[mpiwg-tools] ticket 405: MPI_CHAR as string

[Jeff Squyres (jsquyres)]

Sorry -- nit-picking on my own text:

-----
The use of the datatype MPI_CHAR in the MPI tool information interface implies a null-terminated character array, i.e., a string in the C language. If a variable has type MPI_CHAR, the value of the count parameter returned by MPI_T_CVAR_HANDLE_ALLOC and MPI_T_PVAR_HANDLE_ALLOC must be large enough to include any valid value, including its null-terminating character. The contents of returned MPI_CHAR arrays are only defined from index 0 through the location of the first null-terminating character.
-----

In the 2nd and 3rd sentence, we refer to the "null-terminating character".  Is that correct?  Isn't it just a "null character"?

We use "null-terminated character array" in the first sentence, and that seems fine. But "null-terminated character" means "a character terminated by a null", which doesn't seem to make sense.

In 14.3.3 ("Convention for Returning Strings"), we refer to the "null terminator" and "terminating null character").  

I therefore think we should slightly change the last 2 sentences to:

-----
The use of the datatype MPI_CHAR in the MPI tool information interface implies a null-terminated character array, i.e., a string in the C language. If a variable has type MPI_CHAR, the value of the count parameter returned by MPI_T_CVAR_HANDLE_ALLOC and MPI_T_PVAR_HANDLE_ALLOC must be large enough to include any valid value, including its ***terminating null*** character. The contents of returned MPI_CHAR arrays are only defined from index 0 through the location of the first ***null*** character.
-----




On Feb 14, 2014, at 6:16 PM, "Mohror, Kathryn" <mohror1 at llnl.gov> wrote:

> Thanks for the feedback. I have attached a new draft for this ticket and incorporated your changes. Please give me any additional comments by the start of business Monday.
> 
> Kathryn
> 
> On Feb 14, 2014, at 7:23 AM, Dave Goodell (dgoodell) <dgoodell at cisco.com> wrote:
> 
>> +1 to almost everything that Jeff said.  My only quibble would be to change the last sentence in his text from:
>> 
>> The contents of returned MPI_CHAR values are only defined from index 0 through the location of the first null-terminating character.
>> 
>> To:
>> 
>> The contents of returned MPI_CHAR ***arrays*** are only defined from index 0 through the location of the first null-terminating character.
>> 
>> -Dave
>> 
>> On Feb 14, 2014, at 7:27 AM, Jeff Squyres (jsquyres) <jsquyres at cisco.com> wrote:
>> 
>>> Here's the new text:
>>> 
>>> -----
>>> The use of the datatype MPI_CHAR in the MPI tool information interface implies a null-terminated character array, i.e. a string in the C language. If a variable has type MPI_CHAR, the value of the count parameter returned by MPI_T_CVAR_HANDLE_ALLOC and MPI_T_PVAR_HANDLE_ALLOC will include the null-terminating character. The value returned in count must be large enough to contain any valid value for the given variable. Arrays of type MPI_CHAR are only valid to the first null character.
>>> -----
>>> 
>>> There's a missing "," after "i.e." -- it should be "i.e.,"
>>> 
>>> I'd make some other minor changes:
>>> 
>>> The use of the datatype MPI_CHAR in the MPI tool information interface implies a null-terminated character array, i.e., a string in the C language. If a variable has type MPI_CHAR, the value of the count parameter returned by MPI_T_CVAR_HANDLE_ALLOC and MPI_T_PVAR_HANDLE_ALLOC ***must be large enough to include any valid value, including its null-terminating character.***  ***The contents of returned MPI_CHAR values are only defined from index 0 through the location of the first null-terminating character.***
>>> 
>>> 
>>> 
>>> 
>>> 
>>> On Feb 13, 2014, at 3:24 PM, Kathryn Mohror <kathryn at llnl.gov> wrote:
>>> 
>>>> Hi all,
>>>> 
>>>> Please have a look at this since we  have to decide if it's ready for a reading, and submit it to the forum for review on Monday if we do.
>>>> 
>>>> Kathryn
>>>> 
>>>> On Feb 7, 2014, at 1:05 PM, "Mohror, Kathryn" <mohror1 at llnl.gov> wrote:
>>>> 
>>>>> Hi all,
>>>>> 
>>>>> As we discussed in our call, I have drafted new text for ticket 405 which clarifies our intent for the MPI_CHAR datatype to represent strings in the tools interface. Please take a look at the red text for ticket 405 in the attached document and send your comments.
>>>>> 
>>>>> Kathryn
>>>>> 
>>>>> 
>>>>> _________________________________________________________________
>>>>> Kathryn Mohror, kathryn at llnl.gov, http://scalability.llnl.gov/
>>>>> Scalability Team @ Lawrence Livermore National Laboratory, Livermore, CA, USA
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> <tools-3.pdf>_______________________________________________
>>>>> mpiwg-tools mailing list
>>>>> mpiwg-tools at lists.mpi-forum.org
>>>>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-tools
>>>> 
>>>> _________________________________________________________________
>>>> Kathryn Mohror, kathryn at llnl.gov, http://scalability.llnl.gov/
>>>> Scalability Team @ Lawrence Livermore National Laboratory, Livermore, CA, USA
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> _______________________________________________
>>>> mpiwg-tools mailing list
>>>> mpiwg-tools at lists.mpi-forum.org
>>>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-tools
>>> 
>>> 
>>> -- 
>>> Jeff Squyres
>>> jsquyres at cisco.com
>>> For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/
>>> 
>>> _______________________________________________
>>> mpiwg-tools mailing list
>>> mpiwg-tools at lists.mpi-forum.org
>>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-tools
>> 
>> _______________________________________________
>> mpiwg-tools mailing list
>> mpiwg-tools at lists.mpi-forum.org
>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-tools
> 
> _________________________________________________________________
> Kathryn Mohror, kathryn at llnl.gov, http://scalability.llnl.gov/
> Scalability Team @ Lawrence Livermore National Laboratory, Livermore, CA, USA
> 
> 
> 
> 
> 
> 
> 
> 
> <tools-3.pdf>_______________________________________________
> mpiwg-tools mailing list
> mpiwg-tools at lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-tools


-- 
Jeff Squyres
jsquyres at cisco.com
For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/