<mangix>
same size as typical routers funny enough
torv has joined #openwrt-devel
rua has quit [Quit: Leaving.]
<mrnuke>
svanheule: I need some help with this tplink switch -- I think it's a RTL8380M. The bootloader is stuck with the "BOOTUTILS" u-boot. Can't really do tftp/bootm. I got the GPL sources from the vendor (had to ask). I try to build u-boot from sources, flash it, but the device is stuck in a bootloop. Is there some form os secureboot on this device, or some way to tell why u-boot is crashing and fix it?
torv has quit [Remote host closed the connection]
<Ansuel>
lol....
<Ansuel>
the attention span of a fish. 23 patches sounds like the series would
<Ansuel>
fail on the attention span count.
<Ansuel>
Please optimize the patches for a reviewer with average intelligence and
<mangix>
uhhh, what?
<mangix>
you have a link to this patch series meant for a reviewer with above average intelligence and the attension span greater than a fish?
Tapper has quit [Ping timeout: 480 seconds]
SamantazFox has quit [Remote host closed the connection]
SamantazFox has joined #openwrt-devel
torv has joined #openwrt-devel
torv has quit [Remote host closed the connection]
<Ansuel>
Mangix it's the code split for qca8k
torv has joined #openwrt-devel
<Ansuel>
and as always reviwer asked me to split the changes and i split them too much ahahah
<mrnuke>
Ansuel: thing with "split the changes" -- it's so subjective. Everyone has a different preference on how much to split. You just gotta adapt to each reviewer, sadly
Tapper has joined #openwrt-devel
<Tapper>
So people and Ansuel I have worked out what is bootlooping the r7800
<Tapper>
It's kernel 5.15
<Tapper>
I did a build with kernel 5.10 and used tftp to flash and it works. RC5 will boot to. If I flash a build using kernel 5.15 it bootloops
<Tapper>
I have dun this twice just make shure.
<Tapper>
It's the factory.img I am using to flash over tftp not the sysupgrade.
<Ansuel>
do you have an initramfs image?
<Ansuel>
curious to check what is the output from the serial
<Tapper>
If i sysupgrade from a kernel 5.10 build to a kernel 5.15 build = bootloop. This started in master some time in the lest 10 days or so.
<Tapper>
last*
<Ansuel>
something bad in the flash process?
<Tapper>
Ansuel I don't have a ttl cable
<Tapper>
Ansuel sorry that's all I know.
<Tapper>
Rite now I need sleep. laters
<Ansuel>
gn
danitool has quit [Quit: Cubum autem in duos cubos, aut quadratoquadratum in duos quadratoquadratos]
enyc has quit [Ping timeout: 480 seconds]
enyc has joined #openwrt-devel
Ansuel has quit [Quit: Probably my PC crashed or time to sleep.]
<digitalcircuit>
OpenWRT 22.03.0rc5 (without workaround) _appears_ to fix the issue, no hard CPU resets during roughly 14 test runs of Deja Dup SFTP backups to USB drive (averaging around 2 tests a day, lost track of exact count).
<digitalcircuit>
I'll try the CPU frequency stress-test script next (which might've been an unrelated issue), and after that, I'll try swapping my NBG6817s back (one has serial headers added, other doesn't).
<slh>
:)
hanetzer has quit [Ping timeout: 480 seconds]
Tapper has quit [Ping timeout: 480 seconds]
* digitalcircuit
might still try some basic bisecting to find what fixed the issue, since it's kind of unsatisfying to simply find out "oh it works now" and not know why. But.. that research might not end up happening :)
rua has joined #openwrt-devel
hanetzer has joined #openwrt-devel
<slh>
better make sure that master and kernel v5.15 are also fine, so that it remains fixed
<digitalcircuit>
Good point! Is ipq806x on master with v5.15 switched to DSA yet? (I.e. will I need to redo config, or keep it for testing.)
<slh>
no, still swconfig
<slh>
v5.15 is still guarded by TESTING_KERNEL
<digitalcircuit>
Noted!
Slimey has quit [Read error: Connection reset by peer]
Slimey has joined #openwrt-devel
indy has quit []
Grommish has joined #openwrt-devel
indy has joined #openwrt-devel
Grommish has quit [Read error: Connection reset by peer]
indy has quit []
Grommish has joined #openwrt-devel
indy has joined #openwrt-devel
aiyion_ has quit [Remote host closed the connection]
aiyion_ has joined #openwrt-devel
ekathva has joined #openwrt-devel
ekathva has quit [Remote host closed the connection]
m has joined #openwrt-devel
ptudor has joined #openwrt-devel
ptudor_ has quit [Ping timeout: 480 seconds]
jlsalvador has quit [Quit: jlsalvador]
jlsalvador has joined #openwrt-devel
ptudor_ has joined #openwrt-devel
ptudor has quit [Ping timeout: 480 seconds]
aiyion_ has quit [Remote host closed the connection]
aiyion_ has joined #openwrt-devel
sorinello has quit [Quit: Leaving]
schwicht has joined #openwrt-devel
schwicht has quit [Read error: Connection reset by peer]
<svanheule>
mrnuke: no secure boot on rtl8380, only scrambled register spaces for security-by-obscurity. There was someone who built their own bootloader for a TPL switch on the forums too, maybe you can check with them.
