Re: umodem0, Cisco USB serial console, and quirks
On 02.11.2018 16:59, Warner Losh wrote: > All the commands for USB modems should be described here: > > https://www.usb.org/sites/default/files/CDC_EEM10.pdf > > I'm sorry, but I don't have time currently to decode the Linux traces > and compare with FreeBSD. > > > One thing to also note about the Linux driver. "it just works" is only > because people have inserted a crap-ton of special cases and dealt with > weird quirks of different devices, and devices that are only kinda sorta > standards conforming, but easy enough to hack the driver to make work. > Chances are quite good you'll have to, as hps points out, puzzle through > the traces to find where it's non-conformant and/or our driver is > non-compliant. So, when it "just work" under Linux, Windows and Mac OS, but doesn't work under FreeBSD, I suspect we need fix FreeBSD :) -- WBR, Andrey V. Elsukov signature.asc Description: OpenPGP digital signature
Re: umodem0, Cisco USB serial console, and quirks
On 02.11.2018 12:12, Hans Petter Selasky wrote: >> # cu -l /dev/cuaU0 >> Connected >> >> >> > > Did you try "-s" option to set another baudrate? Yes, the result is the same. # cu -s 115200 -l /dev/cuaU0 Connected If I understand correctly from https://www.bit0.com/tmp/usbmon-ciscoasa-verbose.txt The Leftover Capture Data: 802508 and frame[1] WRITE 7 bytes 80 25 00 00 00 00 08 -- -- -- -- -- -- -- -- -- |.%. is something like configuration data, where 80 25 is 0x2580 == 9600, so it should work with default speed. When I use 115200, I see this: 09:27:52.657367 usbus0.2 SUBM-CTRL-EP=,SPD=FULL,NFR=2,SLEN=16,IVAL=0 frame[0] WRITE 8 bytes 21 20 00 00 00 00 07 00 -- -- -- -- -- -- -- -- |! ..| frame[1] WRITE 7 bytes 80 25 00 00 00 00 08 -- -- -- -- -- -- -- -- -- |.%. | flags 0x10 status 0xca1a3 09:27:52.657711 usbus0.2 DONE-CTRL-EP=,SPD=FULL,NFR=2,SLEN=0,IVAL=0,ERR=0 frame[0] WRITE 8 bytes frame[1] WRITE 7 bytes flags 0x10 status 0xca1a1 09:27:52.760528 usbus0.2 SUBM-CTRL-EP=,SPD=FULL,NFR=2,SLEN=16,IVAL=0 frame[0] WRITE 8 bytes 21 20 00 00 00 00 07 00 -- -- -- -- -- -- -- -- |! ..| frame[1] WRITE 7 bytes 00 C2 01 00 00 00 08 -- -- -- -- -- -- -- -- -- |... | flags 0x10 status 0xea1a3 0x01c200 == 115200. -- WBR, Andrey V. Elsukov signature.asc Description: OpenPGP digital signature
Re: umodem0, Cisco USB serial console, and quirks
On 02.11.2018 11:45, Hans Petter Selasky wrote:
>>> And it seems in Linux it is handled by generic CDC ACM driver without
>>> any quirks. So, I think the problem is with our umodem(4) or ucom(4)
>>> driver.
>>>
>> No, I didn't.
>>
>> The usbmon dumps above are still there, and if someone can suggest
>> anything else I can do to help, lemme know.
>>
>
> Hi,
>
> Are there any errors printed in dmesg for this device?
>
> What does "usbconfig -d X.Y dump_device_desc dump_curr_config_desc"
> where X.Y are numbers after ugenX.Y, output?
Hi,
no errors, it looks like it should work, but cu(1) doesn't show the
console output.
