eem gadget works only after cold reboot

htot commented 3 years ago

This was with your configurations rebased on 5.10.24 (LTS) https://github.com/edison-fw/linux/commits/eds-acpi-5.10.24 I have bitbake apply 2 other (relevant) patches:

SRC_URI_append = " file://0001-HACK-do-not-enumerated-SFI-devices.patch"
SRC_URI_append = " file://0001-usb-dwc3-pci-Enable-dis_uX_susphy_quirk-for-Intel-Me.patch"

I'll highlight the interesting bits and have some conclusions at the end:

This is to test eem ethernet gadget (Connector USB 2 shown as 2 on the photo).

afbeelding

In these tests I leave edison-arduino switch in gadget mode (Switch 1 in the position as shown on the photo). Communication with edison is through console (Connector USB 1 shown as 3 on the photo)

1) Cold boot (press Edison reset button 7 sec), with USB 2 unplugged, wait for prompt, login, plug USB 2

root@yuna:~# connmanctl services
*AO Wired                gadget_aabbccddeef1_usb
...

root@yuna:~# ifconfig
...

usb0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500  metric 1
        inet 10.42.0.221  netmask 255.255.255.0  broadcast 10.42.0.255
        inet6 fe80::a8bb:ccff:fedd:eef1  prefixlen 64  scopeid 0x20<link>
        ether aa:bb:cc:dd:ee:f1  txqueuelen 1000  (Ethernet)
        RX packets 45  bytes 7074 (6.9 KiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 43  bytes 6173 (6.0 KiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

root@yuna:~# iperf3 -s
-----------------------------------------------------------
Server listening on 5201
-----------------------------------------------------------
Accepted connection from 10.42.0.1, port 50840
[  5] local 10.42.0.221 port 5201 connected to 10.42.0.1 port 50842
[ ID] Interval           Transfer     Bitrate
[  5]   0.00-1.00   sec  23.1 MBytes   193 Mbits/sec
[  5]   1.00-2.00   sec  24.1 MBytes   202 Mbits/sec
[  5]   2.00-3.00   sec  24.4 MBytes   205 Mbits/sec
[  5]   3.00-4.00   sec  24.4 MBytes   205 Mbits/sec

On host (run 2 min, at 200Mb/sec Edison doesn't crash in other cases often does)

ferry@delfion:~$ iperf3 -c 10.42.0.221 -t 0
Connecting to host 10.42.0.221, port 5201
[  5] local 10.42.0.1 port 50842 connected to 10.42.0.221 port 5201
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec  26.7 MBytes   224 Mbits/sec    0   1.12 MBytes
[  5]   1.00-2.00   sec  23.8 MBytes   199 Mbits/sec   49   1.13 MBytes
[  5]   2.00-3.00   sec  23.8 MBytes   199 Mbits/sec    0   1.25 MBytes
[  5]   3.00-4.00   sec  25.0 MBytes   210 Mbits/sec    0   1.34 MBytes
[  5]   4.00-5.00   sec  23.8 MBytes   199 Mbits/sec    0   1.40 MBytes
[  5]   5.00-6.00   sec  25.0 MBytes   210 Mbits/sec   14   1.04 MBytes
...
[  5] 157.00-158.00 sec  0.00 Bytes  0.00 bits/sec    0   1.21 MBytes
[  5] 158.00-159.00 sec  0.00 Bytes  0.00 bits/sec    0   1.21 MBytes
^C[  5] 159.00-159.32 sec  0.00 Bytes  0.00 bits/sec    0   1.21 MBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-159.32 sec  3.41 GBytes   184 Mbits/sec  294             sender
[  5]   0.00-159.32 sec  0.00 Bytes  0.00 bits/sec                  receiver
iperf3: interrupt - the client has terminated

2) kill iperf3 on Edison, unplug USB 2, wait 5 sec, plug USB 2

root@yuna:~# connmanctl services
*AO Wired                gadget_aabbccddeef1_usb

