1999-11-18 18:02:44 +00:00
|
|
|
$FreeBSD$
|
|
|
|
|
|
|
|
|
|
/*
|
2000-04-03 14:47:43 +00:00
|
|
|
* Copyright (c) 1998, 1999, 2000 The NetBSD Foundation, Inc.
|
1999-11-18 18:02:44 +00:00
|
|
|
* All rights reserved.
|
|
|
|
|
*
|
|
|
|
|
* This code is derived from software contributed to The NetBSD Foundation
|
2000-05-07 19:44:16 +00:00
|
|
|
* by Lennart Augustsson (lennart@augustsson.net) at
|
1999-11-18 18:02:44 +00:00
|
|
|
* Carlstedt Research & Technology.
|
|
|
|
|
*
|
|
|
|
|
* Redistribution and use in source and binary forms, with or without
|
|
|
|
|
* modification, are permitted provided that the following conditions
|
|
|
|
|
* are met:
|
|
|
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
|
|
|
* notice, this list of conditions and the following disclaimer.
|
|
|
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
|
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
|
|
|
* documentation and/or other materials provided with the distribution.
|
|
|
|
|
* 3. All advertising materials mentioning features or use of this software
|
|
|
|
|
* must display the following acknowledgement:
|
|
|
|
|
* This product includes software developed by the NetBSD
|
|
|
|
|
* Foundation, Inc. and its contributors.
|
|
|
|
|
* 4. Neither the name of The NetBSD Foundation nor the names of its
|
|
|
|
|
* contributors may be used to endorse or promote products derived
|
|
|
|
|
* from this software without specific prior written permission.
|
|
|
|
|
*
|
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
|
|
|
|
|
* ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
|
|
|
|
|
* TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
|
|
|
|
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
|
|
|
|
|
* BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
|
|
|
|
|
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
|
|
|
|
|
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
|
|
|
|
|
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
|
|
|
|
|
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
|
|
|
|
|
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
|
|
|
|
* POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* List of known USB vendors
|
2001-02-06 14:59:03 +00:00
|
|
|
*
|
2001-09-18 22:27:41 +00:00
|
|
|
* Adding an ID and its string description for a device to the usbdevs file,
|
|
|
|
|
* enables the USB stack to print a useful description of the device that was
|
|
|
|
|
* connected.
|
|
|
|
|
*
|
|
|
|
|
* The ID should be added in usbdevs and then the files usbdevs.h and
|
|
|
|
|
* usbdevs_data.h need to be regenerated.
|
|
|
|
|
*
|
|
|
|
|
* # edit usbdevs
|
|
|
|
|
* make -f Makefile.usbdevs
|
|
|
|
|
* # test your change
|
|
|
|
|
* cd ../../modules/usb
|
|
|
|
|
* make
|
|
|
|
|
* # commit, if appropriate
|
|
|
|
|
* cvs -m "ID for device XYZ" commit usbdevs
|
|
|
|
|
* # commit the derived files after the $ FreeBSD $ has been updated.
|
|
|
|
|
* cvs commit usbdevs.h usbdevs_data.h
|
|
|
|
|
*
|
2002-02-02 21:09:54 +00:00
|
|
|
* Please note that these IDs do not do anything. Adding an ID here and
|
2001-02-06 14:59:03 +00:00
|
|
|
* regenerating the usbdevs.h and usbdevs_data.h only makes a symbolic name
|
2002-02-02 21:09:54 +00:00
|
|
|
* available to the source code and does not change any functionality, nor
|
|
|
|
|
* does it make your device available to a specific driver.
|
|
|
|
|
* It will however make the descriptive string available if a device does not
|
|
|
|
|
* provide the string itself.
|
2001-02-06 14:59:03 +00:00
|
|
|
*
|
|
|
|
|
* After adding a vendor ID VNDR and a product ID PRDCT you will have the
|
|
|
|
|
* following extra defines:
|
2001-04-02 13:02:17 +00:00
|
|
|
* #define USB_VENDOR_VNDR 0x????
|
|
|
|
|
* #define USB_PRODUCT_VNDR_PRDCT 0x????
|
2001-02-06 14:59:03 +00:00
|
|
|
*
|
2002-02-02 20:59:43 +00:00
|
|
|
* You may have to add these defines to the respective probe routines to
|
|
|
|
|
* make the device recognised by the appropriate device driver.
|
1999-11-18 18:02:44 +00:00
|
|
|
*/
|
|
|
|
|
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor AOX 0x03e8 AOX
|
|
|
|
|
vendor ATMEL 0x03eb Atmel
|
2001-07-05 09:35:46 +00:00
|
|
|
vendor MITSUMI 0x03ee Mitsumi
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor HP 0x03f0 Hewlett Packard
|
|
|
|
|
vendor ADAPTEC 0x03f3 Adaptec
|
2000-10-25 10:13:47 +00:00
|
|
|
vendor NATIONAL 0x0400 National Semiconductor
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ACERLABS 0x0402 Acer Labs
|
2000-05-07 19:44:16 +00:00
|
|
|
vendor FTDI 0x0403 Future Technology Devices
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor NEC 0x0409 NEC
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor KODAK 0x040a Eastman Kodak
|
|
|
|
|
vendor MELCO 0x0411 Melco
|
2001-01-18 00:15:49 +00:00
|
|
|
vendor CREATIVE 0x041e Creative
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ADI 0x0422 ADI Systems
|
|
|
|
|
vendor CATC 0x0423 Computer Access Technology
|
2002-03-19 10:54:40 +00:00
|
|
|
vendor SMC2 0x0424 Standard Microsystems
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor GRAVIS 0x0428 Advanced Gravis Computer Tech.
|
2000-01-10 22:31:01 +00:00
|
|
|
vendor SUN 0x0430 Sun Microsystems
|
2001-07-05 09:35:46 +00:00
|
|
|
vendor TAUGA 0x0436 Taugagreining HF
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor AMD 0x0438 Advanced Micro Devices
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor LEXMARK 0x043d Lexmark International
|
|
|
|
|
vendor NANAO 0x0440 NANAO
|
|
|
|
|
vendor ALPS 0x044e Alps Electric
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor THRUST 0x044f Thrustmaster
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
vendor TI 0x0451 Texas Instruments
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ANALOGDEVICES 0x0456 Analog Devices
|
|
|
|
|
vendor KYE 0x0458 KYE Systems
|
2000-10-10 23:23:54 +00:00
|
|
|
vendor DIAMOND2 0x045a Diamond (Supra)
|
2000-08-06 23:22:13 +00:00
|
|
|
vendor MICROSOFT 0x045e Microsoft
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor PRIMAX 0x0461 Primax Electronics
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor AMP 0x0464 AMP
|
|
|
|
|
vendor CHERRY 0x046a Cherry Mikroschalter
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor MEGATRENDS 0x046b American Megatrends
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor LOGITECH 0x046d Logitech
|
2000-04-03 00:07:43 +00:00
|
|
|
vendor BTC 0x046e Behavior Tech. Computer
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor PHILIPS 0x0471 Philips
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor CONNECTIX 0x0478 Connectix
|
2000-04-03 00:07:43 +00:00
|
|
|
vendor KENSINGTON 0x047d Kensington
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor LUCENT 0x047e Lucent
|
|
|
|
|
vendor STMICRO 0x0483 STMicroelectronics
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor YAMAHA 0x0499 YAMAHA
|
2000-07-17 09:55:34 +00:00
|
|
|
vendor COMPAQ 0x049f Compaq Computers
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ACERP 0x04a5 Acer Peripherals
|
2001-04-02 13:02:17 +00:00
|
|
|
vendor VISIONEER 0x04a7 Visioneer
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor CANON 0x04a9 Canon
|
2001-01-18 00:15:49 +00:00
|
|
|
vendor NIKON 0x04b0 Nikon
|
2000-08-26 05:06:40 +00:00
|
|
|
vendor IBM 0x04b3 IBM Corporation
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor CYPRESS 0x04b4 Cypress Semiconductor
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor EPSON 0x04b8 Seiko Epson
|
2000-10-07 11:10:53 +00:00
|
|
|
vendor RAINBOW 0x04b9 Rainbow Technologies
|
2000-04-03 00:07:43 +00:00
|
|
|
vendor IODATA 0x04bb I/O Data
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor TDK 0x04bf TDK
|
2000-01-05 04:27:24 +00:00
|
|
|
vendor 3COMUSR 0x04c1 U.S. Robotics
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor METHODE 0x04c2 Methode Electronics Far East
|
|
|
|
|
vendor MAXISWITCH 0x04c3 Maxi Switch
|
2000-10-07 11:10:53 +00:00
|
|
|
vendor LOCKHEEDMER 0x04c4 Lockheed Martin Energy Research
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor FUJITSU 0x04c5 Fujitsu
|
2000-10-07 11:10:53 +00:00
|
|
|
vendor TOSHIBAAM 0x04c6 Toshiba America Electronic Components
|
|
|
|
|
vendor MICROMACRO 0x04c7 Micro Macro Technologies
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor KONICA 0x04c8 Konica
|
|
|
|
|
vendor LITEON 0x04ca Lite-On Technology
|
|
|
|
|
vendor FUJIPHOTO 0x04cb Fuji Photo Film
|
2000-10-07 11:10:53 +00:00
|
|
|
vendor PHILIPSSEMI 0x04cc Philips Semiconductors
|
|
|
|
|
vendor TATUNG 0x04cd Tatung Co. Of America
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor SCANLOGIC 0x04ce ScanLogic
|
|
|
|
|
vendor MYSON 0x04cf Myson Technology
|
2000-10-07 11:10:53 +00:00
|
|
|
vendor DIGI2 0x04d0 Digi International
|
|
|
|
|
vendor ITTCANON 0x04d1 ITT Canon
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ALTEC 0x04d2 Altec Lansing Technologies
|
2001-04-02 13:02:17 +00:00
|
|
|
vendor PANASONIC 0x04da Panasonic (Matsushita)
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor IIYAMA 0x04e1 Iiyama
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor SHUTTLE 0x04e6 Shuttle Technology
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor ANNABOOKS 0x04ed Annabooks
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor CHICONY 0x04f2 Chicony Electronics
|
|
|
|
|
vendor BROTHER 0x04f9 Brother Industries
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor DALLAS 0x04fa Dallas Semiconductor
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ACER 0x0502 Acer
|
|
|
|
|
vendor 3COM 0x0506 3Com
|
|
|
|
|
vendor AZTECH 0x0509 Aztech Systems
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor BELKIN 0x050d Belkin Components
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor KAWATSU 0x050f Kawatsu Semiconductor
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor APC 0x051d American Power Conversion
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor CONNECTEK 0x0522 Advanced Connectek USA
|
2000-01-10 22:31:01 +00:00
|
|
|
vendor NETCHIP 0x0525 NetChip Technology
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor ALTRA 0x0527 ALTRA
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ATI 0x0528 ATI Technologies
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor AKS 0x0529 Aladdin Knowledge Systems
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor UNIACCESS 0x0540 Universal Access
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor XIRLINK 0x0545 Xirlink
|
|
|
|
|
vendor ANCHOR 0x0547 Anchor Chips
|
|
|
|
|
vendor SONY 0x054c Sony
|
|
|
|
|
vendor VISION 0x0553 VLSI Vision
|
|
|
|
|
vendor ASAHIKASEI 0x0556 Asahi Kasei Microsystems
|
|
|
|
|
vendor ATEN 0x0557 ATEN International
|
|
|
|
|
vendor MUSTEK 0x055f Mustek Systems
|
|
|
|
|
vendor TELEX 0x0562 Telex Communications
|
|
|
|
|
vendor PERACOM 0x0565 Peracom Networks
|
|
|
|
|
vendor ALCOR2 0x0566 Alcor Micro
|
|
|
|
|
vendor WACOM 0x056a WACOM
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor ETEK 0x056c e-TEK Labs
|
|
|
|
|
vendor EIZO 0x056d EIZO
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ELECOM 0x056e Elecom
|
2001-04-02 13:02:17 +00:00
|
|
|
vendor HAUPPAUGE 0x0573 Hauppauge Computer Works
|
2000-03-20 19:49:20 +00:00
|
|
|
vendor BAFO 0x0576 BAFO/Quality Computer Accessories
|
2000-03-15 21:53:09 +00:00
|
|
|
vendor YEDATA 0x057b Y-E Data
|
2000-03-20 19:49:20 +00:00
|
|
|
vendor AVM 0x057c AVM GmbH
|
2000-04-03 00:07:43 +00:00
|
|
|
vendor QUICKSHOT 0x057f Quickshot
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ROLAND 0x0582 Roland
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor ROCKFIRE 0x0583 Rockfire
|
2002-04-06 13:01:39 +00:00
|
|
|
vendor RATOC 0x0584 RATOC Systems, Inc.
