#asahi · 2021-12-30 02:53
cth451_desktopSomething broke down the receiving data path. Will do a bit of debugging myself but don't have high hopes#asahi · 2021-12-30 02:52
cth451_desktopmarcan: not sure whether this helps but here's dmesg https://pastebin.aosc.io/paste/uvEd2M4k1K36UjZlzrYzZg#asahi · 2021-12-30 02:33
cth451_desktopand this makes backup and experimenting slightly easier (the LUN is a logical volume on the iSCSI server)#asahi · 2021-12-30 02:32
cth451_desktopout of my hatred for soldered SSD?#asahi · 2021-12-30 02:30
cth451_desktopKnowing I'll be running Linux off iSCSI I opted for a 10G card#asahi · 2021-12-30 02:28
cth451_desktopOr at least that's how it works back in rc2#asahi · 2021-12-30 02:28
cth451_desktopThe 10G card is a aquantia atlantic and should have on-card firmware blob#asahi · 2021-12-30 02:26
cth451_desktopinstead dhcp fails due to timeout#asahi · 2021-12-30 02:26
cth451_desktopnormally the client is supposed to accept that offer with a DHCP request but this never appeared on wire#asahi · 2021-12-30 02:25
cth451_desktopSo a bunch of DHCP discover went out and the DHCP server replied with an offer#asahi · 2021-12-30 02:24
cth451_desktopon a 10GE mac mini#asahi · 2021-12-30 02:24
cth451_desktopPackets replied to the card aren't actually processed#asahi · 2021-12-30 02:23
cth451_desktopIt seems that the eth card isn't fully functional at this point#asahi · 2021-12-30 01:35
cth451_desktopThe offending function is "hw_atl2_utils_soft_reset" in quantia/atlantic/hw_atl2/hw_atl2_utils.c#asahi · 2021-12-30 01:34
cth451_desktopSounds like the card wasn't properly reset before and is already initialized and the driver just picked that up#asahi · 2021-12-30 01:33
cth451_desktopI assume this PCIe reset occurs on PCIe link bring up stage?#asahi · 2021-12-30 01:31
cth451_desktopThe actual time before reset completion was 1.75s on rc6, compared to 0.04s on rc2#asahi · 2021-12-30 01:26
cth451_desktopAny suggestions on why this would be the case given changes from rc2 to rc6?#asahi · 2021-12-30 01:25
cth451_desktopThe driver will report "Boot code hanged", but if we crank the timeout up 0.2s the card actually comes up.#asahi · 2021-12-30 01:23
cth451_desktopFor some reason on 5.16-rc6 with jennau's patchset, soft reset timeout in atlantic driver of 200000us is no longer enough#asahi · 2021-12-23 09:19
cth451_desktopAh.#asahi · 2021-12-23 09:18
cth451_desktopbut otherwise the system starts fine#asahi · 2021-12-23 09:18
cth451_desktopj`ey: That's right#asahi · 2021-12-23 09:17
cth451_desktopTTY> fb init: 640x1136 (32) [s=640] @(nil)#asahi · 2021-12-23 09:17
cth451_desktopraw m1n1.bin on mac os 12 seems to have issue initializing the framebuffer#asahi · 2021-11-24 23:11
cth451_desktopjannau: the error message is still there but ethernet port is back and working.#asahi · 2021-11-24 23:00
cth451_desktopWill try.#asahi · 2021-11-24 22:51
cth451_desktopon 5.16-rc2#asahi · 2021-11-24 22:51
cth451_desktop* again#asahi · 2021-11-24 22:51
cth451_desktopand 10GE is gone from lspci agian#asahi · 2021-11-24 22:51
cth451_desktopI'm getting "powercontroller@3b0: PS dispdfr_be: Failed to reach power state 0xf (now: 0xb4f)"#asahi · 2021-11-24 21:38
cth451_desktopjannau: 0054-fixup-arm64-dts-apple-Add-missing-M1-t8103-devices.patch seems to be empty#asahi · 2021-11-24 21:36
cth451_desktopIt might be helpful to see what happened before aplay got into ppoll#asahi · 2021-11-24 21:33
cth451_desktopOut of ideas.#asahi · 2021-11-24 21:30
cth451_desktopbut on my system any other process try to open a card in use gets a -EBUSY#asahi · 2021-11-24 21:29
cth451_desktopI might be wrong but this sounds like some process already opened the alsa device#asahi · 2021-11-24 21:21
cth451_desktopmight help to strace aplay and see where it got stuck#asahi · 2021-11-24 21:19
cth451_desktopDoes `aplay -l` say anything?#asahi · 2021-11-24 21:03
cth451_desktopIf that works then the problem is probably not in the kernel#asahi · 2021-11-24 21:03
cth451_desktopThe most primal test would be using aplay to make some sound directly via alsa#asahi · 2021-11-24 21:02
cth451_desktopmight need to fiddle with kernel config and grab pulse or any other userspace stuff#asahi · 2021-11-24 21:01
cth451_desktopmps: I'm using an apple c-to-3.5 dongle and it works for me#asahi · 2021-11-23 08:39
cth451_desktopand the audio would get muxed in by some sort of HDMI controller#asahi · 2021-11-23 08:38
cth451_desktopSounds like it would be an independent digital audio output device#asahi · 2021-11-23 08:38
cth451_desktopAh thanks for the explanation#asahi · 2021-11-23 03:54
cth451_desktop* GPU driver#asahi · 2021-11-23 03:54
cth451_desktopI guess M1 minis would not have sound over HDMI until GNU driver gets fully implemented?#asahi · 2021-11-23 00:55
cth451_desktop* if that's mac os it sounds#asahi · 2021-11-23 00:54
cth451_desktop(if that sounds a bit ridiculous#asahi · 2021-11-23 00:47
cth451_desktopUnder mac os or linux?