Re: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Karicheri, Muralidharan m-kariche...@ti.com writes:
Kevin,
I think I have figured it out...
First issue was that I was adding my entry at the end of dm644x_clks[]
array. I need to add it before the CLK(NULL, NULL, NULL)
secondly, your suggestion didn't work as is. This is what I had to
do to get it working...
static struct clk ccdc_master_clk = {
.name = dm644x_ccdc,
.parent = vpss_master_clk,
};
static struct clk ccdc_slave_clk = {
.name = dm644x_ccdc,
.parent = vpss_slave_clk,
};
It doesn't work with out doing this. The cat /proc/davinci_clocks hangs with
your suggestion implemented...
Can you track down the hang. It sounds like a bug in the walking of
the clock tree for davinci_clocks.
You should not need to add new clocks with new names. I don't thinke
the name field of the struct clk is used anywhere in the matching.
I think it's only used in /proc/davinci_clocks
static struct davinci_clk dm365_clks = {
CLK(dm644x_ccdc, master, ccdc_master_clk),
CLK(dm644x_ccdc, slave, ccdc_slave_clk),
Looks like the drivers name is 'dm644x_ccdc', not 'isif'. I'm
guessing just this should work without having to add new clock names.
No. I have mixed up the names. ISIF is for the new ISIF driver on DM365.
Below are for DM644x ccdc driver. With just these entries added, two
things observed
1) Only one clock is shown disabled (usually many are shown disabled) during
bootup
2) cat /proc/davinci_clocks hangs.
So this is the only way I got it working.
Hmm, it worked just fine for me without any of these side effects. I
applied the simple patch below on top of current master branch. It booted
fine showing all the unused clocks being disabled, and I was able to
see davinci_clocks just fine:
diff --git a/arch/arm/mach-davinci/dm644x.c b/arch/arm/mach-davinci/dm644x.c
index e65e29e..e6f3570 100644
--- a/arch/arm/mach-davinci/dm644x.c
+++ b/arch/arm/mach-davinci/dm644x.c
@@ -293,8 +293,8 @@ struct davinci_clk dm644x_clks[] = {
CLK(NULL, dsp, dsp_clk),
CLK(NULL, arm, arm_clk),
CLK(NULL, vicp, vicp_clk),
- CLK(NULL, vpss_master, vpss_master_clk),
- CLK(NULL, vpss_slave, vpss_slave_clk),
+ CLK(dm644x_ccdc, master, vpss_master_clk),
+ CLK(dm644x_ccdc, slave, vpss_slave_clk),
CLK(NULL, arm, arm_clk),
CLK(NULL, uart0, uart0_clk),
CLK(NULL, uart1, uart1_clk),
[...]
Clocks: disable unused uart1
Clocks: disable unused uart2
Clocks: disable unused emac
Clocks: disable unused ide
Clocks: disable unused asp0
Clocks: disable unused mmcsd
Clocks: disable unused spi
Clocks: disable unused usb
Clocks: disable unused vlynq
Clocks: disable unused pwm0
Clocks: disable unused pwm1
Clocks: disable unused pwm2
Clocks: disable unused timer1
[...]
r...@dm644x:~# uname -r
2.6.32-arm-davinci-default-06873-g1a7277b-dirty
r...@dm644x:~# cat /debug/davinci_clocks
ref_clk users= 8 2700 Hz
pll1users= 8 pll 59400 Hz
pll1_sysclk1 users= 0 pll 59400 Hz
dsp users= 1 psc 59400 Hz
pll1_sysclk2 users= 2 pll 29700 Hz
arm users= 2 psc 29700 Hz
pll1_sysclk3 users= 0 pll 19800 Hz
vpss_master users= 0 psc 19800 Hz
vpss_slave users= 0 psc 19800 Hz
pll1_sysclk5 users= 3 pll 9900 Hz
emacusers= 1 psc 9900 Hz
ide users= 0 psc 9900 Hz
asp0users= 0 psc 9900 Hz
mmcsd users= 0 psc 9900 Hz
spi users= 0 psc 9900 Hz
gpiousers= 1 psc 9900 Hz
usb
RE: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Kevin,
That is what I had seen. I will check it again on Monday
and let you know.
Murali Karicheri
Software Design Engineer
Texas Instruments Inc.
Germantown, MD 20874
phone: 301-407-9583
email: m-kariche...@ti.com
-Original Message-
From: Kevin Hilman [mailto:khil...@deeprootsystems.com]
Sent: Friday, December 11, 2009 1:35 PM
To: Karicheri, Muralidharan
Cc: davinci-linux-open-sou...@linux.davincidsp.com; linux-
me...@vger.kernel.org
Subject: Re: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Karicheri, Muralidharan m-kariche...@ti.com writes:
Kevin,
I think I have figured it out...
First issue was that I was adding my entry at the end of dm644x_clks[]
array. I need to add it before the CLK(NULL, NULL, NULL)
secondly, your suggestion didn't work as is. This is what I had to
do to get it working...
static struct clk ccdc_master_clk = {
.name = dm644x_ccdc,
.parent = vpss_master_clk,
};
static struct clk ccdc_slave_clk = {
.name = dm644x_ccdc,
.parent = vpss_slave_clk,
};
It doesn't work with out doing this. The cat /proc/davinci_clocks hangs
with
your suggestion implemented...
Can you track down the hang. It sounds like a bug in the walking of
the clock tree for davinci_clocks.
You should not need to add new clocks with new names. I don't thinke
the name field of the struct clk is used anywhere in the matching.
I think it's only used in /proc/davinci_clocks
static struct davinci_clk dm365_clks = {
CLK(dm644x_ccdc, master, ccdc_master_clk),
CLK(dm644x_ccdc, slave, ccdc_slave_clk),
Looks like the drivers name is 'dm644x_ccdc', not 'isif'. I'm
guessing just this should work without having to add new clock names.
