[PATCH] proc: fix documentation for /proc/stat, line "intr"

[PATCH] proc: fix documentation for /proc/stat, line "intr"

[Jan Moskyto Matejka]

The sum at the beginning of line "intr" includes also unnumbered interrupts.
It implies that the sum at the beginning isn't the sum of the remainder of the
line, not even an estimation.

Fixed the documentation to mention that.

Signed-off-by: Jan Moskyto Matejka <mq@suse.cz>
---
 Documentation/filesystems/proc.txt | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt
index f00bee1..4e6f9d0 100644
--- a/Documentation/filesystems/proc.txt
+++ b/Documentation/filesystems/proc.txt
@@ -1245,8 +1245,9 @@ second).  The meanings of the columns are as follows, from left to right:
 
 The "intr" line gives counts of interrupts  serviced since boot time, for each
 of the  possible system interrupts.   The first  column  is the  total of  all
-interrupts serviced; each  subsequent column is the  total for that particular
-interrupt.
+interrupts serviced  including  unnumbered  architecture specific  interrupts;
+each  subsequent column is the  total for that particular numbered interrupt.
+Unnumbered interrupts are not shown, only summed into the total.
 
 The "ctxt" line gives the total number of context switches across all CPUs.
 
-- 
1.8.4.5

Re: [PATCH] proc: fix documentation for /proc/stat, line "intr"

[Michael Kerrisk (man-pages)]

On 04/22/2014 05:29 PM, Jan Moskyto Matejka wrote:
> The sum at the beginning of line "intr" includes also unnumbered interrupts.
> It implies that the sum at the beginning isn't the sum of the remainder of the
> line, not even an estimation.
> 
> Fixed the documentation to mention that.

I'm not really doubting what you claim, Jan, but it would be really
helpful if you provided some data to support the proposed change
(Kernel source references, or notes about testing you've done, or 
pointers to mailing list threads, authoritative note that you are
the maintainer of some relevant kernel subsystem...). Do you have
something like that? (It's also useful for the change log.)

Thanks,

Michael

> Signed-off-by: Jan Moskyto Matejka <mq@suse.cz>
> ---
>  Documentation/filesystems/proc.txt | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt
> index f00bee1..4e6f9d0 100644
> --- a/Documentation/filesystems/proc.txt
> +++ b/Documentation/filesystems/proc.txt
> @@ -1245,8 +1245,9 @@ second).  The meanings of the columns are as follows, from left to right:
>  
>  The "intr" line gives counts of interrupts  serviced since boot time, for each
>  of the  possible system interrupts.   The first  column  is the  total of  all
> -interrupts serviced; each  subsequent column is the  total for that particular
> -interrupt.
> +interrupts serviced  including  unnumbered  architecture specific  interrupts;
> +each  subsequent column is the  total for that particular numbered interrupt.
> +Unnumbered interrupts are not shown, only summed into the total.
>  
>  The "ctxt" line gives the total number of context switches across all CPUs.
>  
> 


-- 
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
Linux/UNIX System Programming Training: http://man7.org/training/

Re: [PATCH] proc: fix documentation for /proc/stat, line "intr"

[Randy Dunlap]

On 04/22/14 11:19, Michael Kerrisk (man-pages) wrote:
> On 04/22/2014 05:29 PM, Jan Moskyto Matejka wrote:
>> The sum at the beginning of line "intr" includes also unnumbered interrupts.
>> It implies that the sum at the beginning isn't the sum of the remainder of the
>> line, not even an estimation.
>>
>> Fixed the documentation to mention that.
> 
> I'm not really doubting what you claim, Jan, but it would be really
> helpful if you provided some data to support the proposed change
> (Kernel source references, or notes about testing you've done, or 
> pointers to mailing list threads, authoritative note that you are
> the maintainer of some relevant kernel subsystem...). Do you have
> something like that? (It's also useful for the change log.)

I wouldn't mind more of a changelog, but the additional text is
certainly correct.

Thanks.  I'll apply the patch, but more changelog comments are welcome.

> Thanks,
> 
> Michael
> 
>> Signed-off-by: Jan Moskyto Matejka <mq@suse.cz>
>> ---
>>  Documentation/filesystems/proc.txt | 5 +++--
>>  1 file changed, 3 insertions(+), 2 deletions(-)
>>
>> diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt
>> index f00bee1..4e6f9d0 100644
>> --- a/Documentation/filesystems/proc.txt
>> +++ b/Documentation/filesystems/proc.txt
>> @@ -1245,8 +1245,9 @@ second).  The meanings of the columns are as follows, from left to right:
>>  
>>  The "intr" line gives counts of interrupts  serviced since boot time, for each
>>  of the  possible system interrupts.   The first  column  is the  total of  all
>> -interrupts serviced; each  subsequent column is the  total for that particular
>> -interrupt.
>> +interrupts serviced  including  unnumbered  architecture specific  interrupts;
>> +each  subsequent column is the  total for that particular numbered interrupt.
>> +Unnumbered interrupts are not shown, only summed into the total.
>>  
>>  The "ctxt" line gives the total number of context switches across all CPUs.
>>  
>>
> 
> 


