2011-03-11 17:23:45

by Jean-François Dagenais

[permalink] [raw]
Subject: [PATCH] Add 1-wire (w1) reset and resume command API support

This is needed for my next patch about DS2408

>From aba11ba9d50c0fa7a29c81ecb81c13afb986b042 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jean-Fran=C3=A7ois=20Dagenais?= <[email protected]>
Date: Thu, 10 Mar 2011 11:04:16 -0500
Subject: [PATCH] Add 1-wire (w1) reset and resume command API support
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
To: [email protected]
Cc: [email protected]

Signed-off-by: Jean-François Dagenais <[email protected]>
---
drivers/w1/w1.h | 2 ++
drivers/w1/w1_io.c | 26 ++++++++++++++++++++++++++
2 files changed, 28 insertions(+), 0 deletions(-)

diff --git a/drivers/w1/w1.h b/drivers/w1/w1.h
index d8a9709..f804eba 100644
--- a/drivers/w1/w1.h
+++ b/drivers/w1/w1.h
@@ -55,6 +55,7 @@ struct w1_reg_num
#define W1_READ_ROM 0x33
#define W1_READ_PSUPPLY 0xB4
#define W1_MATCH_ROM 0x55
+#define W1_RESUME_CMD 0xA5

#define W1_SLAVE_ACTIVE 0

@@ -213,6 +214,7 @@ void w1_write_block(struct w1_master *, const u8 *,
int);
void w1_touch_block(struct w1_master *, u8 *, int);
u8 w1_read_block(struct w1_master *, u8 *, int);
int w1_reset_select_slave(struct w1_slave *sl);
+int w1_reset_resume_command(struct w1_master *);
void w1_next_pullup(struct w1_master *, int);

static inline struct w1_slave* dev_to_w1_slave(struct device *dev)
diff --git a/drivers/w1/w1_io.c b/drivers/w1/w1_io.c
index 3ebe972..8e8b64c 100644
--- a/drivers/w1/w1_io.c
+++ b/drivers/w1/w1_io.c
@@ -390,6 +390,32 @@ int w1_reset_select_slave(struct w1_slave *sl)
EXPORT_SYMBOL_GPL(w1_reset_select_slave);

/**
+ * When the workflow with a slave amongst many requires several
+ * successive commands a reset between each, this function is similar
+ * to doing a reset then a match ROM for the last matched ROM. The
+ * advantage being that the matched ROM step is skipped in favor of the
+ * resume command. The slave must support the command of course.
+ *
+ * If the bus has only one slave, traditionnaly the match ROM is
skipped
+ * and a "SKIP ROM" is done for efficiency. On multi-slave busses, this
+ * doesn't work of course, but the resume command is the next best
thing.
+ *
+ * The w1 master lock must be held.
+ *
+ * @param dev the master device
+ */
+int w1_reset_resume_command(struct w1_master *dev)
+{
+ if (w1_reset_bus(dev))
+ return -1;
+
+ /* This will make only the last matched slave perform a skip ROM. */
+ w1_write_8(dev, W1_RESUME_CMD);
+ return 0;
+}
+EXPORT_SYMBOL_GPL(w1_reset_resume_command);
+
+/**
* Put out a strong pull-up of the specified duration after the next
write
* operation. Not all hardware supports strong pullups. Hardware that
* doesn't support strong pullups will sleep for the given time after
the
--
1.7.4.1


2011-03-13 00:27:57

by Jean-François Dagenais

[permalink] [raw]
Subject: Re: [PATCH] Add 1-wire (w1) reset and resume command API support

please disregard this patch in favor of my patch trio ([w1 PATCHES 1-2-3/3] ) with header note called "Add 1-wire (w1) support for DS2408 and multi-slave search in ds1wm"

On Mar 11, 2011, at 12:23 PM, Jean-Francois Dagenais wrote:

> This is needed for my next patch about DS2408
>
> From aba11ba9d50c0fa7a29c81ecb81c13afb986b042 Mon Sep 17 00:00:00 2001
> From: =?UTF-8?q?Jean-Fran=C3=A7ois=20Dagenais?= <[email protected]>
> Date: Thu, 10 Mar 2011 11:04:16 -0500
> Subject: [PATCH] Add 1-wire (w1) reset and resume command API support
> MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
> To: [email protected]
> Cc: [email protected]
>
> Signed-off-by: Jean-Fran?ois Dagenais <[email protected]>
> ---
> drivers/w1/w1.h | 2 ++
> drivers/w1/w1_io.c | 26 ++++++++++++++++++++++++++
> 2 files changed, 28 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/w1/w1.h b/drivers/w1/w1.h
> index d8a9709..f804eba 100644
> --- a/drivers/w1/w1.h
> +++ b/drivers/w1/w1.h
> @@ -55,6 +55,7 @@ struct w1_reg_num
> #define W1_READ_ROM 0x33
> #define W1_READ_PSUPPLY 0xB4
> #define W1_MATCH_ROM 0x55
> +#define W1_RESUME_CMD 0xA5
>
> #define W1_SLAVE_ACTIVE 0
>
> @@ -213,6 +214,7 @@ void w1_write_block(struct w1_master *, const u8 *,
> int);
> void w1_touch_block(struct w1_master *, u8 *, int);
> u8 w1_read_block(struct w1_master *, u8 *, int);
> int w1_reset_select_slave(struct w1_slave *sl);
> +int w1_reset_resume_command(struct w1_master *);
> void w1_next_pullup(struct w1_master *, int);
>
> static inline struct w1_slave* dev_to_w1_slave(struct device *dev)
> diff --git a/drivers/w1/w1_io.c b/drivers/w1/w1_io.c
> index 3ebe972..8e8b64c 100644
> --- a/drivers/w1/w1_io.c
> +++ b/drivers/w1/w1_io.c
> @@ -390,6 +390,32 @@ int w1_reset_select_slave(struct w1_slave *sl)
> EXPORT_SYMBOL_GPL(w1_reset_select_slave);
>
> /**
> + * When the workflow with a slave amongst many requires several
> + * successive commands a reset between each, this function is similar
> + * to doing a reset then a match ROM for the last matched ROM. The
> + * advantage being that the matched ROM step is skipped in favor of the
> + * resume command. The slave must support the command of course.
> + *
> + * If the bus has only one slave, traditionnaly the match ROM is
> skipped
> + * and a "SKIP ROM" is done for efficiency. On multi-slave busses, this
> + * doesn't work of course, but the resume command is the next best
> thing.
> + *
> + * The w1 master lock must be held.
> + *
> + * @param dev the master device
> + */
> +int w1_reset_resume_command(struct w1_master *dev)
> +{
> + if (w1_reset_bus(dev))
> + return -1;
> +
> + /* This will make only the last matched slave perform a skip ROM. */
> + w1_write_8(dev, W1_RESUME_CMD);
> + return 0;
> +}
> +EXPORT_SYMBOL_GPL(w1_reset_resume_command);
> +
> +/**
> * Put out a strong pull-up of the specified duration after the next
> write
> * operation. Not all hardware supports strong pullups. Hardware that
> * doesn't support strong pullups will sleep for the given time after
> the
> --
> 1.7.4.1
>
>