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ZYXEL 0x0586 ZyXEL Communication
|
|
|
|
|
vendor ALCOR 0x058f Alcor Micro
|
|
|
|
|
vendor IOMEGA 0x059b Iomega
|
|
|
|
|
vendor ATREND 0x059c A-Trend Technology
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor AID 0x059d Advanced Input Devices
|
2000-10-07 11:10:53 +00:00
|
|
|
vendor LACIE 0x059f LaCie
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor OMNIVISION 0x05a9 OmniVision
|
|
|
|
|
vendor INSYSTEM 0x05ab In-System Design
|
|
|
|
|
vendor APPLE 0x05ac Apple Computer
|
2000-04-03 00:07:43 +00:00
|
|
|
vendor DIGI 0x05c5 Digi International
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor QTRONIX 0x05c7 Qtronix
|
|
|
|
|
vendor ELSA 0x05cc ELSA
|
2000-03-20 19:49:20 +00:00
|
|
|
vendor BRAINBOXES 0x05d1 Brainboxes Limited
|
2001-08-21 21:35:40 +00:00
|
|
|
vendor ULTIMA 0x05d8 Ultima
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor AXIOHM 0x05d9 Axiohm Transaction Solutions
|
2001-04-02 13:02:17 +00:00
|
|
|
vendor MICROTEK 0x05da Microtek
|
2002-03-18 18:23:42 +00:00
|
|
|
vendor SUNTAC 0x05db SUN Corporation
|
2001-04-17 15:34:22 +00:00
|
|
|
vendor LEXAR 0x05dc Lexar Media
|
2000-10-07 11:10:53 +00:00
|
|
|
vendor SYMBOL 0x05e0 Symbol Technologies
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor GENESYS 0x05e3 Genesys Logic
|
|
|
|
|
vendor FUJI 0x05e5 Fuji Electric
|
2000-10-07 11:10:53 +00:00
|
|
|
vendor KEITHLEY 0x05e6 Keithley Instruments
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor EIZONANAO 0x05e7 EIZO Nanao
|
2000-01-05 18:25:53 +00:00
|
|
|
vendor KLSI 0x05e9 Kawasaki LSI
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor FFC 0x05eb FFC
|
|
|
|
|
vendor ANKO 0x05ef Anko Electronic
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
vendor PIENGINEERING 0x05f3 P.I. Engineering
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor AOC 0x05f6 AOC International
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor CHIC 0x05fe Chic Technology
|
2000-03-20 19:49:20 +00:00
|
|
|
vendor BARCO 0x0600 Barco Display Systems
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor BRIDGE 0x0607 Bridge Information
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
vendor SOLIDYEAR 0x060b Solid Year
|
2000-03-20 19:49:20 +00:00
|
|
|
vendor BIORAD 0x0614 Bio-Rad Laboratories
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor MACALLY 0x0618 Macally
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ACTLABS 0x061c Act Labs
|
|
|
|
|
vendor ALARIS 0x0620 Alaris
|
|
|
|
|
vendor APEX 0x0624 Apex
|
|
|
|
|
vendor AVISION 0x0638 Avision
|
2000-07-17 09:55:34 +00:00
|
|
|
vendor TEAC 0x0644 TEAC
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor LINKSYS 0x066b Linksys
|
|
|
|
|
vendor ACERSA 0x066e Acer Semiconductor America
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor SIGMATEL 0x066f Sigmatel
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor AIWA 0x0677 Aiwa
|
|
|
|
|
vendor ACARD 0x0678 ACARD Technology
|
|
|
|
|
vendor PROLIFIC 0x067b Prolific Technology
|
2002-03-31 21:38:05 +00:00
|
|
|
vendor SIEMENS 0x067c Siemens
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ADVANCELOGIC 0x0680 Avance Logic
|
2001-07-05 09:35:46 +00:00
|
|
|
vendor HAGIWARA 0x0693 Hagiwara Sys-Com
|
2002-04-06 12:27:23 +00:00
|
|
|
vendor MINOLTA 0x0686 Minolta
|
2000-07-17 09:55:34 +00:00
|
|
|
vendor CTX 0x0698 Chuntex
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ASKEY 0x069a Askey Computer
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor SAITEK 0x06a3 Saitek
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor ALCATELT 0x06b9 Alcatel Telecom
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor AGFA 0x06bd AGFA-Gevaert
|
|
|
|
|
vendor ASIAMD 0x06be Asia Microelectronic Development
|
|
|
|
|
vendor BIZLINK 0x06c4 Bizlink International
|
2001-04-02 13:02:17 +00:00
|
|
|
vendor KEYSPAN 0x06cd Keyspan
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor AASHIMA 0x06d6 Aashima Technology
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor MULTITECH 0x06e0 MultiTech
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ADS 0x06e1 ADS Technologies
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor ALCATELM 0x06e4 Alcatel Microelectronics
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
vendor SIRIUS 0x06ea Sirius Technologies
|
2000-03-20 19:49:20 +00:00
|
|
|
vendor BOSTON 0x06fd Boston Acoustics
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor SMC 0x0707 Standard Microsystems
|
|
|
|
|
vendor PUTERCOM 0x0708 Putercom
|
|
|
|
|
vendor MCT 0x0711 MCT
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor DIGITALSTREAM 0x074e Digital Stream
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor AUREAL 0x0755 Aureal Semiconductor
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor MIDIMAN 0x0763 Midiman
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor GRIFFIN 0x077d Griffin Technology
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor SANDISK 0x0781 SanDisk Corp
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor BRIMAX 0x078e Brimax
|
|
|
|
|
vendor AXIS 0x0792 Axis Communications
|
|
|
|
|
vendor ABL 0x0794 ABL Electronics
|
|
|
|
|
vendor ALFADATA 0x079d Alfadata Computer
|
2000-10-07 11:10:53 +00:00
|
|
|
vendor NATIONALTECH 0x07a2 National Technical Systems
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ONNTO 0x07a3 Onnto
|
|
|
|
|
vendor BE 0x07a4 Be
|
|
|
|
|
vendor ADMTEK 0x07a6 ADMtek
|
2000-01-13 20:13:58 +00:00
|
|
|
vendor COREGA 0x07aa Corega
|
2000-04-03 14:47:43 +00:00
|
|
|
vendor FREECOM 0x07ab Freecom
|
2000-10-07 11:10:53 +00:00
|
|
|
vendor MICROTECH 0x07af Microtech
|
2001-10-23 22:44:03 +00:00
|
|
|
vendor OLYMPUS 0x07b4 Olympus
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ABOCOM 0x07b8 AboCom Systems
|
|
|
|
|
vendor KEISOKUGIKEN 0x07c1 Keisokugiken
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor APG 0x07c5 APG Cash Drawer
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor BUG 0x07c8 B.U.G.
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor ALLIEDTELESYN 0x07c9 Allied Telesyn International
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor AVERMEDIA 0x07ca AVerMedia Technologies
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor SIIG 0x07cc SIIG
|
2001-04-02 13:02:17 +00:00
|
|
|
vendor CASIO 0x07cf CASIO
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor APTIO 0x07d2 Aptio Products
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor ARASAN 0x07da Arasan Chip Systems
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ALLIEDCABLE 0x07e6 Allied Cable
|
2001-01-03 11:45:03 +00:00
|
|
|
vendor STSN 0x07ef STSN
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ZOOM 0x0803 Zoom Telephonics
|
|
|
|
|
vendor BROADLOGIC 0x0827 BroadLogic
|
|
|
|
|
vendor HANDSPRING 0x082d Handspring
|
|
|
|
|
vendor ACTIONSTAR 0x0835 Action Star Enterprise
|
|
|
|
|
vendor PALM 0x0830 Palm Computing
|
|
|
|
|
vendor ACCTON 0x083a Accton Technology
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
vendor DIAMOND 0x0841 Diamond
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor NETGEAR 0x0846 BayNETGEAR
|
|
|
|
|
vendor ACTIVEWIRE 0x0854 ActiveWire
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor PORTGEAR 0x085a PortGear
|
2000-10-25 10:13:47 +00:00
|
|
|
vendor METRICOM 0x0870 Metricom
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ADESSOKBTEK 0x087c ADESSO/Kbtek America
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor JATON 0x087d Jaton
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor APT 0x0880 APT Technologies
|
|
|
|
|
vendor BOCARESEARCH 0x0885 Boca Research
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor ANDREA 0x08a8 Andrea Electronics
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor BURRBROWN 0x08bb Burr-Brown Japan
|
|
|
|
|
vendor 2WIRE 0x08c8 2Wire
|
|
|
|
|
vendor AIPTEK 0x08ca AIPTEK International
|
|
|
|
|
vendor SMARTBRIDGES 0x08d1 SmartBridges
|
|
|
|
|
vendor BILLIONTON 0x08dd Billionton Systems
|
|
|
|
|
vendor EXTENDED 0x08e9 Extended Systems
|
2001-04-02 13:02:17 +00:00
|
|
|
vendor MSYSTEMS 0x08ec M-Systems
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor AUTHENTEC 0x08ff AuthenTec
|
|
|
|
|
vendor ALATION 0x0910 Alation Systems
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor BIOMETRIC 0x0929 American Biometric Company
|
2000-10-05 23:09:48 +00:00
|
|
|
vendor YANO 0x094f Yano
|
2000-10-10 23:23:54 +00:00
|
|
|
vendor KINGSTON 0x0951 Kingston Technology
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor BLUEWATER 0x0956 BlueWater Systems
|
|
|
|
|
vendor AGILENT 0x0957 Agilent Technologies
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor PORTSMITH 0x095a Portsmith
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor ADIRONDACK 0x0976 Adirondack Wire & Cable
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor BECKHOFF 0x0978 Beckhoff
|
|
|
|
|
vendor ALTIUS 0x09b3 Altius Solutions
|
|
|
|
|
vendor ARRIS 0x09c1 Arris Interactive
|
|
|
|
|
vendor ACTIVCARD 0x09c3 ACTIVCARD
|
|
|
|
|
vendor ACTISYS 0x09c4 ACTiSYS
|
|
|
|
|
vendor AFOURTECH 0x09da A-FOUR TECH
|
|
|
|
|
vendor AIMEX 0x09dc AIMEX
|
|
|
|
|
vendor ADDONICS 0x09df Addonics Technologies
|
|
|
|
|
vendor AKAI 0x09e8 AKAI professional M.I.
