philipp64 is now known as Guest981
philipp64 has joined #openwrt-devel
Guest981 has quit [Ping timeout: 480 seconds]
philipp64 has quit [Quit: philipp64]
philipp64 has joined #openwrt-devel
philipp64 has quit [Quit: philipp64]
philipp64 has joined #openwrt-devel
philipp64 has quit []
minimal has quit [Quit: Leaving]
Emantor has quit [Quit: ZNC - http://znc.in]
Emantor has joined #openwrt-devel
Daanct12 has joined #openwrt-devel
Daanct12 has quit [Ping timeout: 480 seconds]
philipp64 has joined #openwrt-devel
philipp64 has quit []
tSYS has quit [Quit: *squeak*]
tSYS has joined #openwrt-devel
swalker has quit [Read error: No route to host]
swalker has joined #openwrt-devel
Daanct12 has joined #openwrt-devel
Mangix has joined #openwrt-devel
philipp64 has joined #openwrt-devel
cyrozap has quit [Quit: ZNC 1.8.2+deb3.1 - https://znc.in]
cyrozap has joined #openwrt-devel
cyrozap has quit []
rua has quit [Quit: Leaving.]
cyrozap has joined #openwrt-devel
<mrkiko> nbd: I don't seem to find that warning on an AX build (RT3200 / E8450 / mt7622 device)
cyrozap has quit [Ping timeout: 480 seconds]
<dhewg> no such error on my 7320. Just avoid asu and build your own image
<mrkiko> dhewg: I do :D
<mrkiko> dhewg: but even so, wondering what asu may have to do with this
<slh> asu relies on binary package snapshots, which don't always need to be in a consistent state, due to teh two-stage buildbot nature and shared packages for a single ISA. and it's only synced x-times a day anyways, adding more delays to it
tomn has quit [Remote host closed the connection]
<mrkiko> slh: thanks for the explanation, didn't know how things work. Thanks.
<dhewg> Yeah, and that in return can yield asu images where the kernel+modules are newer than the package builds. Which happened recently, where mac80211 was never than the related changes from the hostapd package
<dhewg> but that should be settled by now
tomn has joined #openwrt-devel
philipp64 has quit [Quit: philipp64]
philipp64 has joined #openwrt-devel
cyrozap has joined #openwrt-devel
Danct12 has quit [Remote host closed the connection]
cyrozap has quit [Quit: ZNC 1.8.2+deb3.1 - https://znc.in]
_lore_ has quit [Ping timeout: 480 seconds]
cyrozap has joined #openwrt-devel
cyrozap has quit [Quit: ZNC 1.8.2+deb3.1 - https://znc.in]
cyrozap has joined #openwrt-devel
Danct12 has joined #openwrt-devel
Daanct12 has quit [Read error: Connection reset by peer]
<mrkiko> I am under the impression the problem may lie in mac80211.sh:drv_mac80211_setup(), from a cursory look.
<mrkiko> or maybe the corresponding teardown... but ok, I would need to look closer then :D
rua has joined #openwrt-devel
Borromini has joined #openwrt-devel
Borromini has quit [Quit: Lost terminal]
robimarko has joined #openwrt-devel
hitech95 has joined #openwrt-devel
<hitech95> is there any way to detect an SFP that get inserted from userspace? (no link status etc just inserted)
<PaulFertser> hitech95: you get a message in kernel log for that?
<hitech95> yea the SFP module says that a module is inserted
<PaulFertser> And you want a hotplug event to signal it? Hm.
<hitech95> yes something like that, I need to reboot the device if a sfp is inserted to switch the dts in uboot
<hitech95> yes something like that, I need to reboot the device if a sfp is inserted to switch the dtso in uboot
<PaulFertser> q&d hack would be to run a shell script which would poll ethtool ;)
<hitech95> dammit, I hate switching chat types.. lol this is IRC me dumb!
<hitech95> PaulFertser, the issue is that the SFP is not conncted to a MAC unless I reboot
<hitech95> aMAC is shared between two ports :( I'm waiting for upstream on phy namespaces patch to develop something better
<hauke> hitech95: maybe you can also look at the module detect GPIO
<hauke> what SoC are you working on?
<hitech95> hauke, yea that was the other solution but i'm still using polling which is not the best.
<hitech95> This is also something to sense dynamic phy sense.
<hitech95> hotplud.d cannot sense gpio state changes right? I have to poll the pins
rua has quit [Quit: Leaving.]
<hitech95> btw this is the sort of overlay I have to switch to in uboot: https://pastebin.com/cnXvXPUb
<PaulFertser> hitech95: you can use "gpiomon" to not poll.
<hitech95> PaulFertser, so is keep the script busy until the event is triggered... then I can put s simple reboot and be done. uboot detect the gpio pin and load the correct overlay.
