> On 12 December 2010 04:58, Saravana Kannan wrote:
>> As you and James suggested, having the NS bit set by the secure world is
>> definitely a solution that would work. But IMHO, the explicit cache
>> flush/invalidate approach keeps the design simple and easy to maintain.
>
> That is indeed an a
From: Saravana Kannan
Support locally controllable clocks on the 7x30 target.
Signed-off-by: Saravana Kannan
Signed-off-by: Stephen Boyd
---
arch/arm/mach-msm/Makefile |1 +
arch/arm/mach-msm/clock-7x30.c | 1550
arch/arm/mach-msm/clock-7x30.h
Support the CE_CLK and CODEC_SSBI_CLK. Also add support for sdc,
uart, and i2c clocks on targets which support proc_comm clocks.
Reviewed-by: Saravana Kannan
Signed-off-by: Stephen Boyd
---
arch/arm/mach-msm/clock-pcom.h |4 +++-
arch/arm/mach-msm/devices-msm7x30.c |2 ++
arch/arm/
Not all devices use proc_comm. The soc_clk_set_ops() function was
unused. Determining if a clock is local vs. remote is fragile
when done by comparing clk_ops pointers. Instead, implement an
is_local() function for all clk_ops to determine if the clock is
local. Once we do this we can compile clock
Mark clocks with CLK_7X30 to indicate they have the potential to
be locally controllable. If it's determined at runtime that these
clocks are not locally owned, revert the clocks to proc_comm
control. Mark clocks with CLK_PCOM when we know that they can't
be locally controlled.
Reviewed-by: Sarava
The dummy clock driver was added to make 8x60 compile due to
proc_comm references in clock.c. Now that those references are
gone we can compile the original clock.c with no errors and
support clocks on 8x60.
Reviewed-by: Saravana Kannan
Signed-off-by: Stephen Boyd
---
arch/arm/mach-msm/Makefile
From: Matt Wagantall
Use the SoC's ring oscillator hardware to measure the clock rate
of locally-controlled clocks. This allows for the development of
more comprehensive end-to-end clock tests.
A 'measure' debugfs node is created for each clock to perform the
measurement and retrieve the result.
Remove dead code and push out clock-7x30.h and clock-pcom.h to
where they are actually used.
Reviewed-by: Saravana Kannan
Signed-off-by: Stephen Boyd
---
arch/arm/mach-msm/board-msm8x60.c |5
arch/arm/mach-msm/clock-pcom.c |1 +
arch/arm/mach-msm/clock.c | 21 ++-
There are currently 3 separate directories for clock debugging in
debugfs: clk_enable, clk_rate, and clk_local. Each of these
directories contains a list of clocks. This is rather annoying
when you are focusing on one clock and want to enable/disable it
and then check its rate. You either have to c
From: Matt Wagantall
The clock debugfs code is large enough, and easy enough to separate,
that it deserves its own file which is compiled only when
CONFIG_DEBUG_FS is enabled.
Also, cleanup header file #includes that are no longer required.
Reviewed-by: Saravana Kannan
Signed-off-by: Matt Waga
From: Matt Wagantall
Add debugfs nodes to list all of the supported rates for rate-settable
locally-controlled clocks. This allows a clock's supported rates to
be queried from userspace, and allows more comprehensive userspace
clock tests to be developed.
Reviewed-by: Saravana Kannan
Signed-off
Some clocks have multiple consumers where each consumer requires
a minimum rate. Implement a sub driver to aggregate
clk_set_rate() calls from each consumer and call clk_set_rate()
on the parent clock when the minimum requested rate of all the
voters changes.
Reviewed-by: Saravana Kannan
Signed-o
Voting clocks are linked to the clock they vote on with a parent
pointer. Implement this API for sub drivers which wish to use it.
Reviewed-by: Saravana Kannan
Signed-off-by: Stephen Boyd
---
arch/arm/mach-msm/clock.c |2 ++
arch/arm/mach-msm/clock.h |1 +
2 files changed, 3 insertions(
The pbus clock always runs at half the ebi1 clock frequency.