No. I have mixed up the names. ISIF is for the new ISIF driver on DM365.
Below are for DM644x ccdc driver. With just these entries added, two
things observed
1) Only one clock is shown disabled (usually many are shown disabled)
during bootup
2) cat /proc/davinci_clocks hangs.
So this is the only way I got it working.
Hmm, it worked just fine for me without any of these side effects. I
applied the simple patch below on top of current master branch. It booted
fine showing all the unused clocks being disabled, and I was able to
see davinci_clocks just fine:
diff --git a/arch/arm/mach-davinci/dm644x.c b/arch/arm/mach-
davinci/dm644x.c
index e65e29e..e6f3570 100644
--- a/arch/arm/mach-davinci/dm644x.c
+++ b/arch/arm/mach-davinci/dm644x.c
@@ -293,8 +293,8 @@ struct davinci_clk dm644x_clks[] = {
CLK(NULL, dsp, dsp_clk),
CLK(NULL, arm, arm_clk),
CLK(NULL, vicp, vicp_clk),
- CLK(NULL, vpss_master, vpss_master_clk),
- CLK(NULL, vpss_slave, vpss_slave_clk),
+ CLK(dm644x_ccdc, master, vpss_master_clk),
+ CLK(dm644x_ccdc, slave, vpss_slave_clk),
CLK(NULL, arm, arm_clk),
CLK(NULL, uart0, uart0_clk),
CLK(NULL, uart1, uart1_clk),
[...]
Clocks: disable unused uart1
Clocks: disable unused uart2
Clocks: disable unused emac
Clocks: disable unused ide
Clocks: disable unused asp0
Clocks: disable unused mmcsd
Clocks: disable unused spi
Clocks: disable unused usb
Clocks: disable unused vlynq
Clocks: disable unused pwm0
Clocks: disable unused pwm1
Clocks: disable unused pwm2
Clocks: disable unused timer1
[...]
r...@dm644x:~# uname -r
2.6.32-arm-davinci-default-06873-g1a7277b-dirty
r...@dm644x:~# cat /debug/davinci_clocks
ref_clk users= 8 2700 Hz
pll1users= 8 pll 59400 Hz
pll1_sysclk1 users= 0 pll 59400 Hz
dsp users= 1 psc 59400 Hz
pll1_sysclk2 users= 2 pll 29700 Hz
arm users= 2 psc 29700 Hz
pll1_sysclk3 users= 0 pll 19800 Hz
vpss_master users= 0 psc 19800 Hz
vpss_slave users= 0 psc 19800 Hz
pll1_sysclk5 users= 3 pll 9900 Hz
emacusers= 1 psc 9900 Hz
ide users= 0 psc 9900 Hz
asp0users= 0 psc 9900 Hz
mmcsd users= 0 psc 9900 Hz
spi users= 0 psc 9900 Hz
gpiousers= 1 psc 9900 Hz
usb users= 0 psc 9900 Hz
vlynq users= 0 psc 9900 Hz
aemif users= 1 psc 9900 Hz
pll1_aux_clk users= 3 pll 2700 Hz
uart0 users= 1 psc 2700 Hz
uart1 users= 0 psc 2700 Hz
uart2 users= 0 psc 2700 Hz
i2c users= 1 psc 2700 Hz
pwm0users= 0 psc 2700 Hz
pwm1users= 0 psc 2700 Hz
pwm2users= 0 psc 2700 Hz
timer0 users= 1 psc 2700 Hz
timer1 users= 0 psc 2700 Hz
timer2 users= 1 psc 2700 Hz
pll1_sysclkbp users= 0 pll 2700 Hz
pll2users= 0 pll 64800 Hz
pll2_sysclk1 users= 0 pll 5400 Hz
pll2_sysclk2 users= 0 pll 32400 Hz
pll2_sysclkbp users= 0 pll 1350 Hz
r...@dm644x
Re: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Karicheri, Muralidharan m-kariche...@ti.com writes:
Kevin,
+/**
+ * vpfe_disable_clock() - Disable clocks for vpfe capture driver
+ * @vpfe_dev - ptr to vpfe capture device
+ *
+ * Disables clocks defined in vpfe configuration.
+ */
static void vpfe_disable_clock(struct vpfe_device *vpfe_dev)
{
struct vpfe_config *vpfe_cfg = vpfe_dev-cfg;
+ int i;
- clk_disable(vpfe_cfg-vpssclk);
- clk_put(vpfe_cfg-vpssclk);
- clk_disable(vpfe_cfg-slaveclk);
- clk_put(vpfe_cfg-slaveclk);
- v4l2_info(vpfe_dev-pdev-driver,
-vpfe vpss master slave clocks disabled\n);
+ for (i = 0; i vpfe_cfg-num_clocks; i++) {
+ clk_disable(vpfe_dev-clks[i]);
+ clk_put(vpfe_dev-clks[i]);
While cleaning this up, you should move the clk_put() to module
disable/unload time.
[MK] vpfe_disable_clock() is called from remove(). In the new
patch, from ccdc driver remove() function, clk_put() will be called.
Why do you think it should be moved to exit() function of the module?
You dont' need to put he clock on every disable.
The same for clk_get(). You don't need to get the clock for every
enable. Just do a clk_get() at init time.
Are you suggesting to call clk_get() during init() and call clk_put()
from exit()? What is wrong with calling clk_get() from probe()?
I thought following is correct:-
Probe()
clk_get() followed by clk_enable()
Remove()
clk_disable() followed by clk_put()
Suspend()
clk_disable()
Resume()
clk_enable()
Yes, that is correct.