<PaulFertser> Yes, gpiomon can ask the kernel for interrupt so it's not consuming any CPU time.
<hitech95> got is, but now the issue is how to include the scripts only in tthis specific target :D
<hitech95> any suggestion about: "FATAL ERROR: Unable to parse input tree" on my dtso file?
<hitech95> this is the file: https://pastebin.com/02tDzvMp
<hitech95> Error: ../dts/mt7986a-zyxel-ex5601-t0-sfp.dtso:19.26-27 syntax error
<hitech95> I suppose the error is in line 19 he doesnt like the &i2c0 ref?
<robimarko> Usually that means it cannot find it
<hitech95> so the issue is the base device dts as input?
Danct12 has quit [Quit: WeeChat 4.0.4]
<hitech95> robimarko, this is the full arror and the file that is passed to dtc:https://pastebin.com/xkYSa0xL
Danct12 has joined #openwrt-devel
robimarko has quit [Remote host closed the connection]
robimarko has joined #openwrt-devel
Danct12 has quit [Quit: WeeChat 4.0.4]
Danct12 has joined #openwrt-devel
<hitech95> ok there was a double ", I didnt see it! dumb me!
Danct12 has quit [Quit: WeeChat 4.0.4]
<hitech95> can someone link me something to understand how the bootconf parameter of uboot works? it seems to be related to FIT images. but I cannot find how the configuration is created
<hitech95> for example the bpi-r3 has different confs, but I cannot find the definition
rua has joined #openwrt-devel
rua has quit [Remote host closed the connection]
valku has joined #openwrt-devel
rua has joined #openwrt-devel
<hitech95> [ 0.010137] gpio-421 (sfp-gmac1-mux): hogged as output/low
<hitech95> but pin is input according to the pinmux :(
<hitech95> gpio-421 ( |sfp-gmac1-mux ) in lo
Danct12 has joined #openwrt-devel
rua has quit [Quit: Leaving.]
valku has quit [Quit: valku]
rua has joined #openwrt-devel
<hitech95> WTF my sfp module always have the tx fault pin asserted!
floof58 has quit [Ping timeout: 480 seconds]
<Ansuel> nbd ping?
floof58 has joined #openwrt-devel
<nbd> Ansuel: pong
<Ansuel> fighting with a circular dependency... we recently added luajit2 and that have PROVIDES luajit (same lib they can't coexist so also CONFLICTS)... every package that depends from luajit now emit a circular dependency warning. luajit and luajit2 have the same dependency and they are just some check on the kind of arch...
<Ansuel> Wonder if you wan help or have a clue? I think this is a very corner case since by removing the DEPENDS from luajit and luajit2 the circular dep are removed...
<Ansuel> here the confi-package.in of one of the affected packages
<Ansuel> also notice that by reverting https://github.com/openwrt/openwrt/commit/1fd50531cac9c41334d8f57e2dbc1f50c3572445 the circular dep is solved
<nbd> hm, no idea how to resolve this properly
<nbd> one possible (but ugly) workaround might be to copy the arch dependency as-is to knot-resolver
<nbd> or maybe not copy it, but make a distinct symbol like HAS_LUAJIT_ARCH
<nbd> that both luajit packages depend on
<nbd> and pull that in as an explicit dependency for the other packages
<nbd> not pretty, but should cause package-metadata.pl to omit the conditional deps
<Ansuel> where HAS_LUAJIT_ARCH should be placed? I assume one generic .config
<nbd> i would suggest merging luajit and luajit2 into a single source package first
<nbd> (with build variants and variant dependent source)
<nbd> you can then place the symbol in Package/luajit/config
<Ansuel> mhhhh i wanted to skip bloating the luajit package. do you have by chance an example for the variant dependent source?
<nbd> is the host package of luajit compatible with building luajit2 for the target?
<nbd> if it isn't, then merging unfortunately isn't an option
<nbd> here's how the variant thing works:
<nbd> you define both luajit and luajit2 packages in the same makefile
<nbd> luajit gets VARIANT:=luajit, luajit2 gets VARIANT:=luajit2
<nbd> then you can do ifeq ($(BUILD_VARIANT),luajit2) [set source,version,etc. variables for luajit2] else [same for luajit] endif
<Ansuel> i assume this can't be done for host
<nbd> i don't think we have host build variants support
<Ansuel> well I can only try... in theory luajit2 is just luajit + openresty patch
<nbd> why do we need both?
