From: Bryan Wu <bryan.wu@analog.com>

Subject: [PATCH] Blackfin I2C/TWI driver: update for 2.6.24 merge windows
(Because "i2c-bfin-twi: Remove useless twi_lock mutex" patch was merged,

 my previous 2 twi patch can not be applied. This is the latest one)
 - Add repeat start feature to avoid break of a bundle of i2c master xfer operation

   Create a new mode TWI_I2C_MODE_REPEAT

   No change to smbus operation
 - Add platform_resource interface to support multi-port TWI controllers
Signed-off-by: Sonic Zhang <sonic.zhang@analog.com>

Signed-off-by: Bryan Wu <bryan.wu@analog.com>

---

 drivers/i2c/busses/i2c-bfin-twi.c |  414 +++++++++++++++++++++++--------------

 1 files changed, 263 insertions(+), 151 deletions(-)
diff --git a/drivers/i2c/busses/i2c-bfin-twi.c b/drivers/i2c/busses/i2c-bfin-twi.c

index 67224a4..acae7ef 100644

--- a/drivers/i2c/busses/i2c-bfin-twi.c

+++ b/drivers/i2c/busses/i2c-bfin-twi.c

@@ -7,6 +7,10 @@

  *

  * Copyright (c) 2005-2007 Analog Devices, Inc.

  *

+ * Modified:

+ *	Aug 01, 2007 add platform_resource interface to support multi-port

+ *		     TWI controllers. (Bryan Wu <bryan.wu@analog.com>)

+ *

  * This program is free software; you can redistribute it and/or modify

  * it under the terms of the GNU General Public License as published by

  * the Free Software Foundation; either version 2 of the License, or

@@ -42,6 +46,7 @@

 #define TWI_I2C_MODE_STANDARD		0x01

 #define TWI_I2C_MODE_STANDARDSUB	0x02

 #define TWI_I2C_MODE_COMBINED		0x04

+#define TWI_I2C_MODE_REPEAT		0x08
 struct bfin_twi_iface {

 	int			irq;

@@ -58,39 +63,69 @@ struct bfin_twi_iface {

 	struct timer_list	timeout_timer;

 	struct i2c_adapter	adap;

 	struct completion	complete;

+	struct i2c_msg 		*pmsg;

+	int			msg_num;

+	int			cur_msg;

+	u32			regs_base;

 };
-static struct bfin_twi_iface twi_iface;

+

+#define DEFINE_TWI_REG(reg, off) \

+static inline u16 read_##reg(struct bfin_twi_iface *iface) \

+	{ return bfin_read16(iface->regs_base + ...
[ message continues ]

Hi Jean,
Could you kindly review this patch?
Thanks a lot

-Bryan
-

Hi Bryan,
I'm pretty busy these days, I don't have much spare time for reviews.
BTW, as a rule of thumb, I am ignoring patches that are sent to the

LKML in addition to the i2c list. If you think that your patch is so

important that it has to be send to a list with over 4500 subscribers

that sees 120.000 messages each year, then who am I to dare to comment

on it?
If you want me to consider your patches as something that needs my

attention, send them to the i2c list only, do not add LKML. This is

general advice for everyone sending i2c patches, not just you, of

course.
Back to your patch... Originally you posted two patches, and I see that

they are now merged into a single patch, why? It's much better to have

separate patches, it makes reviews and testing way easier. If you

provide separate patches I'll try to find some time to review them.
Note that this won't be for 2.6.24 anyway, it's too late for that, I've

already sent my i2c patches to Linus. From now on, only bugfixes can go

in 2.6.24, improvements will have to wait for 2.6.25.
--

Jean Delvare

-

Why ? The statement would have made sense if the i2c list was not CC 'd,

but stating that if LK is CC 'd additionally sounds  ...

-

Disagree.  I don't think there's any benefit to anyone for developers to

hide their stuff on remote mailing lists.  Copying lkml increases the

changes that someone will spot a bug or some improvement and it generally

keeps people informed as to what's going on.
yeah, 120,000 messages/year.  But a lot of them are just noise.  Patches

are important.
And I really don't understand this huffy attitude and putting more

roadblocks in the way of our contributors.  People like Bryan are doing a

lot of work to improve Linux and most people are (or should be) grateful

for that and should be prepared to go the extra yard to help them.
-

Hi Andrew,
Sending spam to millions of people increases the chances than someone
Lot of noise, yes. Counting 1 second on average to process posts you're

not interested in, that's 33 hours a year. This is completely insane

and that's why I unsubscribed from LKML long ago. And the situation is

only getting worse every year. You may fail to realize the problem just
Quite on the contrary, I am trying to educate people into targeting

their posts to the right lists so that they get more appropriate

responses in a shorter time.
In a world where we would have infinite work power and time would grow

on trees, I would be fine with everyone sending everything to every

mailing list. But in the real world, time spent to do one thing isn't

spent on another. If we don't learn, collectively, to target our posts

to the right people, then we end up wasting a significant amount of our
My comment is general and absolutely not targeted at Bryan personally.

I thought it was clear. I fully agree that Bryan is a really pleasant

and easy person to work with and I do appreciate his contributions.
--

Jean Delvare

-

I see. Actually this is the first time I was told about it.

I know it missed the merge window.
Thanks again

-Bryan Wu

-

I probably had not been clear about it so far. I've added it to my

"Linux 2.6 I2C development FAQ" now.
--

Jean Delvare

-