freebsd-skq/sys/dev/usb/uhci.c

3026 lines
74 KiB
C
Raw Normal View History

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
/* $NetBSD: uhci.c,v 1.80 2000/01/19 01:16:38 augustss Exp $ */
/* $FreeBSD$ */
/*
* Copyright (c) 1998 The NetBSD Foundation, Inc.
* All rights reserved.
*
* This code is derived from software contributed to The NetBSD Foundation
2000-05-14 16:43:10 +00:00
* by Lennart Augustsson (lennart@augustsson.net) at
* 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.
*/
/*
* USB Universal Host Controller driver.
* Handles e.g. PIIX3 and PIIX4.
*
* UHCI spec: http://www.intel.com/design/usb/uhci11d.pdf
* USB spec: http://www.usb.org/developers/data/usb11.pdf
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
* PIIXn spec: ftp://download.intel.com/design/intarch/datashts/29055002.pdf
* ftp://download.intel.com/design/intarch/datashts/29056201.pdf
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#if defined(__NetBSD__) || defined(__OpenBSD__)
#include <sys/device.h>
#include <sys/select.h>
#elif defined(__FreeBSD__)
#include <sys/module.h>
#include <sys/bus.h>
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
#include <machine/bus_pio.h>
#if defined(DIAGNOSTIC) && defined(__i386__)
#include <machine/cpu.h>
#endif
#endif
#include <sys/queue.h>
#include <machine/bus.h>
#include <machine/endian.h>
#include <dev/usb/usb.h>
#include <dev/usb/usbdi.h>
#include <dev/usb/usbdivar.h>
#include <dev/usb/usb_mem.h>
#include <dev/usb/usb_quirks.h>
#include <dev/usb/uhcireg.h>
#include <dev/usb/uhcivar.h>
#if defined(__FreeBSD__)
#include <machine/clock.h>
#define delay(d) DELAY(d)
#endif
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
#define MS_TO_TICKS(ms) ((ms) * hz / 1000)
#if defined(__OpenBSD__)
struct cfdriver uhci_cd = {
NULL, "uhci", DV_DULL
};
#endif
#ifdef UHCI_DEBUG
#define DPRINTF(x) if (uhcidebug) printf x
#define DPRINTFN(n,x) if (uhcidebug>(n)) printf x
int uhcidebug = 1;
#else
#define DPRINTF(x)
#define DPRINTFN(n,x)
#endif
/*
* The UHCI controller is little endian, so on big endian machines
* the data strored in memory needs to be swapped.
*/
#if BYTE_ORDER == BIG_ENDIAN
#define LE(x) (bswap32(x))
#else
#define LE(x) (x)
#endif
struct uhci_pipe {
struct usbd_pipe pipe;
uhci_intr_info_t *iinfo;
int nexttoggle;
/* Info needed for different pipe kinds. */
union {
/* Control pipe */
struct {
uhci_soft_qh_t *sqh;
usb_dma_t reqdma;
uhci_soft_td_t *setup, *stat;
u_int length;
} ctl;
/* Interrupt pipe */
struct {
int npoll;
uhci_soft_qh_t **qhs;
} intr;
/* Bulk pipe */
struct {
uhci_soft_qh_t *sqh;
u_int length;
int isread;
} bulk;
/* Iso pipe */
struct iso {
uhci_soft_td_t **stds;
int next, inuse;
} iso;
} u;
};
/*
* The uhci_intr_info free list can be global since they contain
* no dma specific data. The other free lists do.
*/
LIST_HEAD(, uhci_intr_info) uhci_ii_free;
Static void uhci_busreset(uhci_softc_t *);
Static usbd_status uhci_run(uhci_softc_t *, int run);
Static uhci_soft_td_t *uhci_alloc_std(uhci_softc_t *);
Static void uhci_free_std(uhci_softc_t *, uhci_soft_td_t *);
Static uhci_soft_qh_t *uhci_alloc_sqh(uhci_softc_t *);
Static void uhci_free_sqh(uhci_softc_t *, uhci_soft_qh_t *);
Static uhci_intr_info_t *uhci_alloc_intr_info(uhci_softc_t *);
Static void uhci_free_intr_info(uhci_intr_info_t *ii);
#if 0
Static void uhci_enter_ctl_q(uhci_softc_t *, uhci_soft_qh_t *,
uhci_intr_info_t *);
Static void uhci_exit_ctl_q(uhci_softc_t *, uhci_soft_qh_t *);
#endif
Static void uhci_free_std_chain(uhci_softc_t *,
uhci_soft_td_t *, uhci_soft_td_t *);
Static usbd_status uhci_alloc_std_chain(struct uhci_pipe *,
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
uhci_softc_t *, int, int, u_int16_t, usb_dma_t *,
uhci_soft_td_t **, uhci_soft_td_t **);
Static void uhci_timo(void *);
Static void uhci_waitintr(uhci_softc_t *, usbd_xfer_handle);
Static void uhci_check_intr(uhci_softc_t *, uhci_intr_info_t *);
Static void uhci_idone(uhci_intr_info_t *);
Static void uhci_abort_xfer(usbd_xfer_handle, usbd_status status);
Static void uhci_abort_xfer_end(void *v);
Static void uhci_timeout(void *);
Static void uhci_lock_frames(uhci_softc_t *);
Static void uhci_unlock_frames(uhci_softc_t *);
Static void uhci_add_ctrl(uhci_softc_t *, uhci_soft_qh_t *);
Static void uhci_add_bulk(uhci_softc_t *, uhci_soft_qh_t *);
Static void uhci_remove_ctrl(uhci_softc_t *,uhci_soft_qh_t *);
Static void uhci_remove_bulk(uhci_softc_t *,uhci_soft_qh_t *);
Static int uhci_str(usb_string_descriptor_t *, int, char *);
Static usbd_status uhci_setup_isoc(usbd_pipe_handle pipe);
Static void uhci_device_isoc_enter(usbd_xfer_handle);
Static usbd_status uhci_allocm(struct usbd_bus *, usb_dma_t *, u_int32_t);
Static void uhci_freem(struct usbd_bus *, usb_dma_t *);
Static usbd_xfer_handle uhci_allocx(struct usbd_bus *);
Static void uhci_freex(struct usbd_bus *, usbd_xfer_handle);
Static usbd_status uhci_device_ctrl_transfer(usbd_xfer_handle);
Static usbd_status uhci_device_ctrl_start(usbd_xfer_handle);
Static void uhci_device_ctrl_abort(usbd_xfer_handle);
Static void uhci_device_ctrl_close(usbd_pipe_handle);
Static void uhci_device_ctrl_done (usbd_xfer_handle);
Static usbd_status uhci_device_intr_transfer(usbd_xfer_handle);
Static usbd_status uhci_device_intr_start(usbd_xfer_handle);
Static void uhci_device_intr_abort(usbd_xfer_handle);
Static void uhci_device_intr_close(usbd_pipe_handle);
Static void uhci_device_intr_done (usbd_xfer_handle);
Static usbd_status uhci_device_bulk_transfer(usbd_xfer_handle);
Static usbd_status uhci_device_bulk_start(usbd_xfer_handle);
Static void uhci_device_bulk_abort(usbd_xfer_handle);
Static void uhci_device_bulk_close(usbd_pipe_handle);
Static void uhci_device_bulk_done (usbd_xfer_handle);
Static usbd_status uhci_device_isoc_transfer(usbd_xfer_handle);
Static usbd_status uhci_device_isoc_start(usbd_xfer_handle);
Static void uhci_device_isoc_abort(usbd_xfer_handle);
Static void uhci_device_isoc_close(usbd_pipe_handle);
Static void uhci_device_isoc_done (usbd_xfer_handle);
Static usbd_status uhci_root_ctrl_transfer(usbd_xfer_handle);
Static usbd_status uhci_root_ctrl_start(usbd_xfer_handle);
Static void uhci_root_ctrl_abort(usbd_xfer_handle);
Static void uhci_root_ctrl_close(usbd_pipe_handle);
Static usbd_status uhci_root_intr_transfer(usbd_xfer_handle);
Static usbd_status uhci_root_intr_start(usbd_xfer_handle);
Static void uhci_root_intr_abort(usbd_xfer_handle);
Static void uhci_root_intr_close(usbd_pipe_handle);
Static void uhci_root_intr_done (usbd_xfer_handle);
Static usbd_status uhci_open(usbd_pipe_handle);
Static void uhci_poll(struct usbd_bus *);
Static usbd_status uhci_device_request(usbd_xfer_handle xfer);
Static void uhci_add_intr(uhci_softc_t *, int, uhci_soft_qh_t *);
Static void uhci_remove_intr(uhci_softc_t *, int, uhci_soft_qh_t *);
Static usbd_status uhci_device_setintr(uhci_softc_t *sc,
struct uhci_pipe *pipe, int ival);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
Static void uhci_device_clear_toggle(usbd_pipe_handle pipe);
Static void uhci_noop(usbd_pipe_handle pipe);
#ifdef UHCI_DEBUG
Static void uhci_dumpregs(uhci_softc_t *);
Static void uhci_dump_qhs(uhci_soft_qh_t *);
Static void uhci_dump_qh(uhci_soft_qh_t *);
Static void uhci_dump_tds(uhci_soft_td_t *);
Static void uhci_dump_td(uhci_soft_td_t *);
#endif
#define UWRITE1(sc, r, x) bus_space_write_1((sc)->iot, (sc)->ioh, (r), (x))
#define UWRITE2(sc, r, x) bus_space_write_2((sc)->iot, (sc)->ioh, (r), (x))
#define UWRITE4(sc, r, x) bus_space_write_4((sc)->iot, (sc)->ioh, (r), (x))
#define UREAD1(sc, r) bus_space_read_1((sc)->iot, (sc)->ioh, (r))
#define UREAD2(sc, r) bus_space_read_2((sc)->iot, (sc)->ioh, (r))
#define UREAD4(sc, r) bus_space_read_4((sc)->iot, (sc)->ioh, (r))
#define UHCICMD(sc, cmd) UWRITE2(sc, UHCI_CMD, cmd)
#define UHCISTS(sc) UREAD2(sc, UHCI_STS)
#define UHCI_RESET_TIMEOUT 100 /* reset timeout */
#define UHCI_CURFRAME(sc) (UREAD2(sc, UHCI_FRNUM) & UHCI_FRNUM_MASK)
#define UHCI_INTR_ENDPT 1
struct usbd_bus_methods uhci_bus_methods = {
uhci_open,
uhci_poll,
uhci_allocm,
uhci_freem,
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
uhci_allocx,
uhci_freex,
};
struct usbd_pipe_methods uhci_root_ctrl_methods = {
uhci_root_ctrl_transfer,
uhci_root_ctrl_start,
uhci_root_ctrl_abort,
uhci_root_ctrl_close,
uhci_noop,
0,
};
struct usbd_pipe_methods uhci_root_intr_methods = {
uhci_root_intr_transfer,
uhci_root_intr_start,
uhci_root_intr_abort,
uhci_root_intr_close,
uhci_noop,
uhci_root_intr_done,
};
struct usbd_pipe_methods uhci_device_ctrl_methods = {
uhci_device_ctrl_transfer,
uhci_device_ctrl_start,
uhci_device_ctrl_abort,
uhci_device_ctrl_close,
uhci_noop,
uhci_device_ctrl_done,
};
struct usbd_pipe_methods uhci_device_intr_methods = {
uhci_device_intr_transfer,
uhci_device_intr_start,
uhci_device_intr_abort,
uhci_device_intr_close,
uhci_device_clear_toggle,
uhci_device_intr_done,
};
struct usbd_pipe_methods uhci_device_bulk_methods = {
uhci_device_bulk_transfer,
uhci_device_bulk_start,
uhci_device_bulk_abort,
uhci_device_bulk_close,
uhci_device_clear_toggle,
uhci_device_bulk_done,
};
struct usbd_pipe_methods uhci_device_isoc_methods = {
uhci_device_isoc_transfer,
uhci_device_isoc_start,
uhci_device_isoc_abort,
uhci_device_isoc_close,
uhci_noop,
uhci_device_isoc_done,
};
void
uhci_busreset(uhci_softc_t *sc)
{
UHCICMD(sc, UHCI_CMD_GRESET); /* global reset */
usb_delay_ms(&sc->sc_bus, USB_BUS_RESET_DELAY); /* wait a little */
UHCICMD(sc, 0); /* do nothing */
}
usbd_status
uhci_init(uhci_softc_t *sc)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
int i, j;
uhci_soft_qh_t *csqh, *bsqh, *sqh;
uhci_soft_td_t *std;
DPRINTFN(1,("uhci_init: start\n"));
#ifdef UHCI_DEBUG
if (uhcidebug > 2)
uhci_dumpregs(sc);
#endif
uhci_run(sc, 0); /* stop the controller */
UWRITE2(sc, UHCI_INTR, 0); /* disable interrupts */
uhci_busreset(sc);
/* Allocate and initialize real frame array. */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = usb_allocmem(&sc->sc_bus,
UHCI_FRAMELIST_COUNT * sizeof(uhci_physaddr_t),
UHCI_FRAMELIST_ALIGN, &sc->sc_dma);
if (err)
return (err);
sc->sc_pframes = KERNADDR(&sc->sc_dma, 0);
UWRITE2(sc, UHCI_FRNUM, 0); /* set frame number to 0 */
UWRITE4(sc, UHCI_FLBASEADDR, DMAADDR(&sc->sc_dma, 0)); /* set frame list*/
/* Allocate the dummy QH where bulk traffic will be queued. */
bsqh = uhci_alloc_sqh(sc);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (bsqh == NULL)
return (USBD_NOMEM);
bsqh->qh.qh_hlink = LE(UHCI_PTR_T); /* end of QH chain */
bsqh->qh.qh_elink = LE(UHCI_PTR_T);
sc->sc_bulk_start = sc->sc_bulk_end = bsqh;
/* Allocate the dummy QH where control traffic will be queued. */
csqh = uhci_alloc_sqh(sc);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (csqh == NULL)
return (USBD_NOMEM);
csqh->hlink = bsqh;
csqh->qh.qh_hlink = LE(bsqh->physaddr | UHCI_PTR_Q);
csqh->qh.qh_elink = LE(UHCI_PTR_T);
sc->sc_ctl_start = sc->sc_ctl_end = csqh;
/*
* Make all (virtual) frame list pointers point to the interrupt
* queue heads and the interrupt queue heads at the control
* queue head and point the physical frame list to the virtual.