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor ARESCOM 0x09f5 ARESCOM
|
2000-03-20 19:49:20 +00:00
|
|
|
vendor BAY 0x09f9 Bay Associates
|
2000-03-20 18:30:04 +00:00
|
|
|
vendor ALTERA 0x09fb Altera
|
2000-10-07 11:10:53 +00:00
|
|
|
vendor TREK 0x0a16 Trek Technology
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ASAHIOPTICAL 0x0a17 Asahi Optical
|
|
|
|
|
vendor BOCASYSTEMS 0x0a43 Boca Systems
|
|
|
|
|
vendor BROADCOM 0x0a5c Broadcom
|
|
|
|
|
vendor GREENHOUSE 0x0a6b GREENHOUSE
|
2000-07-17 09:55:34 +00:00
|
|
|
vendor GEOCAST 0x0a79 Geocast Network Systems
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor TODOS 0x0b0c Todos Data System
|
|
|
|
|
vendor NEC2 0x0b62 NEC
|
|
|
|
|
vendor ATI2 0x0b6f ATI
|
2001-04-02 13:11:59 +00:00
|
|
|
vendor AGATE 0x0c08 Agate Technologies
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor DMI 0x0c0b DMI
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor MOTOROLA 0x1063 Motorola
|
|
|
|
|
vendor PLX 0x10b5 PLX
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor ASANTE 0x10bd Asante
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor BELKIN2 0x1293 Belkin Components
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor MOBILITY 0x1342 Mobility
|
|
|
|
|
vendor SHARK 0x13d2 Shark
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor SILICONPORTALS 0x1527 Silicon Portals
|
2002-02-02 21:02:13 +00:00
|
|
|
vendor SOHOWARE 0x15e8 SOHOware
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor UMAX 0x1606 UMAX Data Systems
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor INSIDEOUT 0x1608 Inside Out Networks
|
|
|
|
|
vendor ENTREGA 0x1645 Entrega
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor ACTIONTEC 0x1668 Actiontec Electronics
|
|
|
|
|
vendor DLINK 0x2001 D-Link
|
2002-01-31 11:39:17 +00:00
|
|
|
vendor DAISY 0x3579 Daisy Technology
|
1999-11-18 18:02:44 +00:00
|
|
|
vendor INTEL 0x8086 Intel
|
2002-02-02 20:59:43 +00:00
|
|
|
vendor HP2 0xf003 Hewlett Packard
|
1999-11-18 18:02:44 +00:00
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* List of known products. Grouped by vendor.
|
|
|
|
|
*/
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* 3Com products */
|
|
|
|
|
product 3COM HOMECONN 0x009d HomeConnect USB Camera
|
|
|
|
|
product 3COM 3C19250 0x03E8 3C19250 Ethernet adapter
|
2001-04-30 17:37:11 +00:00
|
|
|
product 3COM USR56K 0x3021 U.S.Robotics 56000 Voice Faxmodem Pro
|
2002-04-06 13:01:39 +00:00
|
|
|
product 3COM 3C460 0x11f8 HomeConnect 3C460
|
|
|
|
|
product 3COM 3C460B 0x4601 HomeConnect 3C460B
|
2000-04-03 00:07:43 +00:00
|
|
|
|
|
|
|
|
product 3COMUSR OFFICECONN 0x0082 3Com OfficeConnect Analog Modem
|
|
|
|
|
product 3COMUSR USRISDN 0x008f 3Com U.S. Robotics Pro ISDN TA
|
|
|
|
|
product 3COMUSR HOMECONN 0x009d 3Com HomeConnect camera
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product 3COMUSR USR56K 0x3021 U.S.Robotics 56000 Voice Faxmodem Pro
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2001-02-05 11:02:50 +00:00
|
|
|
/* AboCom products */
|
2002-03-31 21:38:05 +00:00
|
|
|
product ABOCOM XX1 0x110c XX1
|
|
|
|
|
product ABOCOM XX2 0x200c XX2
|
2001-02-05 11:02:50 +00:00
|
|
|
product ABOCOM URE450 0x4000 URE450 Ethernet Adapter
|
2002-04-06 13:01:39 +00:00
|
|
|
product ABOCOM UFE1000 0x4002 UFE1000 Fast Ethernet Adapter
|
2001-07-05 09:35:46 +00:00
|
|
|
product ABOCOM DSB650TX_PNA 0x4003 1/10/100 ethernet adapter
|
2002-03-31 21:38:05 +00:00
|
|
|
product ABOCOM XX4 0x4004 XX4
|
|
|
|
|
product ABOCOM XX5 0x4007 XX5
|
|
|
|
|
product ABOCOM XX6 0x400b XX6
|
|
|
|
|
product ABOCOM XX7 0x400c XX7
|
|
|
|
|
product ABOCOM XX8 0x4102 XX8
|
|
|
|
|
product ABOCOM XX9 0x4104 XX9
|
|
|
|
|
product ABOCOM XX10 0xabc1 XX10
|
2000-04-03 00:07:43 +00:00
|
|
|
|
2002-02-02 21:09:54 +00:00
|
|
|
/* Accton products */
|
2001-09-09 03:07:15 +00:00
|
|
|
product ACCTON USB320_EC 0x1046 USB320-EC Ethernet Adapter
|
2002-04-06 13:01:39 +00:00
|
|
|
product ACCTON SS1001 0x5046 SpeedStream Ethernet Adapter
|
2001-09-09 03:07:15 +00:00
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* Acer Peripherals, Inc. products */
|
2000-03-20 18:30:04 +00:00
|
|
|
product ACERP ACERSCAN_C310U 0x12a6 Acerscan C310U
|
2000-10-07 11:10:53 +00:00
|
|
|
product ACERP ACERSCAN_320U 0x2022 Acerscan 320U
|
2001-04-02 13:02:17 +00:00
|
|
|
product ACERP ACERSCAN_640U 0x2040 Acerscan 640U
|
2000-09-05 12:24:05 +00:00
|
|
|
product ACERP ACERSCAN_620U 0x2060 Acerscan 620U
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* ActiveWire, Inc. products */
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product ACTIVEWIRE IOBOARD 0x0100 I/O Board
|
|
|
|
|
product ACTIVEWIRE IOBOARD_FW1 0x0101 I/O Board, rev. 1 firmware
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* ADMtek products */
|
|
|
|
|
product ADMTEK PEGASUS 0x0986 AN986 USB Ethernet adapter
|
2002-03-31 21:38:05 +00:00
|
|
|
product ADMTEK PEGASUSII 0x8511 AN8511 USB Ethernet adapter
|
This commit adds device driver support for the ADMtek AN986 Pegasus
USB ethernet chip. Adapters that use this chip include the LinkSys
USB100TX. There are a few others, but I'm not certain of their
availability in the U.S. I used an ADMtek eval board for development.
Note that while the ADMtek chip is a 100Mbps device, you can't really
get 100Mbps speeds over USB. Regardless, this driver uses miibus to
allow speed and duplex mode selection as well as autonegotiation.
Building and kldloading the driver as a module is also supported.
Note that in order to make this driver work, I had to make what some
may consider an ugly hack to sys/dev/usb/usbdi.c. The usbd_transfer()
function will use tsleep() for synchronous transfers that don't complete
right away. This is a problem since there are times when we need to
do sync transfers from an interrupt context (i.e. when reading registers
from the MAC via the control endpoint), where tsleep() us a no-no.
My hack allows the driver to have the code poll for transfer completion
subject to the xfer->timeout timeout rather that calling tsleep().
This hack is controlled by a quirk entry and is only enabled for the
ADMtek device.
Now, I'm sure there are a few of you out there ready to jump on me
and suggest some other approach that doesn't involve a busy wait. The
only solution that might work is to handle the interrupts in a kernel
thread, where you may have something resembling a process context that
makes it okay to tsleep(). This is lovely, except we don't have any
mechanism like that now, and I'm not about to implement such a thing
myself since it's beyond the scope of driver development. (Translation:
I'll be damned if I know how to do it.) If FreeBSD ever aquires such
a mechanism, I'll be glad to revisit the driver to take advantage of
it. In the meantime, I settled for what I perceived to be the solution
that involved the least amount of code changes. In general, the hit
is pretty light.
Also note that my only USB test box has a UHCI controller: I haven't
I don't have a machine with an OHCI controller available.
Highlights:
- Updated usb_quirks.* to add UQ_NO_TSLEEP quirk for ADMtek part.
- Updated usbdevs and regenerated generated files
- Updated HARDWARE.TXT and RELNOTES.TXT files
- Updated sysinstall/device.c and userconfig.c
- Updated kernel configs -- device aue0 is commented out by default
- Updated /sys/conf/files
- Added new kld module directory
1999-12-28 02:01:18 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* ADS products */
|
|
|
|
|
product ADS UBS10BT 0x0008 UBS-10BT Ethernet adapter
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2001-04-02 13:11:59 +00:00
|
|
|
/* Agate Technologies products */
|
|
|
|
|
product AGATE QDRIVE 0x0378 Q-Drive
|
|
|
|
|
|
2000-09-05 12:24:05 +00:00
|
|
|
/* AGFA products */
|
|
|
|
|
product AGFA SNAPSCAN1212U 0x0001 SnapScan 1212U
|
2001-12-10 11:40:38 +00:00
|
|
|
product AGFA SNAPSCAN1236U 0x0002 SnapScan 1236U
|
2000-09-05 12:24:05 +00:00
|
|
|
product AGFA SNAPSCANTOUCH 0x0100 SnapScan Touch
|
2000-10-07 11:10:53 +00:00
|
|
|
product AGFA SNAPSCAN1212U2 0x2061 SnapScan 1212U
|
2001-12-10 11:40:38 +00:00
|
|
|
product AGFA SNAPSCANE40 0x208d SnapScan e40
|
|
|
|
|
product AGFA SNAPSCANE50 0x208f SnapScan e50
|
|
|
|
|
product AGFA SNAPSCANE20 0x2091 SnapScan e20
|
|
|
|
|
product AGFA SNAPSCANE25 0x2095 SnapScan e25
|
|
|
|
|
product AGFA SNAPSCANE26 0x2097 SnapScan e26
|
|
|
|
|
product AGFA SNAPSCANE52 0x20fd SnapScan e52
|
2000-09-05 12:24:05 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* AKS products */
|
|
|
|
|
product AKS USBHASP 0x0001 USB-HASP 0.06
|
2000-01-10 22:31:01 +00:00
|
|
|
|
2000-07-17 09:55:34 +00:00
|
|
|
/* Alcor Micro, Inc. products */
|
|
|
|
|
product ALCOR2 KBD_HUB 0x2802 Kbd Hub
|
|
|
|
|
|
|
|
|
|
product ALCOR MA_KBD_HUB 0x9213 MacAlly Kbd Hub
|
|
|
|
|
product ALCOR AU9814 0x9215 AU9814 Hub
|
|
|
|
|
product ALCOR SM_KBD 0x9410 MicroConnectors/StrongMan Keyboard
|
|
|
|
|
product ALCOR NEC_KBD_HUB 0x9472 NEC Kbd Hub
|
2000-05-09 20:43:37 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Altec Lansing products */
|
2000-04-03 00:07:43 +00:00
|
|
|
product ALTEC ADA70 0x0070 ADA70 Speakers
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product ALTEC ASC495 0xff05 ASC495 Speakers
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* American Power Conversion products */
|
|
|
|
|
product APC UPSPRO500 0x0002 Back-UPS Pro 500
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Anchor products */
|
|
|
|
|
product ANCHOR EZUSB 0x2131 EZUSB
|
2000-05-01 20:26:12 +00:00
|
|
|
product ANCHOR EZLINK 0x2720 EZLINK
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* AOX, Inc. products */