(A = autoconnected, O connection pinged intel.com)

root@yuna:~# iperf3 -s
-----------------------------------------------------------
Server listening on 5201
-----------------------------------------------------------
Accepted connection from 10.42.0.1, port 50914
[  5] local 10.42.0.221 port 5201 connected to 10.42.0.1 port 50916
[ ID] Interval           Transfer     Bitrate
[  5]   0.00-1.00   sec  4.62 MBytes  38.8 Mbits/sec
[  5]   1.00-2.00   sec  1017 KBytes  8.33 Mbits/sec
[  5]   2.00-3.00   sec   672 KBytes  5.50 Mbits/sec
[  5]   3.00-4.00   sec   342 KBytes  2.80 Mbits/sec
[  5]   4.00-5.00   sec   365 KBytes  2.99 Mbits/sec
[  5]   5.00-6.00   sec   380 KBytes  3.12 Mbits/sec
[  5]   6.00-7.00   sec   329 KBytes  2.70 Mbits/sec
[  5]   7.00-8.00   sec  90.5 KBytes   742 Kbits/sec
[  5]   8.00-9.00   sec  3.14 MBytes  26.3 Mbits/sec
[  5]   9.00-10.00  sec   126 KBytes  1.03 Mbits/sec
[  5]  10.00-11.00  sec   776 KBytes  6.36 Mbits/sec

On host

ferry@delfion:~$ iperf3 -c 10.42.0.221 -t 0
Connecting to host 10.42.0.221, port 5201
[  5] local 10.42.0.1 port 50916 connected to 10.42.0.221 port 5201
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec  6.18 MBytes  51.8 Mbits/sec    2    260 KBytes
[  5]   1.00-2.00   sec   827 KBytes  6.77 Mbits/sec    1    296 KBytes
[  5]   2.00-3.00   sec   827 KBytes  6.77 Mbits/sec    0    327 KBytes
[  5]   3.00-4.00   sec  0.00 Bytes  0.00 bits/sec    1    342 KBytes
[  5]   4.00-5.00   sec   891 KBytes  7.30 Mbits/sec    0    359 KBytes
[  5]   5.00-6.00   sec  0.00 Bytes  0.00 bits/sec    0    378 KBytes
[  5]   6.00-7.00   sec   891 KBytes  7.30 Mbits/sec    0    393 KBytes
[  5]   7.00-8.00   sec  0.00 Bytes  0.00 bits/sec    0    397 KBytes
[  5]   8.00-9.00   sec  4.41 MBytes  37.0 Mbits/sec    1    566 KBytes
[  5]   9.00-10.00  sec  0.00 Bytes  0.00 bits/sec    1    568 KBytes
...
[  5] 120.00-121.00 sec  0.00 Bytes  0.00 bits/sec    0   3.74 MBytes
[  5] 121.00-122.00 sec  0.00 Bytes  0.00 bits/sec    0   3.74 MBytes
[  5] 122.00-123.00 sec  0.00 Bytes  0.00 bits/sec    0   3.75 MBytes
[  5] 123.00-124.00 sec  0.00 Bytes  0.00 bits/sec    0   3.75 MBytes
[  5] 124.00-125.00 sec  1.25 MBytes  10.5 Mbits/sec    0   3.75 MBytes
^C[  5] 125.00-125.41 sec  0.00 Bytes  0.00 bits/sec    0   3.75 MBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-125.41 sec  53.9 MBytes  3.61 Mbits/sec  431             sender
[  5]   0.00-125.41 sec  0.00 Bytes  0.00 bits/sec                  receiver
iperf3: interrupt - the client has terminated

3) warm reboot Edison with USB 2 plugged: no change 4) warm reboot Edison with USB 2 unplugged: plug event seems not detected, other gadgets do not appear (sound, mass storage) But connman thinks gadget is there, powered, but not connected:

root@yuna:~# connmanctl technologies
...
/net/connman/technology/gadget
  Name = Gadget
  Type = gadget
  Powered = True
  Connected = False
  Tethering = False

And device is standbye (R) but not pinging intel.com.

root@yuna:~# connmanctl services
*AR Wired                gadget_aabbccddeef1_usb

Didn't receive correct config from dhcp (or not at all)

root@yuna:~# ifconfig
...

usb0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500  metric 1
        inet 169.254.147.69  netmask 255.255.0.0  broadcast 169.254.255.255
        inet6 fe80::a8bb:ccff:fedd:eef1  prefixlen 64  scopeid 0x20<link>
        ether aa:bb:cc:dd:ee:f1  txqueuelen 1000  (Ethernet)
        RX packets 0  bytes 0 (0.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 0  bytes 0 (0.0 B)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

5) Cold reboot with USB 2 plugged: same as with cable unplugged. 6) Warm reboot with USB 2 plugged (network in good state): device is standbye (R) but not pinging intel.com.

