linux-toradex.git/drivers/usb/host, branch tegra-10.9.7 Linux kernel for Apalis and Colibri modules [ehci-tegra] Fix VBUS handling in OTG mode 2010-10-21T22:35:15+00:00 Venkat Moganty vmoganty@nvidia.com 2010-10-20T06:34:49+00:00 cb800e32170138f00aad5b724dc56c18ce1b07ad When USB cable is removed from the USB1 port side then VBUS is turned off but, if USB cable is removed from the device side then VBUS is not getting turned off and is showing high even when USB cable is removed completely from the USB1 port. Fixed this by checking the OTG state "OTG_STATE_A_SUSPEND" and then turning off the VBUS and phy power down. Bug 741587 Change-Id: Ib8f3dbddc48de0d12179dbd504503621854f44cd Reviewed-on: http://git-master/r/8199 Reviewed-by: Lance Zhao <lazhao@nvidia.com> Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com> Tested-by: Lance Zhao <lazhao@nvidia.com> Reviewed-by: Varun Colbert <vcolbert@nvidia.com> Tested-by: Varun Colbert <vcolbert@nvidia.com> 
When USB cable is removed from the USB1 port side then VBUS is turned
off but, if USB cable is removed from the device side then VBUS is not
getting turned off and is showing high even when USB cable is removed
completely from the USB1 port. Fixed this by checking the OTG state
"OTG_STATE_A_SUSPEND" and then turning off the VBUS and phy power down.

Bug 741587

Change-Id: Ib8f3dbddc48de0d12179dbd504503621854f44cd
Reviewed-on: http://git-master/r/8199
Reviewed-by: Lance Zhao <lazhao@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
Tested-by: Lance Zhao <lazhao@nvidia.com>
Reviewed-by: Varun Colbert <vcolbert@nvidia.com>
Tested-by: Varun Colbert <vcolbert@nvidia.com>
[tegra-ehci]USB device detection for host mode in LP0 2010-10-06T05:16:52+00:00 Venkat Moganty vmoganty@nvidia.com 2010-09-28T16:02:08+00:00 8eaa0afc05248d324ec8d46a557740a043d8e641 USB devices are not recognised when connected to the USB port, when system is already in LP0. Root port power is not getting enabled when resume from LP0. By the time root hub checks for the port connection auto suspend is getting called and device is not recognised. Fixed this by enabling the root port power during the ehci_restart functionality. Bug 741890 (cherry picked from commit 188bf7050751a40d6f2d6a858e2702774fc70f77) Change-Id: Ib88ed3fc4b01719777a0e49e27434ef6425bc53d Reviewed-on: http://git-master/r/7834 Reviewed-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com> Tested-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com> Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com> 
USB devices are not recognised when connected to the USB port, when system is
already in LP0. Root port power is not getting enabled when resume from LP0.
By the time root hub checks for the port connection auto suspend is getting
called and device is not recognised. Fixed this by enabling the root port power
during the ehci_restart functionality.

Bug 741890

(cherry picked from commit 188bf7050751a40d6f2d6a858e2702774fc70f77)

Change-Id: Ib88ed3fc4b01719777a0e49e27434ef6425bc53d
Reviewed-on: http://git-master/r/7834
Reviewed-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com>
Tested-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
[usb-host]Remove controller restart in host mode during resume 2010-10-05T21:51:10+00:00 Venkat Moganty vmoganty@nvidia.com 2010-09-05T01:46:46+00:00 f53faa616294450f1ed05b5bfd249b577b709399 During resume from LP0 controller is getting restarted in host mode due to this hcd state is getting changed from the suspend state. Controller state should not be changed for host mode. Fixed this to restart the controller only in OTG mode. Bug 724437 (cherry picked from commit cbea530301ae3d9a4a7271d11c819bfc8de8f892) Change-Id: Idc2897a2790087e00ee66e30fea734c9d117029e Reviewed-on: http://git-master/r/7833 Reviewed-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com> Tested-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com> Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com> 
During resume from LP0 controller is getting restarted in host mode due to this
hcd state is getting changed from the suspend state. Controller state should not
be changed for host mode. Fixed this to restart the controller only in OTG mode.