<Ansuel> openresty mafia that specifically check the oue of their lib for the nginx lua module
<Ansuel> in theory luajit2 should be always superiori to luajit
<Ansuel> so luajit2 is luajit + mod + fixes and they try to keep the thing synced with luajit changes
<nbd> so maybe just always use luajit2 for the host and use build variants for the target
<Ansuel> soo i think merge it is... didn't want to follow this path as it will also be hell with patching but ok seems to be the most ""clean"" way
<nbd> just make a patches2 directory
<nbd> and override the patch dir for luajit2
<nbd> patches don't need to be shared
<Ansuel> thanks for the hint hope this will solve the circular dep
<nbd> you're welcome
<Ansuel> PKG_NAME can't be conditional right?
<hitech95> is there a way to provide the overlays to load when using uboot bootm and the intramfs image?
floof58 has quit [Ping timeout: 480 seconds]
<nbd> Ansuel: it can be
<Ansuel> nbd can you check ? https://pastebin.com/HtWCLs5F https://pastebin.com/KR9j2B3k doesn't seem to work but probably i'm messing the config part ?
floof58 has joined #openwrt-devel
<nbd> Ansuel: drop this line: Package/luajit2/config = $(Package/luajit/config)
<nbd> the config symbol should be emitted only once
<nbd> also, variables that are the same for both variants should be moved outside of the ifeq
<nbd> you can drop the Build/Patch define
<nbd> just set PATCH_DIR based on the build vairant
<nbd> after the includes
<nbd> sorry, i meant before the includes
<Ansuel> same circular dep... this is the confi-package.in part for the luajit https://pastebin.com/sBqUHpZH
<nbd> did you add @HAS_LUAJIT_ARCH to the dependencies of packages depending on luajit?
<Ansuel> no! let me test
<Ansuel> well very good!
<Ansuel> that was the missing piece
cmonroe has quit [Ping timeout: 480 seconds]
<Ansuel> nbd can confirm also the PATCH_DIR thing... very useful! all fixed... preparing a pr... if you want to give a last check I will post the link here shortly
<ldir> nbd: Have you tried building kernel 6.1 under macos yet? Extract from trying to build x86 arch using TEST_KERNEL_VERSION https://pastebin.com/pySQuVEc
<nbd> ldir: it's been a while since last time i tried
<nbd> should probably do so again
<ldir> If you get bored :-D
<nbd> that doesn't work, because i never do :)
<nbd> my last episode of boredom was years ago
<ldir> That's why I put a smiley at the end, I know you NEVER get bored :-)
<Ansuel> nbd feel free to add more comments... https://github.com/openwrt/packages/pull/22195
<Ansuel> time to go as i have to see oppenheimer at 19:30
<ldir> Ansuel: take your sunglasses. And a lead suit.
<Ansuel> ahahahahahah
<hitech95> who is the uboot guru here? I really have to find out how uboot envs are used to define the boot process.
<hitech95> I'm reading the docs but so far no info
<linusw> hitech95: when compiling U-Boot the boot command is defined in the Kconfig variable CONFIG_BOOTCOMMAND if CONFIG_USE_BOOTCOMMAND=y
<hitech95> linusw, nop its not specified. for reference this is the patch for my HW but the author said it still has issues. se he asked for help: https://github.com/hitech95/openwrt/commit/b8a7ddc31b93409212ba993e34829289062c1c13#diff-add92575f064a0e29112314438eaa879224f037194c1c6a534e6ee1f808d9c83
<hitech95> But other than fixing the dts errors I'm trying to find a way to intercept the boot process and then do some stuff (gpio read) before booting linux to load correct dts
<hitech95> and its overlay
<linusw> this board sets bootcmd=if pstore check ; then run boot_recovery ; else run boot_ubi ; fi
gladiac has joined #openwrt-devel
<hitech95> ok so the whole boot process is defined via envvars
<linusw> yes
<linusw> you can edit bootcmd to change it
<hitech95> ok I was trying to figure out _init_env or _firstboot but on the docs I've found nothing with that match.
<linusw> so maybe you want to alter the env var boot_ubi to do something more in this case
<hitech95> I've started to figure out the bootmenu stuff
<linusw> you got this :)
<hitech95> is there something that is triggered when autobot is stopped?
<KGB-1> https://tests.reproducible-builds.org/openwrt/openwrt_x86.html has been updated. (100.0% images and 99.9% packages reproducible in our current test framework.)
philipp64 is now known as Guest1061
philipp64 has joined #openwrt-devel
Guest1061 has quit [Ping timeout: 480 seconds]
floof58 has quit [Ping timeout: 480 seconds]
floof58 has joined #openwrt-devel
<nbd> Ansuel: LGTM
rmilecki has quit [Quit: Konversation terminated!]
rmilecki has joined #openwrt-devel
goliath has joined #openwrt-devel
cmonroe has joined #openwrt-devel
goliath has quit [Quit: SIGSEGV]
robimarko has quit [Remote host closed the connection]