Since ebi1 clock votes are dropped when the application processor
is power collapsed, the ebi1 clock is reserved for the acpuclock
driver. Other drivers which have ebi1 clock requirements that are
independent of application processor powe
The clock voting code will need to call clk_set_rate() from
within a spinlock so we must push down the refcount and lock into
each sub driver to avoid recursive locking.
Reviewed-by: Saravana Kannan
Signed-off-by: Stephen Boyd
---
arch/arm/mach-msm/clock-local.c |3 +-
arch/arm/mach-msm/clo
Migrating to clkdev has several advantages:
* Less code in mach-msm/clock.c
* A more robust clk_get() implementation
* clk_add_alias() support
* clk_get_sys() support
In general, this will help board authors setup clock aliases and
break the dependency on device pointers in the clock table
Add the appropriate clock voters and integrate clock.c with the
voting driver so that voting works appropriately.
Reviewed-by: Saravana Kannan
Signed-off-by: Stephen Boyd
---
arch/arm/mach-msm/clock.c | 23 +--
arch/arm/mach-msm/devices-msm7x30.c | 10 +
From: Stepan Moskovchenko
Add clock control to the IOMMU driver. The IOMMU bus clock
(and potentially an AXI clock) need to be on to gain access
to IOMMU registers. Actively control these clocks when
needed instead of leaving them on.
Signed-off-by: Stepan Moskovchenko
Signed-off-by: Stephen Bo
From: Stepan Moskovchenko
Add the IOMMU bus clock and the IOMMU bus interconnect
clocks to the platform data for an IOMMU device. These
clocks are needed to access the IOMMU registers.
Signed-off-by: Stepan Moskovchenko
Signed-off-by: Stephen Boyd
---
arch/arm/mach-msm/include/mach/iommu.h |
Now that the IOMMU driver properly handles its clocks we can stop
keeping the clocks it requires to be on during early boot.
Reviewed-by: Saravana Kannan
Signed-off-by: Stephen Boyd
---
arch/arm/mach-msm/clock-8x60.c |9 -
1 files changed, 0 insertions(+), 9 deletions(-)
diff --git
From: Stepan Moskovchenko
Clean up the clock control code in the probe calls, and add
support for controlling the clock for the IOMMU bus
interconnect. With the (proper) clock driver in place, the
clock control logic in the probe function can be made much
cleaner since it does not have to deal wi
This series of patches syncs the MSM clock code in Linus' tree with
the MSM tree @ codeaurora.org. They're prepared against the msm-core
tree (git://codeaurora.org/quic/kernel/davidb/linux-msm.git msm-core).
* 1 - 4 are cleanups and minor additions to the existing clock driver
* 5 - 6 add supp
On Wed, Dec 15, 2010 at 11:37:19PM +0100, Arnd Bergmann wrote:
> Yes, I understand that it's hard for many reasons, but my impression
> is that there is a general agreement on the idea. As I said, I don't
> expect you to fix it all at once, but it shouldn't be too hard
> to take care when adding n
On Thu, Dec 16, 2010 at 01:37:28PM -0800, Greg KH wrote:
> > The board support patches ( [PATCH 1/2] MSM: Add USB suport for QSD8x50)
> > are dependent on USB OTG driver patches ([PATCH V3 00/11] Add MSM USB
> > controller driver). The USB OTG driver patches are in your
> > usb-next tree. Where sh
On Wed, Dec 15, 2010 at 10:54:16PM -0800, Jeff Ohlstein wrote:
> +void __cpuinit platform_secondary_init(unsigned int cpu)
> +{
> + /* Configure edge-triggered PPIs */
> + writel(GIC_PPI_EDGE_MASK, MSM_QGIC_DIST_BASE + GIC_DIST_CONFIG + 4);
At some point we should really clean this up so w
On Thu, Dec 16, 2010 at 09:11:42AM +0530, Pavan Kondeti wrote:
> Hi Greg,
>
> On 12/15/2010 10:47 PM, Greg KH wrote:
> >>> On 12/8/2010 1:37 PM, Pavankumar Kondeti wrote:
> > > > OTG driver takes care of putting hardware into low power mode.
