14:52 <qzed> bamse: is it normal to get a bunch of `Remote returned an error: End of Transfer` in tqftpserv?

14:57 <qzed> I also see an `unhandled op 5`

16:02 <qzed> also what's up with those mcfg_hw/sw files?

16:11 <qzed> ah, I assume modem configs... found them on windows

19:39 <bamse> steev: okay, sounds like something we need to debug

19:40 <bamse> qzed: the wifi hardware is supposed to have a board-id flashed onto it in factory, but unfortunately it's not uncommon for vendors to skip that step :/

19:43 <qzed> bamse: if it's supposed to be in the eeprom, I think there are other issues...

19:43 <qzed> apparently the region ID is also not there and falls back to US, which is what I think blocks channel 12 because that's not available in north america

19:44 <bamse> qzed: i unfortunately don't remember the details :/

19:44 <qzed> I suppose that's probably because I use dummy files instead of those partitions

19:45 <bamse> hmm, i didn't expect that to be the case...but it's not impossible

19:45 <qzed> anyways, it mostly works, I'm sure we can figure out the kinks at some point too

19:45 <bamse> i did think it was some programmable bits in the external wifi module

19:46 <qzed> haven't really looked at it yet in detail, so I'm just speculating

19:48 <qzed> bamse: another thing before I try installing on the nvme:

19:48 <qzed> as far as I can tell, the pro-x uses pcie1 and pcie2 combined for a total of 4 lanes

19:49 <qzed> in the DT, you've set num-lanes = <4> for pcie2 and the nvme works with pcie2 in use

19:49 <qzed> but it kinda looks like pcie2_phy only has two lanes?

19:49 <qzed> is that correct?

19:50 <bamse> i've not looked at sc8180x in this regards for a while, but i think it's the same design as in sc8280xp...

19:50 <qzed> I mean it does seem to work, I'm just wondering a bit whether it's using 2 or 4 lanes and how that whole bonded together thing works

19:50 <bamse> you can either run two 2-lane controllers with the two 2-lane phys...or the first controller in 4-lane mode, wired up to both the associated phys

19:51 <bamse> so i would expect that if this works for you by just setting up the controller to 4 lanes and fire up the first phy, then you're running off the bootloader configuration of the second phy...or something like that

19:52 <qzed> hmm okay, yeah that's kinda what I'd wanted to avoid before installing...

19:52 <bamse> my colleague is working on some patches for doing this properly...

19:52 <qzed> nice!

19:52 <bamse> the only thing i can think of is that it perhaps might break if you do suspend the platform without the support to power things up properly again

19:52 <bamse> but we're not doing that yet...

19:53 <bamse> so just run with it :)

19:53 <qzed> alright

19:53 <qzed> thanks!

20:15 <steev> bamse: i'll wait til rc1 before looking closer

20:27 <qzed> steev: I found a workaround to disable those `chan info` messages

20:27 <qzed> last step in https://gitlab.com/postmarketOS/pmaports/-/issues/494#note_307360685

20:27 <qzed> not sure exactly what it does... but it works

20:28 <steev> ah

20:30 <steev> qzed: https://www.spinics.net/lists/linux-wireless/msg178387.html i think this commit message explains it

20:31 <qzed> oh nice, yes that seems to make sense

20:41 <steev> qzed: and can confirm, do not see them here now

20:43 <bamse> but the firmware-5.bin is generic?

20:44 <steev> it is

20:44 <bamse> or can we somehow make that change stick for all sc8180x boards?

20:44 <steev> if it's pushed upstream, sure

20:44 <steev> upstream's firmware-5 comes with wowlan,mgmt-tx-by-reference,non-bmi

20:44 <bamse> yeah, but wont' that break for other boards using firmware-5.bin?

20:44 <steev> well, the c630 doesn't have that option set

20:44 <qzed> yeah I was wondering that as well... the ath10k-firmware repo from kvalo has firmware-5.bin-s in some subdirectories for certain things

20:44 <steev> so i could throw it on there and see if it breaks?

20:45 <qzed> https://github.com/kvalo/ath10k-firmware/tree/master/WCN3990/hw1.0

20:45 <qzed> so I guess there should be a way to select those via the QC_IMAGE_VERSION_STRING or something?

21:00 <steev> qzed: the one in there seems to be what the c630 uses

21:00 <steev> ath10k_snoc 18800000.wifi: qmi fw_version 0x2009856b fw_build_timestamp 2018-07-19 12:28 fw_build_id QC_IMAGE_VERSION_STRING=WLAN.HL.2.0-01387-QCAHLSWMTPLZ-1

21:04 <qzed> it looks like at least some of the HL3.1 stuff also uses single-chan-info-per-channel

23:51 <steev> qzed: when using your wifi to transmit a decent amount of data (e.g. apt update && apt full-upgrade), does doing something like moving the mouse cursor "lag" ?