nlowe has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
gladiac is now known as Guest6477
gladiac has joined #openwrt-devel
nlowe has joined #openwrt-devel
Guest6477 has quit [Ping timeout: 480 seconds]
nlowe has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
nlowe has joined #openwrt-devel
bluew has joined #openwrt-devel
<victhor>
Tusker: thought I'd let you know - someone got in touch with me about porting OpenWrt to a zf7762 about 2 weeks ago. Unfortunately he hasn't followed up
linusw_ has joined #openwrt-devel
<victhor>
meanwhile, due to "unforeseen circumstances", the R500 GPL dump I asked for has been pushed to December
victhor has quit [Ping timeout: 480 seconds]
nlowe has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<dvn>
i'm trying to understand - in that makefile it doesn't specify a src dir, so it's a default to use the src dir located next to the Makefile?
<dvn>
i.e. $(PKG_BUILD_DIR) is the current dir?
nlowe has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
nlowe has joined #openwrt-devel
Alext has joined #openwrt-devel
<Alext>
Hello, Everyone!
<Alext>
Hello, Everyone! Do someone know where can I found last MTK SDK ?
<jow>
dvn: yes, the default behaviour is to copy src/ into a temporary build scratch directory
<jow>
dvn: which you can refer to using $(PKG_BUILD_DIR(
<dvn>
jow: that's what i inferred so just went with that. thanks for confirming.
Tusker has quit [Quit: Time wasted on IRC: 10 hours 40 minutes 1 second]
<karlp>
mangix: are there still meant to be any gettext package issues? I've just updated openwrt master, ./scripts/feeds clean and reinstall everything, and I get a bunch of "blah has a dep on gettxt, which doesn't exist" still?
<rsalvaterra>
karlp: I don't know if there's an ABI breakage, but I had to make dirclean after the gettext removal. I was getting a weird build failure.
<russell-->
i don't know what did it for me, but possibly some combination of scripts/feeds uninstall -a ; rm -rf tmp ; scripts/feeds install -a ... i am not seeing the dependency complaints anymore
jlsalvador has quit [Quit: jlsalvador]
<karlp>
ok, thanks.
pmelange has joined #openwrt-devel
pmelange has left #openwrt-devel [#openwrt-devel]
guidosarducci has quit [Remote host closed the connection]
guidosarducci has joined #openwrt-devel
chwba has quit [Ping timeout: 480 seconds]
nlowe has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
nitroshift has quit [Quit: Gone that way --->]
<mrkiko>
anyone trying to build ipq4x on current master? I am getting an errore like this: Patch failed! Please fix /mnt/hdd/sdata/openwrt/target/linux/ipq40xx/patches-5.10/901-arm-boot-add-dts-files.patch!
<mrkiko>
*ipq40xx*
kenny has joined #openwrt-devel
<mrkiko>
no, my problem due to bad modification of patch by me
gladiac is now known as Guest6521
gladiac has joined #openwrt-devel
Tapper has quit [Ping timeout: 480 seconds]
Guest6521 has quit [Ping timeout: 480 seconds]
Tapper has joined #openwrt-devel
dedeckeh has quit [Remote host closed the connection]
mrkiko has quit [Quit: leaving]
danitool has quit [Quit: Cubum autem in duos cubos, aut quadratoquadratum in duos quadratoquadratos]
gladiac is now known as Guest6527
gladiac has joined #openwrt-devel
Guest6527 has quit [Ping timeout: 480 seconds]
decke has quit [Quit: Leaving.]
Slimey_ has joined #openwrt-devel
Slimey has quit [Read error: Connection reset by peer]
<hurricos>
I am apparently mistaking AR9160 and AR9170 regarding ath9k compatibility. Looks like AR9170 was USB (meaning it almost certainly needed firmware)
<hurricos>
ah, yes. OK. AR9160 is draft-n but still PCI, so no worries there.
<hurricos>
Well, some worries maybe :^)
<victhor>
hurricos: Yes that one. Nothing extraordinary, shouldn't be that difficult to port if it has the RAM
<PaulFertser>
ath9k_htc firmware is free software btw
nlowe has joined #openwrt-devel
<hurricos>
PaulFertser: This is carl9170
<hurricos>
different fw :P
<hurricos>
or, I was talking about it. But.
valku has joined #openwrt-devel
<hurricos>
> Do we have any boards that have NAND but whose u-boot doesn't support ubi
Alext has quit [Ping timeout: 480 seconds]
<hurricos>
to answer my own question, it appears that package/base-files/files/lib/upgrade/nand.sh's nand_do_upgrade accepts nand_upgrade_tar to do what I want, mostly.
<hurricos>
Is this IRC chat stil monitored and the logs published? Hopefully so, it'd be nice for folks to be able to find this stuff in search engines.
<hurricos>
nand_upgrade_tar checks if there is a 'kernel' mtd partition, and writes to that if so. But it does still allow for writing the rootfs to ubi
<hurricos>
Is Pawel Dembicki on the IRC at all?
nlowe has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
nlowe has joined #openwrt-devel
nlowe has quit []
nlowe has joined #openwrt-devel
hanetzer has quit [Quit: WeeChat 3.3]
hanetzer has joined #openwrt-devel
nlowe has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<PaulFertser>
hurricos: I thought carl9170 isn't similar to ath9k but I do not know much about it.
<slh>
PaulFertser: it's a normal PCI ath9k card, internally connected to a tiny SOC doing all the USB stuff - the (FOSS/ GPLv2) firmware is only for this SOC, a tiny OS with 'ath9k' drivers for the wireless chipset. the drivers are very different, of course
<slh>
biggest problem being the very limited on-device RAM for this SOC, limiting the abilities of the wireless side (not enough memory for more than a few connected STAs)
danitool has joined #openwrt-devel
<PaulFertser>
slh: so same deal as with ath9k_htc essentially, just a different SoC.