-- 
~Randy

Re: [PATCH] proc: fix documentation for /proc/stat, line "intr"

[Michael Kerrisk (man-pages)]

On 04/23/2014 12:35 AM, Randy Dunlap wrote:
> On 04/22/14 11:19, Michael Kerrisk (man-pages) wrote:
>> On 04/22/2014 05:29 PM, Jan Moskyto Matejka wrote:
>>> The sum at the beginning of line "intr" includes also unnumbered interrupts.
>>> It implies that the sum at the beginning isn't the sum of the remainder of the
>>> line, not even an estimation.
>>>
>>> Fixed the documentation to mention that.
>>
>> I'm not really doubting what you claim, Jan, but it would be really
>> helpful if you provided some data to support the proposed change
>> (Kernel source references, or notes about testing you've done, or 
>> pointers to mailing list threads, authoritative note that you are
>> the maintainer of some relevant kernel subsystem...). Do you have
>> something like that? (It's also useful for the change log.)
> 
> I wouldn't mind more of a changelog, but the additional text is
> certainly correct.

Sorry -- I wrote the reply in a hurry, having first mistaken the patch
as being something to apply in man-pages. I should have looked more
closely ;-).

> Thanks.  I'll apply the patch, but more changelog comments are welcome.

Amen.

Cheers,

Michael


>>> Signed-off-by: Jan Moskyto Matejka <mq@suse.cz>
>>> ---
>>>  Documentation/filesystems/proc.txt | 5 +++--
>>>  1 file changed, 3 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt
>>> index f00bee1..4e6f9d0 100644
>>> --- a/Documentation/filesystems/proc.txt
>>> +++ b/Documentation/filesystems/proc.txt
>>> @@ -1245,8 +1245,9 @@ second).  The meanings of the columns are as follows, from left to right:
>>>  
>>>  The "intr" line gives counts of interrupts  serviced since boot time, for each
>>>  of the  possible system interrupts.   The first  column  is the  total of  all
>>> -interrupts serviced; each  subsequent column is the  total for that particular
>>> -interrupt.
>>> +interrupts serviced  including  unnumbered  architecture specific  interrupts;
>>> +each  subsequent column is the  total for that particular numbered interrupt.
>>> +Unnumbered interrupts are not shown, only summed into the total.
>>>  
>>>  The "ctxt" line gives the total number of context switches across all CPUs.
>>>  
>>>
>>
>>
> 
> 


-- 
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
Linux/UNIX System Programming Training: http://man7.org/training/

Re: [PATCH] proc: fix documentation for /proc/stat, line "intr"

[Michael Kerrisk (man-pages)]

On Wed, Apr 23, 2014 at 12:35 AM, Randy Dunlap <rdunlap@infradead.org> wrote:
> On 04/22/14 11:19, Michael Kerrisk (man-pages) wrote:
>> On 04/22/2014 05:29 PM, Jan Moskyto Matejka wrote:
>>> The sum at the beginning of line "intr" includes also unnumbered interrupts.
>>> It implies that the sum at the beginning isn't the sum of the remainder of the
>>> line, not even an estimation.
>>>
>>> Fixed the documentation to mention that.
>>
>> I'm not really doubting what you claim, Jan, but it would be really
>> helpful if you provided some data to support the proposed change
>> (Kernel source references, or notes about testing you've done, or
>> pointers to mailing list threads, authoritative note that you are
>> the maintainer of some relevant kernel subsystem...). Do you have
>> something like that? (It's also useful for the change log.)
>
> I wouldn't mind more of a changelog, but the additional text is
> certainly correct.
>
> Thanks.  I'll apply the patch, but more changelog comments are welcome.

And I see now that this text could also reasonably fit into the
proc(5) man page. So I've added, with a Reviewed-by: tag for Randy.
Thanks for the text, Jan.

Cheers,

Michael


>>> Signed-off-by: Jan Moskyto Matejka <mq@suse.cz>
>>> ---
>>>  Documentation/filesystems/proc.txt | 5 +++--
>>>  1 file changed, 3 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt
>>> index f00bee1..4e6f9d0 100644
>>> --- a/Documentation/filesystems/proc.txt
>>> +++ b/Documentation/filesystems/proc.txt
>>> @@ -1245,8 +1245,9 @@ second).  The meanings of the columns are as follows, from left to right:
>>>
>>>  The "intr" line gives counts of interrupts  serviced since boot time, for each
>>>  of the  possible system interrupts.   The first  column  is the  total of  all
>>> -interrupts serviced; each  subsequent column is the  total for that particular
>>> -interrupt.
>>> +interrupts serviced  including  unnumbered  architecture specific  interrupts;
>>> +each  subsequent column is the  total for that particular numbered interrupt.
>>> +Unnumbered interrupts are not shown, only summed into the total.
>>>
>>>  The "ctxt" line gives the total number of context switches across all CPUs.
>>>
>>>
>>
>>
>
>
> --
> ~Randy