*/
for(i = 0; i < UHCI_VFRAMELIST_COUNT; i++) {
std = uhci_alloc_std(sc);
sqh = uhci_alloc_sqh(sc);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (std == NULL || sqh == NULL)
return (USBD_NOMEM);
std->link.sqh = sqh;
std->td.td_link = LE(sqh->physaddr | UHCI_PTR_Q);
std->td.td_status = LE(UHCI_TD_IOS); /* iso, inactive */
std->td.td_token = LE(0);
std->td.td_buffer = LE(0);
sqh->hlink = csqh;
sqh->qh.qh_hlink = LE(csqh->physaddr | UHCI_PTR_Q);
sqh->elink = 0;
sqh->qh.qh_elink = LE(UHCI_PTR_T);
sc->sc_vframes[i].htd = std;
sc->sc_vframes[i].etd = std;
sc->sc_vframes[i].hqh = sqh;
sc->sc_vframes[i].eqh = sqh;
for (j = i;
j < UHCI_FRAMELIST_COUNT;
j += UHCI_VFRAMELIST_COUNT)
sc->sc_pframes[j] = LE(std->physaddr);
}
LIST_INIT(&sc->sc_intrhead);
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
SIMPLEQ_INIT(&sc->sc_free_xfers);
/* Set up the bus struct. */
sc->sc_bus.methods = &uhci_bus_methods;
sc->sc_bus.pipe_size = sizeof(struct uhci_pipe);
sc->sc_suspend = PWR_RESUME;
#if defined(__NetBSD__)
sc->sc_powerhook = powerhook_establish(uhci_power, sc);
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
sc->sc_shutdownhook = shutdownhook_establish(uhci_shutdown, sc);
#endif
DPRINTFN(1,("uhci_init: enabling\n"));
UWRITE2(sc, UHCI_INTR, UHCI_INTR_TOCRCIE | UHCI_INTR_RIE |
UHCI_INTR_IOCE | UHCI_INTR_SPIE); /* enable interrupts */
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
UHCICMD(sc, UHCI_CMD_MAXP); /* Assume 64 byte packets at frame end */
return (uhci_run(sc, 1)); /* and here we go... */
}
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
#if defined(__NetBSD__) || defined(__OpenBSD__)
int
uhci_activate(device_ptr_t self, enum devact act)
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
{
struct uhci_softc *sc = (struct uhci_softc *)self;
int rv = 0;
switch (act) {
case DVACT_ACTIVATE:
return (EOPNOTSUPP);
break;
case DVACT_DEACTIVATE:
if (sc->sc_child != NULL)
rv = config_deactivate(sc->sc_child);
break;
}
return (rv);
}
int
uhci_detach(struct uhci_softc *sc, int flags)
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +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
usbd_xfer_handle xfer;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
int rv = 0;
if (sc->sc_child != NULL)
rv = config_detach(sc->sc_child, flags);
if (rv != 0)
return (rv);
#if defined(__NetBSD__)
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
powerhook_disestablish(sc->sc_powerhook);
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
shutdownhook_disestablish(sc->sc_shutdownhook);
#endif
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
/* Free all xfers associated with this HC. */
for (;;) {
xfer = SIMPLEQ_FIRST(&sc->sc_free_xfers);
if (xfer == NULL)
break;
SIMPLEQ_REMOVE_HEAD(&sc->sc_free_xfers, xfer, next);
free(xfer, M_USB);
}
/* XXX free other data structures XXX */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
return (rv);
}
#endif
usbd_status
uhci_allocm(struct usbd_bus *bus, usb_dma_t *dma, u_int32_t size)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
return (usb_allocmem(&((struct uhci_softc *)bus)->sc_bus, size, 0,
dma));
}
void
uhci_freem(struct usbd_bus *bus, usb_dma_t *dma)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usb_freemem(&((struct uhci_softc *)bus)->sc_bus, dma);
}
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
usbd_xfer_handle
uhci_allocx(struct usbd_bus *bus)
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
{
struct uhci_softc *sc = (struct uhci_softc *)bus;
usbd_xfer_handle xfer;
xfer = SIMPLEQ_FIRST(&sc->sc_free_xfers);
if (xfer != NULL)
SIMPLEQ_REMOVE_HEAD(&sc->sc_free_xfers, xfer, next);
else
xfer = malloc(sizeof(*xfer), M_USB, M_NOWAIT);
if (xfer != NULL)
memset(xfer, 0, sizeof *xfer);
return (xfer);
}
void
uhci_freex(struct usbd_bus *bus, usbd_xfer_handle xfer)
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
{
struct uhci_softc *sc = (struct uhci_softc *)bus;
SIMPLEQ_INSERT_HEAD(&sc->sc_free_xfers, xfer, next);
}
/*
* Shut down the controller when the system is going down.
*/
void
uhci_shutdown(void *v)
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
{
uhci_softc_t *sc = v;
DPRINTF(("uhci_shutdown: stopping the HC\n"));
uhci_run(sc, 0); /* stop the controller */
}
/*
* Handle suspend/resume.
*
* We need to switch to polling mode here, because this routine is
* called from an intterupt context. This is all right since we
* are almost suspended anyway.
*/
void
uhci_power(int why, void *v)
{
uhci_softc_t *sc = v;
int cmd;
int s;
s = splusb();
cmd = UREAD2(sc, UHCI_CMD);
DPRINTF(("uhci_power: sc=%p, why=%d (was %d), cmd=0x%x\n",
sc, why, sc->sc_suspend, cmd));
if (why != PWR_RESUME) {
#ifdef UHCI_DEBUG
if (uhcidebug > 2)
uhci_dumpregs(sc);
#endif
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
if (sc->sc_has_timo != NULL)
usb_untimeout(uhci_timo, sc->sc_has_timo,
sc->sc_has_timo->timo_handle);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
sc->sc_bus.use_polling++;
uhci_run(sc, 0); /* stop the controller */
/* save some state if BIOS doesn't */
sc->sc_saved_frnum = UREAD2(sc, UHCI_FRNUM);
sc->sc_saved_sof = UREAD1(sc, UHCI_SOF);
UHCICMD(sc, cmd | UHCI_CMD_EGSM); /* enter global suspend */
usb_delay_ms(&sc->sc_bus, USB_RESUME_WAIT);
sc->sc_suspend = why;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
sc->sc_bus.use_polling--;
DPRINTF(("uhci_power: cmd=0x%x\n", UREAD2(sc, UHCI_CMD)));
} else {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
#ifdef DIAGNOSTIC
if (sc->sc_suspend == PWR_RESUME)
printf("uhci_power: weird, resume without suspend.\n");
#endif
sc->sc_bus.use_polling++;
sc->sc_suspend = why;
if (cmd & UHCI_CMD_RS)
uhci_run(sc, 0); /* in case BIOS has started it */
/* restore saved state */
UWRITE4(sc, UHCI_FLBASEADDR, DMAADDR(&sc->sc_dma, 0));
UWRITE2(sc, UHCI_FRNUM, sc->sc_saved_frnum);
UWRITE1(sc, UHCI_SOF, sc->sc_saved_sof);
UHCICMD(sc, cmd | UHCI_CMD_FGR); /* force global resume */
usb_delay_ms(&sc->sc_bus, USB_RESUME_DELAY);
UHCICMD(sc, cmd & ~UHCI_CMD_EGSM); /* back to normal */
UWRITE2(sc, UHCI_INTR, UHCI_INTR_TOCRCIE | UHCI_INTR_RIE |
UHCI_INTR_IOCE | UHCI_INTR_SPIE); /* re-enable intrs */
uhci_run(sc, 1); /* and start traffic again */
usb_delay_ms(&sc->sc_bus, USB_RESUME_RECOVERY);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
sc->sc_bus.use_polling--;
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
if (sc->sc_has_timo != NULL)
usb_timeout(uhci_timo, sc->sc_has_timo,
sc->sc_ival, sc->sc_has_timo->timo_handle);
#ifdef UHCI_DEBUG
if (uhcidebug > 2)
uhci_dumpregs(sc);
#endif
}
splx(s);
}
#ifdef UHCI_DEBUG
Static void
uhci_dumpregs(uhci_softc_t *sc)
{
DPRINTFN(-1,("%s regs: cmd=%04x, sts=%04x, intr=%04x, frnum=%04x, "
"flbase=%08x, sof=%04x, portsc1=%04x, portsc2=%04x\n",
USBDEVNAME(sc->sc_bus.bdev),
UREAD2(sc, UHCI_CMD),
UREAD2(sc, UHCI_STS),
UREAD2(sc, UHCI_INTR),
UREAD2(sc, UHCI_FRNUM),
UREAD4(sc, UHCI_FLBASEADDR),
UREAD1(sc, UHCI_SOF),
UREAD2(sc, UHCI_PORTSC1),
UREAD2(sc, UHCI_PORTSC2)));
}
void
uhci_dump_td(uhci_soft_td_t *p)
{
DPRINTFN(-1,("TD(%p) at %08lx = link=0x%08lx status=0x%08lx "
"token=0x%08lx buffer=0x%08lx\n",
p, (long)p->physaddr,
(long)LE(p->td.td_link),
(long)LE(p->td.td_status),
(long)LE(p->td.td_token),
(long)LE(p->td.td_buffer)));
DPRINTFN(-1,(" %b %b,errcnt=%d,actlen=%d pid=%02x,addr=%d,endpt=%d,"
"D=%d,maxlen=%d\n",
(int)LE(p->td.td_link),
"\20\1T\2Q\3VF",
(int)LE(p->td.td_status),
"\20\22BITSTUFF\23CRCTO\24NAK\25BABBLE\26DBUFFER\27"
"STALLED\30ACTIVE\31IOC\32ISO\33LS\36SPD",
UHCI_TD_GET_ERRCNT(LE(p->td.td_status)),
UHCI_TD_GET_ACTLEN(LE(p->td.td_status)),
UHCI_TD_GET_PID(LE(p->td.td_token)),
UHCI_TD_GET_DEVADDR(LE(p->td.td_token)),
UHCI_TD_GET_ENDPT(LE(p->td.td_token)),
UHCI_TD_GET_DT(LE(p->td.td_token)),
UHCI_TD_GET_MAXLEN(LE(p->td.td_token))));
}
void
uhci_dump_qh(uhci_soft_qh_t *sqh)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(-1,("QH(%p) at %08x: hlink=%08x elink=%08x\n", sqh,
(int)sqh->physaddr, LE(sqh->qh.qh_hlink), LE(sqh->qh.qh_elink)));
}
#if 0
void
uhci_dump()
{
uhci_softc_t *sc = uhci;
uhci_dumpregs(sc);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
printf("intrs=%d\n", sc->sc_bus.no_intrs);
printf("framelist[i].link = %08x\n", sc->sc_framelist[0].link);
uhci_dump_qh(sc->sc_ctl_start->qh.hlink);
}
#endif
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
void
uhci_dump_qhs(uhci_soft_qh_t *sqh)
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
{
uhci_dump_qh(sqh);
/* uhci_dump_qhs displays all the QHs and TDs from the given QH onwards
* Traverses sideways first, then down.
*
* QH1
* QH2
* No QH
* TD2.1
* TD2.2
* TD1.1
* etc.
*
* TD2.x being the TDs queued at QH2 and QH1 being referenced from QH1.