Bug 724437

(cherry picked from commit cbea530301ae3d9a4a7271d11c819bfc8de8f892)

Change-Id: Idc2897a2790087e00ee66e30fea734c9d117029e
Reviewed-on: http://git-master/r/7833
Reviewed-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com>
Tested-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
[usb-hcd]Adding missed memory barrier in the hcd driver 2010-10-05T21:49:18+00:00 Venkat Moganty vmoganty@nvidia.com 2010-09-23T17:44:57+00:00 80833ca384f1a58a207f98bba66905c6e0caac97 There is one wmb missing in the usb host controller driver after the queue head update. Due to this data transaction is not happening on the bus after urb submission by the hcd driver. Register updates/queue heads data in the memory is not reflected on the AHB bus. After adding the wmb after queue head update data transaction the USB bus started with out any delay. Bug 737242 (cherry picked from commit eff18c3f40253f7af0b576126b954d8572cbea64) Change-Id: I9f3ae6d8743a03a58034a82f4fe2850fd7278c4f Reviewed-on: http://git-master/r/7747 Reviewed-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com> Tested-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com> Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com> 
There is one wmb missing in the usb host controller driver after the queue head
update. Due to this data transaction is not happening on the bus after urb
submission by the hcd driver. Register updates/queue heads data in the memory
is not reflected on the AHB bus. After adding the wmb after queue head update
data transaction the USB bus started with out any delay.

Bug 737242

(cherry picked from commit eff18c3f40253f7af0b576126b954d8572cbea64)

Change-Id: I9f3ae6d8743a03a58034a82f4fe2850fd7278c4f
Reviewed-on: http://git-master/r/7747
Reviewed-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com>
Tested-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
[tegra-ehci] Fix Usb device detection for config USB_SUSPEND 2010-08-24T23:53:56+00:00 Venkat Moganty vmoganty@nvidia.com 2010-08-12T11:38:14+00:00 04dbd92636bdf1e4652d00d03c7a5a6ebbad8a4c When CONFIG_USB_SUSPEND is enabled USB device detection is not working. This is due to the wrong API called to resume the hub and HUB resume funtionality not happening properly. Fixed this by calling the correct API to resume the HUB driver when auto suspend is called. Bug 713237 Bug 713966 Change-Id: Ia4d091fd29ea7ebfe5844cf5685fc5a86e66d12a Reviewed-on: http://git-master/r/4984 Reviewed-by: Abhishek Aggarwal <aaggarwal@nvidia.com> Tested-by: Abhishek Aggarwal <aaggarwal@nvidia.com> Reviewed-by: Gary King <gking@nvidia.com> 
When CONFIG_USB_SUSPEND is enabled USB device detection is not working. This is
due to the wrong API called to resume the hub and HUB resume funtionality not
happening properly. Fixed this by calling the correct API to resume the HUB
driver when auto suspend is called.

Bug 713237
Bug 713966

Change-Id: Ia4d091fd29ea7ebfe5844cf5685fc5a86e66d12a
Reviewed-on: http://git-master/r/4984
Reviewed-by: Abhishek Aggarwal <aaggarwal@nvidia.com>
Tested-by: Abhishek Aggarwal <aaggarwal@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
[tegra-hcd]: Fix for LP0 hang in host for CONFIG_USB_SUSPEND 2010-07-16T23:31:02+00:00 Venkat Moganty vmoganty@nvidia.com 2010-07-16T15:13:53+00:00 57c764c72125e0ac38b97c3313e7e2fb2210143d When CONFIG_USB_SUSPEND is enabled auto suspend feature is getting executed by the HUB driver during the suspend/resume, which calls the ehci_hub_control from the hub thread to read the port status. When there is no device connected controller is suspended and can not be accessed the registers which is causing hang during suspend/resume. Fixed this by accessing the controller only if host is resumed. Change-Id: I066ff1e5ef11f3cb15a4dc809cd11ee0345dc31a Reviewed-on: http://git-master/r/4014 Reviewed-by: Iqbal Bhinderwala <iqbalb@nvidia.com> Tested-by: Iqbal Bhinderwala <iqbalb@nvidia.com> Reviewed-by: Gary King <gking@nvidia.com> 
When CONFIG_USB_SUSPEND is enabled auto suspend feature is getting executed by
the HUB driver during the suspend/resume, which calls the ehci_hub_control from
the hub thread to read the port status. When there is no device connected
controller is suspended and can not be accessed the registers which is causing
hang during suspend/resume. Fixed this by accessing the controller only if host
is resumed.