7) Cold reboot, then Warm reboot with USB 2 unplugged, connected but bad:

ferry@delfion:~$ iperf3 -c 10.42.0.221 -t 0
Connecting to host 10.42.0.221, port 5201
[  5] local 10.42.0.1 port 51216 connected to 10.42.0.221 port 5201
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec  6.97 MBytes  58.4 Mbits/sec    2    303 KBytes
[  5]   1.00-2.00   sec  1.43 MBytes  12.0 Mbits/sec    1    361 KBytes
[  5]   2.00-3.00   sec  1.93 MBytes  16.2 Mbits/sec    0    419 KBytes
[  5]   3.00-4.00   sec  4.67 MBytes  39.2 Mbits/sec    0    605 KBytes
[  5]   4.00-5.00   sec  2.50 MBytes  21.0 Mbits/sec    1    707 KBytes
[  5]   5.00-6.00   sec  0.00 Bytes  0.00 bits/sec    0    731 KBytes

8) Cold reboot with USB 2 plugged:

root@yuna:~# connmanctl disable gadget

unplug USB 2, wait 5 sec, plug USB 2:

root@yuna:~# connmanctl enable gadget
ferry@delfion:~$ iperf3 -c 10.42.0.221 -t 0
Connecting to host 10.42.0.221, port 5201
[  5] local 10.42.0.1 port 40422 connected to 10.42.0.221 port 5201
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec  25.6 MBytes   215 Mbits/sec    0   1.10 MBytes       
[  5]   1.00-2.00   sec  25.0 MBytes   210 Mbits/sec   89   1.16 MBytes       
[  5]   2.00-3.00   sec  23.8 MBytes   199 Mbits/sec    0   1.28 MBytes       
[  5]   3.00-4.00   sec  23.8 MBytes   199 Mbits/sec    0   1.38 MBytes       
[  5]   4.00-5.00   sec  25.0 MBytes   210 Mbits/sec   11   1.02 MBytes       
[  5]   5.00-6.00   sec  25.0 MBytes   210 Mbits/sec    0   1.09 MBytes       
[  5]   6.00-7.00   sec  23.8 MBytes   199 Mbits/sec    0   1.14 MBytes       
^C[  5]   7.00-7.39   sec  10.0 MBytes   214 Mbits/sec    0   1.16 MBytes       
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-7.39   sec   182 MBytes   206 Mbits/sec  100             sender
[  5]   0.00-7.39   sec  0.00 Bytes  0.00 bits/sec                  receiver
iperf3: interrupt - the client has terminated

Conclusion

On cold boot dwc3 eem gadget works fine.
Unplug/replug messes up the driver or the hardware, warm reboot does not reset this behavior.
In working state, disabling the eem gadget with connmanctl (I believe this to be the same as ip link set dev usb0 down), plug/replug followed by enabling again, eem gadget works.

htot commented 3 years ago

I'm dropping trace and regdump after a cold and a warm boot captured according to instructions here: https://www.kernel.org/doc/html/latest/driver-api/usb/dwc3.html#reporting-bugs

boot.zip

htot commented 3 years ago

While the above traces in boot.zip allow compare which regs not correctly initialized on warm boot, I have also captured traces of unplug/plug. Here kernel is 5.10.27 (LTS), edison cold booted with USB cable plugged. Loggin via console, verified connection inet 10.42.0.221 netmask 255.255.255.0

root@yuna:~# ifconfig
lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536  metric 1
        inet 127.0.0.1  netmask 255.0.0.0
        inet6 ::1  prefixlen 128  scopeid 0x10<host>
        loop  txqueuelen 1000  (Local Loopback)
        RX packets 44  bytes 3412 (3.3 KiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 44  bytes 3412 (3.3 KiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

usb0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500  metric 1
        inet 10.42.0.221  netmask 255.255.255.0  broadcast 10.42.0.255
        inet6 fe80::a8bb:ccff:fedd:eef1  prefixlen 64  scopeid 0x20<link>
        ether aa:bb:cc:dd:ee:f1  txqueuelen 1000  (Ethernet)
        RX packets 61  bytes 8416 (8.2 KiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 54  bytes 7779 (7.5 KiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

wlan0: flags=-28669<UP,BROADCAST,MULTICAST,DYNAMIC>  mtu 1500  metric 1
        ether 78:4b:87:ad:4d:70  txqueuelen 1000  (Ethernet)
        RX packets 0  bytes 0 (0.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 0  bytes 0 (0.0 B)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

Then trace unplug. Then trace plug.

plug-unplug.zip

htot commented 3 years ago

Now, with host updated to ubuntu kernel ppa 5.12.0-051200rc5-generic and edison to 5.12.0-rc5-edison-acpi-standard vanilla + 2 patches appearing in rc6:

"usb: dwc3: gadget: Clear DEP flags after stop transfers in ep disable"
"usb: dwc3: pci: Enable dis_uX_susphy_quirk for Intel Merrifield" plus one from https://github.com/andy-shev/linux/commits/eds-acpi
"TODO: driver core: Break infinite loop when deferred probe can't be satisfied" captured on good and one bad connection, plus logs on the host side see journalctl-plus-comments.txt. 5.12-rc5.zip

htot commented 3 years ago

I added another 3:

usb: dwc3: Fix DRD mode change sequence following programming guide
usb: dwc3: Trigger a GCTL soft reset when switching modes in DRD
usb: dwc3: gadget: increase BESL baseline to 6

The latter seems to solve the throttling, while the first 2 help to make switching from host -> device mode work (better).

htot commented 3 years ago

I retested usb-next (based 5.12-rc7) + "[PATCH v3] usb: dwc3: core: Do core softreset when switch mode". No throttling, but

if I flip the switch (or unplug/plug the cable) only once, on the PC side I get:

kernel: usb 1-5: new high-speed USB device number 18 using xhci_hcd
kernel: usb 1-5: New USB device found, idVendor=1d6b,
idProduct=0104, bcdDevice= 1.00 kernel: usb 1-5: New USB device
strings: Mfr=1, Product=2, SerialNumber=3 kernel: usb 1-5: Product:
USBArmory Gadget kernel: usb 1-5: Manufacturer: USBArmory kernel:
usb 1-5: SerialNumber: 0123456789abcdef kernel: usb 1-5: can't set
config #1, error -110

In the zip file 1 more trace of just the switch from host->gadget 5.12-rc7+usb-next.zip

htot commented 3 years ago

regdump in zip files empty, resend.. regdump.txt

htot commented 3 years ago

After flipping the first time the trace is empty see first-switch.zip

I found the gadget configuration script was not called, which normally gets called due to a udev rule:

ACTION=="add", KERNEL=="dwc3.0.auto", SUBSYSTEMS=="udc", ATTRS{state}=="not attached", RUN+="/usr/bin/conf-gadget.sh"

So I retried and see with udevadm monitor

flipping the switch from host->gadget

KERNEL[51.824914] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0/net/enp0s17u1u1/queues/rx-0 (queues)
KERNEL[51.825682] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0/net/enp0s17u1u1/queues/tx-0 (queues)
KERNEL[51.826226] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0/net/enp0s17u1u1 (net)
KERNEL[51.836041] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/mdio_bus/usb-001:003/usb-001:003:01 (mdio_bus)
KERNEL[51.836709] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/mdio_bus/usb-001:003/usb-001:003:01 (mdio_bus)
KERNEL[51.837342] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/mdio_bus/usb-001:003 (mdio_bus)
KERNEL[51.837763] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0 (usb)
KERNEL[51.838116] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0 (usb)
KERNEL[51.873712] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1 (usb)
KERNEL[51.874000] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1 (usb)
KERNEL[51.874207] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1:1.0 (usb)
KERNEL[51.874431] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1:1.0 (usb)
KERNEL[51.897175] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1 (usb)
KERNEL[51.897486] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1 (usb)

stopped capture tracepoints here, then switch back to host

KERNEL[253.214406] add      /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1 (usb)
KERNEL[253.263305] change   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1 (usb)
KERNEL[253.263687] add      /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1:1.0 (usb)
KERNEL[253.328354] bind     /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1:1.0 (usb)
KERNEL[253.328734] bind     /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1 (usb)
KERNEL[253.699341] add      /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1 (usb)
KERNEL[253.744911] change   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1 (usb)
KERNEL[253.745804] add      /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0 (usb)
KERNEL[253.805307] add      /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/mdio_bus/usb-001:005 (mdio_bus)
KERNEL[253.812978] add      /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/mdio_bus/usb-001:005/usb-001:005:01 (mdio_bus)
KERNEL[253.814318] bind     /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/mdio_bus/usb-001:005/usb-001:005:01 (mdio_bus)
KERNEL[253.815386] add      /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0/net/eth0 (net)
KERNEL[253.815552] add      /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0/net/eth0/queues/rx-0 (queues)
KERNEL[253.815778] add      /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0/net/eth0/queues/tx-0 (queues)
KERNEL[253.825279] bind     /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0 (usb)
KERNEL[253.825667] bind     /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1 (usb)

switch to gadget again

KERNEL[314.212144] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0/net/eth0/queues/rx-0 (queues)
KERNEL[314.212473] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0/net/eth0/queues/tx-0 (queues)
KERNEL[314.214691] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0/net/eth0 (net)