I didn't look at the whole driver. My concern was that if the driver
was enhanced to more aggressive clock management, you shouldn't do a
clk_get() every time you do a clk_enable(), same for put.
Kevin
--
To unsubscribe from this list: send the line unsubscribe linux-media in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Karicheri, Muralidharan m-kariche...@ti.com writes:
Kevin,
I think I have figured it out...
First issue was that I was adding my entry at the end of dm644x_clks[]
array. I need to add it before the CLK(NULL, NULL, NULL)
secondly, your suggestion didn't work as is. This is what I had to
do to get it working...
static struct clk ccdc_master_clk = {
.name = dm644x_ccdc,
.parent = vpss_master_clk,
};
static struct clk ccdc_slave_clk = {
.name = dm644x_ccdc,
.parent = vpss_slave_clk,
};
You should not need to add new clocks with new names. I don't thinke
the name field of the struct clk is used anywhere in the matching.
I think it's only used in /proc/davinci_clocks
static struct davinci_clk dm365_clks = {
CLK(dm644x_ccdc, master, ccdc_master_clk),
CLK(dm644x_ccdc, slave, ccdc_slave_clk),
Looks like the drivers name is 'dm644x_ccdc', not 'isif'. I'm
guessing just this should work without having to add new clock names.
CLK(dm644x_ccdc, master, vpss_master_clk),
CLK(dm644x_ccdc, slave, vpss_slave_clk),
CLK(NULL, NULL, NULL);
Let me know if you think there is anything wrong with the above scheme.
Kevin
--
To unsubscribe from this list: send the line unsubscribe linux-media in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
RE: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
I thought following is correct:- Probe() clk_get() followed by clk_enable() Remove() clk_disable() followed by clk_put() Suspend() clk_disable() Resume() clk_enable() Yes, that is correct. I didn't look at the whole driver. My concern was that if the driver was enhanced to more aggressive clock management, you shouldn't do a clk_get() every time you do a clk_enable(), same for put. Got you. I am in sync with your thinking. -Murali Kevin -- To unsubscribe from this list: send the line unsubscribe linux-media in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
RE: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Kevin,
I think I have figured it out...
First issue was that I was adding my entry at the end of dm644x_clks[]
array. I need to add it before the CLK(NULL, NULL, NULL)
secondly, your suggestion didn't work as is. This is what I had to
do to get it working...
static struct clk ccdc_master_clk = {
.name = dm644x_ccdc,
.parent = vpss_master_clk,
};
static struct clk ccdc_slave_clk = {
.name = dm644x_ccdc,
.parent = vpss_slave_clk,
};
It doesn't work with out doing this. The cat /proc/davinci_clocks hangs with
your suggestion implemented...
You should not need to add new clocks with new names. I don't thinke
the name field of the struct clk is used anywhere in the matching.
I think it's only used in /proc/davinci_clocks
static struct davinci_clk dm365_clks = {
CLK(dm644x_ccdc, master, ccdc_master_clk),
CLK(dm644x_ccdc, slave, ccdc_slave_clk),
Looks like the drivers name is 'dm644x_ccdc', not 'isif'. I'm
guessing just this should work without having to add new clock names.
No. I have mixed up the names. ISIF is for the new ISIF driver on DM365.
Below are for DM644x ccdc driver. With just these entries added, two
things observed
1) Only one clock is shown disabled (usually many are shown disabled) during
bootup
2) cat /proc/davinci_clocks hangs.
So this is the only way I got it working.
CLK(dm644x_ccdc, master, vpss_master_clk),
CLK(dm644x_ccdc, slave, vpss_slave_clk),
CLK(NULL, NULL, NULL);
Let me know if you think there is anything wrong with the above scheme.
Kevin
--
To unsubscribe from this list: send the line unsubscribe linux-media in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
m-kariche...@ti.com writes:
From: Muralidharan Karicheri m-kariche...@ti.com
v1 - updated based on comments from Vaibhav Hiremath.
On DM365 we use only vpss master clock, where as on DM355 and
DM6446, we use vpss master and slave clocks for vpfe capture and AM3517
we use internal clock and pixel clock. So this patch makes it configurable
on a per platform basis. This is needed for supporting DM365 for which patches
will be available soon.
Tested-by: Vaibhav Hiremath hvaib...@ti.com, Muralidharan Karicheri
m-kariche...@ti.com
Acked-by: Vaibhav Hiremath hvaib...@ti.com
Signed-off-by: Muralidharan Karicheri m-kariche...@ti.com
Sorry for jumping late into this thread, but I have a couple comments.
First, we should *never* pass clock names from platform code to driver
code. We have the clkdevs for this. Using clkdev, the right clock
is determined from the driver being used and any additional info.
Rather than passing the strings along, you should add the driver name
to the 'dev_id' field of the clkdev node, and then use the 'con_id' field
to distinguish between the two clocks:
- CLK(NULL, vpss_master, vpss_master_clk),
- CLK(NULL, vpss_slave, vpss_slave_clk),
+ CLK(vpfe-capture, master, vpss_master_clk),
+ CLK(vpfe-capture, slave, vpss_slave_clk),
NOTE: this assumes clks are used from VPFE. When you move to CCDC, this
should change accordingly.
More on the usage below...
---
drivers/media/video/davinci/vpfe_capture.c | 98 +--
include/media/davinci/vpfe_capture.h | 11 ++-
2 files changed, 70 insertions(+), 39 deletions(-)
diff --git a/drivers/media/video/davinci/vpfe_capture.c
b/drivers/media/video/davinci/vpfe_capture.c
index 12a1b3d..c3468ee 100644
--- a/drivers/media/video/davinci/vpfe_capture.c
+++ b/drivers/media/video/davinci/vpfe_capture.c
@@ -1787,61 +1787,87 @@ static struct vpfe_device *vpfe_initialize(void)
return vpfe_dev;
}
+/**
+ * vpfe_disable_clock() - Disable clocks for vpfe capture driver
+ * @vpfe_dev - ptr to vpfe capture device
+ *
+ * Disables clocks defined in vpfe configuration.