-- 
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
Linux/UNIX System Programming Training: http://man7.org/training/

Re: [PATCH] proc: fix documentation for /proc/stat, line "intr"

[Jan Moskyto Matejka]

[-- Attachment #1: Type: text/plain, Size: 1248 bytes --]

n Tue, Apr 22, 2014 at 03:35:39PM -0700, Randy Dunlap wrote:
> On 04/22/14 11:19, Michael Kerrisk (man-pages) wrote:
> > On 04/22/2014 05:29 PM, Jan Moskyto Matejka wrote:
> >> The sum at the beginning of line "intr" includes also unnumbered interrupts.
> >> It implies that the sum at the beginning isn't the sum of the remainder of the
> >> line, not even an estimation.
> >>
> >> Fixed the documentation to mention that.
> > 
> > I'm not really doubting what you claim, Jan, but it would be really
> > helpful if you provided some data to support the proposed change
> > (Kernel source references, or notes about testing you've done, or 
> > pointers to mailing list threads, authoritative note that you are
> > the maintainer of some relevant kernel subsystem...). Do you have
> > something like that? (It's also useful for the change log.)
> 
> I wouldn't mind more of a changelog, but the additional text is
> certainly correct.
> 
> Thanks.  I'll apply the patch, but more changelog comments are welcome.

Thank you for applying.

You may add following line to the changelog:
This behaviour was added to /proc/stat in commit a2eddfa95919a730e0e5ed17e9c303fe5ba249cd.

MQ
-- 
Jan Matějka <mq@suse.cz>
SUSE Labs

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [PATCH] proc: fix documentation for /proc/stat, line "intr"

[Michael Kerrisk (man-pages)]

On 04/23/2014 11:48 AM, Jan Moskyto Matejka wrote:
> n Tue, Apr 22, 2014 at 03:35:39PM -0700, Randy Dunlap wrote:
>> On 04/22/14 11:19, Michael Kerrisk (man-pages) wrote:
>>> On 04/22/2014 05:29 PM, Jan Moskyto Matejka wrote:
>>>> The sum at the beginning of line "intr" includes also unnumbered interrupts.
>>>> It implies that the sum at the beginning isn't the sum of the remainder of the
>>>> line, not even an estimation.
>>>>
>>>> Fixed the documentation to mention that.
>>>
>>> I'm not really doubting what you claim, Jan, but it would be really
>>> helpful if you provided some data to support the proposed change
>>> (Kernel source references, or notes about testing you've done, or 
>>> pointers to mailing list threads, authoritative note that you are
>>> the maintainer of some relevant kernel subsystem...). Do you have
>>> something like that? (It's also useful for the change log.)
>>
>> I wouldn't mind more of a changelog, but the additional text is
>> certainly correct.
>>
>> Thanks.  I'll apply the patch, but more changelog comments are welcome.
> 
> Thank you for applying.
> 
> You may add following line to the changelog:
> This behaviour was added to /proc/stat in commit a2eddfa95919a730e0e5ed17e9c303fe5ba249cd.

Thanks, Jan -- that's indeed helpful for the man-pages changelog!


-- 
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
Linux/UNIX System Programming Training: http://man7.org/training/

Re: [PATCH] proc: fix documentation for /proc/stat, line "intr"

[Randy Dunlap]

On 04/23/14 02:48, Jan Moskyto Matejka wrote:
> n Tue, Apr 22, 2014 at 03:35:39PM -0700, Randy Dunlap wrote:
>> On 04/22/14 11:19, Michael Kerrisk (man-pages) wrote:
>>> On 04/22/2014 05:29 PM, Jan Moskyto Matejka wrote:
>>>> The sum at the beginning of line "intr" includes also unnumbered interrupts.
>>>> It implies that the sum at the beginning isn't the sum of the remainder of the
>>>> line, not even an estimation.
>>>>
>>>> Fixed the documentation to mention that.
>>>
>>> I'm not really doubting what you claim, Jan, but it would be really
>>> helpful if you provided some data to support the proposed change
>>> (Kernel source references, or notes about testing you've done, or 
>>> pointers to mailing list threads, authoritative note that you are
>>> the maintainer of some relevant kernel subsystem...). Do you have
>>> something like that? (It's also useful for the change log.)
>>
>> I wouldn't mind more of a changelog, but the additional text is
>> certainly correct.
>>
>> Thanks.  I'll apply the patch, but more changelog comments are welcome.
> 
> Thank you for applying.
> 
> You may add following line to the changelog:
> This behaviour was added to /proc/stat in commit a2eddfa95919a730e0e5ed17e9c303fe5ba249cd.
> 
> MQ
> 

I added that and applied your patch.  Thanks.

-- 
~Randy