- 12 Jul, 2007 40 commits
-
-
Pete Zaitcev authored
This patch releases DMA resources if enqueue fails in the HCD. Linux had this bug ever since we converted from virt_to_bus for 2.4. It is difficult to hit. A user would need a significant memory pressure or some other unusual condition. It was reported to me by IBM. They ran a management application for RSA II adapters which sent Bulk requests to an Interrupt endpoint. Submissions got rejected by HCD due to an invalid interval value and the swiotlb pool became depleted in the matter of hours. We fixed the invalid interval issue in devio.c separately, but this seems to be a bug worth fixing as well. Signed-off-by:
Pete Zaitcev <zaitcev@redhat.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@suse.de>
-
Alan Stern authored
A noticeable number of low-speed devices mistakenly include descriptors for Bulk endpoints, which is forbidden by the USB spec. In an attempt to make such devices more usable, this patch (as924) converts the descriptors to Interrupt with an interval of 1 ms. Signed-off-by:
Alan Stern <stern@rowland.harvard.edu> Signed-off-by:
-
Benny Halevy authored
I think there is a race between usb_serial_put() and usb_serial_get_by_index() (and get_free_serial()) with regards to handling the serial port refcount. usb_serial_get_by_index() gets a reference on the serial port under table_lock while return_serial releases all the returned ports from the table under the same lock. However, the table_lock is not taken around the call to kref_put, theoretically allowing to sneak in and grab a reference after kref_put has already determined that the reference count is zero (and before calling destroy_serial) causing use after free. Signed-off-by:
Benny Halevy <bhalevy@ns1.bhalevy.com> Cc: Oliver Neukum <oneukum@suse.de> Signed-off-by:
-
Oliver Neukum authored
this fixes the sleep found with the automatic tool. Signed-off-by:
Oliver Neukum <oneukum@suse.de> Signed-off-by:
-
Christian Engelmayer authored
According to the USB Specification Revision 2.0 chapter 11.12.5 a hub experiencing an over-current condition must place all affected ports in the powered-off state. It seems that some root hubs need port power to be cycled by software in order to get back to normal functionality after an over-current condition ... like the EHCI implementation on an MPC8343E. Signed-off-by:
Christian Engelmayer <christian.engelmayer@frequentis.com> Signed-off-by:
David Brownell <david-b@pacbell.net> Signed-off-by:
-
David Brownell authored
Platforms with PCMCIA support can implement host-side USB with "sl811_cs", so make sure this menu shows up on platforms with PCMCIA. Signed-off-by:
David Brownell <dbrownell@users.sourceforge.net> Signed-off-by:
-
Alan Stern authored
This patch (as923) makes usb-storage's control thread use kthread_should_stop()/kthread_stop(). The scanning thread can't be similarly converted until the core kthread implementation allows threads to call do_exit(). The advantage of this change is that we can now be certain the control thread has terminated before storage_disconnect() returns. This will simplify the locking requirements when autosuspend support is added. Signed-off-by:
Matthew Dharm <mdharm-usb@one-eyed-alien.net> Signed-off-by:
-
Arnaud Patard authored
This patch adds the support for the Usb Device Controller on Samsung S3C24xx SoCs. This driver passes all tests from testusb (including #13) and has been tested on S3C2410, S3C24212, and S3C2440 SoCs. Whitespace updates, minor cleanups by David Signed-off-by:
Arnaud Patard <arnaud.patard@rtp-net.org> Signed-off-by:
Ben Dooks <ben-linux@fluff.org> Cc: Herbert Pötzl <herbert@13thfloor.at> Signed-off-by:
-
Geoff Levand authored
USB HCD glue updates to reflect the new PS3 unifed device support. - Fixed remove() routine. - Added shutdown() routine. - Added request_mem_region() call. - Fixed MODULE_ALIAS(). - Made a proper fix for the hack done to support muti-platform in commit 48fda451. Signed-off-by:
Geoff Levand <geoffrey.levand@am.sony.com> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by:
-
Alan Stern authored