+ */
static void vpfe_disable_clock(struct vpfe_device *vpfe_dev)
{
struct vpfe_config *vpfe_cfg = vpfe_dev-cfg;
+ int i;
- clk_disable(vpfe_cfg-vpssclk);
- clk_put(vpfe_cfg-vpssclk);
- clk_disable(vpfe_cfg-slaveclk);
- clk_put(vpfe_cfg-slaveclk);
- v4l2_info(vpfe_dev-pdev-driver,
- vpfe vpss master slave clocks disabled\n);
+ for (i = 0; i vpfe_cfg-num_clocks; i++) {
+ clk_disable(vpfe_dev-clks[i]);
+ clk_put(vpfe_dev-clks[i]);
While cleaning this up, you should move the clk_put() to module
disable/unload time. You dont' need to put he clock on every disable.
The same for clk_get(). You don't need to get the clock for every
enable. Just do a clk_get() at init time.
+ }
+ kfree(vpfe_dev-clks);
+ v4l2_info(vpfe_dev-pdev-driver, vpfe capture clocks disabled\n);
}
+/**
+ * vpfe_enable_clock() - Enable clocks for vpfe capture driver
+ * @vpfe_dev - ptr to vpfe capture device
+ *
+ * Enables clocks defined in vpfe configuration. The function
+ * assumes that at least one clock is to be defined which is
+ * true as of now. re-visit this if this assumption is not true
+ */
static int vpfe_enable_clock(struct vpfe_device *vpfe_dev)
{
struct vpfe_config *vpfe_cfg = vpfe_dev-cfg;
- int ret = -ENOENT;
+ int i;
- vpfe_cfg-vpssclk = clk_get(vpfe_dev-pdev, vpss_master);
Using my clkdevs from above, you just need to change this to:
vpfe_cfg-vpssclk = clk_get(vpfe_dev-pdev, master);
Now that the device name is in the clkdev node, clk_get() will
find the right clock based on the device name.
- if (NULL == vpfe_cfg-vpssclk) {
- v4l2_err(vpfe_dev-pdev-driver, No clock defined for
- vpss_master\n);
- return ret;
- }
+ if (!vpfe_cfg-num_clocks)
+ return 0;
- if (clk_enable(vpfe_cfg-vpssclk)) {
- v4l2_err(vpfe_dev-pdev-driver,
- vpfe vpss master clock not enabled\n);
- goto out;
- }
- v4l2_info(vpfe_dev-pdev-driver,
- vpfe vpss master clock enabled\n);
+ vpfe_dev-clks = kzalloc(vpfe_cfg-num_clocks *
+sizeof(struct clock *), GFP_KERNEL);
- vpfe_cfg-slaveclk = clk_get(vpfe_dev-pdev, vpss_slave);
And here:
vpfe_cfg-slaveclk = clk_get(vpfe_dev-pdev, slave);
- if (NULL == vpfe_cfg-slaveclk) {
- v4l2_err(vpfe_dev-pdev-driver,
- No clock defined for vpss slave\n);
- goto out;
+ if (NULL == vpfe_dev-clks) {
+ v4l2_err(vpfe_dev-pdev-driver, Memory allocation failed\n);
+ return -ENOMEM;
}
- if (clk_enable(vpfe_cfg-slaveclk)) {
-
RE: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Kevin,
After sending out my last patch (moving clocks to ccdc driver), I thought of
reworking it in similar lines. I will re-send the patch after incorporating
this.
Murali Karicheri
Software Design Engineer
Texas Instruments Inc.
Germantown, MD 20874
phone: 301-407-9583
email: m-kariche...@ti.com
-Original Message-
From: Kevin Hilman [mailto:khil...@deeprootsystems.com]
Sent: Wednesday, December 09, 2009 11:00 AM
To: Karicheri, Muralidharan
Cc: linux-media@vger.kernel.org; hverk...@xs4all.nl; davinci-linux-open-
sou...@linux.davincidsp.com; Hiremath, Vaibhav
Subject: Re: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
m-kariche...@ti.com writes:
From: Muralidharan Karicheri m-kariche...@ti.com
v1 - updated based on comments from Vaibhav Hiremath.
On DM365 we use only vpss master clock, where as on DM355 and
DM6446, we use vpss master and slave clocks for vpfe capture and AM3517
we use internal clock and pixel clock. So this patch makes it
configurable
on a per platform basis. This is needed for supporting DM365 for which
patches
will be available soon.
Tested-by: Vaibhav Hiremath hvaib...@ti.com, Muralidharan Karicheri m-
kariche...@ti.com
Acked-by: Vaibhav Hiremath hvaib...@ti.com
Signed-off-by: Muralidharan Karicheri m-kariche...@ti.com
Sorry for jumping late into this thread, but I have a couple comments.
First, we should *never* pass clock names from platform code to driver
code. We have the clkdevs for this. Using clkdev, the right clock
is determined from the driver being used and any additional info.
Rather than passing the strings along, you should add the driver name
to the 'dev_id' field of the clkdev node, and then use the 'con_id' field
to distinguish between the two clocks:
- CLK(NULL, vpss_master, vpss_master_clk),
- CLK(NULL, vpss_slave, vpss_slave_clk),
+ CLK(vpfe-capture, master, vpss_master_clk),
+ CLK(vpfe-capture, slave, vpss_slave_clk),
NOTE: this assumes clks are used from VPFE. When you move to CCDC, this
should change accordingly.
