Commit graph

60 commits

Author SHA1 Message Date
Wolfgang Denk
3ed16071b0 USB: sync Queue Element Transfer Descriptor against EHCI spec
Appendix B "EHCI 64-Bit Data Structures" of the "Enhanced Host
Controller Interface Specification for Universal Serial Bus" (Rev.
1.0, March 12, 2002) defines additional fields which were missing in
U-Boot's struct qTD; as these are also present in recent versions of
struct ehci_qtd in the Linux kernel, we add them here, too.

This fixes some nasty memory corruption problems.

Reported-by: Dan Lykowski <lykowdk@gmail.com>
See http://permalink.gmane.org/gmane.comp.boot-loaders.u-boot/76942

Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Remy Bohmer <linux@bohmer.net>
Cc: Dan Lykowski <lykowdk@gmail.com>
Cc: Stefano Babic <sbabic@denx.de>
Tested-by: Stefano Babic <sbabic@denx.de>
2010-10-20 00:23:57 +02:00
Sergei Shtylyov
9fb3b50857 EHCI: zero out QH transfer overlay in ehci_submit_async()
ehci_submit_async() doesn't really zero out the QH transfer overlay (as the EHCI
specification suggests) which leads to the controller seeing the "token" field
as the previous call has left it, i.e.:
- if a timeout occured on the previous call (Active bit left as 1), controller
  incorrectly tries to complete a previous transaction on a newly programmed
  endpoint;
- if a halt occured on the previous call (Halted bit set to 1), controller just
  ignores the newly programmed TD(s) and the function then keeps returning error
  ad infinitum.

This turned out to be caused by the wrong orger of the arguments to the memset()
call in ehci_alloc(), so the allocated TDs weren't cleared either.

While at it, stop needlessly initializing the alternate next TD pointer in the
QH transfer overlay...

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Acked-by: Remy Bohmer <linux@bohmer.net>
2010-06-29 23:03:40 +02:00
Sergei Shtylyov
c8b2d1dc0f EHCI: fix port reset reporting
Commit b416191a14 (Fix EHCI port reset.) didn't
move the code that checked for successful clearing of the port reset bit from
ehci_submit_root(), relying on wait_ms() call instead. The mentioned code also
erroneously reported port reset state when the reset was already completed.

Signed-off-by: Sergei Shtylyov <sshtylyov@mvista.com>
2010-04-08 21:39:59 +02:00
Sergei Shtylyov
e06a055bcd EHCI: fix off-by-one error in ehci_submit_root()
USB devices on the 2nd port are not detected and I get the following message:

The request port(1) is not configured

That's with default CONFIG_SYS_USB_EHCI_MAX_ROOT_PORTS value of 2. 'req->index'
is 1-based, so the comparison in ehci_submit_root() can't be correct.

Signed-off-by: Sergei Shtylyov <sshtylyov@mvista.com>
2010-04-08 21:39:59 +02:00
Sergei Shtylyov
6d313c84de EHCI: fix root hub device descriptor
On little endian machines, EHCI root hub's USB revision is reported as 0.2 --
cpu_to_le16() was missed in the initializer for the 'bcdUSB' descriptor field.
The same should be done for the 'bcdDevice' field.

Signed-off-by: Sergei Shtylyov <sshtylyov@mvista.com>
2010-04-08 21:39:58 +02:00
Chris Zhang
b416191a14 Fix EHCI port reset.
In USB ehci driver, the port reset is not terminated. EHCI spec says "A host
 controller must terminate the reset and stabilize the state of the port within
 2 milliseconds". Without termination, a port stays at reset state. This is
 observed on ppc4xx(sequoia) boards.

Signed-off-by: Chris Zhang <chris@seamicro.com>
2010-01-09 10:25:43 +01:00
Tom Rix
8f8bd565f3 USB Consolidate descriptor definitions
The header files usb.h and usbdescriptors.h have the same nameed
structure definitions for

usb_config_descriptor
usb_interface_descriptor
usb_endpoint_descriptor
usb_device_descriptor
usb_string_descriptor

These are out right duplicates in usb.h

usb_device_descriptor
usb_string_descriptor

This one has extra unused elements

usb_endpoint_descriptor

	unsigned char	bRefresh
	unsigned char	bSynchAddress;

These in usb.h have extra elements at the end of the usb 2.0
specified descriptor and are used.

usb_config_descriptor
usb_interface_descriptor

The change is to consolidate the definition of the descriptors
to usbdescriptors.h.  The dublicates in usb.h are removed.
The extra element structure will have their name shorted by
removing the '_descriptor' suffix.

So

usb_config_descriptor -> usb_config
usb_interface_descriptor -> usb_interface

For these, the common descriptor elements are accessed now
by an element 'desc'.

As an example

-	if (iface->bInterfaceClass != USB_CLASS_HUB)
+	if (iface->desc.bInterfaceClass != USB_CLASS_HUB)

This has been compile tested on MAKEALL arm, ppc and mips.

Signed-off-by: Tom Rix <Tom.Rix@windriver.com>
2009-12-20 12:47:37 +01:00
Prafulla Wadaskar
44259bb9e6 usb: bugfix driver/usb/host/ehci-hcd.c function ehci_submit_root
This change is cheked in Linux source and fix found to be in sync.
This patch is tested for USB host interface on Kirkwood based
Sheevaplug platform (ARM little endian board)

Risk: the impact of this patch is not validated on big endian board.
This need to be checked...

Signed-off-by: Prafulla Wadaskar <prafulla@marvell.com>
Signed-off-by: Remy Bohmer <linux@bohmer.net>
2009-07-23 23:27:06 +02:00
Kumar Gala
50243e3e7a usb: Fix compiler warning with gcc4.4
ehci-hcd.c: In function 'ehci_submit_root':
ehci-hcd.c:719: warning: value computed is not used
ehci-hcd.c:748: warning: value computed is not used

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: Remy Bohmer <linux@bohmer.net>
2009-07-11 11:30:17 +02:00
Jean-Christophe PLAGNIOL-VILLARD
2731b9a866 drivers/usb: regorganisation
move to linux usb driver organisation

as following

drivers/usb/gadget
drivers/usb/host
drivers/usb/musb

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Remy Bohmer <linux@bohmer.net>
2009-04-06 20:40:46 +02:00
Renamed from drivers/usb/usb_ehci_core.c (Browse further)