From treumann@XXXXXXXXXX Fri Feb 1 16:36:44 2008 In-Reply-To: <002701c8651e$3b849cb0$860add8c@XXXXXXXXXXX> Subject: RE: Fw: [mpi-21] Ballot 4 - MPI_File_get_info To: "Rajeev Thakur" Cc: bronis@XXXXXXXX,"'Bill Gropp'" ,X-Mailer: Lotus Notes Release 7.0 HF277 June 21, 2006 From: Richard Treumann Date: Fri, 1 Feb 2008 17:32:52 -0500 X-MIMETrack: Serialize by Router on D01ML064/01/M/IBM(Release 7.0.2FP2 IGS702FP2HF5|November 8, 2007) at 02/01/2008 17:32:53 MIME-Version: 1.0 Content-type: multipart/related; Boundary="0__=0ABBF971DFE95B228f9e8a93df938690918c0ABBF971DFE95B22" X-Virus-Scanned: by amavisd-new-20030616-p10 (Debian) at mailgw.mcs.anl.gov --0__=0ABBF971DFE95B228f9e8a93df938690918c0ABBF971DFE95B22 Content-type: multipart/alternative; Boundary="1__=0ABBF971DFE95B228f9e8a93df938690918c0ABBF971DFE95B22" --1__=0ABBF971DFE95B228f9e8a93df938690918c0ABBF971DFE95B22 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: quoted-printable Of course the keyword "filename" never gets "interpreted" in any implementation. The description says it is to be ignored if passed in = so it cannot be a "hint" at all. This pseudo hint is a hack that is useful= but also clearly abuses the "hints" intent of the info interface on files. However - I am persuaded that it is legitimate for there to be no (key,= value) pairs to return for MPI_FILE_GET_INFO and therefore a statement about returning MPI_INFO_NULL may be useful to somebody. Have a good weekend all Dick Treumann - MPI Team/TCEM IBM Systems & Technology Group Dept 0lva / MS P963 -- 2455 South Road -- Poughkeepsie, NY 12601 Tele (845) 433-7846 Fax (845) 433-8363 = "Rajeev Thakur" = = To "'Rolf Rabenseifner'" = 02/01/2008 05:03 , Richard = PM Treumann/Poughkeepsie/IBM@IBMUS = = cc , "'Bill Gropp'=" , "'Rob Ross'=" = Subj= ect RE: Fw: [mpi-21] Ballot 4 - = MPI_File_get_info = = = = = = = Also, the first para under Reserved File Hints says "An implementation = is not required to interpret these key values, but if it does interpret th= e key value, it must provide the functionality described." So an implementati= on is free to not support any hint. Rajeev > -----Original Message----- > From: Rolf Rabenseifner [mailto:rabenseifner@XXXXXXXX > Sent: Friday, February 01, 2008 3:42 PM > To: Richard Treumann > Cc: bronis@XXXXXXXXX Bill Gropp; Rob Ross; Rajeev Thakur > Subject: Re: Fw: [mpi-21] Ballot 4 - MPI_File_get_info > > Dick, > > Yes, looks technical okay but strange in the real wolrd. > But I remember, we allowed also unnamed pipes, > tapes, ... (with MPI_MODE_SEQUENTIAL). > > Those might be a candidate for "no filename". > > Best regards > Rolf > > > On Fri, 1 Feb 2008 15:33:55 -0500 > Richard Treumann wrote:> > > >OK - the standard does say that an implementation should > return filename if > >it can. I guess if an implementation is not able to discover > the filename > >by the time the app calls MPI_FILE_GET_INFO there is no > requirement that > >the "filename" hint be returned. The description of the > "filename" hint on > >page 220 comes close to a requirement but leaves an out. > That means an > >implementation that is not "capable of returning the file name" migh= t > >return MPI_INFO_NULL. > > > >I have a hard time imagining how any implementation could > have been given > >the filename at MPI_FILE_OPEN but not be "capable of > returning the file > >name" later but that is the text > > > > > >Dick Treumann - MPI Team/TCEM > >IBM Systems & Technology Group > >Dept 0lva / MS P963 -- 2455 South Road -- Poughkeepsie, NY 12601 > >Tele (845) 433-7846 Fax (845) 433-8363 > > > Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner@XXXXXXXX > High Performance Computing Center (HLRS) . phone ++49(0)711/685-65530= > University of Stuttgart . . . . . . . . .. fax ++49(0)711 / 685-65832= > Head of Dpmt Parallel Computing . . . www.hlrs.de/people/rabenseifner= > Nobelstr. 19, D-70550 Stuttgart, Germany . (Office: Allmandring 30) > > = --1__=0ABBF971DFE95B228f9e8a93df938690918c0ABBF971DFE95B22 Content-type: text/html; charset=US-ASCII Content-Disposition: inline Content-transfer-encoding: quoted-printable

Of course the keyword "filename" never gets "interpre= ted" in any implementation. The description says it is to be igno= red if passed in so it cannot be a "hint" at all. This pseudo= hint is a hack that is useful but also clearly abuses the "hints&= quot; intent of the info interface on files.