More on the usage below...
---
drivers/media/video/davinci/vpfe_capture.c | 98 +-
-
include/media/davinci/vpfe_capture.h | 11 ++-
2 files changed, 70 insertions(+), 39 deletions(-)
diff --git a/drivers/media/video/davinci/vpfe_capture.c
b/drivers/media/video/davinci/vpfe_capture.c
index 12a1b3d..c3468ee 100644
--- a/drivers/media/video/davinci/vpfe_capture.c
+++ b/drivers/media/video/davinci/vpfe_capture.c
@@ -1787,61 +1787,87 @@ static struct vpfe_device *vpfe_initialize(void)
return vpfe_dev;
}
+/**
+ * vpfe_disable_clock() - Disable clocks for vpfe capture driver
+ * @vpfe_dev - ptr to vpfe capture device
+ *
+ * Disables clocks defined in vpfe configuration.
+ */
static void vpfe_disable_clock(struct vpfe_device *vpfe_dev)
{
struct vpfe_config *vpfe_cfg = vpfe_dev-cfg;
+int i;
-clk_disable(vpfe_cfg-vpssclk);
-clk_put(vpfe_cfg-vpssclk);
-clk_disable(vpfe_cfg-slaveclk);
-clk_put(vpfe_cfg-slaveclk);
-v4l2_info(vpfe_dev-pdev-driver,
- vpfe vpss master slave clocks disabled\n);
+for (i = 0; i vpfe_cfg-num_clocks; i++) {
+clk_disable(vpfe_dev-clks[i]);
+clk_put(vpfe_dev-clks[i]);
While cleaning this up, you should move the clk_put() to module
disable/unload time. You dont' need to put he clock on every disable.
The same for clk_get(). You don't need to get the clock for every
enable. Just do a clk_get() at init time.
+}
+kfree(vpfe_dev-clks);
+v4l2_info(vpfe_dev-pdev-driver, vpfe capture clocks disabled\n);
}
+/**
+ * vpfe_enable_clock() - Enable clocks for vpfe capture driver
+ * @vpfe_dev - ptr to vpfe capture device
+ *
+ * Enables clocks defined in vpfe configuration. The function
+ * assumes that at least one clock is to be defined which is
+ * true as of now. re-visit this if this assumption is not true
+ */
static int vpfe_enable_clock(struct vpfe_device *vpfe_dev)
{
struct vpfe_config *vpfe_cfg = vpfe_dev-cfg;
-int ret = -ENOENT;
+int i;
-vpfe_cfg-vpssclk = clk_get(vpfe_dev-pdev, vpss_master);
Using my clkdevs from above, you just need to change this to:
vpfe_cfg-vpssclk = clk_get(vpfe_dev-pdev, master);
Now that the device name is in the clkdev node, clk_get() will
find the right clock based on the device name.
-if (NULL == vpfe_cfg-vpssclk) {
-v4l2_err(vpfe_dev-pdev-driver, No clock defined for
- vpss_master\n);
-return ret;
-}
+if (!vpfe_cfg-num_clocks)
+return 0;
-if (clk_enable(vpfe_cfg-vpssclk)) {
-v4l2_err(vpfe_dev-pdev-driver,
-vpfe vpss master clock not enabled\n);
-goto out;
-}
-v4l2_info(vpfe_dev-pdev-driver,
- vpfe vpss master clock enabled\n);
+vpfe_dev-clks = kzalloc(vpfe_cfg-num_clocks
RE: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Kevin,
+/**
+ * vpfe_disable_clock() - Disable clocks for vpfe capture driver
+ * @vpfe_dev - ptr to vpfe capture device
+ *
+ * Disables clocks defined in vpfe configuration.
+ */
static void vpfe_disable_clock(struct vpfe_device *vpfe_dev)
{
struct vpfe_config *vpfe_cfg = vpfe_dev-cfg;
+int i;
-clk_disable(vpfe_cfg-vpssclk);
-clk_put(vpfe_cfg-vpssclk);
-clk_disable(vpfe_cfg-slaveclk);
-clk_put(vpfe_cfg-slaveclk);
-v4l2_info(vpfe_dev-pdev-driver,
- vpfe vpss master slave clocks disabled\n);
+for (i = 0; i vpfe_cfg-num_clocks; i++) {
+clk_disable(vpfe_dev-clks[i]);
+clk_put(vpfe_dev-clks[i]);
While cleaning this up, you should move the clk_put() to module
disable/unload time.
[MK] vpfe_disable_clock() is called from remove(). In the new
patch, from ccdc driver remove() function, clk_put() will be called.
Why do you think it should be moved to exit() function of the module?
You dont' need to put he clock on every disable.
The same for clk_get(). You don't need to get the clock for every
enable. Just do a clk_get() at init time.
Are you suggesting to call clk_get() during init() and call clk_put()
from exit()? What is wrong with calling clk_get() from probe()?
I thought following is correct:-
Probe()
clk_get() followed by clk_enable()
Remove()
clk_disable() followed by clk_put()
Suspend()
clk_disable()
Resume()
clk_enable()
Please confirm.
--
To unsubscribe from this list: send the line unsubscribe linux-media in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
RE: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Kevin,
I tried the following and I get error in clk_enable(). Do you know what might
be wrong?
in DM365.c
CLK(isif, master, vpss_master_clk)
The driver name is isif. I call clk_get(pdev-dev, master) from isif driver.
