2010-12-15 10:50:43

by Nicolas Ferre

[permalink] [raw]
Subject: [PATCH] MTD: m25p80: add debugging trace in sst_write()

Add a DEBUG(MTD_DEBUG_LEVEL2, ..) trace at beginning of sst_write() function as
it is done in m25p80_write() function.

Signed-off-by: Nicolas Ferre <[email protected]>
---
drivers/mtd/devices/m25p80.c | 4 ++++
1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/drivers/mtd/devices/m25p80.c b/drivers/mtd/devices/m25p80.c
index bf5a002..e6b5707 100644
--- a/drivers/mtd/devices/m25p80.c
+++ b/drivers/mtd/devices/m25p80.c
@@ -482,6 +482,10 @@ static int sst_write(struct mtd_info *mtd, loff_t to, size_t len,
size_t actual;
int cmd_sz, ret;

+ DEBUG(MTD_DEBUG_LEVEL2, "%s: %s %s 0x%08x, len %zd\n",
+ dev_name(&flash->spi->dev), __func__, "to",
+ (u32)to, len);
+
*retlen = 0;

/* sanity checks */
--
1.7.3


2010-12-15 13:26:31

by Sergei Shtylyov

[permalink] [raw]
Subject: Re: [PATCH] MTD: m25p80: add debugging trace in sst_write()

Hello.

On 15-12-2010 14:59, Nicolas Ferre wrote:

> Add a DEBUG(MTD_DEBUG_LEVEL2, ..) trace at beginning of sst_write() function as
> it is done in m25p80_write() function.

> Signed-off-by: Nicolas Ferre<[email protected]>
[...]

> diff --git a/drivers/mtd/devices/m25p80.c b/drivers/mtd/devices/m25p80.c
> index bf5a002..e6b5707 100644
> --- a/drivers/mtd/devices/m25p80.c
> +++ b/drivers/mtd/devices/m25p80.c
> @@ -482,6 +482,10 @@ static int sst_write(struct mtd_info *mtd, loff_t to, size_t len,
> size_t actual;
> int cmd_sz, ret;
>
> + DEBUG(MTD_DEBUG_LEVEL2, "%s: %s %s 0x%08x, len %zd\n",
> + dev_name(&flash->spi->dev), __func__, "to",

What's the point of printing "to" as variable? :-)

> + (u32)to, len);
> +

WBR, Sergei

2010-12-15 13:44:05

by Nicolas Ferre

[permalink] [raw]
Subject: Re: [PATCH] MTD: m25p80: add debugging trace in sst_write()

Le 15/12/2010 14:25, Sergei Shtylyov :
> Hello.
>
> On 15-12-2010 14:59, Nicolas Ferre wrote:
>
>> Add a DEBUG(MTD_DEBUG_LEVEL2, ..) trace at beginning of sst_write()
>> function as
>> it is done in m25p80_write() function.
>
>> Signed-off-by: Nicolas Ferre<[email protected]>
> [...]
>
>> diff --git a/drivers/mtd/devices/m25p80.c b/drivers/mtd/devices/m25p80.c
>> index bf5a002..e6b5707 100644
>> --- a/drivers/mtd/devices/m25p80.c
>> +++ b/drivers/mtd/devices/m25p80.c
>> @@ -482,6 +482,10 @@ static int sst_write(struct mtd_info *mtd, loff_t
>> to, size_t len,
>> size_t actual;
>> int cmd_sz, ret;
>>
>> + DEBUG(MTD_DEBUG_LEVEL2, "%s: %s %s 0x%08x, len %zd\n",
>> + dev_name(&flash->spi->dev), __func__, "to",
>
> What's the point of printing "to" as variable? :-)

Good question... Maybe a way to have the same shape of messages...

In my patch it is for sure to comply with this driver way of dealing
with DEBUG messages.
It is "to" in this function like its sister m25p80_write(). But it is
also "from" or "at" in others...

>> + (u32)to, len);
>> +

Cheers,
--
Nicolas Ferre

2010-12-15 16:13:10

by Russell King - ARM Linux

[permalink] [raw]
Subject: Re: [PATCH] MTD: m25p80: add debugging trace in sst_write()

On Wed, Dec 15, 2010 at 04:25:28PM +0300, Sergei Shtylyov wrote:
>> + DEBUG(MTD_DEBUG_LEVEL2, "%s: %s %s 0x%08x, len %zd\n",
>> + dev_name(&flash->spi->dev), __func__, "to",
>
> What's the point of printing "to" as variable? :-)

One valid reason to do this is if you have several formatting strings
all the same. GCC can merge identical strings together.

So, if you have:

"%s: %s %s 0x%08x, len %zd", dev_name(), __func__, "to"
"%s: %s %s 0x%08x, len %zd", dev_name(), __func__, "from"
"%s: %s %s 0x%08x, len %zd", dev_name(), __func__, "foo"

Then you end up with one long string and three short strings instead of

"%s: %s to 0x%08x, len %zd", dev_name(), __func__
"%s: %s from 0x%08x, len %zd", dev_name(), __func__
"%s: %s foo 0x%08x, len %zd", dev_name(), __func__

which'd be three long strings.

2010-12-19 17:01:28

by Artem Bityutskiy

[permalink] [raw]
Subject: Re: [PATCH] MTD: m25p80: add debugging trace in sst_write()

On Wed, 2010-12-15 at 12:59 +0100, Nicolas Ferre wrote:
> Add a DEBUG(MTD_DEBUG_LEVEL2, ..) trace at beginning of sst_write() function as
> it is done in m25p80_write() function.
>
> Signed-off-by: Nicolas Ferre <[email protected]>

So do you do this only becaus it is done in m25p80_write()? This is what
the commit message suggests... Have you found these DEBUG macros useful
for you? (I never found them any useful).

Anyway, I've put this to my l2-mtd-2.6.git tree, thanks.

--
Best Regards,
Artem Bityutskiy (Битюцкий Артём)