> > > > Hence
> > > > make peripheral a
Catalin Marinas wrote:
On 16 December 2010 06:54, Jeff Ohlstein wrote:
--- /dev/null
+++ b/arch/arm/mach-msm/hotplug.c
[...]
+static inline void platform_do_lowpower(unsigned int cpu)
+{
+ /* Just enter wfi for now. TODO: Properly shut off the cpu. */
+ for (;;) {
+
On Thu, Dec 16, 2010 at 09:34:19AM +0200, Maya Erez wrote:
> - Add the capability to search for an EP with a required number of streams.
> - Add a gadget_op to allow DCDs to use a proprietary search algorithm
> for required and minimum number of streams.
What do you exactly mean by "proprietary se
On Thu, Dec 16, 2010 at 03:56:01PM +0530, Trilok Soni wrote:
> Hi Dmitry,
>
> gpio_keys.c driver should be converted to use request_any_context_irq(...)
> infrastructure because some of the devices could use gpios from PMIC like
> chips where their irqs gets registered as nested. I don't have any
On 12/16/2010 9:04 PM, Alan Stern wrote:
> On Thu, 16 Dec 2010, Pavankumar Kondeti wrote:
>
>> OTG supplement revision 2.0 spec introduces Attach Detection Protocol
>> (ADP) for detecting peripheral connection without applying power on
>> VBUS. ADP is optional and is included in the OTG descripto
On Thu, 16 Dec 2010, Pavankumar Kondeti wrote:
> OTG supplement revision 2.0 spec introduces Attach Detection Protocol
> (ADP) for detecting peripheral connection without applying power on
> VBUS. ADP is optional and is included in the OTG descriptor along with
> SRP and HNP.
>
> HNP polling is
On 12/16/2010 10:03 AM, Russell King - ARM Linux wrote:
> I've not changed my thoughts on this. PPIs should not be handled by
> genirq - it just doesn't make sense for them to be.
Understood. I will work on a patch to perf events to support the MSM
performance monitor PPI on SMP.
~Stephen
--
On Thu, Dec 16, 2010 at 09:54:23AM -0500, Stephen Caudle wrote:
> On 12/09/2010 11:24 AM, Stephen Caudle wrote:
> >> It is also unreasonable to have one core enabling the PPI on other
> >> cores where the hardware behind the interrupt may not have been
> >> initialized yet. If it is a private inte
On 12/09/2010 11:24 AM, Stephen Caudle wrote:
>> It is also unreasonable to have one core enabling the PPI on other
>> cores where the hardware behind the interrupt may not have been
>> initialized yet. If it is a private interrupt for a private peripheral,
>> then only the associated CPU should b
On 16-12-2010 16:07, Felipe Balbi wrote:
Some USB controllers have common resources (IRQ, register address
space) for Host, Peripheral and OTG. So HCD is added only before
entering into Host mode. Root hub initialization is done in
different steps to decrease boot up time. But this makes B-devic
On 12/16/2010 6:37 PM, Felipe Balbi wrote:
> On Thu, Dec 16, 2010 at 03:39:35PM +0300, Sergei Shtylyov wrote:
>> Hello.
>>
>> On 16-12-2010 14:09, Pavankumar Kondeti wrote:
>>
>>> Some USB controllers have common resources (IRQ, register address
>>> space) for Host, Peripheral and OTG. So HCD is a
On Thu, Dec 16, 2010 at 03:39:35PM +0300, Sergei Shtylyov wrote:
Hello.
On 16-12-2010 14:09, Pavankumar Kondeti wrote:
Some USB controllers have common resources (IRQ, register address
space) for Host, Peripheral and OTG. So HCD is added only before
entering into Host mode. Root hub initiali
On 16 December 2010 06:54, Jeff Ohlstein wrote:
> --- /dev/null
> +++ b/arch/arm/mach-msm/hotplug.c
[...]