*/
if (sqh->hlink != NULL && !(sqh->qh.qh_hlink & UHCI_PTR_T))
uhci_dump_qhs(sqh->hlink);
else
DPRINTF(("No QH\n"));
if (sqh->elink != NULL && !(sqh->qh.qh_elink & UHCI_PTR_T))
uhci_dump_tds(sqh->elink);
else
DPRINTF(("No TD\n"));
}
void
uhci_dump_tds(uhci_soft_td_t *std)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_soft_td_t *td;
for(td = std; td != NULL; td = td->link.std) {
uhci_dump_td(td);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
/* Check whether the link pointer in this TD marks
* the link pointer as end of queue. This avoids
* printing the free list in case the queue/TD has
* already been moved there (seatbelt).
*/
if (td->td.td_link & UHCI_PTR_T ||
td->td.td_link == 0)
break;
}
}
#endif
/*
* This routine is executed periodically and simulates interrupts
* from the root controller interrupt pipe for port status change.
*/
void
uhci_timo(void *addr)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_xfer_handle xfer = addr;
usbd_pipe_handle pipe = xfer->pipe;
uhci_softc_t *sc = (uhci_softc_t *)pipe->device->bus;
int s;
u_char *p;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(20, ("uhci_timo\n"));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usb_timeout(uhci_timo, xfer, sc->sc_ival, xfer->timo_handle);
p = KERNADDR(&xfer->dmabuf, 0);
p[0] = 0;
if (UREAD2(sc, UHCI_PORTSC1) & (UHCI_PORTSC_CSC|UHCI_PORTSC_OCIC))
p[0] |= 1<<1;
if (UREAD2(sc, UHCI_PORTSC2) & (UHCI_PORTSC_CSC|UHCI_PORTSC_OCIC))
p[0] |= 1<<2;
if (p[0] == 0)
/* No change, try again in a while */
return;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->actlen = 1;
xfer->status = USBD_NORMAL_COMPLETION;
s = splusb();
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->hcpriv = 0;
xfer->device->bus->intr_context++;
usb_transfer_complete(xfer);
xfer->device->bus->intr_context--;
splx(s);
}
void
uhci_root_intr_done(usbd_xfer_handle xfer)
{
}
void
uhci_lock_frames(uhci_softc_t *sc)
{
int s = splusb();
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +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
while (sc->sc_vflock & UHCI_HAS_LOCK) {
sc->sc_vflock |= UHCI_WANT_LOCK;
tsleep(&sc->sc_vflock, PRIBIO, "uhcqhl", 0);
}
sc->sc_vflock = UHCI_HAS_LOCK;
splx(s);
}
void
uhci_unlock_frames(uhci_softc_t *sc)
{
int s = splusb();
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
sc->sc_vflock &= ~UHCI_HAS_LOCK;
if (sc->sc_vflock & UHCI_WANT_LOCK)
wakeup(&sc->sc_vflock);
splx(s);
}
/*
* Allocate an interrupt information struct. A free list is kept
* for fast allocation.
*/
uhci_intr_info_t *
uhci_alloc_intr_info(uhci_softc_t *sc)
{
uhci_intr_info_t *ii;
ii = LIST_FIRST(&uhci_ii_free);
if (ii)
LIST_REMOVE(ii, list);
else {
ii = malloc(sizeof(uhci_intr_info_t), M_USBHC, M_NOWAIT);
}
ii->sc = sc;
#if defined(__FreeBSD__)
callout_handle_init(&ii->timeout_handle);
#endif
return ii;
}
void
uhci_free_intr_info(uhci_intr_info_t *ii)
{
LIST_INSERT_HEAD(&uhci_ii_free, ii, list); /* and put on free list */
}
/* Add control QH, called at splusb(). */
void
uhci_add_ctrl(uhci_softc_t *sc, uhci_soft_qh_t *sqh)
{
uhci_soft_qh_t *eqh;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
SPLUSBCHECK;
DPRINTFN(10, ("uhci_add_ctrl: sqh=%p\n", sqh));
eqh = sc->sc_ctl_end;
sqh->hlink = eqh->hlink;
sqh->qh.qh_hlink = eqh->qh.qh_hlink;
eqh->hlink = sqh;
eqh->qh.qh_hlink = LE(sqh->physaddr | UHCI_PTR_Q);
sc->sc_ctl_end = sqh;
}
/* Remove control QH, called at splusb(). */
void
uhci_remove_ctrl(uhci_softc_t *sc, uhci_soft_qh_t *sqh)
{
uhci_soft_qh_t *pqh;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
SPLUSBCHECK;
DPRINTFN(10, ("uhci_remove_ctrl: sqh=%p\n", sqh));
for (pqh = sc->sc_ctl_start; pqh->hlink != sqh; pqh=pqh->hlink)
#if defined(DIAGNOSTIC) || defined(UHCI_DEBUG)
if (LE(pqh->qh.qh_hlink) & UHCI_PTR_T) {
printf("uhci_remove_ctrl: QH not found\n");
return;
}
#else
;
#endif
pqh->hlink = sqh->hlink;
pqh->qh.qh_hlink = sqh->qh.qh_hlink;
if (sc->sc_ctl_end == sqh)
sc->sc_ctl_end = pqh;
}
/* Add bulk QH, called at splusb(). */
void
uhci_add_bulk(uhci_softc_t *sc, uhci_soft_qh_t *sqh)
{
uhci_soft_qh_t *eqh;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
SPLUSBCHECK;
DPRINTFN(10, ("uhci_add_bulk: sqh=%p\n", sqh));
eqh = sc->sc_bulk_end;
sqh->hlink = eqh->hlink;
sqh->qh.qh_hlink = eqh->qh.qh_hlink;
eqh->hlink = sqh;
eqh->qh.qh_hlink = LE(sqh->physaddr | UHCI_PTR_Q);
sc->sc_bulk_end = sqh;
}
/* Remove bulk QH, called at splusb(). */
void
uhci_remove_bulk(uhci_softc_t *sc, uhci_soft_qh_t *sqh)
{
uhci_soft_qh_t *pqh;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
SPLUSBCHECK;
DPRINTFN(10, ("uhci_remove_bulk: sqh=%p\n", sqh));
for (pqh = sc->sc_bulk_start; pqh->hlink != sqh; pqh = pqh->hlink)
#if defined(DIAGNOSTIC) || defined(UHCI_DEBUG)
if (LE(pqh->qh.qh_hlink) & UHCI_PTR_T) {
printf("uhci_remove_bulk: QH not found\n");
return;
}
#else
;
#endif
pqh->hlink = sqh->hlink;
pqh->qh.qh_hlink = sqh->qh.qh_hlink;
if (sc->sc_bulk_end == sqh)
sc->sc_bulk_end = pqh;
}
int
uhci_intr(void *arg)
{
uhci_softc_t *sc = arg;
int status;
int ack;
uhci_intr_info_t *ii;
/*
* It can happen that an interrupt will be delivered to
* us before the device has been fully attached and the
* softc struct has been configured. Usually this happens
* when kldloading the USB support as a module after the
* system has been booted. If we detect this condition,
* we need to squelch the unwanted interrupts until we're
* ready for them.
*/
if (sc->sc_bus.bdev == NULL) {
UWRITE2(sc, UHCI_STS, 0xFFFF); /* ack pending interrupts */
uhci_run(sc, 0); /* stop the controller */
UWRITE2(sc, UHCI_INTR, 0); /* disable interrupts */
return(0);
}
#ifdef UHCI_DEBUG
if (uhcidebug > 15) {
DPRINTF(("%s: uhci_intr\n", USBDEVNAME(sc->sc_bus.bdev)));
uhci_dumpregs(sc);
}
#endif
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
status = UREAD2(sc, UHCI_STS);
if (status == 0) /* The interrupt was not for us. */
return (0);
#if defined(DIAGNOSTIC) && defined(__NetBSD__)
if (sc->sc_suspend != PWR_RESUME)
printf("uhci_intr: suspended sts=0x%x\n", status);
#endif
ack = 0;
if (status & UHCI_STS_USBINT)
ack |= UHCI_STS_USBINT;
if (status & UHCI_STS_USBEI)
ack |= UHCI_STS_USBEI;
if (status & UHCI_STS_RD) {
ack |= UHCI_STS_RD;
printf("%s: resume detect\n", USBDEVNAME(sc->sc_bus.bdev));
}
if (status & UHCI_STS_HSE) {
ack |= UHCI_STS_HSE;
printf("%s: host system error\n", USBDEVNAME(sc->sc_bus.bdev));
}
if (status & UHCI_STS_HCPE) {
ack |= UHCI_STS_HCPE;
printf("%s: host controller process error\n",
USBDEVNAME(sc->sc_bus.bdev));
}
if (status & UHCI_STS_HCH) {
/* no acknowledge needed */
printf("%s: host controller halted\n",
USBDEVNAME(sc->sc_bus.bdev));
}
if (ack) /* acknowledge the ints */
UWRITE2(sc, UHCI_STS, ack);
else /* nothing to acknowledge */
return (0);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
sc->sc_bus.intr_context++;
sc->sc_bus.no_intrs++;
/*
* Interrupts on UHCI really suck. When the host controller
* interrupts because a transfer is completed there is no
* way of knowing which transfer it was. You can scan down
* the TDs and QHs of the previous frame to limit the search,
* but that assumes that the interrupt was not delayed by more
* than 1 ms, which may not always be true (e.g. after debug
* output on a slow console).
* We scan all interrupt descriptors to see if any have
* completed.
*/
LIST_FOREACH(ii, &sc->sc_intrhead, list)
uhci_check_intr(sc, ii);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(10, ("%s: uhci_intr: exit\n", USBDEVNAME(sc->sc_bus.bdev)));
sc->sc_bus.intr_context--;
return (1);
}
/* Check for an interrupt. */
void
uhci_check_intr(uhci_softc_t *sc, uhci_intr_info_t *ii)
{
uhci_soft_td_t *std, *lstd;
u_int32_t status;
DPRINTFN(15, ("uhci_check_intr: ii=%p\n", ii));
#ifdef DIAGNOSTIC
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (ii == NULL) {
printf("uhci_check_intr: no ii? %p\n", ii);
return;
}
#endif
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (ii->stdstart == NULL)
return;
lstd = ii->stdend;
#ifdef DIAGNOSTIC
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (lstd == NULL) {
printf("uhci_check_intr: std==0\n");
return;
}
#endif
/*
* If the last TD is still active we need to check whether there
* is a an error somewhere in the middle, or whether there was a
* short packet (SPD and not ACTIVE).
*/
if (LE(lstd->td.td_status) & UHCI_TD_ACTIVE) {
DPRINTFN(15, ("uhci_check_intr: active ii=%p\n", ii));
for (std = ii->stdstart; std != lstd; std = std->link.std) {
status = LE(std->td.td_status);
/* If there's an active TD the xfer isn't done. */
if (status & UHCI_TD_ACTIVE)
break;
/* Any kind of error makes the xfer done. */
if (status & UHCI_TD_STALLED)
goto done;
/*
* We want short packets,
* and it is short: it's done
*/
if ((status & UHCI_TD_SPD) &&
UHCI_TD_GET_ACTLEN(status) <
UHCI_TD_GET_MAXLEN(LE(std->td.td_token)))
goto done;
}
DPRINTFN(15, ("uhci_check_intr: ii=%p std=%p still active\n",
ii, ii->stdstart));
return;
}
done:
usb_untimeout(uhci_timeout, ii, ii->timeout_handle);
uhci_idone(ii);
}
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
/* Called at splusb() */
void
uhci_idone(uhci_intr_info_t *ii)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_xfer_handle xfer = ii->xfer;
struct uhci_pipe *upipe = (struct uhci_pipe *)xfer->pipe;
uhci_soft_td_t *std;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
u_int32_t status = 0, nstatus;
int actlen;
#ifdef DIAGNOSTIC
{
int s = splhigh();
if (ii->isdone) {
splx(s);
printf("uhci_idone: ii=%p is done!\n", ii);
return;
}
ii->isdone = 1;
splx(s);
}
#endif
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->status == USBD_CANCELLED ||
xfer->status == USBD_TIMEOUT) {
DPRINTF(("uhci_idone: aborted xfer=%p\n", xfer));
return;
}
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->nframes != 0) {
/* Isoc transfer, do things differently. */
uhci_soft_td_t **stds = upipe->u.iso.stds;
int i, n, nframes;
DPRINTFN(5,("uhci_idone: ii=%p isoc ready\n", ii));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
nframes = xfer->nframes;
actlen = 0;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
n = xfer->hcprivint;
for (i = 0; i < nframes; i++) {
std = stds[n];
#ifdef UHCI_DEBUG
if (uhcidebug > 5) {
DPRINTFN(-1,("uhci_idone: isoc TD %d\n", i));
uhci_dump_td(std);
}
#endif
if (++n >= UHCI_VFRAMELIST_COUNT)
n = 0;
status = LE(std->td.td_status);
actlen += UHCI_TD_GET_ACTLEN(status);
}
upipe->u.iso.inuse -= nframes;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->actlen = actlen;
xfer->status = USBD_NORMAL_COMPLETION;
xfer->hcpriv = ii;
usb_transfer_complete(xfer);
return;
}
#ifdef UHCI_DEBUG
DPRINTFN(10, ("uhci_idone: ii=%p, xfer=%p, pipe=%p ready\n",
ii, xfer, upipe));
if (uhcidebug > 10)
uhci_dump_tds(ii->stdstart);
#endif
/* The transfer is done, compute actual length and status. */
actlen = 0;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
for (std = ii->stdstart; std != NULL; std = std->link.std) {
nstatus = LE(std->td.td_status);
if (nstatus & UHCI_TD_ACTIVE)
break;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
status = nstatus;
if (UHCI_TD_GET_PID(LE(std->td.td_token)) != UHCI_TD_PID_SETUP)
actlen += UHCI_TD_GET_ACTLEN(status);
}
/* If there are left over TDs we need to update the toggle. */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (std != NULL)
upipe->nexttoggle = UHCI_TD_GET_DT(LE(std->td.td_token));
status &= UHCI_TD_ERROR;
DPRINTFN(10, ("uhci_check_intr: actlen=%d, status=0x%x\n",
actlen, status));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->actlen = actlen;
if (status != 0) {
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
DPRINTFN((status == UHCI_TD_STALLED)*10,
("uhci_idone: error, addr=%d, endpt=0x%02x, "
"status 0x%b\n",
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->pipe->device->address,
xfer->pipe->endpoint->edesc->bEndpointAddress,
(int)status,
"\20\22BITSTUFF\23CRCTO\24NAK\25BABBLE\26DBUFFER\27"
"STALLED\30ACTIVE"));
if (status == UHCI_TD_STALLED)
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->status = USBD_STALLED;
else
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->status = USBD_IOERROR; /* more info XXX */
} else {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->status = USBD_NORMAL_COMPLETION;
}
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->hcpriv = ii;
usb_transfer_complete(xfer);
}
/*
* Called when a request does not complete.