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product AOX USB101 0x0008 USB ethernet controller engine
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2000-10-07 11:10:53 +00:00
|
|
|
/* Apple Computer products */
|
2001-01-03 11:45:03 +00:00
|
|
|
product APPLE OPTMOUSE 0x0302 Optical mouse
|
2000-10-07 11:10:53 +00:00
|
|
|
product APPLE SPEAKERS 0x1101 Speakers
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* ATen products */
|
|
|
|
|
product ATEN UC1284 0x2001 Parallel printer adapter
|
|
|
|
|
product ATEN UC10T 0x2002 10Mbps ethernet adapter
|
2000-08-06 23:22:13 +00:00
|
|
|
product ATEN UC232A 0x2008 Serial adapter
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2001-01-03 11:45:03 +00:00
|
|
|
/* Atmel Comp. products */
|
|
|
|
|
product ATMEL UHB124 0x3301 UHB124 hub
|
2001-04-02 13:02:17 +00:00
|
|
|
product ATMEL DWL120 0x7602 DWL-120 Wireless adapter
|
2001-01-03 11:45:03 +00:00
|
|
|
|
2000-09-05 12:24:05 +00:00
|
|
|
/* Avision products */
|
|
|
|
|
product AVISION 1200U 0x0268 1200U scanner
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Belkin products */
|
|
|
|
|
/*product BELKIN F5U111 0x???? F5U111 Ethernet adapter*/
|
2001-04-02 13:02:17 +00:00
|
|
|
product BELKIN2 F5U002 0x0002 F5U002 Parallel printer adapter
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Billionton products */
|
|
|
|
|
product BILLIONTON USB100 0x0986 USB100N 10/100 FastEthernet Adapter
|
2002-03-31 21:38:05 +00:00
|
|
|
product BILLIONTON USBLP100 0x0987 USB100LP
|
|
|
|
|
product BILLIONTON USBEL100 0x0988 USB100EL
|
|
|
|
|
product BILLIONTON USBE100 0x8511 USBE100
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
|
|
|
|
|
/* Brother Industries products */
|
|
|
|
|
product BROTHER HL1050 0x0002 HL-1050 laser printer
|
1999-11-18 18:02:44 +00:00
|
|
|
|
|
|
|
|
/* Behavior Technology Computer products */
|
|
|
|
|
product BTC BTC7932 0x6782 Keyboard with mouse port
|
|
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* Canon, Inc. products */
|
2001-04-02 13:02:17 +00:00
|
|
|
product CANON N656U 0x2206 CANOSCAN N656U
|
2001-04-12 10:59:30 +00:00
|
|
|
product CANON S10 0x3041 PowerShot S10
|
|
|
|
|
product CANON S100 0x3045 PowerShot S100
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* CATC products */
|
|
|
|
|
product CATC NETMATE 0x000a Netmate ethernet adapter
|
|
|
|
|
product CATC NETMATE2 0x000c Netmate2 ethernet adapter
|
|
|
|
|
product CATC CHIEF 0x000d USB Chief Bus & Protocol Analyzer
|
|
|
|
|
product CATC ANDROMEDA 0x1237 Andromeda hub
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2001-04-02 13:02:17 +00:00
|
|
|
/* CASIO products */
|
|
|
|
|
product CASIO NAMELAND 0x4001 CASIO Nameland EZ-USB
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Cherry products */
|
|
|
|
|
product CHERRY MY3000KBD 0x0001 My3000 keyboard
|
|
|
|
|
product CHERRY MY3000HUB 0x0003 My3000 hub
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Chic Technology products */
|
|
|
|
|
product CHIC MOUSE1 0x0001 mouse
|
2000-04-03 00:07:43 +00:00
|
|
|
product CHIC CYPRESS 0x0003 Cypress USB Mouse
|
1999-11-18 18:02:44 +00:00
|
|
|
|
|
|
|
|
/* Chicony products */
|
|
|
|
|
product CHICONY KB8933 0x0001 KB-8933 keyboard
|
|
|
|
|
|
2000-07-17 09:55:34 +00:00
|
|
|
/* Compaq products */
|
|
|
|
|
product COMPAQ PJB100 0x504a Personal Jukebox PJB100
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Connectix products */
|
|
|
|
|
product CONNECTIX QUICKCAM 0x0001 QuickCam
|
2000-01-05 04:27:24 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Corega products */
|
|
|
|
|
product COREGA ETHER_USB_T 0x0001 Ether USB-T
|
2000-03-09 16:28:58 +00:00
|
|
|
product COREGA FETHER_USB_TX 0x0004 FEther USB-TX
|
2002-04-06 13:01:39 +00:00
|
|
|
product COREGA FETHER_USB_TXS 0x000d FEther USB-TXS
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2001-01-18 00:15:49 +00:00
|
|
|
/* Creative products */
|
2001-02-05 11:00:13 +00:00
|
|
|
product CREATIVE NOMAD_II 0x1002 Nomad II MP3 player
|
2001-01-18 00:15:49 +00:00
|
|
|
|
2000-07-17 09:55:34 +00:00
|
|
|
/* CTX products */
|
|
|
|
|
product CTX EX1300 0x9999 Ex1300 hub
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Cypress Semiconductor products */
|
|
|
|
|
product CYPRESS MOUSE 0x0001 mouse
|
|
|
|
|
product CYPRESS THERMO 0x0002 thermometer
|
2000-05-07 19:44:16 +00:00
|
|
|
product CYPRESS FMRADIO 0x1002 FM Radio
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2002-01-31 11:39:17 +00:00
|
|
|
/* Daisy Technology products */
|
2002-05-20 12:01:58 +00:00
|
|
|
product DAISY DMC 0x6901 USB MultiMedia Reader
|
2002-01-31 11:39:17 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Dallas Semiconductor products */
|
|
|
|
|
product DALLAS J6502 0x4201 J-6502 speakers
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Diamond products */
|
|
|
|
|
product DIAMOND RIO500USB 0x0001 Rio 500 USB
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* Digi International products */
|
|
|
|
|
product DIGI ACCELEPORT2 0x0002 AccelePort USB 2
|
|
|
|
|
product DIGI ACCELEPORT4 0x0004 AccelePort USB 4
|
|
|
|
|
product DIGI ACCELEPORT8 0x0008 AccelePort USB 8
|
|
|
|
|
|
2002-04-06 13:01:39 +00:00
|
|
|
/* D-Link products */
|
|
|
|
|
/*product DLINK DSBS25 0x0100 DSB-S25 serial adapter*/
|
|
|
|
|
product DLINK DSB650TX4 0x200c 10/100 ethernet adapter
|
|
|
|
|
product DLINK DSB650C 0x4000 10Mbps ethernet adapter
|
|
|
|
|
product DLINK DSB650TX1 0x4001 10/100 ethernet adapter
|
|
|
|
|
product DLINK DSB650TX 0x4002 10/100 ethernet adapter
|
|
|
|
|
product DLINK DSB650TX_PNA 0x4003 1/10/100 ethernet adapter
|
|
|
|
|
product DLINK DSB650TX3 0x400b 10/100 ethernet adapter
|
|
|
|
|
product DLINK DSB650TX2 0x4102 10/100 ethernet adapter
|
|
|
|
|
product DLINK DSB650 0xabc1 10/100 ethernet adapter
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* EIZO products */
|
|
|
|
|
product EIZO HUB 0x0000 hub
|
|
|
|
|
product EIZO MONITOR 0x0001 monitor
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2002-04-06 13:01:39 +00:00
|
|
|
/* Elecom products */
|
|
|
|
|
product ELECOM MOUSE29UO 0x0002 mouse 29UO
|
|
|
|
|
product ELECOM LDUSBTX0 0x200c LD-USB/TX
|
|
|
|
|
product ELECOM LDUSBTX1 0x4002 LD-USB/TX
|
2002-05-27 00:00:48 +00:00
|
|
|
product ELECOM LDUSBLTX 0x4005 LD-USBL/TX
|
2002-04-06 13:01:39 +00:00
|
|
|
product ELECOM LDUSBTX2 0x400b LD-USB/TX
|
|
|
|
|
product ELECOM LDUSBTX3 0xabc1 LD-USB/TX
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Elsa products */
|
|
|
|
|
product ELSA MODEM1 0x2265 ELSA Modem Board
|
2002-03-31 21:38:05 +00:00
|
|
|
product ELSA USB2ETHERNET 0x3000 Microlink USB2Ethernet
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Entrega products */
|
|
|
|
|
product ENTREGA 1S 0x0001 1S serial connector
|
|
|
|
|
product ENTREGA 2S 0x0002 2S serial connector
|
|
|
|
|
product ENTREGA 1S25 0x0003 1S25 serial connector
|
|
|
|
|
product ENTREGA 4S 0x0004 4S serial connector
|
|
|
|
|
product ENTREGA E45 0x0005 E45 Ethernet adapter
|
|
|
|
|
product ENTREGA CENTRONICS 0x0006 Centronics connector
|
|
|
|
|
product ENTREGA 1S9 0x0093 1S9 serial connector
|
|
|
|
|
product ENTREGA EZUSB 0x8000 EZ-USB
|
|
|
|
|
/*product ENTREGA SERIAL 0x8001 DB25 Serial connector*/
|
2000-10-07 11:10:53 +00:00
|
|
|
product ENTREGA 2U4S 0x8004 2U4S serial connector/usb hub
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/*product ENTREGA SERIAL_DB9 0x8093 DB9 Serial connector*/
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Epson products */
|
2000-10-07 11:10:53 +00:00
|
|
|
product EPSON PRINTER1 0x0001 USB Printer
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product EPSON PRINTER2 0x0002 ISD USB Smart Cable for Mac
|
|
|
|
|
product EPSON PRINTER3 0x0003 ISD USB Smart Cable
|
2001-01-03 11:45:03 +00:00
|
|
|
product EPSON PRINTER5 0x0005 USB Printer
|
2000-09-05 12:24:05 +00:00
|
|
|
product EPSON 636 0x0101 Perfection 636U / 636Photo scanner
|
|
|
|
|
product EPSON 610 0x0103 Perfection 610 scanner
|
|
|
|
|
product EPSON 1200 0x0104 Perfection 1200U / 1200Photo scanner
|
2001-01-03 11:45:03 +00:00
|
|
|
product EPSON 1600 0x0107 Expression 1600 scanner
|
2001-07-05 09:35:46 +00:00
|
|
|
product EPSON 1640 0x010a Perfection 1640SU scanner
|
|
|
|
|
product EPSON 1240 0x010b Perfection 1240U / 1240Photo scanner
|
2002-02-11 03:35:53 +00:00
|
|
|
product EPSON 640U 0x010c Perfection 640U scanner
|
|
|
|
|
product EPSON 1650 0x0110 Perfection 1650 scanner
|
|
|
|
|
product EPSON GT9700F 0x0112 GT-9700F scanner
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2000-07-17 09:55:34 +00:00
|
|
|
/* e-TEK Labs products */
|
|
|
|
|
product ETEK 1COM 0x8007 Serial port
|
|
|
|
|
|
|
|
|
|
/* Extended Systems products */
|
|
|
|
|
product EXTENDED XTNDACCESS 0x0100 XTNDAccess IrDA
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Gravis products */
|
|
|
|
|
product GRAVIS GAMEPADPRO 0x4001 GamePad Pro
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2001-04-02 13:02:17 +00:00
|
|
|
/* GREENHOUSE products */
|
|
|
|
|
product GREENHOUSE KANA21 0x0001 CF-writer with Portable MP3 Player
|
|
|
|
|
|
2000-10-07 11:10:53 +00:00
|
|
|
/* Griffin Technology */
|
2001-02-05 11:02:50 +00:00
|
|
|
product GRIFFIN IMATE 0x0405 iMate, ADB adapter
|
2000-10-07 11:10:53 +00:00
|
|
|
|
2000-04-03 14:47:43 +00:00
|
|
|
/* Freecom products */
|
|
|
|
|
product FREECOM DVD 0xfc01 Connector for DVD drive
|
|
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* Future Technology Devices products */
|
2000-05-07 19:44:16 +00:00
|
|
|
product FTDI SERIAL 0x8372 Serial converter
|
2000-04-03 00:07:43 +00:00
|
|
|
|
2000-10-25 10:13:47 +00:00
|
|
|
/* Fuji photo products */