<slh>
yes
<slh>
same as ar5523 as well
<slh>
the only difference to ath9k_htc is that the later is slightly more integrated into the rest of the ath9k driver suite
<slh>
while carl9170 and ar5523 have been developed pretty much standalone
Tapper has quit [Ping timeout: 480 seconds]
pmelange has joined #openwrt-devel
Borromini has quit [Quit: leaving]
arifred has joined #openwrt-devel
arifred has quit []
shibboleth has joined #openwrt-devel
<hurricos>
If there is no rootfs partition and no rootfs ubi volume, and no command-line option for rootfs mounting, how does OpenWrt know what to mount as rootfs?
goliath has quit [Quit: SIGSEGV]
<PaulFertser>
hurricos: it tries mounting partition named "rootfs", and e.g. one of the mtd splitters can create one after splitting "kernel" from "firmware".
<hurricos>
PaulFertser: mtd splitters?
<hurricos>
I have seen similarly named config options
<Habbie>
while we're on that, can somebody explain to me why my ubnt erl has /overlay on /loop?
<Habbie>
where is that loop backend coming from?
<hurricos>
Habbie? squashfs, maybe? You can get losetup installed and inspect it that way.
<Habbie>
in /proc it says the file is /sda2
<hurricos>
Yes, it's squashfs
<Habbie>
squashfs is rw?
<hurricos>
A portion of yours is
<Habbie>
ah right
<Habbie>
i suspected something like that
<hurricos>
x86 does the same: you install to a USB. ERL is installed by modifying the USB, if I remember correctly
<Habbie>
correct, i stick a kernel on partition one and dd a thing to partition two
<hurricos>
I would use sysupgrade-tar to build the sysupgrade, but I need the dtb to be the 128K appended at the beginning of the firmware partition
<hurricos>
I have NAND and want to minimize non-UBI area
<hurricos>
I should double-check whether appended dts works
Tusker has joined #openwrt-devel
<Tusker>
hurricos: nice find with the PoE settings
nlowe has joined #openwrt-devel
rua has quit [Ping timeout: 480 seconds]
<hurricos>
Yes, but still struggling.
<hurricos>
FIT images re-confirmed non-working
rua has joined #openwrt-devel
<Tusker>
that's what I thought. dtb prepend and dtb append also not work? (ie, not FIT, but just append)
<hurricos>
Neither nor. Well, I haven't tried dtb preprend.
<hurricos>
I was going to use that plus a pad to just hard-code, say, 256k free for the dtb
<hurricos>
but now I have hit a hard wall: if I do that, I cannot use sysupgrade-tar
<Tusker>
yeah, you'd have to point the kernel manually like that, correct
<hurricos>
(because you cannot layer steps in the p1020.mk)
<Tusker>
just suck it up and use my approach :)
<hurricos>
by which I mean, when you are building sysupgrade.bin you cannot specify how to construct the input to the tar
<hurricos>
other than just giving the standard ones
<hurricos>
My goal is to do this without touching any other boards
<hurricos>
I just need to figure out how to use a .mk definition to describe building an interim file
<hurricos>
That, or I need to figure out how to apply Aerohive's patches and just add FIT images to U-boot
<hurricos>
speaking of, I have no idea how Enterasys WS-AP3710 has a rootfs
<hurricos>
they don't even use the config options to cause the mtd splitter to activate
<Tusker>
maybe it's on the kernel command line, hardcoded
<hurricos>
No
<Tusker>
not really sure then
<Tusker>
maybe it doesn't work, and the target is not used
<hurricos>
Oh!
<hurricos>
No, it is in config-default
<hurricos>
So really, I could sacrifice this all and just accept an mtd split for rootfs
<hurricos>
it weakens the NAND but I'm over it tbh
<Tusker>
just be pragmatic about it
<hurricos>
I think I will smush together the Aerohives, though
<hurricos>
I do need to be able to write dtb | pad | kernel to a firmware partition
<Tusker>
yeah, they seem close enough
<hurricos>
they're basically identical, in terms of the code they invoke ...
<hurricos>
and both will need the hard-coded boot code patched
<hurricos>
and both have a command (ggtest) to check that
<Tusker>
will you build the ggtest binary to be bundledd in with it ?
<hurricos>
Tusker: No need, it comes with u-boot
<hurricos>
it's only for the three Golden-Gate like boards that we care about: Millau, Goldengate and BR200
<Tusker>
ah, uboot command, not linux binary
<hurricos>
Yep
<hurricos>
Millau is AP370 / AP390 (officially AP3X0); Goldengate is AP330 / AP350; and BR200 is some LTE-capable branch office routering board thing
<hurricos>
which should be equally hackable and ported using the same code, tbh, the only problem is that they don't really make mPCIe LTE cards these days
<hurricos>
Oh, looks like the BR-200 has many variants actually
<hurricos>
some LTE some not
<hurricos>
The other thing that weirds me out is that apparently AP370 has a Marvell PHY?
pmelange has left #openwrt-devel [#openwrt-devel]
<Tusker>
oh, I didn't even notice
<hurricos>
If I'm right about Millau being AP3x0, then according to u-boot, it does, but I don't have good enough photos to confirm.
<hurricos>
I could tear one down. Do you already have photos, Tusker
<Tusker>
i have a bricked one with an exposed board
<hurricos>
OK, gonna go add the other todos to my issue tracker
<hurricos>
Tusker: Is the ubiformatting step something that is taken care of by something you changed, or is it `nand.sh` standard operating procedure?