*/
void
uhci_timeout(void *addr)
{
uhci_intr_info_t *ii = addr;
DPRINTF(("uhci_timeout: ii=%p\n", ii));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
#ifdef UHCI_DEBUG
if (uhcidebug > 10)
uhci_dump_tds(ii->stdstart);
#endif
ii->xfer->device->bus->intr_context++;
uhci_abort_xfer(ii->xfer, USBD_TIMEOUT);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
ii->xfer->device->bus->intr_context--;
}
/*
* Wait here until controller claims to have an interrupt.
* Then call uhci_intr and return. Use timeout to avoid waiting
* too long.
* Only used during boot when interrupts are not enabled yet.
*/
void
uhci_waitintr(uhci_softc_t *sc, usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
int timo = xfer->timeout;
uhci_intr_info_t *ii;
DPRINTFN(10,("uhci_waitintr: timeout = %dms\n", timo));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->status = USBD_IN_PROGRESS;
for (; timo >= 0; timo--) {
usb_delay_ms(&sc->sc_bus, 1);
DPRINTFN(20,("uhci_waitintr: 0x%04x\n", UREAD2(sc, UHCI_STS)));
if (UREAD2(sc, UHCI_STS) & UHCI_STS_USBINT) {
uhci_intr(sc);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->status != USBD_IN_PROGRESS)
return;
}
}
/* Timeout */
DPRINTF(("uhci_waitintr: timeout\n"));
for (ii = LIST_FIRST(&sc->sc_intrhead);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
ii != NULL && ii->xfer != xfer;
ii = LIST_NEXT(ii, list))
;
#ifdef DIAGNOSTIC
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (ii == NULL)
panic("uhci_waitintr: lost intr_info\n");
#endif
uhci_idone(ii);
}
void
uhci_poll(struct usbd_bus *bus)
{
uhci_softc_t *sc = (uhci_softc_t *)bus;
if (UREAD2(sc, UHCI_STS) & UHCI_STS_USBINT)
uhci_intr(sc);
}
#if 0
void
uhci_reset(void *p)
{
uhci_softc_t *sc = p;
int n;
UHCICMD(sc, UHCI_CMD_HCRESET);
/* The reset bit goes low when the controller is done. */
for (n = 0; n < UHCI_RESET_TIMEOUT &&
(UREAD2(sc, UHCI_CMD) & UHCI_CMD_HCRESET); n++)
delay(100);
if (n >= UHCI_RESET_TIMEOUT)
printf("%s: controller did not reset\n",
USBDEVNAME(sc->sc_bus.bdev));
}
#endif
usbd_status
uhci_run(uhci_softc_t *sc, int run)
{
int s, n, running;
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
u_int16_t cmd;
run = run != 0;
s = splusb();
DPRINTF(("uhci_run: setting run=%d\n", run));
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
cmd = UREAD2(sc, UHCI_CMD);
if (run)
cmd |= UHCI_CMD_RS;
else
cmd &= ~UHCI_CMD_RS;
UHCICMD(sc, cmd);
for(n = 0; n < 10; n++) {
running = !(UREAD2(sc, UHCI_STS) & UHCI_STS_HCH);
/* return when we've entered the state we want */
if (run == running) {
splx(s);
DPRINTF(("uhci_run: done cmd=0x%x sts=0x%x\n",
UREAD2(sc, UHCI_CMD), UREAD2(sc, UHCI_STS)));
return (USBD_NORMAL_COMPLETION);
}
usb_delay_ms(&sc->sc_bus, 1);
}
splx(s);
printf("%s: cannot %s\n", USBDEVNAME(sc->sc_bus.bdev),
run ? "start" : "stop");
return (USBD_IOERROR);
}
/*
* Memory management routines.
* uhci_alloc_std allocates TDs
* uhci_alloc_sqh allocates QHs
* These two routines do their own free list management,
* partly for speed, partly because allocating DMAable memory
* has page size granularaity so much memory would be wasted if
* only one TD/QH (32 bytes) was placed in each allocated chunk.
*/
uhci_soft_td_t *
uhci_alloc_std(uhci_softc_t *sc)
{
uhci_soft_td_t *std;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
int i, offs;
usb_dma_t dma;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (sc->sc_freetds == NULL) {
DPRINTFN(2,("uhci_alloc_std: allocating chunk\n"));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = usb_allocmem(&sc->sc_bus, UHCI_STD_SIZE * UHCI_STD_CHUNK,
UHCI_TD_ALIGN, &dma);
if (err)
return (0);
for(i = 0; i < UHCI_STD_CHUNK; i++) {
offs = i * UHCI_STD_SIZE;
std = (uhci_soft_td_t *)((char *)KERNADDR(&dma, offs));
std->physaddr = DMAADDR(&dma, offs);
std->link.std = sc->sc_freetds;
sc->sc_freetds = std;
}
}
std = sc->sc_freetds;
sc->sc_freetds = std->link.std;
memset(&std->td, 0, sizeof(uhci_td_t));
return std;
}
void
uhci_free_std(uhci_softc_t *sc, uhci_soft_td_t *std)
{
#ifdef DIAGNOSTIC
#define TD_IS_FREE 0x12345678
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (std->td.td_token == LE(TD_IS_FREE)) {
printf("uhci_free_std: freeing free TD %p\n", std);
return;
}
std->td.td_token = LE(TD_IS_FREE);
#endif
std->link.std = sc->sc_freetds;
sc->sc_freetds = std;
}
uhci_soft_qh_t *
uhci_alloc_sqh(uhci_softc_t *sc)
{
uhci_soft_qh_t *sqh;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
int i, offs;
usb_dma_t dma;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (sc->sc_freeqhs == NULL) {
DPRINTFN(2, ("uhci_alloc_sqh: allocating chunk\n"));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = usb_allocmem(&sc->sc_bus, UHCI_SQH_SIZE * UHCI_SQH_CHUNK,
UHCI_QH_ALIGN, &dma);
if (err)
return (0);
for(i = 0; i < UHCI_SQH_CHUNK; i++) {
offs = i * UHCI_SQH_SIZE;
sqh = (uhci_soft_qh_t *)((char *)KERNADDR(&dma, offs));
sqh->physaddr = DMAADDR(&dma, offs);
sqh->hlink = sc->sc_freeqhs;
sc->sc_freeqhs = sqh;
}
}
sqh = sc->sc_freeqhs;
sc->sc_freeqhs = sqh->hlink;
memset(&sqh->qh, 0, sizeof(uhci_qh_t));
return (sqh);
}
void
uhci_free_sqh(uhci_softc_t *sc, uhci_soft_qh_t *sqh)
{
sqh->hlink = sc->sc_freeqhs;
sc->sc_freeqhs = sqh;
}
#if 0
/*
* Enter a list of transfers onto a control queue.
* Called at splusb()
*/
void
uhci_enter_ctl_q(uhci_softc_t *sc, uhci_soft_qh_t *sqh, uhci_intr_info_t *ii)
{
DPRINTFN(5, ("uhci_enter_ctl_q: sqh=%p\n", sqh));
}
#endif
void
uhci_free_std_chain(uhci_softc_t *sc, uhci_soft_td_t *std,
uhci_soft_td_t *stdend)
{
uhci_soft_td_t *p;
for (; std != stdend; std = p) {
p = std->link.std;
uhci_free_std(sc, std);
}
}
usbd_status
uhci_alloc_std_chain(struct uhci_pipe *upipe, uhci_softc_t *sc,
int len, int rd, u_int16_t flags, usb_dma_t *dma,
uhci_soft_td_t **sp, uhci_soft_td_t **ep)
{
uhci_soft_td_t *p, *lastp;
uhci_physaddr_t lastlink;
int i, ntd, l, tog, maxp;
u_int32_t status;
int addr = upipe->pipe.device->address;
int endpt = upipe->pipe.endpoint->edesc->bEndpointAddress;
DPRINTFN(8, ("uhci_alloc_std_chain: addr=%d endpt=%d len=%d ls=%d "
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
"flags=0x%x\n", addr, UE_GET_ADDR(endpt), len,
upipe->pipe.device->lowspeed, flags));
maxp = UGETW(upipe->pipe.endpoint->edesc->wMaxPacketSize);
if (maxp == 0) {
printf("uhci_alloc_std_chain: maxp=0\n");
return (USBD_INVAL);
}
ntd = (len + maxp - 1) / maxp;
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
if ((flags & USBD_FORCE_SHORT_XFER) && len % maxp == 0)
ntd++;
DPRINTFN(10, ("uhci_alloc_std_chain: maxp=%d ntd=%d\n", maxp, ntd));
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
if (ntd == 0) {
*sp = *ep = 0;
DPRINTFN(-1,("uhci_alloc_std_chain: ntd=0\n"));
return (USBD_NORMAL_COMPLETION);
}
tog = upipe->nexttoggle;
if (ntd % 2 == 0)
tog ^= 1;
upipe->nexttoggle = tog ^ 1;
lastp = 0;
lastlink = UHCI_PTR_T;
ntd--;
status = UHCI_TD_ZERO_ACTLEN(UHCI_TD_SET_ERRCNT(3) | UHCI_TD_ACTIVE);
if (upipe->pipe.device->lowspeed)
status |= UHCI_TD_LS;
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
if (flags & USBD_SHORT_XFER_OK)
status |= UHCI_TD_SPD;
for (i = ntd; i >= 0; i--) {
p = uhci_alloc_std(sc);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (p == NULL) {
uhci_free_std_chain(sc, lastp, 0);
return (USBD_NOMEM);
}
p->link.std = lastp;
if (lastlink == UHCI_PTR_T)
p->td.td_link = LE(lastlink);
else
p->td.td_link = LE(lastlink|UHCI_PTR_VF);
lastp = p;
lastlink = p->physaddr;
p->td.td_status = LE(status);
if (i == ntd) {
/* last TD */
l = len % maxp;
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
if (l == 0 && !(flags & USBD_FORCE_SHORT_XFER))
l = maxp;
*ep = p;
} else
l = maxp;
p->td.td_token =
LE(rd ? UHCI_TD_IN (l, endpt, addr, tog) :
UHCI_TD_OUT(l, endpt, addr, tog));
p->td.td_buffer = LE(DMAADDR(dma, i * maxp));
tog ^= 1;
}
*sp = lastp;
DPRINTFN(10, ("uhci_alloc_std_chain: nexttog=%d\n",
upipe->nexttoggle));
return (USBD_NORMAL_COMPLETION);
}
void
uhci_device_clear_toggle(usbd_pipe_handle pipe)
{
struct uhci_pipe *upipe = (struct uhci_pipe *)pipe;
upipe->nexttoggle = 0;
}
void
uhci_noop(usbd_pipe_handle pipe)
{
}
usbd_status
uhci_device_bulk_transfer(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
/* Insert last in queue. */
err = usb_insert_transfer(xfer);
if (err)
return (err);
/* Pipe isn't running (otherwise err would be USBD_INPROG),
* start first
*/
return (uhci_device_bulk_start(SIMPLEQ_FIRST(&xfer->pipe->queue)));
}
usbd_status
uhci_device_bulk_start(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
struct uhci_pipe *upipe = (struct uhci_pipe *)xfer->pipe;
usbd_device_handle dev = upipe->pipe.device;
uhci_softc_t *sc = (uhci_softc_t *)dev->bus;
uhci_intr_info_t *ii = upipe->iinfo;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_soft_td_t *data, *dataend;
uhci_soft_qh_t *sqh;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
int len, isread, endpt;
int s;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(3, ("uhci_device_bulk_transfer: xfer=%p len=%d flags=%d\n",
