recursive dependency limitations"
Mar 28 19:35:08 <ldir> tmp/.config-package.in:154933:	symbol PACKAGE_prometheus-node-exporter-lua-bmx7 depends on PACKAGE_bmx7
Mar 28 19:35:10 <ldir> For a resolution refer to Documentation/kbuild/kconfig-language.txt
Mar 28 19:35:12 <ldir> subsection "Kconfig recursive dependency limitations"
Mar 28 19:35:14 <ldir> tmp/.config-package.in:116096:	symbol PACKAGE_bmx7 is selected by PACKAGE_bmx7-dnsupdate
Mar 28 19:35:16 <ldir> For a resolution refer to Documentation/kbuild/kconfig-language.txt
Mar 28 19:35:18 <ldir> subsection "Kconfig recursive dependency limitations"
Mar 28 19:35:20 <ldir> tmp/.config-package.in:146739:	symbol PACKAGE_bmx7-dnsupdate depends on PACKAGE_jshn
Mar 28 19:35:22 <ldir> For a resolution refer to Documentation/kbuild/kconfig-language.txt
Mar 28 19:35:24 <ldir> subsection "Kconfig recursive dependency limitations"
Mar 28 19:35:26 <ldir> tmp/.config-package.in:150980:	symbol PACKAGE_jshn is selected by PACKAGE_adblock
Mar 28 20:46:18 <owrt-snap-builds> build #316 of ath79/nand is complete: Failure [failed pkgbuild]  Build details are at http://buildbot.openwrt.org/master/images/builders/ath79%2Fnand/builds/316  blamelist: Leon M. George <leon@georgemail.eu>, Petr ?tetiar <ynezz@true.cz>, Nick Bowler <nbowler@draconx.ca>, Paul Spooren <mail@aparcar.org>, Baptiste Jonglez <git@bitsofnetworks.org>,
Mar 28 20:46:18 <owrt-snap-builds> Aleksander Jan Bajkowski <A.Bajkowski@stud.elka.pw.edu.pl>, Eneas U de Queiroz <cotequeiroz@gmail.com>
Mar 28 21:14:45 <owrt-snap-builds> build #301 of mpc85xx/generic is complete: Failure [failed pkgbuild]  Build details are at http://buildbot.openwrt.org/master/images/builders/mpc85xx%2Fgeneric/builds/301  blamelist: Leon M. George <leon@georgemail.eu>, Petr ?tetiar <ynezz@true.cz>, Nick Bowler <nbowler@draconx.ca>, Paul Spooren <mail@aparcar.org>, Baptiste Jonglez
Mar 28 21:14:45 <owrt-snap-builds> <git@bitsofnetworks.org>, Aleksander Jan Bajkowski <A.Bajkowski@stud.elka.pw.edu.pl>, Eneas U de Queiroz <cotequeiroz@gmail.com>
Mar 28 21:42:57 <owrt-snap-builds> build #294 of apm821xx/nand is complete: Failure [failed pkgbuild]  Build details are at http://buildbot.openwrt.org/master/images/builders/apm821xx%2Fnand/builds/294  blamelist: Leon M. George <leon@georgemail.eu>, Petr ?tetiar <ynezz@true.cz>, Nick Bowler <nbowler@draconx.ca>, Paul Spooren <mail@aparcar.org>, Baptiste Jonglez <git@bitsofnetworks.org>,
Mar 28 21:42:57 <owrt-snap-builds> Aleksander Jan Bajkowski <A.Bajkowski@stud.elka.pw.edu.pl>, Eneas U de Queiroz <cotequeiroz@gmail.com>
Mar 28 21:52:49 <LoRez> trying to spin this AP up with something close, but I'm running into boot issues.  https://pastebin.com/ibBpue9Q  anybody got some pointers?
Mar 28 22:10:16 <owrt-snap-builds> build #314 of ramips/rt3883 is complete: Failure [failed pkgbuild]  Build details are at http://buildbot.openwrt.org/master/images/builders/ramips%2Frt3883/builds/314  blamelist: Leon M. George <leon@georgemail.eu>, Petr ?tetiar <ynezz@true.cz>, Nick Bowler <nbowler@draconx.ca>, Paul Spooren <mail@aparcar.org>, Baptiste Jonglez <git@bitsofnetworks.org>,
Mar 28 22:10:16 <owrt-snap-builds> Aleksander Jan Bajkowski <A.Bajkowski@stud.elka.pw.edu.pl>, Eneas U de Queiroz <cotequeiroz@gmail.com>
Mar 29 01:06:53 <mangix> ldir: will be fixed soon
**** ENDING LOGGING AT Sun Mar 29 03:01:48 2020