[PATCH v3 0/2] clocksource: Add BE APIs support.

[PATCH v3 0/2] clocksource: Add BE APIs support.

[Xiubo Li]

Change in V3:
- Provide a separate patch for using newer memory accessors.

Change in V2:
- Using ioread{16,32}[be]() instead of readl_relaxed().
- Add clocksource_mmio_readX_Y_be() supports only.


Xiubo Li (2):
  clocksource: Update to the newer memory functions.
  clocksource: Add BE APIs support for clocksource counter reading.

 drivers/clocksource/mmio.c  | 28 ++++++++++++++++++++++++----
 include/linux/clocksource.h |  4 ++++
 2 files changed, 28 insertions(+), 4 deletions(-)

-- 
2.1.0.27.g96db324

[PATCH v3 1/2] clocksource: Update to the newer memory functions.

[Xiubo Li]

Use ioread{16,32} instead of read{w,l}_relaxed.

For read{w,l}_relaxed accessor, if one arch has its own defination,
then used it. Or will use the generic one, which will be read as LE
endian as default.

For some ARCHes, such PowerPC, if using the clocksource mmio, the
read{w,l}_relaxed will couldn't be used directly. There must be its
own BE accessor in its individual clocksource driver.

The ioread{16,32} will be read as LE endian as default if the arch
hasn't any defination of it too, like read{w,l}_relaxed do.

To prepare supporting the BE APIs and to be more unified with them,
using ioread{16,32}[be] is a good chioce here.

Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
---
 drivers/clocksource/mmio.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/clocksource/mmio.c b/drivers/clocksource/mmio.c
index 1593ade..d3cbf70 100644
--- a/drivers/clocksource/mmio.c
+++ b/drivers/clocksource/mmio.c
@@ -22,22 +22,22 @@ static inline struct clocksource_mmio *to_mmio_clksrc(struct clocksource *c)
 
 cycle_t clocksource_mmio_readl_up(struct clocksource *c)
 {
-	return (cycle_t)readl_relaxed(to_mmio_clksrc(c)->reg);
+	return (cycle_t)ioread32(to_mmio_clksrc(c)->reg);
 }
 
 cycle_t clocksource_mmio_readl_down(struct clocksource *c)
 {
-	return ~(cycle_t)readl_relaxed(to_mmio_clksrc(c)->reg) & c->mask;
+	return ~(cycle_t)ioread32(to_mmio_clksrc(c)->reg) & c->mask;
 }
 
 cycle_t clocksource_mmio_readw_up(struct clocksource *c)
 {
-	return (cycle_t)readw_relaxed(to_mmio_clksrc(c)->reg);
+	return (cycle_t)ioread16(to_mmio_clksrc(c)->reg);
 }
 
 cycle_t clocksource_mmio_readw_down(struct clocksource *c)
 {
-	return ~(cycle_t)readw_relaxed(to_mmio_clksrc(c)->reg) & c->mask;
+	return ~(cycle_t)ioread16(to_mmio_clksrc(c)->reg) & c->mask;
 }
 
 /**
-- 
2.1.0.27.g96db324

[PATCH v3 2/2] clocksource: Add BE APIs support for clocksource counter reading.

[Xiubo Li]

Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
---
 drivers/clocksource/mmio.c  | 20 ++++++++++++++++++++
 include/linux/clocksource.h |  4 ++++
 2 files changed, 24 insertions(+)

diff --git a/drivers/clocksource/mmio.c b/drivers/clocksource/mmio.c
index d3cbf70..d35a407 100644
--- a/drivers/clocksource/mmio.c
+++ b/drivers/clocksource/mmio.c
@@ -20,21 +20,41 @@ static inline struct clocksource_mmio *to_mmio_clksrc(struct clocksource *c)
 	return container_of(c, struct clocksource_mmio, clksrc);
 }
 