Change-Id: I066ff1e5ef11f3cb15a4dc809cd11ee0345dc31a
Reviewed-on: http://git-master/r/4014
Reviewed-by: Iqbal Bhinderwala <iqbalb@nvidia.com>
Tested-by: Iqbal Bhinderwala <iqbalb@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
[tegra-usb] Fix USB-OTG for CONFIG_USB_SUSPEND 2010-07-09T22:59:16+00:00 Venkat Moganty vmoganty@nvidia.com 2010-07-08T04:59:21+00:00 0a78303054f9fd89d382cc14f0b8dff69ac937cb When CONFIG_USB_SUSPEND is enabled, after auto suspend by the usb core driver, USB devices(mouse,keyboard etc.) are not recognised. After auto suspend device state goes to suspend and not resuming to the address. This is fixed by changing the device state to address state on USB cable insertion. Bug 674938 Bug 698734 Change-Id: I837562fd1eb020f400824f2d475289df5fbc1f83 Reviewed-on: http://git-master/r/3494 Reviewed-by: Gary King <gking@nvidia.com> Tested-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com> 
When CONFIG_USB_SUSPEND is enabled, after auto suspend by the usb core
driver, USB devices(mouse,keyboard etc.) are not recognised. After auto
suspend device state goes to suspend and not resuming to the address.
This is fixed by changing the device state to address state on USB cable
insertion.

Bug 674938
Bug 698734

Change-Id: I837562fd1eb020f400824f2d475289df5fbc1f83
Reviewed-on: http://git-master/r/3494
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com>
tegra-ehci: enable DMABOUNCE for tegra EHCI controller 2010-06-22T23:39:53+00:00 Gary King gking@nvidia.com 2010-06-21T17:41:14+00:00 db47e4a3e890c34cb6fd73cb817ef4edfbd5f9d9 tegra 2 EHCI controllers need transfers aligned to 32B boundaries; enable DMABOUNCE by default whenever USB_TEGRA_HCD is selected, and create DMA pools for the EHCI devices Change-Id: I6c391a52370e19c904cab714a4ebd223c7647329 Reviewed-on: http://git-master.nvidia.com/r/2959 Reviewed-by: Gary King <gking@nvidia.com> Tested-by: Gary King <gking@nvidia.com> 
tegra 2 EHCI controllers need transfers aligned to 32B boundaries;
enable DMABOUNCE by default whenever USB_TEGRA_HCD is selected,
and create DMA pools for the EHCI devices

Change-Id: I6c391a52370e19c904cab714a4ebd223c7647329
Reviewed-on: http://git-master.nvidia.com/r/2959
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Gary King <gking@nvidia.com>
ehci-tegra: halting the controller before powering down the PHY 2010-06-22T17:19:42+00:00 Abhishek Aggarwal aaggarwal@nvidia.com 2010-06-22T09:49:32+00:00 b51a545ccd32f85e3c37d87ae3888b9ee7ef82a2 It was observed that when trying to power down the PHY, the RS bit(0:0) in USBCMD register was still set to RUN(1). Calling ehci_halt() before powering down the PHY in tegra_ehci_suspend() to set RS bit to STOP(0) before PHY is powered down. Bug: 695655 Change-Id: I677550be63eebd5140ca466096cb796a21a1167 Reviewed-on: http://git-master/r/2929 Tested-by: Abhishek Aggarwal <aaggarwal@nvidia.com> Reviewed-by: Gary King <gking@nvidia.com> 
It was observed that when trying to power down the PHY, the RS bit(0:0)
in USBCMD register was still set to RUN(1). Calling ehci_halt() before
powering down the PHY in tegra_ehci_suspend() to set RS bit to STOP(0)
before PHY is powered down.

Bug: 695655

Change-Id: I677550be63eebd5140ca466096cb796a21a1167
Reviewed-on: http://git-master/r/2929
Tested-by: Abhishek Aggarwal <aaggarwal@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
tegra: EHCI bus suspend/resume requires CONFIG_PM 2010-06-11T23:11:24+00:00 Scott Williams scwilliams@nvidia.com 2010-06-11T19:11:11+00:00 283d990f999660b588ccfc917b197656ab72742d Change-Id: I86faa2f56dbfd0ed145f40b04346a85f64bab12c Reviewed-on: http://git-master/r/2508 Reviewed-by: Scott Williams <scwilliams@nvidia.com> Tested-by: Scott Williams <scwilliams@nvidia.com> Reviewed-by: Gary King <gking@nvidia.com> 
Change-Id: I86faa2f56dbfd0ed145f40b04346a85f64bab12c
Reviewed-on: http://git-master/r/2508
Reviewed-by: Scott Williams <scwilliams@nvidia.com>
Tested-by: Scott Williams <scwilliams@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>