However - I am persuaded that it is legitimate for there to be no (key,= value) pairs to return for MPI_FILE_GET_INFO and therefore a statement= about returning MPI_INFO_NULL may be useful to somebody.

Have a good weekend all

Dick Treumann - MPI Team/TCEM
IBM Systems & Technology Group
Dept 0lva / MS P963 -- 2455 South Road -- Poughkeepsie, NY 12601
Tele (845) 433-7846 Fax (845) 433-8363

3D"Inactive
=
          "Rajeev Thakur" <thakur@XXXXXXXXXXX v>

          02/01/2008 05:03 PM

 =
3D=
To
3D""
"'Rolf Rabenseifner'" <rabenseifner@XXXXXX de>, Richard Treumann/Poughkeepsie/IBM@IBMUS
3D=
cc
3D""
<bronis@XXXXXXXXXXXXX "'Bill Gropp'" <= gropp@XXXXXXXXXXXXXXXX "'Rob Ross'" <rro
3D=
Subject
3D""
RE: Fw: [mpi-21] Ballot 4 - MPI_File_get_info
3D=3D""

Also, the first para under Reserved File Hints says "An implem= entation is
not required to interpret these key values, but if it does interpret th= e key
value, it must provide the functionality described." So an impleme= ntation is
free to not support any hint.

Rajeev


> -----Original Message-----
> From: Rolf Rabenseifner [ > Sent: Friday, February 01, 2008 3:42 PM
> To: Richard Treumann
> Cc: bronis@XXXXXXXXX Bill Gropp; Rob Ross; Rajeev Thakur
> Subject: Re: Fw: [mpi-21] Ballot 4 - MPI_File_get_info
>
> Dick,
>
> Yes, looks technical okay but strange in the real wolrd.
> But I remember, we allowed also unnamed pipes,
> tapes, ... (with MPI_MODE_SEQUENTIAL).
>
> Those might be a candidate for "no filename".
>
> Best regards
> Rolf
>
>
> On Fri, 1 Feb 2008 15:33:55 -0500
>  Richard Treumann <treumann@XXXXXXXXXXXXXX wrote:
> >
> >OK - the standard does say that an implementation should
> return filename if
> >it can. I guess if an implementation is not able to discover <= br> > the filename
> >by the time the app calls MPI_FILE_GET_INFO there is no
> requirement that
> >the "filename" hint be returned.  The descripti= on of the
> "filename" hint on
> >page 220 comes close to a requirement but leaves an out.  = ;
> That means an
> >implementation that is not "capable of returning the file= name" might
> >return MPI_INFO_NULL.
> >
> >I have a hard time imagining how any implementation could
= > have been given
> >the filename at MPI_FILE_OPEN but not be "capable of
= > returning the file
> >name" later but that is the text
> >
> >
> >Dick Treumann  -  MPI Team/TCEM
> >IBM Systems & Technology Group
> >Dept 0lva / MS P963 -- 2455 South Road -- Poughkeepsie, NY 126= 01
> >Tele (845) 433-7846         Fax (845) 433-= 8363
>
>
> Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner@hlrs= .de
> High Performance Computing Center (HLRS) . phone ++49(0)711/685-65= 530
> University of Stuttgart . . . . . . . . .. fax ++49(0)711 / 685-65= 832
> Head of Dpmt Parallel Computing . . . www.hlrs.de/people/rabenseif= ner
> Nobelstr. 19, D-70550 Stuttgart, Germany . (Office: Allmandring 30= )
>
>