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Tapper has joined #openwrt-devel
xavier_ has joined #openwrt-devel
cmonroe has quit [Ping timeout: 480 seconds]
<xavier_>
There is a commit from march where it states the need to execute the sync command after boot (https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=f44f8b07b0277d38fafb3d613a303454d3b9faaf). I have had some issues related to this, loosing recently modified files after a power cut. I would like to know if it was not an option to mount the ubifs partitions with the sync flag, instead of executing the sync command only after
<aiyion_>
the ath79-nand target does build a tar file for this device.
sorinello has joined #openwrt-devel
<aiyion_>
What would be a good starting point on altering the tar file in openwrets buildprocess?
<aiyion_>
*OpenWrt's
<aiyion_>
I'd like to add a symlink to it in order for it to allow upgrade from ar71xx
linusw____ has joined #openwrt-devel
<enyc>
digitalcircuit: I know what you mean... e.g. "I can't get it to go wrong now!?!?!?" goes on ;o
aiyion_ is now known as aiyion
<aiyion>
like the image ext
<aiyion>
*currently extracts to sysupgrade-aerohive_hiveap-121 I'd like it to do so, but to extract a symlink to this directory called sysupgrade-hiveap-121
<aiyion>
where would I inject such a symlink?
<EqUaTe>
aiyion: iirc the config structure has changed quite drastically, I wouldn't try doing a sysupgrade approach.
Tapper has quit [Ping timeout: 480 seconds]
torv has quit [Remote host closed the connection]
torv has joined #openwrt-devel
torv has quit [Remote host closed the connection]
m has quit [Remote host closed the connection]
torv has joined #openwrt-devel
bluew has quit [Quit: Leaving]
<aiyion>
EqUaTe: mhm, I know it did. Rebuilding the network section from scratch works sufficiently well though.
<aiyion>
So if I'd like to go on, how would I add such a symlink in a reasonable way?
Grommish has quit [Read error: Connection reset by peer]
* f00b4r0
roasts. 40C outside
Grommish has joined #openwrt-devel
Grommish has quit []
<tmn505>
aiyion: look at BOARD_NAME var also sysupgrade-tar in include/images-commands.mk and in scripts. sysupgrade-tar accepts only two files as input, You'll need to modify it.
torv has quit [Remote host closed the connection]
torv has joined #openwrt-devel
Tapper has joined #openwrt-devel
jlsalvador has quit [Quit: jlsalvador]
jlsalvador has joined #openwrt-devel
indy has quit []
indy has joined #openwrt-devel
Grommish has joined #openwrt-devel
Grommish has quit []
torv_ has joined #openwrt-devel
torv has quit [Remote host closed the connection]
schwicht has joined #openwrt-devel
schwicht_ has joined #openwrt-devel
schwicht has quit [Read error: Connection reset by peer]
schwicht has joined #openwrt-devel
schwicht_ has quit [Read error: Connection reset by peer]
valku has joined #openwrt-devel
snh has joined #openwrt-devel
snh_ has quit [Ping timeout: 480 seconds]
hexagonwin has joined #openwrt-devel
schwicht_ has joined #openwrt-devel
schwicht has quit [Read error: Connection reset by peer]
<enyc>
OpenWRT have a pragmatic solution/workaround ??
<stintel>
it means the problem happens on release/22.03
goliath has joined #openwrt-devel
<valku>
Mangix : after commit c0ba4201f8372eee579584b7f5900e073b207b0c, 155 packages fail to build with 'package/feeds/packages/<PKGNAME>/Makefile' has a build dependency on 'libiconv', which does not exist
<stintel>
rm -rf tmp/
<f00b4r0>
hi stintel ! How do you like your heat? Have you been toasted today? :)
<stintel>
f00b4r0: I went out for a walk when the temp was the highest :)
<stintel>
valku: yeah that takes a while even on a not so slow machine
<f00b4r0>
i stepped outside to fix a flapping shutter, and I felt like I was being choked ;P
<stintel>
but should silence the warnings
<valku>
stintel: that worked, thank you! Jee, I thought that "make clean" would clean tmp/ too. Turns out that it does not.