|
|
|
|
|
product FUJIPHOTO MASS0100 0x0100 Mass Storage
|
|
|
|
|
|
2001-07-05 09:35:46 +00:00
|
|
|
/* Hagiwara products */
|
|
|
|
|
product HAGIWARA FGSM 0x0002 FlashGate SmartMedia Card Reader
|
|
|
|
|
product HAGIWARA FGCF 0x0003 FlashGate CompactFlash Card Reader
|
|
|
|
|
product HAGIWARA FG 0x0005 FlashGate
|
|
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* Handspring, Inc. */
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product HANDSPRING VISOR 0x0100 Handspring Visor
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2001-04-02 13:02:17 +00:00
|
|
|
/* Hauppauge Computer Works */
|
|
|
|
|
product HAUPPAUGE WINTV_USB_FM 0x4d12 WinTV USB FM
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* HP products */
|
2001-07-05 09:35:46 +00:00
|
|
|
product HP 895C 0x0004 DeskJet 895C
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product HP 4100C 0x0101 Scanjet 4100C
|
2000-10-25 10:13:47 +00:00
|
|
|
product HP S20 0x0102 Photosmart S20
|
2001-07-05 09:35:46 +00:00
|
|
|
product HP 880C 0x0104 DeskJet 880C
|
2000-04-03 00:07:43 +00:00
|
|
|
product HP 4200C 0x0105 ScanJet 4200C
|
2001-07-05 09:35:46 +00:00
|
|
|
product HP CDWRITERPLUS 0x0107 CD-Writer Plus
|
|
|
|
|
product HP KBDHUB 0x010c Multimedia Keyboard Hub
|
2000-04-03 00:07:43 +00:00
|
|
|
product HP 6200C 0x0201 ScanJet 6200C
|
|
|
|
|
product HP S20b 0x0202 PhotoSmart S20
|
2001-07-05 09:35:46 +00:00
|
|
|
product HP 815C 0x0204 DeskJet 815C
|
2000-04-03 00:07:43 +00:00
|
|
|
product HP 3300C 0x0205 ScanJet 3300C
|
2001-07-05 09:35:46 +00:00
|
|
|
product HP CDW8200 0x0207 CD-Writer Plus 8200e
|
2002-04-06 13:28:02 +00:00
|
|
|
product HP 1220C 0x0212 DeskJet 1220C
|
2001-07-05 09:35:46 +00:00
|
|
|
product HP 810C 0x0304 DeskJet 810C/812C
|
|
|
|
|
product HP 4300C 0x0305 Scanjet 4300C
|
|
|
|
|
product HP G85XI 0x0311 OfficeJet G85xi
|
2002-04-08 10:16:19 +00:00
|
|
|
product HP 1200 0x0317 LaserJet 1200
|
2000-10-25 10:13:47 +00:00
|
|
|
product HP 5200C 0x0401 Scanjet 5200C
|
2001-07-05 09:35:46 +00:00
|
|
|
product HP 830C 0x0404 DeskJet 830C
|
|
|
|
|
product HP 3400CSE 0x0405 ScanJet 3400cse
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product HP 6300C 0x0601 Scanjet 6300C
|
2001-07-05 09:35:46 +00:00
|
|
|
product HP 840C 0x0604 DeskJet 840c
|
2000-10-25 10:13:47 +00:00
|
|
|
product HP 5300C 0x0701 Scanjet 5300C
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product HP 970CSE 0x1004 Deskjet 970Cse
|
2002-04-06 13:06:07 +00:00
|
|
|
product HP 5400C 0x1005 Scanjet 5400C
|
2001-07-05 09:35:46 +00:00
|
|
|
product HP 930C 0x1204 DeskJet 930c
|
2002-04-06 13:06:07 +00:00
|
|
|
product HP P2000U 0x1801 Inkjet P-2000U
|
2001-07-05 09:35:46 +00:00
|
|
|
product HP 640C 0x2004 DeskJet 640c
|
2000-10-25 10:13:47 +00:00
|
|
|
product HP P1100 0x3102 Photosmart P1100
|
2000-10-08 21:31:39 +00:00
|
|
|
|
2000-10-25 10:13:47 +00:00
|
|
|
/* HP products */
|
2000-10-08 21:31:39 +00:00
|
|
|
product HP2 C500 0x6002 PhotoSmart C500
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2000-08-26 05:06:40 +00:00
|
|
|
/* IBM Corporation */
|
2001-04-02 13:02:17 +00:00
|
|
|
product IBM USBCDROMDRIVE 0x4427 USB CD-ROM Drive
|
2000-08-26 05:06:40 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Inside Out Networks products */
|
|
|
|
|
product INSIDEOUT EDGEPORT4 0x0001 EdgePort/4 serial ports
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* In-System products */
|
|
|
|
|
product INSYSTEM F5U002 0x0002 Parallel printer adapter
|
2001-07-05 09:35:46 +00:00
|
|
|
product INSYSTEM ATAPI 0x0031 ATAPI adapter
|
2001-04-02 13:02:17 +00:00
|
|
|
product INSYSTEM ISD110 0x0200 IDE adapter ISD110
|
|
|
|
|
product INSYSTEM ISD105 0x0202 IDE adapter ISD105
|
2000-06-08 19:26:16 +00:00
|
|
|
product INSYSTEM USBCABLE 0x081a USB cable
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Intel products */
|
2001-01-03 11:45:03 +00:00
|
|
|
product INTEL EASYPC_CAMERA 0x0110 Easy PC Camera
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product INTEL TESTBOARD 0x9890 82930 test board
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* I/O DATA products */
|
2001-01-03 11:45:03 +00:00
|
|
|
product IODATA USBETT 0x0901 USB ETT
|
2000-04-03 00:07:43 +00:00
|
|
|
product IODATA USBETTX 0x0904 USB ETTX
|
2002-05-27 00:00:48 +00:00
|
|
|
product IODATA USBETTXS 0x0913 USB ETTX
|
2001-02-05 11:02:50 +00:00
|
|
|
product IODATA USBRSAQ 0x0a03 USB serial adapter USB-RSAQ1
|
2000-04-03 00:07:43 +00:00
|
|
|
|
1999-11-18 18:02:44 +00:00
|
|
|
/* Iomega products */
|
|
|
|
|
product IOMEGA ZIP100 0x0001 Zip 100
|
2000-04-06 10:16:24 +00:00
|
|
|
product IOMEGA ZIP250 0x0030 Zip 250
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* Kawasaki products */
|
|
|
|
|
product KLSI DUH3E10BT 0x0008 USB ethernet controller engine
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Kawatsu products */
|
|
|
|
|
product KAWATSU MH4000P 0x0003 MiniHub 4000P
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2000-03-29 15:07:20 +00:00
|
|
|
/* Keisokugiken Corp. products */
|
|
|
|
|
product KEISOKUGIKEN USBDAQ 0x0068 HKS-0200 USBDAQ
|
|
|
|
|
|
2002-02-02 21:09:54 +00:00
|
|
|
/* Kawasaki products */
|
2000-03-29 19:44:48 +00:00
|
|
|
product KLSI DUH3E10BT 0x0008 10BT Ethernet adapter, in the DU-H3E
|
|
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* Kensington products */
|
|
|
|
|
product KENSINGTON ORBIT 0x1003 Orbit USB/PS2 trackball
|
|
|
|
|
product KENSINGTON TURBOBALL 0x1005 TurboBall
|
|
|
|
|
|
2001-04-02 13:02:17 +00:00
|
|
|
/* Keyspan products */
|
|
|
|
|
product KEYSPAN USA28 0x0101 USA-28 serial adapter
|
|
|
|
|
product KEYSPAN USA28X 0x0102 USA-28X serial adapter
|
|
|
|
|
product KEYSPAN USA19 0x0103 USA-19 serial adapter
|
|
|
|
|
product KEYSPAN USA18X 0x0105 USA-18X serial adapter
|
|
|
|
|
product KEYSPAN USA19W 0x0106 USA-19W serial adapter
|
|
|
|
|
product KEYSPAN USA49W 0x0109 USA-49W serial adapter
|
|
|
|
|
|
2000-10-10 23:23:54 +00:00
|
|
|
/* Kingston products */
|
2000-10-25 10:13:47 +00:00
|
|
|
product KINGSTON KNU101TX 0x000a KNU101TX USB Ethernet
|
2000-10-10 23:23:54 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Kodak products */
|
2000-04-03 00:07:43 +00:00
|
|
|
product KODAK DC220 0x0100 Digital Science DC220
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product KODAK DC260 0x0110 Digital Science DC260
|
|
|
|
|
product KODAK DC265 0x0111 Digital Science DC265
|
2000-01-28 10:26:29 +00:00
|
|
|
product KODAK DC290 0x0112 Digital Science DC290
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product KODAK DC240 0x0120 Digital Science DC240
|
|
|
|
|
product KODAK DC280 0x0130 Digital Science DC280
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Konica Corp. Products */
|
|
|
|
|
product KONICA CAMERA 0x0720 Digital Color Camera
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* KYE products */
|
|
|
|
|
product KYE NICHE 0x0001 Niche mouse
|
2000-04-03 00:07:43 +00:00
|
|
|
product KYE NETSCROLL 0x0003 Genius NetScroll mouse
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product KYE FLIGHT2000 0x1004 Flight 2000 joystick
|
2000-09-05 12:24:05 +00:00
|
|
|
product KYE VIVIDPRO 0x2001 ColorPage Vivid-Pro scanner
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
|
2000-10-07 11:10:53 +00:00
|
|
|
/* LaCie products */
|
2001-07-05 09:35:46 +00:00
|
|
|
product LACIE HD 0xa601 Hard Disk
|
2000-10-07 11:10:53 +00:00
|
|
|
product LACIE CDRW 0xa602 CD R/W
|
|
|
|
|
|
2001-04-17 15:34:22 +00:00
|
|
|
/* Lexar products */
|
|
|
|
|
product LEXAR JUMPSHOT 0x0001 jumpSHOT CompactFlash Reader
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Lexmark products */
|
|
|
|
|
product LEXMARK S2450 0x0009 Optra S 2450
|
|
|
|
|
|
|
|
|
|
/* Linksys products */
|
2001-07-05 09:35:46 +00:00
|
|
|
product LINKSYS MAUSB2 0x0105 Camedia MAUSB-2
|
2002-03-31 21:38:05 +00:00
|
|
|
product LINKSYS USB10TX1 0x200c USB10TX
|
2000-03-21 15:18:51 +00:00
|
|
|
product LINKSYS USB10T 0x2202 USB10T Ethernet
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product LINKSYS USB100TX 0x2203 USB100TX Ethernet
|
2000-05-07 19:44:16 +00:00
|
|
|
product LINKSYS USB100H1 0x2204 USB100H1 Ethernet/HPNA
|
2000-04-02 12:43:34 +00:00
|
|
|
product LINKSYS USB10TA 0x2206 USB10TA Ethernet
|
2002-04-06 13:01:39 +00:00
|
|
|
product LINKSYS USB10TX2 0x400b USB10TX
|
1999-11-18 18:02:44 +00:00
|
|
|
|
|
|
|
|
/* Logitech products */
|
|
|
|
|
product LOGITECH M2452 0x0203 M2452 keyboard
|
|
|
|
|
product LOGITECH M4848 0x0301 M4848 mouse
|
2000-07-17 09:55:34 +00:00
|
|
|
product LOGITECH PAGESCAN 0x040f PageScan
|
|
|
|
|
product LOGITECH QUICKCAMWEB 0x0801 QuickCam Web
|
1999-11-18 18:02:44 +00:00
|
|
|
product LOGITECH QUICKCAMPRO 0x0810 QuickCam Pro
|
2000-07-17 09:55:34 +00:00
|
|
|
product LOGITECH QUICKCAMEXP 0x0840 QuickCam Express
|
|
|
|
|
product LOGITECH QUICKCAM 0x0850 QuickCam
|
2000-04-03 00:07:43 +00:00
|
|
|
product LOGITECH N43 0xc000 N43
|
1999-11-18 18:02:44 +00:00
|
|
|
product LOGITECH N48 0xc001 N48 mouse
|
|
|
|
|
product LOGITECH MBA47 0xc002 M-BA47 mouse
|
2000-04-03 00:07:43 +00:00
|
|
|
product LOGITECH WMMOUSE 0xc004 WingMan Gaming Mouse
|
2001-01-03 11:45:03 +00:00
|
|
|
product LOGITECH BD58 0xc00c BD58 mouse
|
2001-07-05 09:35:46 +00:00
|
|
|
product LOGITECH UN58A 0xc030 iFeel Mouse
|
2000-10-07 11:10:53 +00:00
|
|
|
product LOGITECH BB13 0xc401 USB-PS/2 Trackball
|
2000-04-03 00:07:43 +00:00
|
|
|
product LOGITECH WMPAD 0xc208 WingMan GamePad Extreme
|
|
|
|
|
product LOGITECH WMJOY 0xc281 WingMan Force joystick
|
2001-04-02 13:02:17 +00:00
|
|
|
product LOGITECH RK53 0xc501 Cordless mouse
|
|
|
|
|