= --1__=0ABBF971DFE95B228f9e8a93df938690918c0ABBF971DFE95B22-- --0__=0ABBF971DFE95B228f9e8a93df938690918c0ABBF971DFE95B22 Content-type: image/gif; name="graycol.gif" Content-Disposition: inline; filename="graycol.gif" Content-ID: <1__=0ABBF971DFE95B228f9e8a93df938@XXXXXXXXXX>Content-transfer-encoding: base64 R0lGODlhEAAQAKECAMzMzAAAAP///wAAACH5BAEAAAIALAAAAAAQABAAAAIXlI+py+0PopwxUbpu ZRfKZ2zgSJbmSRYAIf4fT3B0aW1pemVkIGJ5IFVsZWFkIFNtYXJ0U2F2ZXIhAAA7 --0__=0ABBF971DFE95B228f9e8a93df938690918c0ABBF971DFE95B22 Content-type: image/gif; name="pic10974.gif" Content-Disposition: inline; filename="pic10974.gif" Content-ID: <2__=0ABBF971DFE95B228f9e8a93df938@XXXXXXXXXX>Content-transfer-encoding: base64 R0lGODlhWABDALP/AAAAAK04Qf79/o+Gm7WuwlNObwoJFCsoSMDAwGFsmIuezf///wAAAAAAAAAA AAAAACH5BAEAAAgALAAAAABYAEMAQAT/EMlJq704682770RiFMRinqggEUNSHIchG0BCfHhOjAuh EDeUqTASLCbBhQrhG7xis2j0lssNDopE4jfIJhDaggI8YB1sZeZgLVA9YVCpnGagVjV171aRVrYR RghXcAGFhoUETwYxcXNyADJ3GlcSKGAwLwllVC1vjIUHBWsFilKQdI8GA5IcpApeJQt8L09lmgkH LZikoU5wjqcyAMMFrJIDPAKvCFletKSev1HBw8KrxtjZ2tvc3d5VyKtCKW3jfz4uMKmq3xu4N0nK BVoJQmx2LGVOmrqNjjJf2hHAQo/eDwJGTKhQMcgQEEAnEjFS98+RnW3smGkZU6ncCWav/4wYOnAI TihRL/4FEwbp28BXMMcoscQCVxlepL4IGDSCyJyVQOu0o7CjmLN50OZlqWmyFy5/6yBBuji0AxFR M00oQAqNIstqI6qKHUsWRAEAvagsmfUEAImyxgbmUpJk3IklNUtJOUAVLoUr1+wqDGTE4zk+T6FG uQb3SizBCwatiiUgCBN8vrz+zFjVyQ8FWkOlg4NQiZMB5QS8QO3mpOaKnL0Z2EKvNMSILEThKhCg zMKPVxYJh23qm9KNW7pArPynMqZDiErsTMqI+LRi3QAgkFUbXpuFKhSYZALd0O5RKa2z9EYKBbpb qxIKsjUPRgD7I2XYV6wyrOw92ykExP8NW4URhknC5dKGE4v4NENQj2jXjmfNgOZDaXb5glRmXQ33 YEWQYNcZFnrYcIQLNzyTFDQNkXIff0ExVlY4srziQk43inZgL4rwxxINMvpFFAz1KOODHiu+4aEw NEjFl5B3JIKWKF3k6I9bfUGp5ZZcdunll5IA4cuHvQQJ5gcsoCWOOUwgltIwAKRxJgbIkJAQZEq0 2YliZnpZZ4BH3CnYOXldOUOfQoYDqF1LFHbXCrO8xmRsfoXDXJ6ChjCAH3QlhJcT6VWE6FCkfCco CgrMFsROrIEX3o2whVjWDjoJccN3LdggSGXLCdLEgHr1lyU3O3QxhgohNKXJCWv8JQr/PDdaqd6w 2rj1inLiGeiCJoDspAoQlYE6QWLSECehcWIYxIQES6zhbn1iImTHEQyqJ4eIxJJoUBc+3CbBuwZE V5cJPPkIjFDdeEabQbd6WgICTxiiz0f5dBKquXF6k4senwEhYGnKEFJeGrxUZy8dB8gmAXI/sPvH ESfCwVt5hTgYiqQqtdRNHQIU1PJ33ZqmzgE90OwLaoJcnMop1WiMmgkPHQRIrwgFuNV90A3doNKT mrKIN07AnGcI9BQjhCBN4RfA1qIZnMqorJCogKfGQnxSCDilTVIA0yl5ciTovgLuBDKFUDE9aQcw 9SA+rjSNf9/M1gxrj6VwDTS0IUSElMzBfsj0NFXR2kwsV1A5IF1grLgLL/r1R40BZEnuBWgmQEyb jqRwSAt6bqMCOFkvKFN2GPPkUzIm/SCF8z8pVzpbjVnMsy0vOr1hw3SaSRUhpY09v0z0J1FnwzPl fmh+xl4WtR0zGu24I4KbMQm3lnVu2oNWxI9W/lcyzA+mCKF4DBikxb/+UWtOGRiFP8qEwAayIgIA Ow== --0__=0ABBF971DFE95B228f9e8a93df938690918c0ABBF971DFE95B22 Content-type: image/gif; name="ecblank.gif" Content-Disposition: inline; filename="ecblank.gif" Content-ID: <3__=0ABBF971DFE95B228f9e8a93df938@XXXXXXXXXX>Content-transfer-encoding: base64 R0lGODlhEAABAIAAAAAAAP///yH5BAEAAAEALAAAAAAQAAEAAAIEjI8ZBQA7 --0__=0ABBF971DFE95B228f9e8a93df938690918c0ABBF971DFE95B22--