The platform device name is isif. This call succeeds, but clk_enable()
fails...
clk_ptr = clk_get(pdev-dev, master);
clk_enable(clk_ptr);
r...@dm355-evm:~# cat /proc/davinci_clocks
ref_clk users= 7 2400 Hz
pll1users= 6 pll 48600 Hz
pll1_aux_clk users= 3 pll 2400 Hz
uart0 users= 1 psc 2400 Hz
i2c users= 1 psc 2400 Hz
spi4users= 0 psc 2400 Hz
pwm0users= 0 psc 2400 Hz
pwm1users= 0 psc 2400 Hz
pwm2users= 0 psc 2400 Hz
timer0 users= 1 psc 2400 Hz
timer1 users= 0 psc 2400 Hz
timer2 users= 1 psc 2400 Hz
timer3 users= 0 psc 2400 Hz
usb users= 0 psc 2400 Hz
pll1_sysclkbp users= 0 pll 2400 Hz
clkout0 users= 0 pll 2400 Hz
pll1_sysclk1 users= 0 pll 48600 Hz
pll1_sysclk2 users= 0 pll 24300 Hz
pll1_sysclk3 users= 0 pll 24300 Hz
vpss_dacusers= 0 psc 24300 Hz
mjcpusers= 0 psc 24300 Hz
pll1_sysclk4 users= 3 pll 12150 Hz
uart1 users= 0 psc 12150 Hz
mmcsd1 users= 0 psc 12150 Hz
spi0users= 0 psc 12150 Hz
spi1users= 0 psc 12150 Hz
spi2users= 0 psc 12150 Hz
spi3users= 0 psc 12150 Hz
gpiousers= 1 psc 12150 Hz
aemif users= 1 psc 12150 Hz
emacusers= 1 psc 12150 Hz
asp0users= 0 psc 12150 Hz
rto users= 0 psc 12150 Hz
pll1_sysclk5 users= 0 pll 24300 Hz
vpss_master users= 0 psc 24300 Hz
pll1_sysclk6 users= 0 pll 2700 Hz
pll1_sysclk7 users= 0 pll 48600 Hz
pll1_sysclk8 users= 0 pll 12150 Hz
mmcsd0 users= 0 psc 12150 Hz
pll1_sysclk9 users= 0 pll 24300 Hz
pll2users= 1 pll 59400 Hz
pll2_aux_clk users= 0 pll 2400 Hz
clkout1 users= 0 pll 2400 Hz
pll2_sysclk1 users= 0 pll 59400 Hz
pll2_sysclk2 users= 1 pll 29700 Hz
arm_clk users= 1 psc 29700 Hz
pll2_sysclk3 users= 0 pll 59400 Hz
pll2_sysclk4 users= 0 pll 20482758 Hz
voice_codec users= 0 psc 20482758 Hz
pll2_sysclk5 users= 0 pll 7425 Hz
pll2_sysclk6 users= 0 pll 59400 Hz
pll2_sysclk7 users= 0 pll 59400 Hz
pll2_sysclk8 users= 0 pll 59400 Hz
pll2_sysclk9 users= 0 pll 59400 Hz
pwm3users= 0 psc 2400 Hz
r...@dm355-evm:~#
Murali Karicheri
Software Design Engineer
Texas Instruments Inc.
Germantown, MD 20874
phone: 301-407-9583
email: m-kariche...@ti.com
-Original Message-
From: davinci-linux-open-source-boun...@linux.davincidsp.com
[mailto:davinci-linux-open-source-boun...@linux.davincidsp.com] On Behalf
Of Karicheri, Muralidharan
Sent: Wednesday, December 09, 2009 12:45 PM
To: Kevin Hilman
Cc: davinci-linux-open-sou...@linux.davincidsp.com; linux-
me...@vger.kernel.org
Subject: RE: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Kevin,
+/**
+ * vpfe_disable_clock() - Disable clocks for vpfe capture driver
+ * @vpfe_dev - ptr to vpfe capture device
+ *
+ * Disables clocks defined in vpfe configuration.
+ */
static void vpfe_disable_clock(struct vpfe_device *vpfe_dev)
{
struct vpfe_config *vpfe_cfg = vpfe_dev-cfg;
+ int i;
- clk_disable(vpfe_cfg-vpssclk);
- clk_put(vpfe_cfg-vpssclk);
- clk_disable(vpfe_cfg-slaveclk);
- clk_put(vpfe_cfg-slaveclk);
- v4l2_info(vpfe_dev-pdev-driver,
-vpfe vpss master slave clocks disabled\n);
+ for (i = 0; i vpfe_cfg-num_clocks; i++) {
+ clk_disable(vpfe_dev-clks[i]);
+ clk_put(vpfe_dev-clks[i]);
While cleaning this up, you should move the clk_put() to module
disable/unload time.
[MK] vpfe_disable_clock() is called from remove(). In the new
patch, from ccdc driver remove() function, clk_put() will be called.
Why do you think it should be moved to exit() function of the module?
You dont' need to put he clock on every disable.
The same for clk_get(). You don't need to get the clock for every
enable. Just do a clk_get() at init time.
Are you suggesting to call clk_get() during init() and call clk_put()
from exit()? What is wrong with calling clk_get() from probe()?
I thought following is correct:-
Probe()
clk_get() followed by clk_enable()
Remove()
clk_disable() followed by clk_put()
Suspend()
clk_disable()
Resume()
clk_enable()
Please confirm.
___
Davinci-linux-open-source mailing list
davinci-linux-open-sou...@linux.davincidsp.com
http
RE: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Kevin,
I think I have figured it out...
First issue was that I was adding my entry at the end of dm644x_clks[]
array. I need to add it before the CLK(NULL, NULL, NULL)
secondly, your suggestion didn't work as is. This is what I had to
do to get it working...
static struct clk ccdc_master_clk = {
.name = dm644x_ccdc,
.parent = vpss_master_clk,
};
static struct clk ccdc_slave_clk = {
.name = dm644x_ccdc,
.parent = vpss_slave_clk,
};
static struct davinci_clk dm365_clks = {
CLK(dm644x_ccdc, master, ccdc_master_clk),
CLK(dm644x_ccdc, slave, ccdc_slave_clk),
CLK(NULL, NULL, NULL);
Let me know if you think there is anything wrong with the above scheme.