+cycle_t clocksource_mmio_readl_up_be(struct clocksource *c)
+{
+	return (cycle_t)ioread32be(to_mmio_clksrc(c)->reg);
+}
+
 cycle_t clocksource_mmio_readl_up(struct clocksource *c)
 {
 	return (cycle_t)ioread32(to_mmio_clksrc(c)->reg);
 }
 
+cycle_t clocksource_mmio_readl_down_be(struct clocksource *c)
+{
+	return ~(cycle_t)ioread32be(to_mmio_clksrc(c)->reg) & c->mask;
+}
+
 cycle_t clocksource_mmio_readl_down(struct clocksource *c)
 {
 	return ~(cycle_t)ioread32(to_mmio_clksrc(c)->reg) & c->mask;
 }
 
+cycle_t clocksource_mmio_readw_up_be(struct clocksource *c)
+{
+	return (cycle_t)ioread16be(to_mmio_clksrc(c)->reg);
+}
+
 cycle_t clocksource_mmio_readw_up(struct clocksource *c)
 {
 	return (cycle_t)ioread16(to_mmio_clksrc(c)->reg);
 }
 
+cycle_t clocksource_mmio_readw_down_be(struct clocksource *c)
+{
+	return ~(cycle_t)ioread16be(to_mmio_clksrc(c)->reg) & c->mask;
+}
+
 cycle_t clocksource_mmio_readw_down(struct clocksource *c)
 {
 	return ~(cycle_t)ioread16(to_mmio_clksrc(c)->reg) & c->mask;
diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h
index 653f0e2..931a037 100644
--- a/include/linux/clocksource.h
+++ b/include/linux/clocksource.h
@@ -327,9 +327,13 @@ static inline void __clocksource_updatefreq_khz(struct clocksource *cs, u32 khz)
 
 extern int timekeeping_notify(struct clocksource *clock);
 
+extern cycle_t clocksource_mmio_readl_up_be(struct clocksource *);
 extern cycle_t clocksource_mmio_readl_up(struct clocksource *);
+extern cycle_t clocksource_mmio_readl_down_be(struct clocksource *);
 extern cycle_t clocksource_mmio_readl_down(struct clocksource *);
+extern cycle_t clocksource_mmio_readw_up_be(struct clocksource *);
 extern cycle_t clocksource_mmio_readw_up(struct clocksource *);
+extern cycle_t clocksource_mmio_readw_down_be(struct clocksource *);
 extern cycle_t clocksource_mmio_readw_down(struct clocksource *);
 
 extern int clocksource_mmio_init(void __iomem *, const char *,
-- 
2.1.0.27.g96db324

Re: [PATCH v3 1/2] clocksource: Update to the newer memory functions.

[Thomas Gleixner]

On Tue, 30 Sep 2014, Xiubo Li wrote:

> Use ioread{16,32} instead of read{w,l}_relaxed.
> 
> For read{w,l}_relaxed accessor, if one arch has its own defination,
> then used it. Or will use the generic one, which will be read as LE
> endian as default.
>  
> For some ARCHes, such PowerPC, if using the clocksource mmio, the
> read{w,l}_relaxed will couldn't be used directly. There must be its
> own BE accessor in its individual clocksource driver.
>
> The ioread{16,32} will be read as LE endian as default if the arch
> hasn't any defination of it too, like read{w,l}_relaxed do.
> 
> To prepare supporting the BE APIs and to be more unified with them,
> using ioread{16,32}[be] is a good chioce here.

This is complete nonsense. I asked you to split this change out and
provide a very reasonable changelog for it in the hope that you would
start to think about what the difference between read*, read*_relaxed
and ioread* are.

I bet you did not even try to think about it for a split second.

You just mechanically split out the changes and dumped your completely
incompetent opinion how these interfaces work and why they are there
in the first place into a uncomprehensible changelog.

Try again, once you understood the semantics of the interfaces and a
proper explanation why inflicting a massive performance hit on most
existing users seems to be a brilliant idea.

Once you've done that use that acquired knowledge to look at your
proposed BE functions and figure out why they are wrong as well.

Thanks,

	tglx