- 01 Feb, 2008 40 commits
-
-
David Brownell authored
While most isochronous endpoints have short polling intervals, the EHCI driver won't necessarily handle larger ones correctly. This patch switches to use a "u16" to represent those periods, not a u8, since it can always work: the largest expressible period is 2^15 units ... not the previous too-short limit of 128 frames (full or low speeds) or microframes (high speed, 32 frames). This bug is essentially theoretical, since the few ISO endpoints I've seen which don't use one transfer per frame are high speed ones using more than that (including high bandwidth, 24 KB/msec). Signed-off-by:
David Brownell <dbrownell@users.sourceforge.net> Signed-off-by:
Greg Kroah-Hartman <gregkh@suse.de>
-
David Brownell authored
Some of the "EHCI ports reset forever" problems may be explained by code paths which wrongly flagged resets as complete. This removes two such paths; the ehci_hub_status_data() path should be the only one to have an effect, since it was already properly flagged on the other path. (Issue noted by Minhyoung Kim <a9a9@lge.com>.) Signed-off-by:
-
Denis Cheng authored
commit 04d06ad0 have added menuconfig support for the whole USB Kconfig, but there are still menuconfig need for usb/serial, usb/atm, and usb/gadget, so that the user can disable all the options in that menu at once instead of having to disable each option separately. Signed-off-by:
Denis Cheng <crquan@gmail.com> Signed-off-by:
-
Magnus Damm authored
usb: ohci-sm501 driver V2 This patch adds sm501 ohci support. It's all very straightforward with the exception of dma_declare_coherent_memory() and HCD_LOCAL_MEM. Together they are used to ensure that usb data is allocated using dma_alloc_coherent(), and that only valid dma memory is used to allocate from. This driver is a platform device, and the mfd driver sm501.c is already creating one usb host controller instance per sm501. Signed-off-by:
Magnus Damm <damm@igel.co.jp> Cc: David Brownell <david-b@pacbell.net> Signed-off-by:
-
Magnus Damm authored
usb: dma bounce buffer support V4 This patch adds dma bounce buffer support to the usb core. These buffers can be enabled with the HCD_LOCAL_MEM flag, and they make sure that all data passed to the host controller is allocated using dma_alloc_coherent(). Signed-off-by:
Alan Stern <stern@rowland.harvard.edu> Acked-by:
-
Oliver Neukum authored
these drivers abused intfdata in close() as flags for binding. That races with reprobing of those devices. This patch fixes that by using the flag and the locks introduced with the patch against mos7720. Signed-off-by:
Oliver Neukum <oneukum@suse.de> Signed-off-by:
-
Oliver Neukum authored
If disconnect() is called for a logical disconnect, no more IO must be done after disconnect() returns, or the old and new drivers may conflict. This patch avoids this by using the flag and lock introduced by the earlier patch for the mos7720 driver. Signed-off-by:
-
Alan Cox authored
- Rename the copied buffer functions from pl2303 to oti6858 to avodi confusion - Initialise speeds properly - Use modern baud rate handling - Remove GSERIAL/SSERIAL ioctl hacks that reference termios unlocked Signed-off-by:
Alan Cox <alan@redhat.com> Signed-off-by:
-
Oliver Neukum authored
this covers the rest of the obvious cases by using the flags and locks to guard against disconnect which were introduced in the earlier patch against mos7720. Signed-off-by:
-
Oliver Neukum authored
If disconnect() is called for a logical disconnect, no more IO must be done after disconnect() returns, or the old and new drivers may conflict. This patch avoids this by using the flag and lock introduced by the earlier patch for the mos7720 driver. Signed-off-by:
-
Oliver Neukum authored
in an error case memory already allocated must be freed again. Signed-off-by:
-
Oliver Neukum authored
this driver uses usb_get_intfdata() == NULL as a test for disconnect(). You must not do that as this races with probe(). By the time you test your erstwhile interface may already be somebody else's interface. This fixes the close() method of cypress_m8 to use the recently introduced flag and use locking against disconnect() where required in close(). Signed-off-by:
-
Jan Engelhardt authored
Signed-off-by:
Jan Engelhardt <jengelh@computergmbh.de> Signed-off-by:
-
Oliver Neukum authored
This fixes a problem where the mos7720 driver will make io to a device from which it has been logically disconnected. It does so by introducing a flag by which the generic usb serial code can signal the subdrivers their disconnection and appropriate locking. Signed-off-by:
-
Yoshihiro Shimoda authored
add support for SuperH OHCI. supported CPU are: - SH7720 - SH7721 - SH7763 Signed-off-by:
Yoshihiro Shimoda <shimoda.yoshihiro@renesas.com> Acked-by:
David Brownell <david-b@pacbell.net> Signed-off-by:
-
Greg Kroah-Hartman authored
People keep trying to add entries to this section of the driver for things. That's what the Changelog is supposed to be for, not the .c file. Signed-off-by: -
Paul Chavent authored
When building an external module, the ezusb_* functions are not defined if we haven't loaded any built'in module that use them (whiteheat, keyspan, ...). This patch allow to build those functions even if we only have selected the usbserial generic driver. Signed-off-by:
Paul Chavent <paul.chavent@fnac.net> Signed-off-by:
-
Alan Stern authored
This patch (as1031) adds a short delay to the bus-suspend routine in ehci-hcd. Without it some devices disconnect when they should suspend. Signed-off-by:
-
Alan Stern authored
This patch (as1030b) moves a del_timer_sync() call outside the scope of a spinlock, where it could cause a deadlock, and adds a new del_timer_sync() call for the new IAA watchdog timer (it was omitted by mistake). Signed-off-by:
-
Pete Zaitcev authored
The ISO descriptors are allocated separately in proc_submiturb for a fetch from user mode, then tucked at the end of URB. This seems like a dead code. Signed-off-by:
Pete Zaitcev <zaitcev@redhat.com> Signed-off-by:
-
David Brownell authored
This has some bugfixes for the EHCI driver's ISO transfer scanning logic. It was leaving ITDs and SITDs on the schedule too long, for a few different reasons, which caused trouble. (a) Look at all microframes for high speed transfers, not just the ones we expect to have finished. This way transfers ending mid-frame will complete without needing another IRQ. This also minimizes bogus scheduling underruns (e.g. EL2NSYNC). (b) When we encounter an ISO transfer (either speed, but this hits mostly at full speed) that's not yet been completed, immediately stop scanning; we've caught up to the hardware, no matter what other indications might say. (c) Always clean up ITDs (for high speed transfers) when the HC is no longer running. I'm not sure whether the last one has been observed before, but both the others have been reported with "real world" audio and video code. Signed-off-by:
-
David Brownell authored
Small updates to the EHCI driver's ISO support: - Get rid of the Kconfig option for full speed ISO. It may not be perfect yet, but it hasn't appeared to be dangerous and pretty much every configuration wants it. - Instead of two places to disable an empty periodic schedule after an ISO transfer completes, just have one. - After the periodic schedule is disabled, we can short-circuit the schedule scan ... it can't possibly have more work to do. Assuming a typical config with split iso enabled, the only change in behavior should be almost unobservable: quicker termination of periodic scans when the schedule gets emptied. Signed-off-by:
-
Karsten Wiese authored
In the EHCI driver, itd->usecs[8] is used in periodic_usecs(), indexed by uframe. For an ITD's unused uframes it is 0, else it contains the same value as itd->stream->usecs. To check if an ITD's uframe is used, we can instead test itd->hw_transaction[uframe]: if used, it will be nonzero no matter what endianess is used. This patch replaces those two uses, eliminates itd->usecs[], and saves eight bytes from each ITD. Signed-off-by:
Karsten Wiese <fzu@wemgehoertderstaat.de> Signed-off-by:
-
Vladimir Barinov authored
EHCI Glue driver for Intel IXP4XX EHCI USB controller Signed-off-by:
Vladimir Barinov <vbarinov@ru.mvista.com> Signed-off-by:
-
Daniel Walker authored
The dev->sem conforms to mutex style usage. This patch converts it to use the struct mutex type, and new API. There is also a small style fix around this comment, /* unlock here as tower_delete frees dev */ Where I broke the line up to meet the 80 char limit. Signed-off-by:
Daniel Walker <dwalker@mvista.com> Signed-off-by:
-
Valentine Barshak authored
This adds device-tree-aware ehci-ppc-of driver. The code is based on the ehci-ppc-soc driver by Stefan Roese <sr@denx.de>. Signed-off-by:
Valentine Barshak <vbarshak@ru.mvista.com> Acked-by:
Arnd Bergmann <arnd@arndb.de> Acked-by:
Stefan Roese <sr@denx.de> Signed-off-by:
-
Dmitry Baryshkov authored
Use debugfs instead of /proc/driver/udc Signed-off-by:
Dmitry Baryshkov <dbaryshkov@gmail.com> Signed-off-by:
-
David Brownell authored
Various small at91_udc cleanups: - Use generic GPIO calls, not older platform-specific ones - Use gpio_request()/gpio_free() - Use VERBOSE_DEBUG convention, not older VERBOSE - Fix sparse complaint about parameter type (changed to gfp_t) - Add missing newline to some rarely-seen debug messages - Fix some old cleanup bugs on probe() fault paths Also add a mechanism whereby rm9200 gpios can drive the D+ pullup through an inverting transistor, based on a patch from Steve Birtles. Most UDC drivers supporting a GPIO based pullup should probably have such an option, but testing it requries such a board in hand! Signed-off-by:
-
David Brownell authored
Update the ohci-at91 bus glue to start understanding about the per-port power switch GPIOs it's given (on the sam9263-ek and potentially other boards). For the moment this just claims them and forces them active (assuming active-low power enables) whenever the HCD is loaded. The assumption is still that board setup configures the GPIOs. Using gpio_request() tracks actual usage and guards against conflict. Signed-off-by:
-
Sarah Sharp authored
This documents two newly created files: /sys/bus/usb/device/.../power/connected_duration /sys/bus/usb/device/.../power/active_duration Documentation was placed in Documentation/ABI/testing, since that's where the documentation is for the other USB sysfs power files. Signed-off-by:
Sarah Sharp <sarah.a.sharp@intel.com> Signed-off-by:
-
Sarah Sharp authored
This patch exports two statistics to userspace: /sys/bus/usb/device/.../power/connected_duration /sys/bus/usb/device/.../power/active_duration connected_duration is the total time (in msec) that the device has been connected. active_duration is the total time the device has not been suspended. With these two statistics, tools like PowerTOP can calculate the percentage time that a device is active, i.e. not suspended or auto-suspended. Users can also use the active_duration to check if a device is actually autosuspended. Currently, they can set power/level to auto and power/autosuspend to a positive timeout, but there's no way to know from userspace if a device was actually autosuspended without looking at the dmesg output. These statistics will be useful in creating an automated userspace script to test autosuspend for USB devices. Signed-off-by:
-
Kevin Lloyd authored
The following improvements were made: - Fixed control line issue where asserting DTR on ep5 would close ep2 - Added support for calc_num_ports (will help support future composite devices) Signed-off-by:
Kevin Lloyd <linux@sierrawireless.com> Signed-off-by:
-
Harald Welte authored
[PATCH] ftdi_sio: add support for more FTDI based JTAG adaptors There are more devices similar to the Olimex JTAG adaptor, in that the first port of the FT2232C is used for JTAG, and only the second port is available as UART. I have thus renamed ftdi_olimex_{probe,quirk} to ftdi_jtag_{probe,quirk} and added vendor/product ID's for the OpenMoko Neo1973 Debug Board as well as the OOCDlink device. I've also updated the KERN_INFO message sent to userspace to remove the word 'olimex' and an extra '\n' that was causing an empty line in dmesg. Signed-off-by:Harald Welte <laforge@openmoko.org> Signed-off-by:
-
Alan Cox authored
Don't pass NULL into termios functions when calling them internally Remove all the crap which then checks for NULL which can't occur now Clear CMSPAR as it is not supported Report the baud rate back to the caller properly (See FIXME someone with the docs) Signed-off-by:
-
Alan Cox authored
Remove internal NULL passing in termios code Remove all the if checks it causes Encode the baud rate back properly Clear CMSPAR as it is not supported Signed-off-by:
-
Alan Cox authored
Remove lots of NULL checks that can no longer occur Encode the baud rate back into the termios (again someone with docs see FIXME to improve this further) Signed-off-by:
-
Dmitry Baryshkov authored
Use gpio_vbus instead of udc_is_connected for udc on tosa. Signed-off-by:
Russell King <rmk+kernel@arm.linux.org.uk> Signed-off-by:
-
David Brownell authored
Add a missing dependency which goofs up the xconfig display. Signed-off-by:
-
David Brownell authored
Fix a small glitch noted by Yannick Cote. There is no endpoint number six, so if a (broken) host wrongly tried to change or read status of that endpoint, the driver could access reserved register space. Signed-off-by:
-
Oliver Neukum authored
Am Sonntag, 16. Dezember 2007 05:23:47 schrieb Andrew Morton: > On Sat, 15 Dec 2007 07:08:52 -0800 (PST) bugme-daemon@bugzilla.kernel.org wrote: > > > http://bugzilla.kernel.org/show_bug.cgi?id=9573 > > There's a null pointer dereference on drivers/usb/serial/whiteheat.c as > > follows: > > > > (1) line 613: the test "if (port->tty)" implies that null is a legal value > > for "port->tty" at that point > > > > (2) neither firm_open nor firm_purge initialize "port->tty" drivers/usb/serial/usb-serial::serial_open() sets port->tty The check for NULL is bogus. This patch removes the check. Signed-off-by:
-