extcon event didn't show the first time

KERNEL[314.238385] change   /devices/pci0000:00/0000:00:13.0/INTC100E:00/mrfld_bcove_pwrsrc/extcon/extcon0 (extcon)
KERNEL[314.238677] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/mdio_bus/usb-001:005/usb-001:005:01 (mdio_bus)
KERNEL[314.238863] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/mdio_bus/usb-001:005/usb-001:005:01 (mdio_bus)
KERNEL[314.239015] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/mdio_bus/usb-001:005 (mdio_bus)
KERNEL[314.239205] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0 (usb)
KERNEL[314.239429] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1/1-1.1:1.0 (usb)
KERNEL[314.239666] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb2/2-0:1.0 (usb)
KERNEL[314.239933] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb2/2-0:1.0 (usb)
KERNEL[314.262713] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb2 (usb)
KERNEL[314.263030] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1 (usb)
KERNEL[314.263298] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb2 (usb)
KERNEL[314.263569] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1.1 (usb)
KERNEL[314.263815] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1:1.0 (usb)
KERNEL[314.264042] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1/1-1:1.0 (usb)
KERNEL[314.264753] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usbmon/usbmon2 (usbmon)
KERNEL[314.265019] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1 (usb)
KERNEL[314.265289] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-1 (usb)
KERNEL[314.288792] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-0:1.0 (usb)
KERNEL[314.289057] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1/1-0:1.0 (usb)
KERNEL[314.289327] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1 (usb)
KERNEL[314.289661] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usb1 (usb)
KERNEL[314.647375] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto/usbmon/usbmon1 (usbmon)
KERNEL[314.647816] unbind   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto (platform)
KERNEL[314.648143] remove   /kernel/software_nodes/node1 (software_nodes)
KERNEL[314.648672] remove   /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/xhci-hcd.2.auto (platform)

here is the event we were waiting for

KERNEL[314.649158] add      /devices/pci0000:00/0000:00:11.0/dwc3.0.auto/udc/dwc3.0.auto (udc)

after this gadget devices appear normally

htot commented 3 years ago

Also after some time (a few minutes) the connection drop but automatically reconnects. The following trace shows that near the end. lost-connection.zip

htot commented 3 years ago

In total there have appeared to be 5 separate issues, summury here:

Throttling

https://lore.kernel.org/linux-usb Caused by LPM

Fixed by “PROPERTY_ENTRY_BOOL("snps,dis_enblslpm_quirk")”

Also fixed by “BESL baseline to 6”:


diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c index 60e850a395a2..423533df8927 100644
--- a/drivers/usb/dwc3/gadget.c
+++ b/drivers/usb/dwc3/gadget.c
@@ -2895,7 +2895,7 @@ static void dwc3_gadget_config_params(struct usb_gadget *g,

recommended BESL baseline to 1 and clamp the BESL deep to be within 2 to 15. /
params->besl_baseline = 1;

params->besl_baseline = 6; if (dwc->is_utmi_l1_suspend) params->besl_deep = clamp_t(u8, dwc->hird_threshold, 2, 15);


### Problems switching DRD
[https://lore.kernel.org/linux-usb](https://lore.kernel.org/linux-usb/96c64e6a788552371081f37f544041b7ee046ef5.1618452732.git.Thinh.Nguyen@synopsys.com/)

Need to quickly flip switch back and forth to switch to gadget mode
Fixed by “usb: dwc3: core: Do core softreset when switch mode“

Failed plug

https://lore.kernel.org/linux-usb

appears as "can't set config #1, error -110" on the host side
fixed by "snps,usb2-gadget-lpm-disable" or
fixed by "usb: dwc3: gadget: Fix START_TRANSFER link state check"

USB port resets

Happens also due to LPM
Fixed by "usb: dwc3: pci: Enable usb2-gadget-lpm-disable for Intel Merrifield"
This requires patch "usb: dwc3: gadget: Check for disabled LPM quirk"
"snps,dis_enblslpm_quirk" doesn't work to resolve port resets
"snps,usb2-gadget-lpm-disable" disables LPM entirely so “BESL baseline to 6” no longer required.

Initially switch needs flipping 2x to go to device mode

Extcon has status variable that needs initialization
Fixed by “extcon-intel-mrfkd: initialize mrfld_extcon status“

andy-shev commented 3 years ago

So, as far as I understand we need only two patches now to fix all 5 issues, i.e. extcon + snps,usb2-gadget-lpm-disable. Correct?

htot commented 3 years ago

Yes. Thanks for reviewing, sent to ml.

andy-shev / linux