Might fix bug 8561 On Mon, 4 Jun 2007, Paulo Pereira wrote: > The patch that you send is not resolving the problem... :( > I stil have Kernel panic after 45/60 min of work with Ktorrent/Amule... > > The Drump is: > > Call Trace: > [<c055fb36>] usb_hcd_submit+0xb1/0x763 > [<f9276488>] ipt_do_table+0x2c7/0x2ef [ip_tables] > [<f929a6d7>] nf_ct_deliver_cached_events+0x41/0x96 [nf_conntrak] > [<f9288254>] ipv4_confirm+0x36/0c3b [nf_conntrack_ipv4] > [<c05ce7c2>] tcp_v4_rcv+0x827/0x899 > [<c05afcc0>] nf_hook_slow+0x4d/0xb5 > [<c042826f>] irq_enter+0x19/0x23 > [<c042826f>] irq_enter+0x19/0x23 > [<c040794c>] do_IRQ+0xbd/0xd1 > [<f90893c9>] option_write+0xa7/0xef [option] Okay, from this it looks like there's a problem in the option.c serial driver. Glancing at the code, it's obvious why: The thing totally abuses the USB API. Try applying this patch; it should help. From: Alan Stern <stern@rowland.harvard.edu> Cc: Paulo Pereira <pfmp.404@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
-
Li Yang authored
Currently the driver is expecting max ep number in platform data which isn't passing this information. This patch fix the problem by reading it from DCCPARAMS(Device Controller Capability Parameters) register. The change also need some reordering of the probe code. Signed-off-by:
Li Yang <leoli@freescale.com> Signed-off-by:
-
Li Yang authored
Signed-off-by:
-
Alan Stern authored
This patch (as920) adds an extra level of protection to the USB-Persist facility. Now it will apply by default only to hubs; for all other devices the user must enable it explicitly by setting the power/persist device attribute. The disconnect_all_children() routine in hub.c has been removed and its code placed inline. This is the way it was originally as part of hub_pre_reset(); the revised usage in hub_reset_resume() is sufficiently different that the code can no longer be shared. Likewise, mark_children_for_reset() is now inline as part of hub_reset_resume(). The end result looks much cleaner than before. The sysfs interface is updated to add the new attribute file, and there are corresponding documentation updates. Signed-off-by:
-
Alan Stern authored
This patch (as919) unifies the code paths used for normal resume and for reset-resume. Earlier I had failed to note a section in the USB spec which requires the host to resume a suspended port before resetting it if the attached device is enabled for remote wakeup. Since the port has to be resumed anyway, we might as well reuse the existing code. The main changes are: usb_reset_suspended_device() is eliminated. usb_root_hub_lost_power() is moved down next to the hub_reset_resume() routine, to which it is logically related. finish_port_resume() does a port reset() if the device's reset_resume flag is set. usb_port_resume() doesn't check whether the port is initially enabled if this is a USB-Persist sort of resume. Code to perform the port reset is added to the resume pathway for the non-CONFIG_USB_SUSPEND case. Signed-off-by:
-
Alan Stern authored
This patch (as918) introduces a new USB driver method: reset_resume. It is called when a device needs to be reset as part of a resume procedure (whether because of a device quirk or because of the USB-Persist facility), thereby taking over a role formerly assigned to the post_reset method. As a consequence, post_reset no longer needs an argument indicating whether it is being called as part of a reset-resume. This separation of functions makes the code clearer. In addition, the pre_reset and post_reset method return types are changed; they now must return an error code. The return value is unused at present, but at some later time we may unbind drivers and re-probe if they encounter an error during reset handling. The existing pre_reset and post_reset methods in the usbhid, usb-storage, and hub drivers are updated to match the new requirements. For usbhid the post_reset routine is also used for reset_resume (duplicate method pointers); for the other drivers a new reset_resume routine is added. The change to hub.c looks bigger than it really is, because mark_children_for_reset_resume() gets moved down next to the new hub_reset_resume() routine. A minor change to usb-storage makes the usb_stor_report_bus_reset() routine acquire the host lock instead of requiring the caller to hold it already. Signed-off-by:
Jiri Kosina <jkosina@suse.cz> CC: Matthew Dharm <mdharm-usb@one-eyed-alien.net> Signed-off-by:
-
Alan Stern authored
This patch (as917) removes a now-unnecessary level of subroutine nesting from hub.c. Since usb_port_suspend() does nothing but call hub_port_suspend(), and usb_port_resume() does nothing but call hub_port_resume(), there's no reason to keep the routines separate. Also included in the patch are a few cosmetic changes involving whitespace and use of braces. Signed-off-by:
-
Alan Stern authored
This patch (as916) completes the separation of code paths for suspend and resume of root hubs as opposed to non-root devices. Root hubs will be power-managed through their bus_suspend and bus_resume methods, whereas normal devices will use usb_port_suspend() and usb_port_resume(). Changes to the hcd_bus_{suspend,resume} routines mostly represent motion of code that was already present elsewhere. They include: Adding debugging log messages, Setting the device state appropriately, and Adding a resume recovery time delay. Changes to the port-suspend and port-resume routines in hub.c include: Removal of checks for root devices (since they will never be triggered), and Removal of checks for NULL or invalid device pointers (these were left over from earlier kernel versions and aren't needed at all). Signed-off-by: -
Alan Stern authored
This patch (as915b) combines the public routine usb_port_suspend() and the private routine __usb_port_suspend() into a single function. By removing the explicit mention of otg_port in the call to __usb_port_suspend(), we prevent a possible error in which the system tries to perform HNP on the wrong port when a non-targeted device is plugged into a non-OTG port. Signed-off-by:
-
Michael Hanselmann authored
This patch fixes a silicon bug in some NEC OHCI chips. The bug appears at random times and is very, very difficult to reproduce. Without the following patch, Linux would shut the chip and its associated devices down. In Apple PowerBooks this leads to an unusable keyboard and mouse (SSH still working). The idea of restarting the chip is taken from public Darwin code. Signed-off-by:
Michael Hanselmann <linux-kernel@hansmi.ch> Cc: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by:
-
Tony Lindgren authored
Disable file_storage USB_CONFIG_ATT_WAKEUP as it requires user interaction during Chapter 9 tests. Signed-off-by: Tony Lindgren <tony@atomide.com Signed-off-by:
Felipe Balbi <felipe.lima@indt.org.br> Acked-by:
-
Vladimir Barinov authored
This patch fixes the endianness select for transfer buffers in EHCI controllers that have Transaction Translator built in the hub. Also I cleaned it up to make rid of magic numbers. Signed-off-by:
Vladimir Barinov <vbarinov@ru.mvista.com> Cc: <david-b@pacbell.net> Signed-off-by:
-
Al Borchers authored
Some further cleanup after Oliver's patch to update the tty buffering. The input buffer is not used at all anymore, so I removed it. Signed-off-by:
Al Borchers <alborchers@steinerpoint.com> Signed-off-by:
-
Oliver Neukum authored
this is an update of the whiteheat driver. It fixes: - switch from spinlocks to mutexes to prevent sleeping with a spinlock held - locking to stop races with disconnect - error handling for commands that time out Signed-off-by:
-
Oliver Neukum authored
This patch set introduces usb_anchor and uses it to implement all modern APIs in the skeleton driver. - proper error reporting in the skeleton driver - implementation of flush() Signed-off-by:
-
Oliver Neukum authored
- introduction of usb_anchor and its methods Signed-off-by:
-
Vikram Pandita authored
Patch is to prevent the OTG host of doing 3 times enumeration of device when the Host suspends for HNP. The error code used in this case is ENOTSUPP. Signed-off-by:
Vikram Pandita <vikram.pandita@ti.com> Acked-by:
-
David Brownell authored
Fix a few serial gadget issues reported by the latest "sparse": some functions should have been defined as static, not just declared that way. Signed-off-by:
-
t.sefzick authored
in order to be able to switch back to 'flow-control none' after having activated 'flow-control rts/cts', I made a small change to 'pl2303.c'. Signed-off-by: -
Oliver Neukum authored
the generic driver also had its own buffering. Signed-off-by: Oliver Neukum <oneukum@suse.de_ Signed-off-by: -
Oliver Neukum authored
this fixes the flushing trouble due to its own buffering for this driver. Signed-off-by:
-
Oliver Neukum authored
the new tty buffering code allows usb drivers to stop private buffering. In fact we must do so to allow flushing to work correctly. This does so for the visor driver. Signed-off-by:
-
David Brownell authored
Make sure gadgetfs userspace interface is properly exported: - Move <linux/usb_gadgetfs.h> to <linux/usb/gadgetfs.h>; - Export it using Kbuild; - Add an #include guard; - Correct some internal documentation; - Update struct layout so it's the same on 32/64 bit kernels. Signed-off-by:
-
David Rientjes authored
Substitute USB instances of __attribute__ ((unused)) functions with the newly introduced __maybe_unused. Signed-off-by:
David Rientjes <rientjes@google.com> Signed-off-by:
-
Li Yang authored
For MPC831x support, change the ehci-fsl driver to preserve bits set in platform code. Add a common CONFIG_USB_EHCI_FSL to indicate presence of Freescale EHCI SOC. Add FSL_USB2_DR_OTG operating mode support, thus both host and device can work for the mini-ab receptacle. Note: this doesn't enable OTG protocol support. Signed-off-by:
-
Stefan Roese authored
Now select the big-endian configuration options CONFIG_USB_EHCI_BIG_ENDIAN_MMIO and CONFIG_USB_EHCI_BIG_ENDIAN_DESC in the usb host Kconfig file and not in the platform Kconfig files. Signed-off-by:
Stefan Roese <sr@denx.de> Signed-off-by:
-
Stefan Roese authored
This patch adds support for the AMCC 440EPx EHCI controller whose in-memory data structures and the registers are represented in big- endian format. Signed-off-by:
-
Stepan Moskovchenko authored
The new FT232RL allows setting and getting the value of the latency timer, like on the FT232BM. However, the driver will not create the sysfs entries for the RL without this one-line patch. I have tested it on two systems with successful results. From: Stepan Moskovchenko <stevenm86@gmail.com> Signed-off-by: -
Alan Stern authored
This patch (as911) replaces some C++-style commented-out debugging lines in driver.c with a new "verbose debugging" macro. It makes the code look cleaner, and it's easier to turn the debugging on or off. Signed-off-by:
-
Alan Stern authored
This patch (as908) adds central protection in usbcore for the prototypical race between opening and unregistering a char device. The spinlock used to protect the minor-numbers array is replaced with an rwsem, which can remain locked across a call to a driver's open() method. This guarantees that open() and deregister() will be mutually exclusive. The private locks currently used in several individual drivers for this purpose are no longer necessary, and the patch removes them. The following USB drivers are affected: usblcd, idmouse, auerswald, legousbtower, sisusbvga/sisusb, ldusb, adutux, iowarrior, and usb-skeleton. As a side effect of this change, usb_deregister_dev() must not be called while holding a lock that is acquired by open(). Unfortunately a number of drivers do this, but luckily the solution is simple: call usb_deregister_dev() before acquiring the lock. In addition to these changes (and their consequent code simplifications), the patch fixes a use-after-free bug in adutux and a race between open() and release() in iowarrior. Signed-off-by:
-
Mark Lord authored
Okay, found it. The root cause here was a missing CONFIG_USB_SUSPEND=y, which means the hci_usb device never got marked as USB_STATE_SUSPENDED, which then caused the loop to go on forever. The system works fine now with CONFIG_USB_SUSPEND=y in the .config. Here's the patch to prevent future lockups for this or other causes. I no longer need it, but it does still seem a good idea. Signed-off-by:
Mark Lord <mlord@pobox.com> Signed-off-by:
-