product LOGITECH RB6 0xc503 Cordless keyboard
|
2000-07-17 09:55:34 +00:00
|
|
|
product LOGITECH QUICKCAMPRO2 0xd001 QuickCam Pro
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Lucent products */
|
|
|
|
|
product LUCENT EVALKIT 0x1001 USS-720 evaluation kit
|
1999-11-18 18:02:44 +00:00
|
|
|
|
|
|
|
|
/* Macally products */
|
|
|
|
|
product MACALLY MOUSE1 0x0101 mouse
|
|
|
|
|
|
2000-05-07 19:44:16 +00:00
|
|
|
/* MCT Corp. */
|
|
|
|
|
product MCT HUB0100 0x0100 Hub
|
2002-02-02 21:09:54 +00:00
|
|
|
product MCT DU_H3SP_USB232 0x0200 D-Link DU-H3SP USB BAY Hub
|
2000-05-07 19:44:16 +00:00
|
|
|
product MCT USB232 0x0210 USB-232 Interface
|
2001-04-02 13:02:17 +00:00
|
|
|
product MCT SITECOM_USB232 0x0230 Sitecom USB-232 Products
|
2000-05-07 19:44:16 +00:00
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* Melco, Inc products */
|
2001-01-03 11:45:03 +00:00
|
|
|
product MELCO LUATX1 0x0001 LUA-TX Ethernet
|
|
|
|
|
product MELCO LUATX5 0x0005 LUA-TX Ethernet
|
2001-11-19 18:40:17 +00:00
|
|
|
product MELCO LUA2TX5 0x0009 LUA2-TX Ethernet
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
|
2000-10-25 10:13:47 +00:00
|
|
|
/* Metricom products */
|
|
|
|
|
product METRICOM RICOCHET_GS 0x0001 Ricochet GS
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Microsoft products */
|
2000-04-03 00:07:43 +00:00
|
|
|
product MICROSOFT SIDEPREC 0x0008 SideWinder Precision Pro
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product MICROSOFT INTELLIMOUSE 0x0009 IntelliMouse
|
|
|
|
|
product MICROSOFT NATURALKBD 0x000b Natural Keyboard Elite
|
|
|
|
|
product MICROSOFT DDS80 0x0014 Digital Sound System 80
|
|
|
|
|
product MICROSOFT SIDEWINDER 0x001a Sidewinder Precision Racing Wheel
|
2000-04-03 00:07:43 +00:00
|
|
|
product MICROSOFT INTELLIEYE 0x0025 IntelliEye mouse
|
|
|
|
|
product MICROSOFT INETPRO 0x002b Internet Keyboard Pro
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
|
2001-04-02 13:02:17 +00:00
|
|
|
/* Microtech products */
|
2001-07-05 09:35:46 +00:00
|
|
|
product MICROTECH SCSIDB25 0x0004 USB-SCSI-DB25
|
|
|
|
|
product MICROTECH SCSIHD50 0x0005 USB-SCSI-HD50
|
2001-04-02 13:02:17 +00:00
|
|
|
product MICROTECH DPCM 0x0006 USB CameraMate
|
2001-07-05 09:35:46 +00:00
|
|
|
product MICROTECH FREECOM 0xfc01 Freecom USB-IDE
|
2001-04-02 13:02:17 +00:00
|
|
|
|
2000-09-05 12:24:05 +00:00
|
|
|
/* Microtek products */
|
|
|
|
|
product MICROTEK 336CX 0x0094 Phantom 336CX - C3 scanner
|
2001-01-03 11:45:03 +00:00
|
|
|
product MICROTEK X6U 0x0099 ScanMaker X6 - X6U
|
2000-09-05 12:24:05 +00:00
|
|
|
product MICROTEK C6 0x009a Phantom C6 scanner
|
2001-01-03 11:45:03 +00:00
|
|
|
product MICROTEK 336CX2 0x00a0 Phantom 336CX - C3 scanner
|
2000-09-05 12:24:05 +00:00
|
|
|
product MICROTEK V6USL 0x00a3 ScanMaker V6USL
|
2000-09-05 12:26:37 +00:00
|
|
|
product MICROTEK V6USL2 0x80a3 ScanMaker V6USL
|
2000-09-05 12:24:05 +00:00
|
|
|
product MICROTEK V6UL 0x80ac ScanMaker V6UL
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Midiman products */
|
|
|
|
|
product MIDIMAN MIDISPORT2X2 0x1001 Midisport 2x2
|
|
|
|
|
|
2002-04-06 12:27:23 +00:00
|
|
|
/* Minolta Co., Ltd. */
|
|
|
|
|
product MINOLTA 2300 0x4001 Dimage 2300
|
|
|
|
|
product MINOLTA S304 0x4007 Dimage S304
|
|
|
|
|
product MINOLTA X 0x4009 Dimage X
|
|
|
|
|
|
2002-04-06 13:01:39 +00:00
|
|
|
/* Mitsumi products */
|
|
|
|
|
product MITSUMI CDRRW 0x0000 CD-R/RW Drive
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Motorola products */
|
|
|
|
|
product MOTOROLA MC141555 0x1555 MC141555 hub controller
|
2000-01-08 07:32:48 +00:00
|
|
|
|
1999-11-18 18:02:44 +00:00
|
|
|
/* MultiTech products */
|
|
|
|
|
product MULTITECH ATLAS 0xf101 MT5634ZBA-USB modem
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Mustek products */
|
2000-09-05 12:24:05 +00:00
|
|
|
product MUSTEK 1200CU 0x0001 1200 CU scanner
|
2000-10-25 10:13:47 +00:00
|
|
|
product MUSTEK 600CU 0x0002 600 CU scanner
|
|
|
|
|
product MUSTEK 1200USB 0x0003 1200 USB scanner
|
|
|
|
|
product MUSTEK 1200UB 0x0006 1200 UB scanner
|
2001-08-21 21:35:40 +00:00
|
|
|
product MUSTEK 1200USBPLUS 0x0007 1200 USB Plus scanner
|
2001-11-28 12:37:04 +00:00
|
|
|
product MUSTEK 1200CUPLUS 0x0008 1200 CU Plus scanner
|
2001-08-21 21:35:40 +00:00
|
|
|
product MUSTEK BEARPAW1200F 0x0010 BearPaw 1200F scanner
|
|
|
|
|
product MUSTEK 600USB 0x0873 600 USB scanner
|
2000-10-07 11:10:53 +00:00
|
|
|
product MUSTEK MDC800 0xa800 MDC-800 digital camera
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2001-04-02 13:02:17 +00:00
|
|
|
/* M-Systems products */
|
|
|
|
|
product MSYSTEMS DISKONKEY 0x0010 DiskOnKey
|
|
|
|
|
|
2000-10-25 10:13:47 +00:00
|
|
|
/* National Semiconductor */
|
2002-02-02 21:09:54 +00:00
|
|
|
product NATIONAL BEARPAW1200 0x1000 BearPaw 1200
|
|
|
|
|
product NATIONAL BEARPAW2400 0x1001 BearPaw 2400
|
2000-10-25 10:13:47 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* NEC products */
|
|
|
|
|
product NEC HUB 0x55aa hub
|
|
|
|
|
product NEC HUB_B 0x55ab hub
|
Add the vendor and device IDs for a whole bunch of additional USB
ethernet adapters that are supported by the aue and kue drivers.
There are actually a couple more out there from Accton, Asante and
EXP Computer, however I was not able to find any Windows device
drivers for these on their servers, and hence could not harvest
their vendor/device ID info. If somebody has one of these things
and can look in the .inf file that comes with the Windows driver,
I'd appreciate knowing what it says for 'VID' and 'PID.'
Additional adapters include: the D-Link DSB-650 and DSB-650TX, the
SMC 2102USB, 2104USB and 2202USB, the ATen UC10T, and the Netgear EA101.
These are all mentioned in the man pages, relnotes and LINT.
Also correct the date in the kue(4) man page. I wrote this thing
on Jan, 4 2000, not 1999.
2000-01-07 22:18:49 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* NetChip Technology Products */
|
|
|
|
|
product NETCHIP TURBOCONNECT 0x1080 Turbo-Connect
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Netgear products */
|
|
|
|
|
product NETGEAR EA101 0x1001 Ethernet adapter
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2001-01-18 00:15:49 +00:00
|
|
|
/* Nikon products */
|
|
|
|
|
product NIKON E990 0x0102 Digital Camera E990
|
|
|
|
|
|
2001-10-23 22:44:03 +00:00
|
|
|
/* Olympus products */
|
2001-11-23 16:35:40 +00:00
|
|
|
product OLYMPUS C1 0x0102 C-1 Digital Camera
|
2001-10-23 22:44:03 +00:00
|
|
|
product OLYMPUS C700 0x0105 C-700 Ultra Zoom
|
|
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* OmniVision Technologies, Inc. products */
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product OMNIVISION OV511 0x0511 OV511 Camera
|
2000-07-17 09:55:34 +00:00
|
|
|
product OMNIVISION OV511PLUS 0xa511 OV511+ Camera
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2001-04-02 13:02:17 +00:00
|
|
|
/* Palm Computing, Inc. product */
|
2000-04-03 00:07:43 +00:00
|
|
|
product PALM SERIAL 0x0080 USB Serial Adaptor
|
|
|
|
|
|
2001-04-02 13:02:17 +00:00
|
|
|
/* Panasonic products */
|
|
|
|
|
product PANASONIC SDCAAE 0x1b00 MultiMediaCard Adapter
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Peracom products */
|
|
|
|
|
product PERACOM SERIAL1 0x0001 Serial Converter
|
|
|
|
|
product PERACOM ENET 0x0002 Ethernet adapter
|
2000-03-29 19:44:48 +00:00
|
|
|
product PERACOM ENET3 0x0003 At Home Ethernet Adapter
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product PERACOM ENET2 0x0005 Ethernet adapter
|
This commit adds device driver support for the ADMtek AN986 Pegasus
USB ethernet chip. Adapters that use this chip include the LinkSys
USB100TX. There are a few others, but I'm not certain of their
availability in the U.S. I used an ADMtek eval board for development.
Note that while the ADMtek chip is a 100Mbps device, you can't really
get 100Mbps speeds over USB. Regardless, this driver uses miibus to
allow speed and duplex mode selection as well as autonegotiation.
Building and kldloading the driver as a module is also supported.
Note that in order to make this driver work, I had to make what some
may consider an ugly hack to sys/dev/usb/usbdi.c. The usbd_transfer()
function will use tsleep() for synchronous transfers that don't complete
right away. This is a problem since there are times when we need to
do sync transfers from an interrupt context (i.e. when reading registers
from the MAC via the control endpoint), where tsleep() us a no-no.
My hack allows the driver to have the code poll for transfer completion
subject to the xfer->timeout timeout rather that calling tsleep().
This hack is controlled by a quirk entry and is only enabled for the
ADMtek device.
Now, I'm sure there are a few of you out there ready to jump on me
and suggest some other approach that doesn't involve a busy wait. The
only solution that might work is to handle the interrupts in a kernel
thread, where you may have something resembling a process context that
makes it okay to tsleep(). This is lovely, except we don't have any
mechanism like that now, and I'm not about to implement such a thing
myself since it's beyond the scope of driver development. (Translation:
I'll be damned if I know how to do it.) If FreeBSD ever aquires such
a mechanism, I'll be glad to revisit the driver to take advantage of
it. In the meantime, I settled for what I perceived to be the solution
that involved the least amount of code changes. In general, the hit
is pretty light.
Also note that my only USB test box has a UHCI controller: I haven't
I don't have a machine with an OHCI controller available.
Highlights:
- Updated usb_quirks.* to add UQ_NO_TSLEEP quirk for ADMtek part.
