Skip to contentSkip to conversation

sunxi clock fixes

PATCH New Emilio López 2 participants 9 messages Sep 21, 2013 – Oct 31, 2013

Subsystems (3)
lore
Share & tools
lore

Emilio López

author maintainer
#0 lore raw

cover sunxi clock fixes

Hi,

This small series contains two bug fixes and a patch to solve the issue of accidental clock gating, as discussed earlier on my RFC patch, "of:
add a basic memory driver".

I believe the two first patches should go into -rcN, although I have not noticed any issues so far coming from the bugs themselves.

The third patch should be a no-op for now, and I'd like to have it merged before these clocks actually appear on the driver; in this way they will never be gated accidentally, and it remains bisectable. I will be sending PLL5/6 support patches in short order.

Cheers,

Emilio

Emilio L?pez (3):
clk: sunxi: factors: fix off-by-one masks
clk: sunxi: factors: clear variables before using them
clk: sunxi: protect core clocks from accidental shutdown

 drivers/clk/sunxi/clk-factors.c |  4 ++--
 drivers/clk/sunxi/clk-sunxi.c   | 28 ++++++++++++++++++++++++++++
 2 files changed, 30 insertions(+), 2 deletions(-)
Signature · 1 line
1.8.4
author maintainer
#1 lore raw

PATCH 1/3 clk: sunxi: factors: fix off-by-one masks

The previous code would generate one bit too long masks, and was needlessly complicated. This patch replaces it by simpler code that can generate the masks correctly.

---
 drivers/clk/sunxi/clk-factors.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clk/sunxi/clk-factors.c +1 -1
index 88523f9..5687ac9 100644
@@ -40,7 +40,7 @@ struct clk_factors {
 
 #define to_clk_factors(_hw) container_of(_hw, struct clk_factors, hw)
 
-#define SETMASK(len, pos)		(((-1U) >> (31-len))  << (pos))
+#define SETMASK(len, pos)		(((1U << (len)) - 1) << (pos))
 #define CLRMASK(len, pos)		(~(SETMASK(len, pos)))
 #define FACTOR_GET(bit, len, reg)	(((reg) & SETMASK(len, bit)) >> (bit))
 
Signature · 1 line
1.8.4
#2 lore raw

PATCH 1/3 clk: sunxi: factors: fix off-by-one masks

Quoted 5 lines · Emilio L?pez

The previous code would generate one bit too long masks, and was needlessly complicated. This patch replaces it by simpler code that can generate the masks correctly.

Thanks!
Maxime

Signature · 10 lines
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20130923/73614c4b/attachment.sig>
author maintainer
#3 lore raw

PATCH 2/3 clk: sunxi: factors: clear variables before using them

In reply to Emilio López

Random bits may get into our factors if we don't clear n, k, m and p.

---
 drivers/clk/sunxi/clk-factors.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clk/sunxi/clk-factors.c +1 -1
index 5687ac9..f05207a 100644
@@ -88,7 +88,7 @@ static long clk_factors_round_rate(struct clk_hw *hw, unsigned long rate,
 static int clk_factors_set_rate(struct clk_hw *hw, unsigned long rate,
 				unsigned long parent_rate)
 {
-	u8 n, k, m, p;
+	u8 n = 0, k = 0, m = 0, p = 0;
 	u32 reg;
 	struct clk_factors *factors = to_clk_factors(hw);
 	struct clk_factors_config *config = factors->config;
Signature · 1 line
1.8.4
#4 lore raw

PATCH 2/3 clk: sunxi: factors: clear variables before using them

Quoted 3 lines · Emilio L?pez

Random bits may get into our factors if we don't clear n, k, m and p.

Thanks!
Maxime

Signature · 10 lines
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20130923/b34eb042/attachment.sig>
author maintainer
#5 lore raw

PATCH 3/3 clk: sunxi: protect core clocks from accidental shutdown

In reply to Emilio López

Some important clocks may get disabled as a side effect of another clock being disabled, because they have no consumers. This patch implements a mechanism so those clocks can be claimed by the driver and therefore remain enabled at all times.