> +static inline void platform_do_lowpower(unsigned int cpu)
> +{
> + /* Just enter wfi for now. TODO: Properly shut off the cpu. */
> + for (;;) {
> + /*
> +
On 12/16/2010 5:45 PM, Felipe Balbi wrote:
> On Thu, Dec 16, 2010 at 04:56:15PM +0530, Pavan Kondeti wrote:
>> On 12/15/2010 5:51 PM, Felipe Balbi wrote:
>>> On Wed, Dec 15, 2010 at 04:44:14PM +0530, Pavankumar Kondeti wrote:
Some USB controllers have common resources (IRQ, register address
>>
On Wed, Dec 15, 2010 at 05:49:05PM -0800, Brandon Leong wrote:
I just noticed that you only posted this to the MSM kernel list -
please always post generic patches to the relevant generic list, lkml in
this case. I've CCed lkml.
> Allow the user to read and edit regulator information in user spa
Hello.
On 16-12-2010 14:09, Pavankumar Kondeti wrote:
Some USB controllers have common resources (IRQ, register address
space) for Host, Peripheral and OTG. So HCD is added only before
entering into Host mode. Root hub initialization is done in
different steps to decrease boot up time. But t
On Thu, Dec 16, 2010 at 04:56:15PM +0530, Pavan Kondeti wrote:
On 12/15/2010 5:51 PM, Felipe Balbi wrote:
On Wed, Dec 15, 2010 at 04:44:14PM +0530, Pavankumar Kondeti wrote:
Some USB controllers have common resources (IRQ, register address
space) for Host, Peripheral and OTG. So HCD is added o
On 12/15/2010 5:51 PM, Felipe Balbi wrote:
> On Wed, Dec 15, 2010 at 04:44:14PM +0530, Pavankumar Kondeti wrote:
>> Some USB controllers have common resources (IRQ, register address
>> space) for Host, Peripheral and OTG. So HCD is added only before
>> entering into Host mode. Root hub initializa
Introduce otg_version field in usb_gadget struct. UDC can
advertise OTG spec version compatibility by setting otg_version
field appropriately. Gadget drivers fill the bcdOTG field in
OTG descriptor based on UDC's OTG version.
Add sysfs file for host_request and UDC returns the same when
HNP poll
Introduce start_hnp callback function for HCD to receive notification
from EHCI core that HNP enabled port is suspended. HCD may initiate
HNP or notify the same to OTG via otg_start_hnp().
This patch is inspired by "USB: Hook start_hnp into ohci struct"
(e8b24450).
Signed-off-by: Pavankumar Kond
Some USB controllers have common resources (IRQ, register address
space) for Host, Peripheral and OTG. So HCD is added only before
entering into Host mode. Root hub initialization is done in
different steps to decrease boot up time. But this makes B-device
difficult to meet HNP timings. Hence e
OTG supplement revision 2.0 spec introduces Attach Detection Protocol
(ADP) for detecting peripheral connection without applying power on
VBUS. ADP is optional and is included in the OTG descriptor along with
SRP and HNP.
HNP polling is introduced for peripheral to notify its wish to become
host.
This patch series adds OTG 2.0 enhancements and misc changes
required for supporting SRP & HNP in MSM OTG driver. As these
patches modify the USB core code, I would like to receive
feedback before posting driver patches. The driver patches are
available at
https://www.codeaurora.org/gitweb/quic/la
Hi Dmitry,
gpio_keys.c driver should be converted to use request_any_context_irq(...)
infrastructure because some of the devices could use gpios from PMIC like
chips where their irqs gets registered as nested. I don't have any
device with such configuration but I was looking at this driver and I t
Hi,
On Thu, Dec 16, 2010 at 02:24:05PM +0530, Pavan Kondeti wrote:
@@ -1270,6 +1283,43 @@ static int usb_resume_both(struct usb_device *udev,
pm_message_t msg)
return status;
}
+#ifdef CONFIG_USB_OTG
+void usb_hnp_polling_work(struct work_struct *work)
do you really need this to be e
cdev struct is accessed in suspended sysfs show function. So
remove sysfs file before freeing the cdev in composite_unbind.