xfer, xfer->length, xfer->flags));
#ifdef DIAGNOSTIC
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->rqflags & URQ_REQUEST)
panic("uhci_device_bulk_transfer: a request\n");
#endif
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
len = xfer->length;
endpt = xfer->pipe->endpoint->edesc->bEndpointAddress;
isread = UE_GET_DIR(endpt) == UE_DIR_IN;
sqh = upipe->u.bulk.sqh;
upipe->u.bulk.isread = isread;
upipe->u.bulk.length = len;
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
err = uhci_alloc_std_chain(upipe, sc, len, isread, xfer->flags,
&xfer->dmabuf, &data, &dataend);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (err)
return (err);
dataend->td.td_status |= LE(UHCI_TD_IOC);
#ifdef UHCI_DEBUG
if (uhcidebug > 8) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTF(("uhci_device_bulk_transfer: data(1)\n"));
uhci_dump_tds(data);
}
#endif
/* Set up interrupt info. */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
ii->xfer = xfer;
ii->stdstart = data;
ii->stdend = dataend;
#if defined(__FreeBSD__)
callout_handle_init(&ii->timeout_handle);
#endif
#ifdef DIAGNOSTIC
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
if (!ii->isdone) {
printf("uhci_device_bulk_transfer: not done, ii=%p\n", ii);
}
ii->isdone = 0;
#endif
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
sqh->elink = data;
sqh->qh.qh_elink = LE(data->physaddr);
sqh->intr_info = ii;
s = splusb();
uhci_add_bulk(sc, sqh);
LIST_INSERT_HEAD(&sc->sc_intrhead, ii, list);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->timeout && !sc->sc_bus.use_polling) {
usb_timeout(uhci_timeout, ii, MS_TO_TICKS(xfer->timeout),
ii->timeout_handle);
}
splx(s);
#ifdef UHCI_DEBUG
if (uhcidebug > 10) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTF(("uhci_device_bulk_transfer: data(2)\n"));
uhci_dump_tds(data);
}
#endif
if (sc->sc_bus.use_polling)
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_waitintr(sc, xfer);
return (USBD_IN_PROGRESS);
}
/* Abort a device bulk request. */
void
uhci_device_bulk_abort(usbd_xfer_handle xfer)
{
DPRINTF(("uhci_device_bulk_abort:\n"));
uhci_abort_xfer(xfer, USBD_CANCELLED);
}
void
uhci_abort_xfer(usbd_xfer_handle xfer, usbd_status status)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
struct uhci_pipe *upipe = (struct uhci_pipe *)xfer->pipe;
uhci_intr_info_t *ii = upipe->iinfo;
uhci_soft_td_t *std;
DPRINTFN(1,("uhci_abort_xfer: xfer=%p, status=%d\n", xfer, status));
/* Make interrupt routine ignore it, */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->status = status;
/* don't timeout, */
usb_untimeout(uhci_timeout, ii, ii->timeout_handle);
/* make hardware ignore it, */
for (std = ii->stdstart; std != 0; std = std->link.std)
std->td.td_status &= LE(~(UHCI_TD_ACTIVE | UHCI_TD_IOC));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->hcpriv = ii;
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
#if 1
/* Make sure hardware has completed. */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->device->bus->intr_context) {
/* We have no process context, so we can't use tsleep(). */
timeout(uhci_abort_xfer_end, xfer, hz / USB_FRAMES_PER_SECOND);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
} else {
#if defined(DIAGNOSTIC) && defined(__i386__) && defined(__FreeBSD__)
KASSERT(curproc->p_intr_nesting_level == 0,
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
("ohci_abort_req in interrupt context"));
#endif
usb_delay_ms(xfer->pipe->device->bus, 1);
/* and call final part of interrupt handler. */
uhci_abort_xfer_end(xfer);
}
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
#else
delay(1000);
uhci_abort_xfer_end(xfer);
#endif
}
void
uhci_abort_xfer_end(void *v)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_xfer_handle xfer = v;
int s;
s = splusb();
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usb_transfer_complete(xfer);
splx(s);
}
/* Close a device bulk pipe. */
void
uhci_device_bulk_close(usbd_pipe_handle pipe)
{
struct uhci_pipe *upipe = (struct uhci_pipe *)pipe;
usbd_device_handle dev = upipe->pipe.device;
uhci_softc_t *sc = (uhci_softc_t *)dev->bus;
uhci_free_sqh(sc, upipe->u.bulk.sqh);
uhci_free_intr_info(upipe->iinfo);
/* XXX free other resources */
}
usbd_status
uhci_device_ctrl_transfer(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
/* Insert last in queue. */
err = usb_insert_transfer(xfer);
if (err)
return (err);
/* Pipe isn't running (otherwise err would be USBD_INPROG),
* start first
*/
return (uhci_device_ctrl_start(SIMPLEQ_FIRST(&xfer->pipe->queue)));
}
usbd_status
uhci_device_ctrl_start(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_softc_t *sc = (uhci_softc_t *)xfer->pipe->device->bus;
usbd_status err;
#ifdef DIAGNOSTIC
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (!(xfer->rqflags & URQ_REQUEST))
panic("uhci_device_ctrl_transfer: not a request\n");
#endif
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = uhci_device_request(xfer);
if (err)
return (err);
if (sc->sc_bus.use_polling)
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_waitintr(sc, xfer);
return (USBD_IN_PROGRESS);
}
usbd_status
uhci_device_intr_transfer(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
/* Insert last in queue. */
err = usb_insert_transfer(xfer);
if (err)
return (err);
/* Pipe isn't running (otherwise err would be USBD_INPROG),
* start first
*/
return (uhci_device_intr_start(SIMPLEQ_FIRST(&xfer->pipe->queue)));
}
usbd_status
uhci_device_intr_start(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
struct uhci_pipe *upipe = (struct uhci_pipe *)xfer->pipe;
usbd_device_handle dev = upipe->pipe.device;
uhci_softc_t *sc = (uhci_softc_t *)dev->bus;
uhci_intr_info_t *ii = upipe->iinfo;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_soft_td_t *data, *dataend;
uhci_soft_qh_t *sqh;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
int i, s;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(3,("uhci_device_intr_transfer: xfer=%p len=%d flags=%d\n",
xfer, xfer->length, xfer->flags));
#ifdef DIAGNOSTIC
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->rqflags & URQ_REQUEST)
panic("uhci_device_intr_transfer: a request\n");
#endif
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
err = uhci_alloc_std_chain(upipe, sc, xfer->length, 1, xfer->flags,
&xfer->dmabuf, &data, &dataend);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (err)
return (err);
dataend->td.td_status |= LE(UHCI_TD_IOC);
#ifdef UHCI_DEBUG
if (uhcidebug > 10) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTF(("uhci_device_intr_transfer: data(1)\n"));
uhci_dump_tds(data);
uhci_dump_qh(upipe->u.intr.qhs[0]);
}
#endif
s = splusb();
/* Set up interrupt info. */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
ii->xfer = xfer;
ii->stdstart = data;
ii->stdend = dataend;
#if defined(__FreeBSD__)
callout_handle_init(&ii->timeout_handle);
#endif
#ifdef DIAGNOSTIC
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
if (!ii->isdone) {
printf("uhci_device_intr_transfer: not done, ii=%p\n", ii);
}
ii->isdone = 0;
#endif
DPRINTFN(10,("uhci_device_intr_transfer: qhs[0]=%p\n",
upipe->u.intr.qhs[0]));
for (i = 0; i < upipe->u.intr.npoll; i++) {
sqh = upipe->u.intr.qhs[i];
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
sqh->elink = data;
sqh->qh.qh_elink = LE(data->physaddr);
}
splx(s);
#ifdef UHCI_DEBUG
if (uhcidebug > 10) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTF(("uhci_device_intr_transfer: data(2)\n"));
uhci_dump_tds(data);
uhci_dump_qh(upipe->u.intr.qhs[0]);
}
#endif
return (USBD_IN_PROGRESS);
}
/* Abort a device control request. */
void
uhci_device_ctrl_abort(usbd_xfer_handle xfer)
{
DPRINTF(("uhci_device_ctrl_abort:\n"));
uhci_abort_xfer(xfer, USBD_CANCELLED);
}
/* Close a device control pipe. */
void
uhci_device_ctrl_close(usbd_pipe_handle pipe)
{
struct uhci_pipe *upipe = (struct uhci_pipe *)pipe;
uhci_free_intr_info(upipe->iinfo);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
/* XXX free other resources? */
}
/* Abort a device interrupt request. */
void
uhci_device_intr_abort(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(1,("uhci_device_intr_abort: xfer=%p\n", xfer));
if (xfer->pipe->intrxfer == xfer) {
DPRINTFN(1,("uhci_device_intr_abort: remove\n"));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->pipe->intrxfer = 0;
}
uhci_abort_xfer(xfer, USBD_CANCELLED);
}
/* Close a device interrupt pipe. */
void
uhci_device_intr_close(usbd_pipe_handle pipe)
{
struct uhci_pipe *upipe = (struct uhci_pipe *)pipe;
uhci_softc_t *sc = (uhci_softc_t *)pipe->device->bus;
int i, s, npoll;
upipe->iinfo->stdstart = 0; /* inactive */
/* Unlink descriptors from controller data structures. */
npoll = upipe->u.intr.npoll;
uhci_lock_frames(sc);
for (i = 0; i < npoll; i++)
uhci_remove_intr(sc, upipe->u.intr.qhs[i]->pos,
upipe->u.intr.qhs[i]);
uhci_unlock_frames(sc);
/*
* We now have to wait for any activity on the physical
* descriptors to stop.