- Updated usbdevs and regenerated generated files
- Updated HARDWARE.TXT and RELNOTES.TXT files
- Updated sysinstall/device.c and userconfig.c
- Updated kernel configs -- device aue0 is commented out by default
- Updated /sys/conf/files
- Added new kld module directory
1999-12-28 02:01:18 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Philips products */
|
|
|
|
|
product PHILIPS DSS350 0x0101 DSS 350 Digital Speaker System
|
|
|
|
|
product PHILIPS DSS 0x0104 DSS XXX Digital Speaker System
|
|
|
|
|
product PHILIPS HUB 0x0201 hub
|
2001-07-05 09:35:46 +00:00
|
|
|
product PHILIPS PCA646VC 0x0303 PCA646VC PC Camera
|
2000-05-07 19:44:16 +00:00
|
|
|
product PHILIPS PCVC680K 0x0308 PCVC680K Vesta Pro PC Camera
|
2000-04-03 00:07:43 +00:00
|
|
|
product PHILIPS DSS150 0x0471 DSS 150 Digital Speaker System
|
2001-12-12 15:58:23 +00:00
|
|
|
product PHILIPS DIVAUSB 0x1801 DIVA USB mp3 player
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
|
2001-02-05 11:02:50 +00:00
|
|
|
/* Philips Semiconductor products */
|
|
|
|
|
product PHILIPSSEMI HUB1122 0x1122 hub
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* P.I. Engineering products */
|
|
|
|
|
product PIENGINEERING PS2USB 0x020b PS2 to Mac USB Adapter
|
|
|
|
|
|
|
|
|
|
/* PLX products */
|
|
|
|
|
product PLX TESTBOARD 0x9060 test board
|
|
|
|
|
|
|
|
|
|
/* Primax products */
|
2000-10-25 10:13:47 +00:00
|
|
|
product PRIMAX G2X300 0x0300 G2-200 scanner
|
2000-09-05 12:24:05 +00:00
|
|
|
product PRIMAX G2E300 0x0301 G2E-300 scanner
|
|
|
|
|
product PRIMAX G2300 0x0302 G2-300 scanner
|
|
|
|
|
product PRIMAX G2E3002 0x0303 G2E-300 scanner
|
|
|
|
|
product PRIMAX 9600 0x0340 Colorado USB 9600 scanner
|
|
|
|
|
product PRIMAX 600U 0x0341 Colorado 600u scanner
|
2000-10-07 11:10:53 +00:00
|
|
|
product PRIMAX 19200 0x0360 Colorado USB 19200 scanner
|
2000-09-05 12:24:05 +00:00
|
|
|
product PRIMAX 1200U 0x0361 Colorado 1200u scanner
|
2000-10-25 10:13:47 +00:00
|
|
|
product PRIMAX G600 0x0380 G2-600 scanner
|
2000-10-07 11:10:53 +00:00
|
|
|
product PRIMAX 636I 0x0381 ReadyScan 636i
|
|
|
|
|
product PRIMAX G2600 0x0382 G2-600 scanner
|
|
|
|
|
product PRIMAX G2E600 0x0383 G2E-600 scanner
|
|
|
|
|
product PRIMAX COMFORT 0x4d01 Comfort
|
|
|
|
|
product PRIMAX MOUSEINABOX 0x4d02 Mouse-in-a-Box
|
|
|
|
|
product PRIMAX PCGAUMS1 0x4d04 Sony PCGA-UMS1
|
2000-04-03 00:07:43 +00:00
|
|
|
|
2000-05-07 19:44:16 +00:00
|
|
|
/* Prolific products */
|
|
|
|
|
product PROLIFIC PL2301 0x0000 PL2301 Host-Host interface
|
|
|
|
|
product PROLIFIC PL2302 0x0001 PL2302 Host-Host interface
|
2001-07-05 09:35:46 +00:00
|
|
|
product PROLIFIC RSAQ2 0x04bb PL2303 Serial adapter (IODATA USB-RSAQ2)
|
|
|
|
|
product PROLIFIC PL2303 0x2303 PL2303 Serial adapter (ATEN/IOGEAR UC232A)
|
2001-01-03 11:45:03 +00:00
|
|
|
product PROLIFIC PL2305 0x2305 Parallel printer adapter
|
2000-09-05 16:00:55 +00:00
|
|
|
product PROLIFIC ATAPI4 0x2307 ATAPI-4 Bridge Controller
|
2000-05-07 19:44:16 +00:00
|
|
|
|
2001-04-12 10:59:30 +00:00
|
|
|
/* Putercom products */
|
|
|
|
|
product PUTERCOM UPA100 0x047e USB-1284 BRIDGE
|
|
|
|
|
|
2000-10-07 11:10:53 +00:00
|
|
|
/* Qtronix products */
|
|
|
|
|
product QTRONIX 980N 0x2011 Scorpion-980N keyboard
|
|
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* Quickshot products */
|
|
|
|
|
product QUICKSHOT STRIKEPAD 0x6238 USB StrikePad
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
|
2000-10-07 11:10:53 +00:00
|
|
|
/* Rainbow Technologies products */
|
|
|
|
|
product RAINBOW IKEY2000 0x1200 i-Key 2000
|
|
|
|
|
|
2001-02-05 11:02:50 +00:00
|
|
|
/* Roland products */
|
|
|
|
|
product ROLAND UM1 0x0009 UM-1 MIDI I/F
|
2002-04-06 13:01:39 +00:00
|
|
|
product ROLAND UM880N 0x0014 EDIROL UM-880 MIDI I/F (native)
|
|
|
|
|
product ROLAND UM880G 0x0015 EDIROL UM-880 MIDI I/F (generic)
|
2001-02-05 11:02:50 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Rockfire products */
|
|
|
|
|
product ROCKFIRE GAMEPAD 0x2033 gamepad 203USB
|
|
|
|
|
|
2002-04-06 13:01:39 +00:00
|
|
|
/* RATOC Systems products */
|
2002-03-18 18:23:42 +00:00
|
|
|
product RATOC REXUSB60 0xb000 USB serial adapter REX-USB60
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* SanDisk products */
|
2001-07-05 09:35:46 +00:00
|
|
|
product SANDISK SDDR05A 0x0001 ImageMate SDDR-05a
|
|
|
|
|
product SANDISK SDDR31 0x0002 ImageMate SDDR-31
|
|
|
|
|
product SANDISK SDDR12 0x0100 ImageMate SDDR-12
|
|
|
|
|
product SANDISK SDDR09 0x0200 ImageMate SDDR-09
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
|
2001-01-03 11:45:03 +00:00
|
|
|
/* ScanLogic products */
|
2002-04-07 13:45:24 +00:00
|
|
|
product SCANLOGIC SL11R 0x0002 SL11R IDE Adapter
|
2001-01-03 11:45:03 +00:00
|
|
|
product SCANLOGIC 336CX 0x0300 Phantom 336CX - C3 scanner
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Shuttle Technology products */
|
|
|
|
|
product SHUTTLE EUSB 0x0001 E-USB Bridge
|
2001-07-05 09:35:46 +00:00
|
|
|
product SHUTTLE EUSCSI 0x0002 eUSCSI Bridge
|
|
|
|
|
product SHUTTLE SDDR09 0x0003 ImageMate SDDR09
|
2001-04-02 13:02:17 +00:00
|
|
|
product SHUTTLE ZIOMMC 0x0006 eUSB MultiMediaCard Adapter
|
2001-07-05 09:35:46 +00:00
|
|
|
product SHUTTLE HIFD 0x0007 Sony Hifd
|
|
|
|
|
product SHUTTLE EUSBATAPI 0x0009 eUSB ATA/ATAPI Adapter
|
|
|
|
|
product SHUTTLE CF 0x000a eUSB CompactFlash Adapter
|
|
|
|
|
product SHUTTLE EUSCSI_B 0x000b eUSCSI Bridge
|
|
|
|
|
product SHUTTLE EUSCSI_C 0x000c eUSCSI Bridge
|
|
|
|
|
product SHUTTLE CDRW 0x0101 CD-RW Device
|
2000-01-13 20:13:58 +00:00
|
|
|
|
2002-03-31 21:38:05 +00:00
|
|
|
/* Siemens products */
|
2002-04-06 13:01:39 +00:00
|
|
|
product SIEMENS SPEEDSTREAM 0x1001 SpeedStream USB
|
2002-03-31 21:38:05 +00:00
|
|
|
|
1999-11-18 18:02:44 +00:00
|
|
|
/* SIIG products */
|
|
|
|
|
product SIIG DIGIFILMREADER 0x0004 DigiFilm-Combo Reader
|
|
|
|
|
|
2001-01-03 11:45:03 +00:00
|
|
|
/* Silicon Portals Inc. */
|
|
|
|
|
product SILICONPORTALS YAPPH_NF 0x0200 YAP Phone (no firmware)
|
|
|
|
|
product SILICONPORTALS YAPPHONE 0x0201 YAP Phone
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Sirius Technologies products */
|
|
|
|
|
product SIRIUS ROADSTER 0x0001 NetComm Roadster II 56 USB
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2000-04-14 14:43:34 +00:00
|
|
|
/* SmartBridges products */
|
|
|
|
|
product SMARTBRIDGES SMARTLINK 0x0001 SmartLink USB ethernet adapter
|
2002-03-31 21:38:05 +00:00
|
|
|
product SMARTBRIDGES SMARTNIC 0x0003 smartNIC 2 PnP Adapter
|
2000-04-14 14:43:34 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* SMC products */
|
|
|
|
|
product SMC 2102USB 0x0100 10Mbps ethernet adapter
|
|
|
|
|
product SMC 2202USB 0x0200 10/100 ethernet adapter
|
2002-03-19 10:54:40 +00:00
|
|
|
product SMC2 2020HUB 0x2020 USB Hub
|
|
|
|
|
|
2002-03-31 21:38:05 +00:00
|
|
|
/* SOHOware products */
|
|
|
|
|
product SOHOWARE NUB100 0x9100 10/100 USB Ethernet
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* SOLID YEAR products */
|
|
|
|
|
product SOLIDYEAR KEYBOARD 0x2101 Solid Year USB keyboard
|
Add the vendor and device IDs for a whole bunch of additional USB
ethernet adapters that are supported by the aue and kue drivers.
There are actually a couple more out there from Accton, Asante and
EXP Computer, however I was not able to find any Windows device
drivers for these on their servers, and hence could not harvest
their vendor/device ID info. If somebody has one of these things
and can look in the .inf file that comes with the Windows driver,
I'd appreciate knowing what it says for 'VID' and 'PID.'
Additional adapters include: the D-Link DSB-650 and DSB-650TX, the
SMC 2102USB, 2104USB and 2202USB, the ATen UC10T, and the Netgear EA101.
These are all mentioned in the man pages, relnotes and LINT.
Also correct the date in the kue(4) man page. I wrote this thing
on Jan, 4 2000, not 1999.
2000-01-07 22:18:49 +00:00
|
|
|
|
2000-09-04 15:29:42 +00:00
|
|
|
/* SONY products */
|
2001-07-05 09:35:46 +00:00
|
|
|
product SONY DSC 0x0010 DSC cameras
|
|
|
|
|
product SONY MSACUS1 0x002d Memorystick MSAC-US1
|
|
|
|
|
product SONY MSC 0x0032 MSC memory stick slot
|
2000-09-04 15:29:42 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* STMicroelectronics products */
|
|
|
|
|
product STMICRO COMMUNICATOR 0x7554 USB Communicator
|
This commit adds device driver support for the ADMtek AN986 Pegasus
USB ethernet chip. Adapters that use this chip include the LinkSys
USB100TX. There are a few others, but I'm not certain of their
availability in the U.S. I used an ADMtek eval board for development.
Note that while the ADMtek chip is a 100Mbps device, you can't really
get 100Mbps speeds over USB. Regardless, this driver uses miibus to
allow speed and duplex mode selection as well as autonegotiation.
Building and kldloading the driver as a module is also supported.
Note that in order to make this driver work, I had to make what some
may consider an ugly hack to sys/dev/usb/usbdi.c. The usbd_transfer()
function will use tsleep() for synchronous transfers that don't complete
right away. This is a problem since there are times when we need to
do sync transfers from an interrupt context (i.e. when reading registers
from the MAC via the control endpoint), where tsleep() us a no-no.