Signed-off-by: Pavankumar Kondeti
---
drivers/usb/gadget/composite.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/usb/gadget/composite
bd6882 commit (usb: gadget: fix Kconfig warning) removes
the duplicate USB_OTG config from gadget/Kconfig. But
does not copy the input prompt and help text to the original
config defined in core/Kconfig. Add them now.
Signed-off-by: Pavankumar Kondeti
---
drivers/usb/core/Kconfig | 12 +++
On 12/16/2010 3:08 AM, Alan Stern wrote:
> On Wed, 15 Dec 2010, Pavankumar Kondeti wrote:
>
>> This patch series adds OTG 2.0 enhancements and misc changes
>> required for supporting SRP & HNP in MSM OTG driver. As these
>> patches modify the USB core code, I would like to receive
>> feedback bef
On 12/15/2010 5:49 PM, Felipe Balbi wrote:
> Hi,
>
> On Wed, Dec 15, 2010 at 04:44:13PM +0530, Pavankumar Kondeti wrote:
>> Introduce start_hnp callback function for HCD to receive notification
>>from EHCI core that HNP enabled port is suspended. HCD may initiate
>> HNP or notify the same to OTG
On 12/15/2010 5:48 PM, Felipe Balbi wrote:
> Hi,
>
> On Wed, Dec 15, 2010 at 04:44:12PM +0530, Pavankumar Kondeti wrote:
>> @@ -1047,9 +1065,10 @@ composite_unbind(struct usb_gadget *gadget)
>> kfree(cdev->req->buf);
>> usb_ep_free_request(gadget->ep0, cdev->req);
>>
On 12/16/2010 3:23 AM, Greg KH wrote:
> On Wed, Dec 15, 2010 at 04:44:11PM +0530, Pavankumar Kondeti wrote:
>> --- a/drivers/usb/core/driver.c
>> +++ b/drivers/usb/core/driver.c
>> @@ -1207,6 +1207,19 @@ static int usb_suspend_both(struct usb_device *udev,
>> pm_message_t msg)
>> * and flush
On 12/16/2010 2:59 AM, Alan Stern wrote:
> On Wed, 15 Dec 2010, Pavankumar Kondeti wrote:
>
>> OTG supplement revision 2.0 spec introduces Attach Detection Protocol
>> (ADP) for detecting peripheral connection without applying power on
>> VBUS. ADP is optional and is included in the OTG descripto
On 12/15/2010 5:46 PM, Felipe Balbi wrote:
> Hi,
>
> On Wed, Dec 15, 2010 at 04:44:11PM +0530, Pavankumar Kondeti wrote:
>> OTG supplement revision 2.0 spec introduces Attach Detection Protocol
>> (ADP) for detecting peripheral connection without applying power on
>> VBUS. ADP is optional and is
On 12/16/2010 3:22 AM, Greg KH wrote:
> On Wed, Dec 15, 2010 at 04:44:10PM +0530, Pavankumar Kondeti wrote:
>> bd6882 commit removes the duplicate USB_OTG config from
>> gadget/Kconfig. But does not copy the input prompt and
>> help text to the original config defined in core/Kconfig.
>> Add them
On Wed, Dec 15, 2010 at 04:39:13PM -0500, Alan Stern wrote:
On Wed, 15 Dec 2010, Felipe Balbi wrote:
>+{
>+ int ret;
>+ struct usb_bus *bus =
>+ container_of(work, struct usb_bus, hnp_polling.work);
>+ struct usb_device *udev = bus->root_hub->children[bus->otg_port - 1];
>+ u8
60 matches
Mail list logo