*/
usb_delay_ms(&sc->sc_bus, 2);
for(i = 0; i < npoll; i++)
uhci_free_sqh(sc, upipe->u.intr.qhs[i]);
free(upipe->u.intr.qhs, M_USBHC);
s = splusb();
LIST_REMOVE(upipe->iinfo, list); /* remove from active list */
splx(s);
uhci_free_intr_info(upipe->iinfo);
/* XXX free other resources */
}
usbd_status
uhci_device_request(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
struct uhci_pipe *upipe = (struct uhci_pipe *)xfer->pipe;
usb_device_request_t *req = &xfer->request;
usbd_device_handle dev = upipe->pipe.device;
uhci_softc_t *sc = (uhci_softc_t *)dev->bus;
int addr = dev->address;
int endpt = upipe->pipe.endpoint->edesc->bEndpointAddress;
uhci_intr_info_t *ii = upipe->iinfo;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_soft_td_t *setup, *data, *stat, *next, *dataend;
uhci_soft_qh_t *sqh;
int len;
u_int32_t ls;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
int isread;
int s;
DPRINTFN(3,("uhci_device_control type=0x%02x, request=0x%02x, "
"wValue=0x%04x, wIndex=0x%04x len=%d, addr=%d, endpt=%d\n",
req->bmRequestType, req->bRequest, UGETW(req->wValue),
UGETW(req->wIndex), UGETW(req->wLength),
addr, endpt));
ls = dev->lowspeed ? UHCI_TD_LS : 0;
isread = req->bmRequestType & UT_READ;
len = UGETW(req->wLength);
setup = upipe->u.ctl.setup;
stat = upipe->u.ctl.stat;
sqh = upipe->u.ctl.sqh;
/* Set up data transaction */
if (len != 0) {
upipe->nexttoggle = 1;
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
err = uhci_alloc_std_chain(upipe, sc, len, isread, xfer->flags,
&xfer->dmabuf, &data, &dataend);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (err)
return (err);
next = data;
dataend->link.std = stat;
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
dataend->td.td_link = LE(stat->physaddr | UHCI_PTR_VF);
} else {
next = stat;
}
upipe->u.ctl.length = len;
memcpy(KERNADDR(&upipe->u.ctl.reqdma, 0), req, sizeof *req);
setup->link.std = next;
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
setup->td.td_link = LE(next->physaddr | UHCI_PTR_VF);
setup->td.td_status = LE(UHCI_TD_SET_ERRCNT(3) | ls | UHCI_TD_ACTIVE);
setup->td.td_token = LE(UHCI_TD_SETUP(sizeof *req, endpt, addr));
setup->td.td_buffer = LE(DMAADDR(&upipe->u.ctl.reqdma, 0));
stat->link.std = 0;
stat->td.td_link = LE(UHCI_PTR_T);
stat->td.td_status = LE(UHCI_TD_SET_ERRCNT(3) | ls |
UHCI_TD_ACTIVE | UHCI_TD_IOC);
stat->td.td_token =
LE(isread ? UHCI_TD_OUT(0, endpt, addr, 1) :
UHCI_TD_IN (0, endpt, addr, 1));
stat->td.td_buffer = LE(0);
#ifdef UHCI_DEBUG
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (uhcidebug > 10) {
DPRINTF(("uhci_device_request: before transfer\n"));
uhci_dump_tds(setup);
}
#endif
/* Set up interrupt info. */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
ii->xfer = xfer;
ii->stdstart = setup;
ii->stdend = stat;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
#if defined(__FreeBSD__)
callout_handle_init(&ii->timeout_handle);
#endif
#ifdef DIAGNOSTIC
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
if (!ii->isdone) {
printf("uhci_device_request: not done, ii=%p\n", ii);
}
ii->isdone = 0;
#endif
sqh->elink = setup;
sqh->qh.qh_elink = LE(setup->physaddr);
sqh->intr_info = ii;
s = splusb();
uhci_add_ctrl(sc, sqh);
LIST_INSERT_HEAD(&sc->sc_intrhead, ii, list);
#ifdef UHCI_DEBUG
if (uhcidebug > 12) {
uhci_soft_td_t *std;
uhci_soft_qh_t *xqh;
uhci_soft_qh_t *sxqh;
int maxqh = 0;
uhci_physaddr_t link;
DPRINTF(("uhci_enter_ctl_q: follow from [0]\n"));
for (std = sc->sc_vframes[0].htd, link = 0;
(link & UHCI_PTR_Q) == 0;
std = std->link.std) {
link = LE(std->td.td_link);
uhci_dump_td(std);
}
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
sxqh = (uhci_soft_qh_t *)std;
uhci_dump_qh(sxqh);
for (xqh = sxqh;
xqh != NULL;
xqh = (maxqh++ == 5 || xqh->hlink==sxqh ||
xqh->hlink==xqh ? NULL : xqh->hlink)) {
uhci_dump_qh(xqh);
}
DPRINTF(("Enqueued QH:\n"));
uhci_dump_qh(sqh);
uhci_dump_tds(sqh->elink);
}
#endif
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->timeout && !sc->sc_bus.use_polling) {
usb_timeout(uhci_timeout, ii,
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
MS_TO_TICKS(xfer->timeout), ii->timeout_handle);
}
splx(s);
return (USBD_NORMAL_COMPLETION);
}
usbd_status
uhci_device_isoc_transfer(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(5,("uhci_device_isoc_transfer: xfer=%p\n", xfer));
/* Put it on our queue, */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = usb_insert_transfer(xfer);
/* bail out on error, */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (err && err != USBD_IN_PROGRESS)
return (err);
/* XXX should check inuse here */
/* insert into schedule, */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_device_isoc_enter(xfer);
/* and put on interrupt list if the pipe wasn't running */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (!err)
uhci_device_isoc_start(SIMPLEQ_FIRST(&xfer->pipe->queue));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
return (err);
}
void
uhci_device_isoc_enter(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
struct uhci_pipe *upipe = (struct uhci_pipe *)xfer->pipe;
usbd_device_handle dev = upipe->pipe.device;
uhci_softc_t *sc = (uhci_softc_t *)dev->bus;
struct iso *iso = &upipe->u.iso;
uhci_soft_td_t *std;
u_int32_t buf, len, status;
int s, i, next, nframes;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(5,("uhci_device_isoc_enter: used=%d next=%d xfer=%p "
"nframes=%d\n",
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
iso->inuse, iso->next, xfer, xfer->nframes));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->status == USBD_IN_PROGRESS) {
/* This request has already been entered into the frame list */
/* XXX */
}
#ifdef DIAGNOSTIC
if (iso->inuse >= UHCI_VFRAMELIST_COUNT)
printf("uhci_device_isoc_enter: overflow!\n");
#endif
next = iso->next;
if (next == -1) {
/* Not in use yet, schedule it a few frames ahead. */
next = (UREAD2(sc, UHCI_FRNUM) + 3) % UHCI_VFRAMELIST_COUNT;
DPRINTFN(2,("uhci_device_isoc_enter: start next=%d\n", next));
}
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->status = USBD_IN_PROGRESS;
xfer->hcprivint = next;
buf = DMAADDR(&xfer->dmabuf, 0);
status = LE(UHCI_TD_ZERO_ACTLEN(UHCI_TD_SET_ERRCNT(0) |
UHCI_TD_ACTIVE |
UHCI_TD_IOS));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
nframes = xfer->nframes;
s = splusb();
for (i = 0; i < nframes; i++) {
std = iso->stds[next];
if (++next >= UHCI_VFRAMELIST_COUNT)
next = 0;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
len = xfer->frlengths[i];
std->td.td_buffer = LE(buf);
if (i == nframes - 1)
status |= LE(UHCI_TD_IOC);
std->td.td_status = status;
std->td.td_token &= LE(~UHCI_TD_MAXLEN_MASK);
std->td.td_token |= LE(UHCI_TD_SET_MAXLEN(len));
#ifdef UHCI_DEBUG
if (uhcidebug > 5) {
DPRINTFN(5,("uhci_device_isoc_enter: TD %d\n", i));
uhci_dump_td(std);
}
#endif
buf += len;
}
iso->next = next;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
iso->inuse += xfer->nframes;
splx(s);
}
usbd_status
uhci_device_isoc_start(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
struct uhci_pipe *upipe = (struct uhci_pipe *)xfer->pipe;
uhci_softc_t *sc = (uhci_softc_t *)upipe->pipe.device->bus;
uhci_intr_info_t *ii = upipe->iinfo;
uhci_soft_td_t *end;
int s, i;
#ifdef DIAGNOSTIC
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->status != USBD_IN_PROGRESS)
printf("uhci_device_isoc_start: not in progress %p\n", xfer);
#endif
/* Find the last TD */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
i = xfer->hcprivint + xfer->nframes;
if (i >= UHCI_VFRAMELIST_COUNT)
i -= UHCI_VFRAMELIST_COUNT;
end = upipe->u.iso.stds[i];
s = splusb();
/* Set up interrupt info. */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
ii->xfer = xfer;
ii->stdstart = end;
ii->stdend = end;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
#if defined(__FreeBSD__)
callout_handle_init(&ii->timeout_handle);
#endif
#ifdef DIAGNOSTIC
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
if (!ii->isdone) {
printf("uhci_device_isoc_start: not done, ii=%p\n", ii);
}
ii->isdone = 0;
#endif
LIST_INSERT_HEAD(&sc->sc_intrhead, ii, list);
splx(s);
return (USBD_IN_PROGRESS);
}
void
uhci_device_isoc_abort(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
struct uhci_pipe *upipe = (struct uhci_pipe *)xfer->pipe;
uhci_intr_info_t *ii = upipe->iinfo;
uhci_soft_td_t **stds = upipe->u.iso.stds;
uhci_soft_td_t *std;
int i, n, nframes;
/* Make interrupt routine ignore it, */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->status = USBD_CANCELLED;
/* make hardware ignore it, */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
nframes = xfer->nframes;
n = xfer->hcprivint;
for (i = 0; i < nframes; i++) {
std = stds[n];
std->td.td_status &= LE(~(UHCI_TD_ACTIVE | UHCI_TD_IOC));
if (++n >= UHCI_VFRAMELIST_COUNT)
n = 0;
}
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->hcpriv = ii;
/* make sure hardware has completed, */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->device->bus->intr_context) {
/* We have no process context, so we can't use tsleep(). */
timeout(uhci_abort_xfer_end, xfer, hz / USB_FRAMES_PER_SECOND);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
} else {
usb_delay_ms(xfer->pipe->device->bus, 1);
/* and call final part of interrupt handler. */
uhci_abort_xfer_end(xfer);
}
}
void
uhci_device_isoc_close(usbd_pipe_handle pipe)
{
struct uhci_pipe *upipe = (struct uhci_pipe *)pipe;
usbd_device_handle dev = upipe->pipe.device;
uhci_softc_t *sc = (uhci_softc_t *)dev->bus;
uhci_soft_td_t *std, *vstd;
struct iso *iso;
int i;
/*
* Make sure all TDs are marked as inactive.
* Wait for completion.
* Unschedule.
* Deallocate.
*/
iso = &upipe->u.iso;
for (i = 0; i < UHCI_VFRAMELIST_COUNT; i++)
iso->stds[i]->td.td_status &= LE(~UHCI_TD_ACTIVE);
usb_delay_ms(&sc->sc_bus, 2); /* wait for completion */
uhci_lock_frames(sc);
for (i = 0; i < UHCI_VFRAMELIST_COUNT; i++) {
std = iso->stds[i];
for (vstd = sc->sc_vframes[i].htd;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
vstd != NULL && vstd->link.std != std;
vstd = vstd->link.std)
;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (vstd == NULL) {
/*panic*/
printf("uhci_device_isoc_close: %p not found\n", std);
uhci_unlock_frames(sc);
return;
}
vstd->link = std->link;
vstd->td.td_link = std->td.td_link;
uhci_free_std(sc, std);
}
uhci_unlock_frames(sc);
free(iso->stds, M_USBHC);
}
usbd_status
uhci_setup_isoc(usbd_pipe_handle pipe)
{
struct uhci_pipe *upipe = (struct uhci_pipe *)pipe;
usbd_device_handle dev = upipe->pipe.device;
uhci_softc_t *sc = (uhci_softc_t *)dev->bus;
int addr = upipe->pipe.device->address;
int endpt = upipe->pipe.endpoint->edesc->bEndpointAddress;
int rd = UE_GET_DIR(endpt) == UE_DIR_IN;
uhci_soft_td_t *std, *vstd;
u_int32_t token;
struct iso *iso;
int i;
iso = &upipe->u.iso;
iso->stds = malloc(UHCI_VFRAMELIST_COUNT * sizeof (uhci_soft_td_t *),
M_USBHC, M_WAITOK);
token = LE(rd ? UHCI_TD_IN (0, endpt, addr, 0) :
UHCI_TD_OUT(0, endpt, addr, 0));
/* Allocate the TDs and mark as inactive; */
for (i = 0; i < UHCI_VFRAMELIST_COUNT; i++) {
std = uhci_alloc_std(sc);
if (std == 0)
goto bad;
std->td.td_status = LE(UHCI_TD_IOS); /* iso, inactive */
std->td.td_token = token;
iso->stds[i] = std;
}
/* Insert TDs into schedule. */
uhci_lock_frames(sc);
for (i = 0; i < UHCI_VFRAMELIST_COUNT; i++) {
std = iso->stds[i];
vstd = sc->sc_vframes[i].htd;
std->link = vstd->link;
std->td.td_link = vstd->td.td_link;
vstd->link.std = std;
vstd->td.td_link = LE(std->physaddr);
}