Murali Karicheri
Software Design Engineer
Texas Instruments Inc.
Germantown, MD 20874
phone: 301-407-9583
email: m-kariche...@ti.com
-Original Message-
From: Karicheri, Muralidharan
Sent: Wednesday, December 09, 2009 1:22 PM
To: Karicheri, Muralidharan; Kevin Hilman
Cc: davinci-linux-open-sou...@linux.davincidsp.com; linux-
me...@vger.kernel.org
Subject: RE: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Kevin,
I tried the following and I get error in clk_enable(). Do you know what
might be wrong?
in DM365.c
CLK(isif, master, vpss_master_clk)
The driver name is isif. I call clk_get(pdev-dev, master) from isif
driver. The platform device name is isif. This call succeeds, but
clk_enable() fails...
clk_ptr = clk_get(pdev-dev, master);
clk_enable(clk_ptr);
r...@dm355-evm:~# cat /proc/davinci_clocks
ref_clk users= 7 2400 Hz
pll1users= 6 pll 48600 Hz
pll1_aux_clk users= 3 pll 2400 Hz
uart0 users= 1 psc 2400 Hz
i2c users= 1 psc 2400 Hz
spi4users= 0 psc 2400 Hz
pwm0users= 0 psc 2400 Hz
pwm1users= 0 psc 2400 Hz
pwm2users= 0 psc 2400 Hz
timer0 users= 1 psc 2400 Hz
timer1 users= 0 psc 2400 Hz
timer2 users= 1 psc 2400 Hz
timer3 users= 0 psc 2400 Hz
usb users= 0 psc 2400 Hz
pll1_sysclkbp users= 0 pll 2400 Hz
clkout0 users= 0 pll 2400 Hz
pll1_sysclk1 users= 0 pll 48600 Hz
pll1_sysclk2 users= 0 pll 24300 Hz
pll1_sysclk3 users= 0 pll 24300 Hz
vpss_dacusers= 0 psc 24300 Hz
mjcpusers= 0 psc 24300 Hz
pll1_sysclk4 users= 3 pll 12150 Hz
uart1 users= 0 psc 12150 Hz
mmcsd1 users= 0 psc 12150 Hz
spi0users= 0 psc 12150 Hz
spi1users= 0 psc 12150 Hz
spi2users= 0 psc 12150 Hz
spi3users= 0 psc 12150 Hz
gpiousers= 1 psc 12150 Hz
aemif users= 1 psc 12150 Hz
emacusers= 1 psc 12150 Hz
asp0users= 0 psc 12150 Hz
rto users= 0 psc 12150 Hz
pll1_sysclk5 users= 0 pll 24300 Hz
vpss_master users= 0 psc 24300 Hz
pll1_sysclk6 users= 0 pll 2700 Hz
pll1_sysclk7 users= 0 pll 48600 Hz
pll1_sysclk8 users= 0 pll 12150 Hz
mmcsd0 users= 0 psc 12150 Hz
pll1_sysclk9 users= 0 pll 24300 Hz
pll2users= 1 pll 59400 Hz
pll2_aux_clk users= 0 pll 2400 Hz
clkout1 users= 0 pll 2400 Hz
pll2_sysclk1 users= 0 pll 59400 Hz
pll2_sysclk2 users= 1 pll 29700 Hz
arm_clk users= 1 psc 29700 Hz
pll2_sysclk3 users= 0 pll 59400 Hz
pll2_sysclk4 users= 0 pll 20482758 Hz
voice_codec users= 0 psc 20482758 Hz
pll2_sysclk5 users= 0 pll 7425 Hz
pll2_sysclk6 users= 0 pll 59400 Hz
pll2_sysclk7 users= 0 pll 59400 Hz
pll2_sysclk8 users= 0 pll 59400 Hz
pll2_sysclk9 users= 0 pll 59400 Hz
pwm3users= 0 psc 2400 Hz
r...@dm355-evm:~#
Murali Karicheri
Software Design Engineer
Texas Instruments Inc.
Germantown, MD 20874
phone: 301-407-9583
email: m-kariche...@ti.com
-Original Message-
From: davinci-linux-open-source-boun...@linux.davincidsp.com
[mailto:davinci-linux-open-source-boun...@linux.davincidsp.com] On Behalf
Of Karicheri, Muralidharan
Sent: Wednesday, December 09, 2009 12:45 PM
To: Kevin Hilman
Cc: davinci-linux-open-sou...@linux.davincidsp.com; linux-
me...@vger.kernel.org
Subject: RE: [PATCH - v1 1/2] V4L - vpfe capture - make clocks
configurable
Kevin,
+/**
+ * vpfe_disable_clock() - Disable clocks for vpfe capture driver
+ * @vpfe_dev - ptr to vpfe capture device
+ *
+ * Disables clocks defined in vpfe configuration.
+ */
static void vpfe_disable_clock(struct vpfe_device *vpfe_dev)
{
struct vpfe_config *vpfe_cfg = vpfe_dev-cfg;
+ int i;
- clk_disable(vpfe_cfg-vpssclk);
- clk_put(vpfe_cfg-vpssclk);
- clk_disable(vpfe_cfg-slaveclk);
- clk_put(vpfe_cfg-slaveclk);
- v4l2_info
RE: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Hans,
Please hold on to this. I will send you an updated patch.
Murali Karicheri
Software Design Engineer
Texas Instruments Inc.
