Date: Wed, 14 May 2008 11:07:34 -0400 From: Steve Clark <sclark@netwolves.com> To: Hans Petter Selasky <hselasky@c2i.net> Cc: freebsd-usb@freebsd.org Subject: Re: In on control endpoint Message-ID: <482B0036.5060106@netwolves.com> In-Reply-To: <200805131755.54080.hselasky@c2i.net> References: <4828AA64.7060306@netwolves.com> <200805131755.54080.hselasky@c2i.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Hans Petter Selasky wrote: > On Monday 12 May 2008, Steve Clark wrote: > >>Hello List, >> >>I have spent the afternoon going thru the usb code trying to figure out how >>to do a read on the control port (endpoint 80 ? ) instead of a write ( >>endpoint 0 ). I am still trying to emulate what the linux sierra.c usb >>serial driver does. >> >>Any pointers would be greatly appreciated. >> >>Thanks, >>Steve >>_______________________________________________ >>freebsd-usb@freebsd.org mailing list >>http://lists.freebsd.org/mailman/listinfo/freebsd-usb >>To unsubscribe, send any mail to "freebsd-usb-unsubscribe@freebsd.org" > > > Hi, > > All transactions on the control endpoint (0) consist of three parts: > > SETUP > DATA, if any > STATUS > > The two most common variants are: > > 1) SETUP > DATA OUT > STATUS IN > > 2) SETUP > DATA IN > STATUS OUT > > Rules: > > The MSB of the first byte in the SETUP decides wheter the data is OUT (0x00) > or IN (0x80). IN and OUT is relative to the USB Host. > > See: usbd_do_request and /sys/dev/usb/usb.h > > typedef struct { > uByte bmRequestType; > uByte bRequest; > uWord wValue; > uWord wIndex; > uWord wLength; > uByte bData[0]; > } __packed usb_device_request_t; > > #define UT_WRITE 0x00 > #define UT_READ 0x80 > #define UT_STANDARD 0x00 > #define UT_CLASS 0x20 > #define UT_VENDOR 0x40 > #define UT_DEVICE 0x00 > #define UT_INTERFACE 0x01 > #define UT_ENDPOINT 0x02 > #define UT_OTHER 0x03 > > --HPS > > Hi Hans, I have decided to try your new usb stack. I am running on FreeBSD 6.1. The device I am trying to get working is a sierra wireless usb 597 EVDO modem. It is also has a slot for a micro-sd memory card, plus when it is first powered up it doesn't look like a modem it looks like a different device that has both the umass device plus a cdrom device that has windows software drivers on it. In this mode it has a VID of 0x1199 and a PID of 0xfff, after a message on the control pipe it changes its identity to VID = 0x1199, PID=0x0023. What I had done previously with the original 6.1 usba.c was in the match code look for the 1199:fff and then in the attach code send a message on the control pipe to set it in modem mode like the code below: { ubsa_cfg_request( sc, 0x0b, 1); ucom->sc_dying = 1; goto error; } I tried to do something similar in the usba.c that is part of your new code as follows: { ubsa_cfg_request( sc, 0x0b, 1); goto detach; } but I immediately get a panic: Fatal trap 12 page fault in kernel mode ... Stopped at usbd_config_td_is_gone bt gives: ucom_cfg_is_gone ubsa_cfg_request ubsa_attach ... Which looks like it doesn't like the device disappearing when it switched modes. Any ideas how this should be handled? Or where I should be trying to switch the device into modem mode? Thanks, Steve
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?482B0036.5060106>