uhci_unlock_frames(sc);
iso->next = -1;
iso->inuse = 0;
return (USBD_NORMAL_COMPLETION);
bad:
while (--i >= 0)
uhci_free_std(sc, iso->stds[i]);
free(iso->stds, M_USBHC);
return (USBD_NOMEM);
}
void
uhci_device_isoc_done(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_intr_info_t *ii = xfer->hcpriv;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(4, ("uhci_isoc_done: length=%d\n", xfer->actlen));
/* Turn off the interrupt since it is active even if the TD is not. */
ii->stdend->td.td_status &= LE(~UHCI_TD_IOC);
LIST_REMOVE(ii, list); /* remove from active list */
}
void
uhci_device_intr_done(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_intr_info_t *ii = xfer->hcpriv;
uhci_softc_t *sc = ii->sc;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
struct uhci_pipe *upipe = (struct uhci_pipe *)xfer->pipe;
uhci_soft_qh_t *sqh;
int i, npoll;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(5, ("uhci_intr_done: length=%d\n", xfer->actlen));
npoll = upipe->u.intr.npoll;
for(i = 0; i < npoll; i++) {
sqh = upipe->u.intr.qhs[i];
sqh->elink = 0;
sqh->qh.qh_elink = LE(UHCI_PTR_T);
}
uhci_free_std_chain(sc, ii->stdstart, 0);
/* XXX Wasteful. */
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->pipe->repeat) {
uhci_soft_td_t *data, *dataend;
/* This alloc cannot fail since we freed the chain above. */
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
uhci_alloc_std_chain(upipe, sc, xfer->length, 1, xfer->flags,
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
&xfer->dmabuf, &data, &dataend);
dataend->td.td_status |= LE(UHCI_TD_IOC);
#ifdef UHCI_DEBUG
if (uhcidebug > 10) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTF(("uhci_device_intr_done: data(1)\n"));
uhci_dump_tds(data);
uhci_dump_qh(upipe->u.intr.qhs[0]);
}
#endif
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
ii->stdstart = data;
ii->stdend = dataend;
#if defined(__FreeBSD__)
callout_handle_init(&ii->timeout_handle);
#endif
#ifdef DIAGNOSTIC
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
if (!ii->isdone) {
printf("uhci_device_intr_done: not done, ii=%p\n", ii);
}
ii->isdone = 0;
#endif
for (i = 0; i < npoll; i++) {
sqh = upipe->u.intr.qhs[i];
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
sqh->elink = data;
sqh->qh.qh_elink = LE(data->physaddr);
}
} else {
ii->stdstart = 0; /* mark as inactive */
}
}
/* Deallocate request data structures */
void
uhci_device_ctrl_done(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_intr_info_t *ii = xfer->hcpriv;
uhci_softc_t *sc = ii->sc;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
struct uhci_pipe *upipe = (struct uhci_pipe *)xfer->pipe;
#ifdef DIAGNOSTIC
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (!(xfer->rqflags & URQ_REQUEST))
panic("uhci_ctrl_done: not a request\n");
#endif
LIST_REMOVE(ii, list); /* remove from active list */
uhci_remove_ctrl(sc, upipe->u.ctl.sqh);
if (upipe->u.ctl.length != 0)
uhci_free_std_chain(sc, ii->stdstart->link.std, ii->stdend);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(5, ("uhci_ctrl_done: length=%d\n", xfer->actlen));
}
/* Deallocate request data structures */
void
uhci_device_bulk_done(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_intr_info_t *ii = xfer->hcpriv;
uhci_softc_t *sc = ii->sc;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
struct uhci_pipe *upipe = (struct uhci_pipe *)xfer->pipe;
LIST_REMOVE(ii, list); /* remove from active list */
uhci_remove_bulk(sc, upipe->u.bulk.sqh);
uhci_free_std_chain(sc, ii->stdstart, 0);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(5, ("uhci_bulk_done: length=%d\n", xfer->actlen));
}
/* Add interrupt QH, called with vflock. */
void
uhci_add_intr(uhci_softc_t *sc, int n, uhci_soft_qh_t *sqh)
{
struct uhci_vframe *vf = &sc->sc_vframes[n];
uhci_soft_qh_t *eqh;
DPRINTFN(4, ("uhci_add_intr: n=%d sqh=%p\n", n, sqh));
eqh = vf->eqh;
sqh->hlink = eqh->hlink;
sqh->qh.qh_hlink = eqh->qh.qh_hlink;
eqh->hlink = sqh;
eqh->qh.qh_hlink = LE(sqh->physaddr | UHCI_PTR_Q);
vf->eqh = sqh;
vf->bandwidth++;
}
/* Remove interrupt QH, called with vflock. */
void
uhci_remove_intr(uhci_softc_t *sc, int n, uhci_soft_qh_t *sqh)
{
struct uhci_vframe *vf = &sc->sc_vframes[n];
uhci_soft_qh_t *pqh;
DPRINTFN(4, ("uhci_remove_intr: n=%d sqh=%p\n", n, sqh));
for (pqh = vf->hqh; pqh->hlink != sqh; pqh = pqh->hlink)
#if defined(DIAGNOSTIC) || defined(UHCI_DEBUG)
if (LE(pqh->qh.qh_hlink) & UHCI_PTR_T) {
DPRINTF(("uhci_remove_intr: QH not found\n"));
return;
}
#else
;
#endif
pqh->hlink = sqh->hlink;
pqh->qh.qh_hlink = sqh->qh.qh_hlink;
if (vf->eqh == sqh)
vf->eqh = pqh;
vf->bandwidth--;
}
usbd_status
uhci_device_setintr(uhci_softc_t *sc, struct uhci_pipe *upipe, int ival)
{
uhci_soft_qh_t *sqh;
int i, npoll, s;
u_int bestbw, bw, bestoffs, offs;
DPRINTFN(2, ("uhci_setintr: pipe=%p\n", upipe));
if (ival == 0) {
printf("uhci_setintr: 0 interval\n");
return (USBD_INVAL);
}
if (ival > UHCI_VFRAMELIST_COUNT)
ival = UHCI_VFRAMELIST_COUNT;
npoll = (UHCI_VFRAMELIST_COUNT + ival - 1) / ival;
DPRINTFN(2, ("uhci_setintr: ival=%d npoll=%d\n", ival, npoll));
upipe->u.intr.npoll = npoll;
upipe->u.intr.qhs =
malloc(npoll * sizeof(uhci_soft_qh_t *), M_USBHC, M_WAITOK);
/*
* Figure out which offset in the schedule that has most
* bandwidth left over.
*/
#define MOD(i) ((i) & (UHCI_VFRAMELIST_COUNT-1))
for (bestoffs = offs = 0, bestbw = ~0; offs < ival; offs++) {
for (bw = i = 0; i < npoll; i++)
bw += sc->sc_vframes[MOD(i * ival + offs)].bandwidth;
if (bw < bestbw) {
bestbw = bw;
bestoffs = offs;
}
}
DPRINTFN(1, ("uhci_setintr: bw=%d offs=%d\n", bestbw, bestoffs));
upipe->iinfo->stdstart = 0;
for(i = 0; i < npoll; i++) {
upipe->u.intr.qhs[i] = sqh = uhci_alloc_sqh(sc);
sqh->elink = 0;
sqh->qh.qh_elink = LE(UHCI_PTR_T);
sqh->pos = MOD(i * ival + bestoffs);
sqh->intr_info = upipe->iinfo;
}
#undef MOD
s = splusb();
LIST_INSERT_HEAD(&sc->sc_intrhead, upipe->iinfo, list);
splx(s);
uhci_lock_frames(sc);
/* Enter QHs into the controller data structures. */
for(i = 0; i < npoll; i++)
uhci_add_intr(sc, upipe->u.intr.qhs[i]->pos,
upipe->u.intr.qhs[i]);
uhci_unlock_frames(sc);
DPRINTFN(5, ("uhci_setintr: returns %p\n", upipe));
return (USBD_NORMAL_COMPLETION);
}
/* Open a new pipe. */
usbd_status
uhci_open(usbd_pipe_handle pipe)
{
uhci_softc_t *sc = (uhci_softc_t *)pipe->device->bus;
struct uhci_pipe *upipe = (struct uhci_pipe *)pipe;
usb_endpoint_descriptor_t *ed = pipe->endpoint->edesc;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
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
int ival;
DPRINTFN(1, ("uhci_open: pipe=%p, addr=%d, endpt=%d (%d)\n",
pipe, pipe->device->address,
ed->bEndpointAddress, sc->sc_addr));
if (pipe->device->address == sc->sc_addr) {
switch (ed->bEndpointAddress) {
case USB_CONTROL_ENDPOINT:
pipe->methods = &uhci_root_ctrl_methods;
break;
case UE_DIR_IN | UHCI_INTR_ENDPT:
pipe->methods = &uhci_root_intr_methods;
break;
default:
return (USBD_INVAL);
}
} else {
upipe->iinfo = uhci_alloc_intr_info(sc);
if (upipe->iinfo == 0)
return (USBD_NOMEM);
switch (ed->bmAttributes & UE_XFERTYPE) {
case UE_CONTROL:
pipe->methods = &uhci_device_ctrl_methods;
upipe->u.ctl.sqh = uhci_alloc_sqh(sc);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (upipe->u.ctl.sqh == NULL)
goto bad;
upipe->u.ctl.setup = uhci_alloc_std(sc);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (upipe->u.ctl.setup == NULL) {
uhci_free_sqh(sc, upipe->u.ctl.sqh);
goto bad;
}
upipe->u.ctl.stat = uhci_alloc_std(sc);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (upipe->u.ctl.stat == NULL) {
uhci_free_sqh(sc, upipe->u.ctl.sqh);
uhci_free_std(sc, upipe->u.ctl.setup);
goto bad;
}
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = usb_allocmem(&sc->sc_bus,
sizeof(usb_device_request_t),
0, &upipe->u.ctl.reqdma);
if (err) {
uhci_free_sqh(sc, upipe->u.ctl.sqh);
uhci_free_std(sc, upipe->u.ctl.setup);
uhci_free_std(sc, upipe->u.ctl.stat);
goto bad;
}
break;
case UE_INTERRUPT:
pipe->methods = &uhci_device_intr_methods;
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
ival = pipe->interval;
if (ival == USBD_DEFAULT_INTERVAL)
ival = ed->bInterval;
return (uhci_device_setintr(sc, upipe, ival));
case UE_ISOCHRONOUS:
pipe->methods = &uhci_device_isoc_methods;
return (uhci_setup_isoc(pipe));
case UE_BULK:
pipe->methods = &uhci_device_bulk_methods;
upipe->u.bulk.sqh = uhci_alloc_sqh(sc);
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (upipe->u.bulk.sqh == NULL)
goto bad;
break;
}
}
return (USBD_NORMAL_COMPLETION);
bad:
uhci_free_intr_info(upipe->iinfo);
return (USBD_NOMEM);
}
/*
* Data structures and routines to emulate the root hub.
*/
usb_device_descriptor_t uhci_devd = {
USB_DEVICE_DESCRIPTOR_SIZE,
UDESC_DEVICE, /* type */
{0x00, 0x01}, /* USB version */
UCLASS_HUB, /* class */
USUBCLASS_HUB, /* subclass */
0, /* protocol */
64, /* max packet */
{0},{0},{0x00,0x01}, /* device id */
1,2,0, /* string indicies */
1 /* # of configurations */
};
usb_config_descriptor_t uhci_confd = {
USB_CONFIG_DESCRIPTOR_SIZE,
UDESC_CONFIG,
{USB_CONFIG_DESCRIPTOR_SIZE +
USB_INTERFACE_DESCRIPTOR_SIZE +
USB_ENDPOINT_DESCRIPTOR_SIZE},
1,
1,
0,
UC_SELF_POWERED,
0 /* max power */
};
usb_interface_descriptor_t uhci_ifcd = {
USB_INTERFACE_DESCRIPTOR_SIZE,
UDESC_INTERFACE,
0,
0,
1,
UCLASS_HUB,
USUBCLASS_HUB,
0,
0
};
usb_endpoint_descriptor_t uhci_endpd = {
USB_ENDPOINT_DESCRIPTOR_SIZE,
UDESC_ENDPOINT,
UE_DIR_IN | UHCI_INTR_ENDPT,
UE_INTERRUPT,
{8},
255
};
usb_hub_descriptor_t uhci_hubd_piix = {
USB_HUB_DESCRIPTOR_SIZE,
UDESC_HUB,
2,
{ UHD_PWR_NO_SWITCH | UHD_OC_INDIVIDUAL, 0 },
50, /* power on to power good */
0,
{ 0x00 }, /* both ports are removable */
};
int
uhci_str(usb_string_descriptor_t *p, int l, char *s)
{
int i;
if (l == 0)
return (0);
p->bLength = 2 * strlen(s) + 2;
if (l == 1)
return (1);
p->bDescriptorType = UDESC_STRING;
l -= 2;
for (i = 0; s[i] && l > 1; i++, l -= 2)
USETW2(p->bString[i], 0, s[i]);
return (2*i+2);
}
/*
* Simulate a hardware hub by handling all the necessary requests.
*/
usbd_status
uhci_root_ctrl_transfer(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
/* Insert last in queue. */
err = usb_insert_transfer(xfer);
if (err)
return (err);
/* Pipe isn't running (otherwise err would be USBD_INPROG),
* start first
*/
return (uhci_root_ctrl_start(SIMPLEQ_FIRST(&xfer->pipe->queue)));
}
usbd_status
uhci_root_ctrl_start(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_softc_t *sc = (uhci_softc_t *)xfer->pipe->device->bus;
usb_device_request_t *req;
void *buf = NULL;
int port, x;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
int s, len, value, index, status, change, l, totlen = 0;
usb_port_status_t ps;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
#ifdef DIAGNOSTIC
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (!(xfer->rqflags & URQ_REQUEST))
panic("uhci_root_ctrl_transfer: not a request\n");
#endif
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
req = &xfer->request;
DPRINTFN(2,("uhci_root_ctrl_control type=0x%02x request=%02x\n",
req->bmRequestType, req->bRequest));
len = UGETW(req->wLength);
value = UGETW(req->wValue);
index = UGETW(req->wIndex);
if (len != 0)
buf = KERNADDR(&xfer->dmabuf, 0);
#define C(x,y) ((x) | ((y) << 8))
switch(C(req->bRequest, req->bmRequestType)) {
case C(UR_CLEAR_FEATURE, UT_WRITE_DEVICE):
case C(UR_CLEAR_FEATURE, UT_WRITE_INTERFACE):
case C(UR_CLEAR_FEATURE, UT_WRITE_ENDPOINT):
/*
* DEVICE_REMOTE_WAKEUP and ENDPOINT_HALT are no-ops
* for the integrated root hub.