# kldload umodem
umodem0 on uhub1
umodem0: on
usbus0
umodem0: data interface 1, has no CM over data, has no break
# usbconfig -d ugen0.2 dump_device_desc dump_curr_config_desc
ugen0.2: at usbus0, cfg=0 md=HOST spd=FULL
(12Mbps) pwr=ON (100mA)
bLength = 0x0012
bDescriptorType = 0x0001
bcdUSB = 0x0200
bDeviceClass = 0x0002
bDeviceSubClass = 0x
bDeviceProtocol = 0x
bMaxPacketSize0 = 0x0008
idVendor = 0x05a6
idProduct = 0x0009
bcdDevice = 0x
iManufacturer = 0x0001
iProduct = 0x0002
iSerialNumber = 0x
bNumConfigurations = 0x0001
Configuration index 0
bLength = 0x0009
bDescriptorType = 0x0002
wTotalLength = 0x0043
bNumInterfaces = 0x0002
bConfigurationValue = 0x0001
iConfiguration = 0x0003
bmAttributes = 0x00c0
bMaxPower = 0x0032
Interface 0
bLength = 0x0009
bDescriptorType = 0x0004
bInterfaceNumber = 0x
bAlternateSetting = 0x
bNumEndpoints = 0x0001
bInterfaceClass = 0x0002
bInterfaceSubClass = 0x0002
bInterfaceProtocol = 0x0001
iInterface = 0x
Additional Descriptor
bLength = 0x05
bDescriptorType = 0x24
bDescriptorSubType = 0x00
RAW dump:
0x00 | 0x05, 0x24, 0x00, 0x10, 0x01
Additional Descriptor
bLength = 0x04
bDescriptorType = 0x24
bDescriptorSubType = 0x02
RAW dump:
0x00 | 0x04, 0x24, 0x02, 0x02
Additional Descriptor
bLength = 0x05
bDescriptorType = 0x24
bDescriptorSubType = 0x06
RAW dump:
0x00 | 0x05, 0x24, 0x06, 0x00, 0x01
Additional Descriptor
bLength = 0x05
bDescriptorType = 0x24
bDescriptorSubType = 0x01
RAW dump:
0x00 | 0x05, 0x24, 0x01, 0x00, 0x01
Endpoint 0
bLength = 0x0007
bDescriptorType = 0x0005
bEndpointAddress = 0x0081
bmAttributes = 0x0003
wMaxPacketSize = 0x0008
bInterval = 0x0002
bRefresh = 0x
bSynchAddress = 0x
Interface 1
bLength = 0x0009
bDescriptorType = 0x0004
bInterfaceNumber = 0x0001
bAlternateSetting = 0x
bNumEndpoints = 0x0002
bInterfaceClass = 0x000a
bInterfaceSubClass = 0x
bInterfaceProtocol = 0x
iInterface = 0x
Endpoint 0
bLength = 0x0007
bDescriptorType = 0x0005
bEndpointAddress = 0x0082
bmAttributes = 0x0002
wMaxPacketSize = 0x0040
bInterval = 0x
bRefresh = 0x
bSynchAddress = 0x
Endpoint 1
bLength = 0x0007
bDescriptorType = 0x0005
bEndpointAddress = 0x0003
bmAttributes = 0x0002
wMaxPacketSize = 0x0020
bInterval = 0x
bRefresh = 0x
bSynchAddress = 0x
# cu -l /dev/cuaU0
Connected
--
WBR, Andrey V. Elsukov
signature.asc
Description: OpenPGP digital signature
Re: umodem0, Cisco USB serial console, and quirks
On 27.09.2017 05:11, Mike Andrews wrote: >> This is the USB read transfer. The other side is not returning any >> data, so neither is "cu" printing anything. Maybe you can try to get a >> similar trace from the system where it works? >> >> Looks to me like some kind of protocol error, that the Cisco USB >> serial console needs some kind of activation. > > Took me a few days to get around to getting usbmon going on Linux. The > verbose dump is too large for the list, so: > > https://www.bit0.com/tmp/usbmon-ciscoasa-verbose.txt > https://www.bit0.com/tmp/usbmon-ciscoasa.txt Hi, did you find some solution? It seems this device still doesn't work with recent FreeBSD. I found how it should work on linux: https://community.cisco.com/t5/network-architecture-documents/cisco-console-connection-via-usb-serial/ta-p/3163540 And it seems in Linux it is handled by generic CDC ACM driver without any quirks. So, I think the problem is with our umodem(4) or ucom(4) driver. -- WBR, Andrey V. Elsukov signature.asc Description: OpenPGP digital signature
Re: How do I prevent host usb wifi driver from attaching?