My hack allows the driver to have the code poll for transfer completion
subject to the xfer->timeout timeout rather that calling tsleep().
This hack is controlled by a quirk entry and is only enabled for the
ADMtek device.
Now, I'm sure there are a few of you out there ready to jump on me
and suggest some other approach that doesn't involve a busy wait. The
only solution that might work is to handle the interrupts in a kernel
thread, where you may have something resembling a process context that
makes it okay to tsleep(). This is lovely, except we don't have any
mechanism like that now, and I'm not about to implement such a thing
myself since it's beyond the scope of driver development. (Translation:
I'll be damned if I know how to do it.) If FreeBSD ever aquires such
a mechanism, I'll be glad to revisit the driver to take advantage of
it. In the meantime, I settled for what I perceived to be the solution
that involved the least amount of code changes. In general, the hit
is pretty light.
Also note that my only USB test box has a UHCI controller: I haven't
I don't have a machine with an OHCI controller available.
Highlights:
- Updated usb_quirks.* to add UQ_NO_TSLEEP quirk for ADMtek part.
- Updated usbdevs and regenerated generated files
- Updated HARDWARE.TXT and RELNOTES.TXT files
- Updated sysinstall/device.c and userconfig.c
- Updated kernel configs -- device aue0 is commented out by default
- Updated /sys/conf/files
- Added new kld module directory
1999-12-28 02:01:18 +00:00
|
|
|
|
2001-01-03 11:45:03 +00:00
|
|
|
/* STSN products */
|
|
|
|
|
product STSN STSN0001 0x0001 Internet Access Device
|
|
|
|
|
|
2002-04-06 13:01:39 +00:00
|
|
|
/* SUN Corporation products */
|
|
|
|
|
product SUNTAC PS64P1 0x0005 SUNTAC U-Cable type P1
|
|
|
|
|
product SUNTAC VS10U 0x0009 SUNTAC Slipper U
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Sun Microsystems products */
|
2001-04-12 10:59:30 +00:00
|
|
|
product SUN KEYBOARD 0x0005 Type 6 USB keyboard
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* XXX The above is a North American PC style keyboard possibly */
|
2001-04-12 10:59:30 +00:00
|
|
|
product SUN MOUSE 0x0100 Type 6 USB mouse
|
1999-11-18 18:02:44 +00:00
|
|
|
|
2000-05-25 22:27:55 +00:00
|
|
|
/* Supra products */
|
2000-10-07 11:10:53 +00:00
|
|
|
product DIAMOND2 SUPRAEXPRESS56K 0x07da Supra Express 56K modem
|
|
|
|
|
product DIAMOND2 SUPRA2890 0x0b4a SupraMax 2890 56K Modem
|
|
|
|
|
product DIAMOND2 RIO600USB 0x5001 Rio 600 USB
|
2001-02-05 11:00:13 +00:00
|
|
|
product DIAMOND2 RIO800USB 0x5002 Rio 800 USB
|
2001-04-12 11:04:08 +00:00
|
|
|
|
2001-07-05 09:35:46 +00:00
|
|
|
/* Taugagreining products */
|
|
|
|
|
product TAUGA CAMERAMATE 0x0005 CameraMate (DPCM_USB)
|
|
|
|
|
|
2001-04-12 11:04:08 +00:00
|
|
|
/* TDK products */
|
2001-07-05 09:35:46 +00:00
|
|
|
product TDK UPA9664 0x0115 USB-PDC Adapter UPA9664
|
|
|
|
|
product TDK UCA1464 0x0116 USB-cdmaOne Adapter UCA1464
|
|
|
|
|
product TDK UHA6400 0x0117 USB-PHS Adapter UHA6400
|
|
|
|
|
product TDK UPA6400 0x0118 USB-PHS Adapter UPA6400
|
2001-04-12 11:04:08 +00:00
|
|
|
|
2000-07-17 09:55:34 +00:00
|
|
|
/* TEAC products */
|
|
|
|
|
product TEAC FD05PUB 0x0000 FD-05PUB floppy
|
|
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Telex Communications products */
|
|
|
|
|
product TELEX MIC1 0x0001 Enhanced USB Microphone
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Texas Intel products */
|
|
|
|
|
product TI UTUSB41 0x1446 UT-USB41 hub
|
2001-01-03 11:45:03 +00:00
|
|
|
product TI TUSB2046 0x2046 TUSB2046 hub
|
1999-11-18 18:02:44 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Thrustmaster products */
|
|
|
|
|
product THRUST FUSION_PAD 0xa0a3 Fusion Digital Gamepad
|
Add the vendor and device IDs for a whole bunch of additional USB
ethernet adapters that are supported by the aue and kue drivers.
There are actually a couple more out there from Accton, Asante and
EXP Computer, however I was not able to find any Windows device
drivers for these on their servers, and hence could not harvest
their vendor/device ID info. If somebody has one of these things
and can look in the .inf file that comes with the Windows driver,
I'd appreciate knowing what it says for 'VID' and 'PID.'
Additional adapters include: the D-Link DSB-650 and DSB-650TX, the
SMC 2102USB, 2104USB and 2202USB, the ATen UC10T, and the Netgear EA101.
These are all mentioned in the man pages, relnotes and LINT.
Also correct the date in the kue(4) man page. I wrote this thing
on Jan, 4 2000, not 1999.
2000-01-07 22:18:49 +00:00
|
|
|
|
2000-10-07 11:10:53 +00:00
|
|
|
/* Trek Technology products */
|
|
|
|
|
product TREK THUMBDRIVE 0x1111 ThumbDrive
|
|
|
|
|
|
2001-08-21 21:35:40 +00:00
|
|
|
/* Ultima products */
|
|
|
|
|
product ULTIMA 1200UBPLUS 0x4002 1200 UB Plus scanner
|
|
|
|
|
|
2000-08-06 23:26:58 +00:00
|
|
|
/* UMAX products */
|
2000-09-05 12:24:05 +00:00
|
|
|
product UMAX ASTRA1236U 0x0002 Astra 1236U Scanner
|
|
|
|
|
product UMAX ASTRA1220U 0x0010 Astra 1220U Scanner
|
|
|
|
|
product UMAX ASTRA2000U 0x0030 Astra 2000U Scanner
|
|
|
|
|
product UMAX ASTRA2100U 0x0130 Astra 2100U Scanner
|
|
|
|
|
product UMAX ASTRA2200U 0x0230 Astra 2200U Scanner
|
2002-02-11 03:35:53 +00:00
|
|
|
product UMAX ASTRA3400 0x0060 Astra 3400 Scanner
|
2000-08-06 23:26:58 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Universal Access products */
|
|
|
|
|
product UNIACCESS PANACHE 0x0101 Panache Surf USB ISDN Adapter
|
|
|
|
|
|
|
|
|
|
/* Vision products */
|
2000-07-17 09:55:34 +00:00
|
|
|
product VISION VC6452V002 0x0002 CPiA Camera
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
|
2000-09-05 12:24:05 +00:00
|
|
|
/* Visioneer products */
|
2000-10-25 10:13:47 +00:00
|
|
|
product VISIONEER 7600 0x0211 OneTouch 7600
|
|
|
|
|
product VISIONEER 5300 0x0221 OneTouch 5300
|
2000-09-05 12:24:05 +00:00
|
|
|
product VISIONEER 6100 0x0231 OneTouch 6100
|
2000-10-25 10:13:47 +00:00
|
|
|
product VISIONEER 6200 0x0311 OneTouch 6200
|
|
|
|
|
product VISIONEER 8100 0x0321 OneTouch 8100
|
|
|
|
|
product VISIONEER 8600 0x0331 OneTouch 8600
|
2001-04-02 13:02:17 +00:00
|
|
|
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
/* Wacom products */
|
|
|
|
|
product WACOM CT0405U 0x0000 CT-0405-U Tablet
|
2000-04-03 00:07:43 +00:00
|
|
|
product WACOM GRAPHIRE 0x0010 Graphire
|
|
|
|
|
product WACOM INTUOSA5 0x0021 Intuos A5
|
|
|
|
|
|
2001-07-05 09:35:46 +00:00
|
|
|
/* Xirlink products */
|
|
|
|
|
product XIRLINK PCCAM 0x8080 IBM PC Camera
|
|
|
|
|
|
2000-03-15 21:53:09 +00:00
|
|
|
/* Y-E Data products */
|
|
|
|
|
product YEDATA FLASHBUSTERU 0x0000 Flashbuster-U
|
|
|
|
|
|
2001-02-05 11:02:50 +00:00
|
|
|
/* Yamaha products */
|
|
|
|
|
product YAMAHA UX256 0x1000 UX256 MIDI I/F
|
2001-07-05 09:35:46 +00:00
|
|
|
product YAMAHA UX96 0x1008 UX96 MIDI I/F
|
2001-02-05 11:02:50 +00:00
|
|
|
|
2002-04-06 13:01:39 +00:00
|
|
|
/* Yano products */
|
|
|
|
|
product YANO U640MO 0x0101 U640MO-03
|
|
|
|
|
|
2000-04-03 00:07:43 +00:00
|
|
|
/* Zoom Telephonics, Inc. products */
|
More USB ethernet tweaks:
- Sync ohci, uhci and usbdi modules with NetBSD in order to obtain the
following improvements:
o New USBD_NO_TSLEEP flag can be used in place of UQ_NO_TSLEEP
quirk. This allows drivers to specify busy waiting only for
certain transfers (namely control transfers for reading/writing
registers and stuff).
o New USBD_FORCE_SHORT_XFER flag can be used to deal with
devices like the ADMtek Pegasus that sense the end of bulk OUT
transfers in a special way (if a transfer is exactly a multiple
of 64 bytes in size, you need to send an extra empty packet
to terminate the transfer).
o usbd_open_pipe_intr() now accepts an interval argument which
can be used to change the rate at which the interrupt callback
routine is invoked. Specifying USBD_DEFAULT_INTERVAL uses the
value specified in the device's config data, but drivers can
override it if needed.
- Change if_aue to use USBD_FORCE_SHORT_XFER for packet transmissions.
- Change if_aue, if_kue and if_cue to use USBD_NO_TSLEEP for all
control transfers. We no longer force the non-tsleep hack for
bulk transfers since these are done asynchronously anyway.
- Removed quirk entry fiddling from if_aue and if_kue since we don't
need it anymore now that we have the USBD_NO_TSLEEP flag.
- Tweak ulpt, uhid, ums and ukbd drivers to use the new arg to
usbd_open_pipe_intr().
- Add a flag to the softc struct in the ethernet drivers to indicate
when a device has been detached, and use this flag to perform
tests to prevent the drivers from trying to do control transfers
if this is the case. This is necessary because calling if_detach()
with INET6 enabled will eventually result in a call to the driver's
ioctl() routine to delete the multicast groups on the interface,
which will result in attempts to perform control transfers. (It's
possible this also happens even without INET6 support enabled.) This
is pointless since we know that if the detach method has been called,
the hardware has been unplugged.
- Changed watchdog timeout routines to just call the driver init routines
to initialize the device states without trying to close and re-open the
pipes. This is partly because we don't want to frob things at interrupt
context, but also because this doesn't seem to work right and I don't
want to panic the system just because a USB device may have stopped
responding.
- Fix aue_rxeof() to be a little smarter about detecting when a double
transfer is needed. Unfortunately, the design of the chip makes it hard
to get this exactly right. Hopefully, this will go away once either
Nick or Lennart finds the bug in the uhci driver that makes this ugly
hack necessary.
- Also sync usbdevs with NetBSD.
2000-01-20 07:38:33 +00:00
|
|
|
product ZOOM 2986L 0x9700 2986L Fax modem
|
2000-10-07 11:01:54 +00:00
|
|
|
|
|
|
|
|
/* ZyXEL Communication Co. products */
|
2001-04-09 18:19:20 +00:00
|
|
|
product ZYXEL OMNI56K 0x1500 Omni 56K Plus
|
2000-10-07 11:01:54 +00:00
|
|
|
product ZYXEL 980N 0x2011 Scorpion-980N keyboard
|