---
 drivers/clk/sunxi/clk-sunxi.c | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)
drivers/clk/sunxi/clk-sunxi.c +28 -0
index 34ee69f..83d3eac 100644
@@ -617,6 +617,31 @@ static void __init of_sunxi_table_clock_setup(const struct of_device_id *clk_mat
 	}
 }
 
+/**
+ * System clock protection
+ *
+ * By enabling these critical clocks, we prevent their accidental gating
+ * by the framework
+ */
+static void __init sunxi_clock_protect(void)
+{
+	struct clk *clk;
+
+	/* memory bus clock - sun5i+ */
+	clk = clk_get(NULL, "mbus");
+	if (!IS_ERR(clk)) {
+		clk_prepare_enable(clk);
+		clk_put(clk);
+	}
+
+	/* DDR clock - sun4i+ */
+	clk = clk_get(NULL, "pll5_ddr");
+	if (!IS_ERR(clk)) {
+		clk_prepare_enable(clk);
+		clk_put(clk);
+	}
+}
+
 void __init sunxi_init_clocks(void)
 {
 	/* Register all the simple and basic clocks on DT */
@@ -633,4 +658,7 @@ void __init sunxi_init_clocks(void)
 
 	/* Register gate clocks */
 	of_sunxi_table_clock_setup(clk_gates_match, sunxi_gates_clk_setup);
+
+	/* Enable core system clocks */
+	sunxi_clock_protect();
 }
Signature · 1 line
1.8.4
#6 lore raw

PATCH 3/3 clk: sunxi: protect core clocks from accidental shutdown

Quoted 6 lines · Emilio L?pez

Some important clocks may get disabled as a side effect of another clock being disabled, because they have no consumers. This patch implements a mechanism so those clocks can be claimed by the driver and therefore remain enabled at all times.

Thanks!
Maxime

Signature · 10 lines
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20130923/878c7f57/attachment.sig>
author maintainer
#7 lore raw

cover sunxi clock fixes

In reply to Emilio López

Hi Mike,

El 20/09/13 22:03, Emilio L?pez escribi?:

Quoted 18 lines

Hi,

This small series contains two bug fixes and a patch to solve the issue of accidental clock gating, as discussed earlier on my RFC patch, "of:
add a basic memory driver".

I believe the two first patches should go into -rcN, although I have not noticed any issues so far coming from the bugs themselves.

The third patch should be a no-op for now, and I'd like to have it merged before these clocks actually appear on the driver; in this way they will never be gated accidentally, and it remains bisectable. I will be sending PLL5/6 support patches in short order.

Cheers,

Emilio

Quoted 9 lines

Emilio L?pez (3):
clk: sunxi: factors: fix off-by-one masks
clk: sunxi: factors: clear variables before using them
clk: sunxi: protect core clocks from accidental shutdown

  drivers/clk/sunxi/clk-factors.c |  4 ++--
  drivers/clk/sunxi/clk-sunxi.c   | 28 ++++++++++++++++++++++++++++
  2 files changed, 30 insertions(+), 2 deletions(-)

Can you please take a look at this series and pick the patches if you are ok with them? It would be nice to see the first two fixes on -rcN by the way.

Thanks!

Emilio

author maintainer
#8 lore raw

cover sunxi clock fixes

El 09/10/13 09:51, Emilio L?pez escribi?:

Quoted 26 lines

Hi Mike,

El 20/09/13 22:03, Emilio L?pez escribi?:

Quoted 23 lines

Hi,

This small series contains two bug fixes and a patch to solve the issue of accidental clock gating, as discussed earlier on my RFC patch, "of:
add a basic memory driver".

I believe the two first patches should go into -rcN, although I have not noticed any issues so far coming from the bugs themselves.

The third patch should be a no-op for now, and I'd like to have it merged before these clocks actually appear on the driver; in this way they will never be gated accidentally, and it remains bisectable. I will be sending PLL5/6 support patches in short order.

Cheers,

Emilio

Emilio L?pez (3):
clk: sunxi: factors: fix off-by-one masks
clk: sunxi: factors: clear variables before using them
clk: sunxi: protect core clocks from accidental shutdown

Quoted 8 lines
Quoted 4 lines
  drivers/clk/sunxi/clk-factors.c |  4 ++--
  drivers/clk/sunxi/clk-sunxi.c   | 28 ++++++++++++++++++++++++++++
  2 files changed, 30 insertions(+), 2 deletions(-)

Can you please take a look at this series and pick the patches if you are ok with them? It would be nice to see the first two fixes on -rcN by the way.

$ ping Mike