<stintel>
valku: tmp/ is annoying
<stintel>
when something weird is going on, rm -rf tmp/ should probably be your first or second action
<valku>
Perhaps "make clean" **should** clean up tmp/
KevinDB has joined #openwrt-devel
KevinDB is now known as ldir
<stintel>
valku: dirclean cleans it up
ldir has quit [Quit: Out of Cheese Error]
<stintel>
maybe we should move that to clean, yes
ldir has joined #openwrt-devel
<stintel>
it's in toplevel Makefile, feel free to send a patch for that
ldir has quit []
ldir has joined #openwrt-devel
<hurricos>
stintel: I have no idea where my BSAP is. o_o
<stintel>
:(
<valku>
stintel: dirclean is "nuclear" in the sense that it deletes everything except .config, which puts it in the same category as distclean. Yes, I agree, cleaning tmp/ belongs to "make clean" as "rm -r tmp/" is relatively harmless, it only increases exec time for defconfig
<hurricos>
I reorganized my room and packed things.
<hurricos>
This box is an FPGA+M68K that will drive JTAG for pretty much anything. Abatron is out of business but their hardware and firmware are everywhere
<mrnuke>
hurricos: ahead? That was my plan for weeks :D
<hurricos>
well
<hurricos>
behind you, yes, but seconds ahead of your comment.
<hurricos>
*seconds*.
<hurricos>
Wars have been won over less!!
<hurricos>
coughs. anyways.
<hurricos>
ahem
ldir has joined #openwrt-devel
<mrnuke>
good thing I'm not fighting a war :p
<hurricos>
I'd hvae lost. You are, after all, mr nuke
<mrnuke>
svanheule: Thanks! Do you remember who it was? I'll search the thread. I do feel like a nobody on that thread. You guys are working on 10G SoCs, and here come I with a puny 8-port contraption :p
<ldir>
it's not the size it's the ....
Tapper has joined #openwrt-devel
<mrnuke>
ldir: ... the bootloader?
<svanheule>
mrnuke: your wallet will thank you for not having those 10G devices
Grommish has joined #openwrt-devel
Grommish has quit []
ldir has quit [Quit: Out of Cheese Error]
ldir has joined #openwrt-devel
<mrnuke>
svanheule: wallet or electricity provider?
<svanheule>
mrnuke: actually I have the impression these realtek switches are rather low power, even for providing 10G
<enyc>
stintel: aah, hrrm, i see, still raises question in faffy situations like that who comes up with pragmatic workarounds for releases to work usefully ;o
ldir has quit []
<svanheule>
mrnuke: the XGS1250-12 for example only has the tiniest fan (but there a slab of aluminium in the housing as a heatsing)
ldir has joined #openwrt-devel
<svanheule>
best thing is, that wee fan doesn't even start properly on my device
<hurricos>
stintel: brought all of my USB to RJ45 RS232 cables to the Lab =_=
<hurricos>
I'll hop over there.
<mrnuke>
svanheule: nothing gets me to jump off a bridge fatser that a fan in a network switch
<svanheule>
mrnuke: best stick with the 1G models then :)
ldir has quit [Quit: Out of Cheese Error]
ldir has joined #openwrt-devel
<stintel>
enyc: I guess none of the developers use that hardware
<mrnuke>
hurricos: okay, I think I can un-draft PR19, given that you've already tested it :)
danitool has quit [Remote host closed the connection]
<digitalcircuit>
enyc: I'm not sure what you are asking - do you mean.. is it known why the ipq806x CPU crash stopped happening? (I don't know yet.) Or something else..?
<enyc>
digitalcircuit: you replied "yup" appranetly to me "still raises question in faffy situations like that who comes up with pragmatic workarounds for releases to work usefully" which was acutally after stintels's replies following my mention of https://github.com/openwrt/openwrt/issues/9956
<digitalcircuit>
enyc: Whoops, my bad - I meant my "yup" to be responding to your much earlier ping (where you remarked on knowing the feeling of something working but wanting to know /why/ it started working)
<digitalcircuit>
Next time I'll fit in more context than a single word
<mrnuke>
hurricos: BTW if the tek-poe logs only show failures after commands 26, 30, or 28, then the PSE is properly configured.
<mrnuke>
so my morbid curiosity is if we'll see any failures outside of those commands. If not, we drop command 26, and that should solve the "spamming" issue that svanheule identified
<mrnuke>
hurricos: okay, I guess add 23 to that list. That also comes after PSE is configured (23 == poe_cmd_power_stats)
<mrnuke>
Hehe. Those failure rates align nicely with how often we're sending a specific command. So it does appear that the probablility of getting a failure on a command is not related to the command itself :D