On 26.12.2012 01:22, G. Paul Ziemba wrote: I'm running FreeBSD 8.2-STABLE (host) and VirtualBox 4.2.0. I have followed the steps in http://wiki.freebsd.org/VirtualBox to make USB devices available to VirtualBox guests, and that works (e.g., a Win2K guest can use a USB scanner I have attached to the host's USB interface). I'm planning another guest running FreeBSD (probably 9.X) and I'd like to have a USB wifi dongle (probably a run(4) device) that is visible to the guest as a network interface, but that is not visible to the host. I think this can be accomplished by preventing the run(4) driver on the host from seizing the usb device, and then configuring VirtualBox to let that USB device be visible to the guest. I think I can prevent the host's usb driver from loading by editing (or overriding?) the matching entry in /etc/devd/usb.conf which would prevent the run(4) module from loading, but that's kind of a big hammer. Has anyone else solved a similar problem? Is there a finer-grained approach? +freebsd-usb@, +hselasky@ Hi, I also find it uncomfortable. In general, automatic kldload of needed kernel module when new device plugged in is useful feature for desktop. But it will be more useful if we could disable it without hacking configs in the /etc/devd/ every time when we updating the system. -- WBR, Andrey V. Elsukov signature.asc Description: OpenPGP digital signature
Re: USB camera not detected by 8.1, was working in 7
On 10.11.2010 3:07, Alexander Churanov wrote: I am software engineer and don't hesitate to investigate into the code if it is necessary to fix the issue. However, I basically know nothing about USB and SCSI. Tell me, if I may look into specific places in source code, perform some debugging, compare results with specifications, etc. Hi, Alexander did you try to setup quirks for your usb device with usbconfig? -- WBR, Andrey V. Elsukov signature.asc Description: OpenPGP digital signature
Re: usb/108509: [hang] FreeBSD hang at startup after ehci0 detected (CD Loader 1.2) (regression)
[EMAIL PROTECTED] wrote: To submitter: Is this still a problem with newer versions of FreeBSD (6.3 or 7.0)? If so, can you confirm that disabling EHCI or USB (from either the BIOS or laoder) allows FreeBSD to boot? I saw similar problem with 7.0 on different hardware. When I disabled USB in BIOS it worked. After upgrade to CURRENT problem gone. Also I saw another problem related to USB. Several days ago i installed 7.0 on C2D system with ICH9. System worked very slow. vmstat -i showed very hight interrupt rate for usb. After disabling USB and upgrading to CURRENT it works well now. -- WBR, Andrey V. Elsukov ___ freebsd-usb@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: usb/118571: [boot] [request] fix BTX issues when booting FreeBSD 7 from a USB CD-ROM
The following reply was made to PR usb/118571; it has been noted by GNATS. From: Andrey V. Elsukov [EMAIL PROTECTED] To: [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] Cc: Subject: Re: usb/118571: [boot] [request] fix BTX issues when booting FreeBSD 7 from a USB CD-ROM Date: Mon, 28 Jan 2008 14:11:59 +0300 Seems that Konstantin (kib@) already had a patch that you can try. Message: http://lists.freebsd.org/pipermail/freebsd-current/2007-December/081161.html Patch: http://people.freebsd.org/~kib/realbtx/realbtx.2.patch -- WBR, Andrey V. Elsukov ___ freebsd-usb@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: USB goes away
Andrey V. Elsukov wrote: Oct 1 20:26:16 btr-nb kernel: ugen0: at uhub1 port 8 (addr 2) disconnected Oct 1 20:26:16 btr-nb kernel: ugen0: detached Oct 1 20:26:19 btr-nb kernel: ugen0: vendor 0x0402 USB2.0 Camera, class 0/0, rev 2.00/1.00, addr 2 on uhub1 The hardware is MCP51 USB Controller (ohci/ehci). Forgot to mention: this is fresh CURRENT. -- WBR, Andrey V. Elsukov ___ freebsd-usb@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to [EMAIL PROTECTED]