*/
break;
case C(UR_GET_CONFIG, UT_READ_DEVICE):
if (len > 0) {
*(u_int8_t *)buf = sc->sc_conf;
totlen = 1;
}
break;
case C(UR_GET_DESCRIPTOR, UT_READ_DEVICE):
DPRINTFN(2,("uhci_root_ctrl_control wValue=0x%04x\n", value));
switch(value >> 8) {
case UDESC_DEVICE:
if ((value & 0xff) != 0) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
totlen = l = min(len, USB_DEVICE_DESCRIPTOR_SIZE);
USETW(uhci_devd.idVendor, sc->sc_id_vendor);
memcpy(buf, &uhci_devd, l);
break;
case UDESC_CONFIG:
if ((value & 0xff) != 0) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
totlen = l = min(len, USB_CONFIG_DESCRIPTOR_SIZE);
memcpy(buf, &uhci_confd, l);
buf = (char *)buf + l;
len -= l;
l = min(len, USB_INTERFACE_DESCRIPTOR_SIZE);
totlen += l;
memcpy(buf, &uhci_ifcd, l);
buf = (char *)buf + l;
len -= l;
l = min(len, USB_ENDPOINT_DESCRIPTOR_SIZE);
totlen += l;
memcpy(buf, &uhci_endpd, l);
break;
case UDESC_STRING:
if (len == 0)
break;
*(u_int8_t *)buf = 0;
totlen = 1;
switch (value & 0xff) {
case 1: /* Vendor */
totlen = uhci_str(buf, len, sc->sc_vendor);
break;
case 2: /* Product */
totlen = uhci_str(buf, len, "UHCI root hub");
break;
}
break;
default:
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
break;
case C(UR_GET_INTERFACE, UT_READ_INTERFACE):
if (len > 0) {
*(u_int8_t *)buf = 0;
totlen = 1;
}
break;
case C(UR_GET_STATUS, UT_READ_DEVICE):
if (len > 1) {
USETW(((usb_status_t *)buf)->wStatus,UDS_SELF_POWERED);
totlen = 2;
}
break;
case C(UR_GET_STATUS, UT_READ_INTERFACE):
case C(UR_GET_STATUS, UT_READ_ENDPOINT):
if (len > 1) {
USETW(((usb_status_t *)buf)->wStatus, 0);
totlen = 2;
}
break;
case C(UR_SET_ADDRESS, UT_WRITE_DEVICE):
if (value >= USB_MAX_DEVICES) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
sc->sc_addr = value;
break;
case C(UR_SET_CONFIG, UT_WRITE_DEVICE):
if (value != 0 && value != 1) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
sc->sc_conf = value;
break;
case C(UR_SET_DESCRIPTOR, UT_WRITE_DEVICE):
break;
case C(UR_SET_FEATURE, UT_WRITE_DEVICE):
case C(UR_SET_FEATURE, UT_WRITE_INTERFACE):
case C(UR_SET_FEATURE, UT_WRITE_ENDPOINT):
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
case C(UR_SET_INTERFACE, UT_WRITE_INTERFACE):
break;
case C(UR_SYNCH_FRAME, UT_WRITE_ENDPOINT):
break;
/* Hub requests */
case C(UR_CLEAR_FEATURE, UT_WRITE_CLASS_DEVICE):
break;
case C(UR_CLEAR_FEATURE, UT_WRITE_CLASS_OTHER):
DPRINTFN(3, ("uhci_root_ctrl_control: UR_CLEAR_PORT_FEATURE "
"port=%d feature=%d\n",
index, value));
if (index == 1)
port = UHCI_PORTSC1;
else if (index == 2)
port = UHCI_PORTSC2;
else {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
switch(value) {
case UHF_PORT_ENABLE:
x = UREAD2(sc, port);
UWRITE2(sc, port, x & ~UHCI_PORTSC_PE);
break;
case UHF_PORT_SUSPEND:
x = UREAD2(sc, port);
UWRITE2(sc, port, x & ~UHCI_PORTSC_SUSP);
break;
case UHF_PORT_RESET:
x = UREAD2(sc, port);
UWRITE2(sc, port, x & ~UHCI_PORTSC_PR);
break;
case UHF_C_PORT_CONNECTION:
x = UREAD2(sc, port);
UWRITE2(sc, port, x | UHCI_PORTSC_CSC);
break;
case UHF_C_PORT_ENABLE:
x = UREAD2(sc, port);
UWRITE2(sc, port, x | UHCI_PORTSC_POEDC);
break;
case UHF_C_PORT_OVER_CURRENT:
x = UREAD2(sc, port);
UWRITE2(sc, port, x | UHCI_PORTSC_OCIC);
break;
case UHF_C_PORT_RESET:
sc->sc_isreset = 0;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_NORMAL_COMPLETION;
goto ret;
case UHF_PORT_CONNECTION:
case UHF_PORT_OVER_CURRENT:
case UHF_PORT_POWER:
case UHF_PORT_LOW_SPEED:
case UHF_C_PORT_SUSPEND:
default:
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
break;
case C(UR_GET_BUS_STATE, UT_READ_CLASS_OTHER):
if (index == 1)
port = UHCI_PORTSC1;
else if (index == 2)
port = UHCI_PORTSC2;
else {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
if (len > 0) {
*(u_int8_t *)buf =
(UREAD2(sc, port) & UHCI_PORTSC_LS) >>
UHCI_PORTSC_LS_SHIFT;
totlen = 1;
}
break;
case C(UR_GET_DESCRIPTOR, UT_READ_CLASS_DEVICE):
if (value != 0) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
l = min(len, USB_HUB_DESCRIPTOR_SIZE);
totlen = l;
memcpy(buf, &uhci_hubd_piix, l);
break;
case C(UR_GET_STATUS, UT_READ_CLASS_DEVICE):
if (len != 4) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
memset(buf, 0, len);
totlen = len;
break;
case C(UR_GET_STATUS, UT_READ_CLASS_OTHER):
if (index == 1)
port = UHCI_PORTSC1;
else if (index == 2)
port = UHCI_PORTSC2;
else {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
if (len != 4) {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
x = UREAD2(sc, port);
status = change = 0;
if (x & UHCI_PORTSC_CCS )
status |= UPS_CURRENT_CONNECT_STATUS;
if (x & UHCI_PORTSC_CSC )
change |= UPS_C_CONNECT_STATUS;
if (x & UHCI_PORTSC_PE )
status |= UPS_PORT_ENABLED;
if (x & UHCI_PORTSC_POEDC)
change |= UPS_C_PORT_ENABLED;
if (x & UHCI_PORTSC_OCI )
status |= UPS_OVERCURRENT_INDICATOR;
if (x & UHCI_PORTSC_OCIC )
change |= UPS_C_OVERCURRENT_INDICATOR;
if (x & UHCI_PORTSC_SUSP )
status |= UPS_SUSPEND;
if (x & UHCI_PORTSC_LSDA )
status |= UPS_LOW_SPEED;
status |= UPS_PORT_POWER;
if (sc->sc_isreset)
change |= UPS_C_PORT_RESET;
USETW(ps.wPortStatus, status);
USETW(ps.wPortChange, change);
l = min(len, sizeof ps);
memcpy(buf, &ps, l);
totlen = l;
break;
case C(UR_SET_DESCRIPTOR, UT_WRITE_CLASS_DEVICE):
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
case C(UR_SET_FEATURE, UT_WRITE_CLASS_DEVICE):
break;
case C(UR_SET_FEATURE, UT_WRITE_CLASS_OTHER):
if (index == 1)
port = UHCI_PORTSC1;
else if (index == 2)
port = UHCI_PORTSC2;
else {
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
switch(value) {
case UHF_PORT_ENABLE:
x = UREAD2(sc, port);
UWRITE2(sc, port, x | UHCI_PORTSC_PE);
break;
case UHF_PORT_SUSPEND:
x = UREAD2(sc, port);
UWRITE2(sc, port, x | UHCI_PORTSC_SUSP);
break;
case UHF_PORT_RESET:
x = UREAD2(sc, port);
UWRITE2(sc, port, x | UHCI_PORTSC_PR);
usb_delay_ms(&sc->sc_bus, 10);
UWRITE2(sc, port, x & ~UHCI_PORTSC_PR);
delay(100);
x = UREAD2(sc, port);
UWRITE2(sc, port, x | UHCI_PORTSC_PE);
delay(100);
DPRINTFN(3,("uhci port %d reset, status = 0x%04x\n",
index, UREAD2(sc, port)));
sc->sc_isreset = 1;
break;
case UHF_PORT_POWER:
/* Pretend we turned on power */
err = USBD_NORMAL_COMPLETION;
goto ret;
case UHF_C_PORT_CONNECTION:
case UHF_C_PORT_ENABLE:
case UHF_C_PORT_OVER_CURRENT:
case UHF_PORT_CONNECTION:
case UHF_PORT_OVER_CURRENT:
case UHF_PORT_LOW_SPEED:
case UHF_C_PORT_SUSPEND:
case UHF_C_PORT_RESET:
default:
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
break;
default:
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
err = USBD_IOERROR;
goto ret;
}
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->actlen = totlen;
err = USBD_NORMAL_COMPLETION;
ret:
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
xfer->status = err;
xfer->hcpriv = 0;
s = splusb();
usb_transfer_complete(xfer);
splx(s);
return (USBD_IN_PROGRESS);
}
/* Abort a root control request. */
void
uhci_root_ctrl_abort(usbd_xfer_handle xfer)
{
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
/* Nothing to do, all transfers are synchronous. */
}
/* Close the root pipe. */
void
uhci_root_ctrl_close(usbd_pipe_handle pipe)
{
DPRINTF(("uhci_root_ctrl_close\n"));
}
/* Abort a root interrupt request. */
void
uhci_root_intr_abort(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
uhci_softc_t *sc = (uhci_softc_t *)xfer->pipe->device->bus;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usb_untimeout(uhci_timo, xfer, xfer->timo_handle);
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
sc->sc_has_timo = NULL;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
if (xfer->pipe->intrxfer == xfer) {
DPRINTF(("uhci_root_intr_abort: remove\n"));
xfer->pipe->intrxfer = 0;
}
xfer->status = USBD_CANCELLED;
usb_transfer_complete(xfer);
}
usbd_status
uhci_root_intr_transfer(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_status err;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
/* Insert last in queue. */
err = usb_insert_transfer(xfer);
if (err)
return (err);
/* Pipe isn't running (otherwise err would be USBD_INPROG),
* start first
*/
return (uhci_root_intr_start(SIMPLEQ_FIRST(&xfer->pipe->queue)));
}
/* Start a transfer on the root interrupt pipe */
usbd_status
uhci_root_intr_start(usbd_xfer_handle xfer)
{
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usbd_pipe_handle pipe = xfer->pipe;
uhci_softc_t *sc = (uhci_softc_t *)pipe->device->bus;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
DPRINTFN(3, ("uhci_root_intr_transfer: xfer=%p len=%d flags=%d\n",
xfer, xfer->length, xfer->flags));
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
sc->sc_ival = MS_TO_TICKS(xfer->pipe->endpoint->edesc->bInterval);
usb_timeout(uhci_timo, xfer, sc->sc_ival, xfer->timo_handle);
sc->sc_has_timo = xfer;
return (USBD_IN_PROGRESS);
}
/* Close the root interrupt pipe. */
void
uhci_root_intr_close(usbd_pipe_handle pipe)
{
uhci_softc_t *sc = (uhci_softc_t *)pipe->device->bus;
Synchronisation with NetBSD as of 1999/11/16: Cleaning up the code: - Declare many functions static - Change variable names to make them more self explanatory - Change usbd_request_handle -> usbd_xfer_handle - Syntactical changes - Remove some unused code - Other KNF changes Interrupt context handling - Change delay to usbd_delay_ms were possible (takes polling mode into account) - Change detection mechanism for interrupt context Add support for pre-allocation DMA-able memory by device driver Add preliminary support for isochronous to the UHCI driver (not for OHCI yet). usb.c, uhci.c, ohci.c - Initial attempt at detachable USB host controllers - Handle the use_polling flag with a lttle more care and only set it if we are cold booting. usb.c, uhci.c ohci.c, usbdi.c usbdi_util.c usb_subr.c - Make sure an aborted pipe is marked as not running. - Start queued request in the right order. - Insert some more DIAGNOSTIC sanity checks. - Remove (almost) unused definitions USBD_XFER_OUT and USBD_XFER_IN. usb.c, usb_subr.c - Add an event mechanism so that a userland process can watch devices come and go. ohci.c - Handle the case when a USB transfer is so long that it crosses two page (4K) boundaries. OHCI cannot do that with a single TD so we make a chain. ulpt.c - Use a bigger buffer when transferring data. - Pre-allocate the DMA buffer. This makes the driver slightly more efficient. - Comment out the GET_DEVICE_ID code, because for some unknown reason it causes printing to fail sometimes. usb.h - Add a macro to extract the isoc type. - Add a macro to check whether the routine has been entered after splusb and if not, complain. usbdi.c - Fix a glitch in dequeueing and aborting requests on interrupt pipes. - Add a flag in the request to determine if the data copying is done by the driver or the usbdi layer.
1999-11-17 22:33:51 +00:00
usb_untimeout(uhci_timo, pipe->intrxfer, pipe->intrxfer->timo_handle);
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
sc->sc_has_timo = NULL;
DPRINTF(("uhci_root_intr_close\n"));
}