Germantown, MD 20874
phone: 301-407-9583
email: m-kariche...@ti.com
-Original Message-
From: Karicheri, Muralidharan
Sent: Tuesday, December 01, 2009 12:22 PM
To: 'Hans Verkuil'
Cc: Hiremath, Vaibhav; linux-media@vger.kernel.org
Subject: FW: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
Hans,
Could you please add this to your hg tree and send a pull
request to Mauro? This was reviewed by Vaibhav and tested on
DM355, DM6446, AM3517 DM365. I will request Kevin to pull
the Architecture part of this patch.
Thanks.
Murali Karicheri
Software Design Engineer
Texas Instruments Inc.
Germantown, MD 20874
phone: 301-407-9583
email: m-kariche...@ti.com
-Original Message-
From: Karicheri, Muralidharan
Sent: Tuesday, December 01, 2009 12:19 PM
To: linux-media@vger.kernel.org; hverk...@xs4all.nl;
khil...@deeprootsystems.com
Cc: davinci-linux-open-sou...@linux.davincidsp.com; Hiremath, Vaibhav;
Karicheri, Muralidharan
Subject: [PATCH - v1 1/2] V4L - vpfe capture - make clocks configurable
From: Muralidharan Karicheri m-kariche...@ti.com
v1 - updated based on comments from Vaibhav Hiremath.
On DM365 we use only vpss master clock, where as on DM355 and
DM6446, we use vpss master and slave clocks for vpfe capture and AM3517
we use internal clock and pixel clock. So this patch makes it configurable
on a per platform basis. This is needed for supporting DM365 for which
patches
will be available soon.
Tested-by: Vaibhav Hiremath hvaib...@ti.com, Muralidharan Karicheri m-
kariche...@ti.com
Acked-by: Vaibhav Hiremath hvaib...@ti.com
Signed-off-by: Muralidharan Karicheri m-kariche...@ti.com
---
drivers/media/video/davinci/vpfe_capture.c | 98 +--
-
---
include/media/davinci/vpfe_capture.h | 11 ++-
2 files changed, 70 insertions(+), 39 deletions(-)
diff --git a/drivers/media/video/davinci/vpfe_capture.c
b/drivers/media/video/davinci/vpfe_capture.c
index 12a1b3d..c3468ee 100644
--- a/drivers/media/video/davinci/vpfe_capture.c
+++ b/drivers/media/video/davinci/vpfe_capture.c
@@ -1787,61 +1787,87 @@ static struct vpfe_device *vpfe_initialize(void)
return vpfe_dev;
}
+/**
+ * vpfe_disable_clock() - Disable clocks for vpfe capture driver
+ * @vpfe_dev - ptr to vpfe capture device
+ *
+ * Disables clocks defined in vpfe configuration.
+ */
static void vpfe_disable_clock(struct vpfe_device *vpfe_dev)
{
struct vpfe_config *vpfe_cfg = vpfe_dev-cfg;
+ int i;
- clk_disable(vpfe_cfg-vpssclk);
- clk_put(vpfe_cfg-vpssclk);
- clk_disable(vpfe_cfg-slaveclk);
- clk_put(vpfe_cfg-slaveclk);
- v4l2_info(vpfe_dev-pdev-driver,
- vpfe vpss master slave clocks disabled\n);
+ for (i = 0; i vpfe_cfg-num_clocks; i++) {
+ clk_disable(vpfe_dev-clks[i]);
+ clk_put(vpfe_dev-clks[i]);
+ }
+ kfree(vpfe_dev-clks);
+ v4l2_info(vpfe_dev-pdev-driver, vpfe capture clocks disabled\n);
}
+/**
+ * vpfe_enable_clock() - Enable clocks for vpfe capture driver
+ * @vpfe_dev - ptr to vpfe capture device
+ *
+ * Enables clocks defined in vpfe configuration. The function
+ * assumes that at least one clock is to be defined which is
+ * true as of now. re-visit this if this assumption is not true
+ */
static int vpfe_enable_clock(struct vpfe_device *vpfe_dev)
{
struct vpfe_config *vpfe_cfg = vpfe_dev-cfg;
- int ret = -ENOENT;
+ int i;
- vpfe_cfg-vpssclk = clk_get(vpfe_dev-pdev, vpss_master);
- if (NULL == vpfe_cfg-vpssclk) {
- v4l2_err(vpfe_dev-pdev-driver, No clock defined for
- vpss_master\n);
- return ret;
- }
+ if (!vpfe_cfg-num_clocks)
+ return 0;
- if (clk_enable(vpfe_cfg-vpssclk)) {
- v4l2_err(vpfe_dev-pdev-driver,
- vpfe vpss master clock not enabled\n);
- goto out;
- }
- v4l2_info(vpfe_dev-pdev-driver,
- vpfe vpss master clock enabled\n);
+ vpfe_dev-clks = kzalloc(vpfe_cfg-num_clocks *
+sizeof(struct clock *), GFP_KERNEL);
- vpfe_cfg-slaveclk = clk_get(vpfe_dev-pdev, vpss_slave);
- if (NULL == vpfe_cfg-slaveclk) {
- v4l2_err(vpfe_dev-pdev-driver,
- No clock defined for vpss slave\n);
- goto out;
+ if (NULL == vpfe_dev-clks) {
+ v4l2_err(vpfe_dev-pdev-driver, Memory allocation failed\n);
+ return -ENOMEM;
}
- if (clk_enable(vpfe_cfg-slaveclk)) {
- v4l2_err(vpfe_dev-pdev-driver,
- vpfe vpss slave clock not enabled\n);
- goto out;
+ for (i = 0; i vpfe_cfg-num_clocks; i++) {
+ if (NULL == vpfe_cfg-clocks[i]) {
+ v4l2_err(vpfe_dev-pdev-driver,
+ clock %s is not defined in vpfe
