Received: from msr.epm.ornl.gov (root@msr.epm.ornl.gov [128.219.8.1]) by rios2.EPM.ORNL.GOV (8.6.10/8.6.10) with ESMTP id HAA17449 for <walker@rios2.epm.ornl.gov>; Wed, 21 Jun 1995 07:31:52 -0400
Received: from CS.UTK.EDU (CS.UTK.EDU [128.169.94.1]) by msr.epm.ornl.gov (8.6.10/8.6.10) with ESMTP id HAA20303 for <walker@msr.epm.ornl.gov>; Wed, 21 Jun 1995 07:31:45 -0400
Received: from elegabalus.cs.qub.ac.uk by CS.UTK.EDU with SMTP (cf v2.9s-UTK)
id HAA07533; Wed, 21 Jun 1995 07:28:04 -0400
Received: from domitian.cs.qub.ac.uk by elegabalus.cs.qub.ac.uk (NeXT-1.0 (From Sendmail 5.52)/NeXT-2.0)
id AA14283; Wed, 21 Jun 95 12:27:58 BST
From: sfitzp@cs.qub.ac.uk (Stephen Fitzpatrick)
Message-Id: <9506211127.AA14283@ elegabalus.cs.qub.ac.uk >
Received: by domitian.cs.qub.ac.uk (NX5.67c/NX3.0X)
id AA01875; Wed, 21 Jun 95 12:26:44 +0100
Date: Wed, 21 Jun 95 12:26:44 +0100
Received: by NeXT.Mailer (1.87.1)
Received: by NeXT Mailer (1.87.1)
To: mpi-comments@cs.utk.edu
Subject: All-gather description
Reply-To: S.Fitzpatrick@cs.qub.ac.uk
A small suggestion...
The description of mpi_allgather is, perhaps, misleading:
(5 May 1994 edition of definition, page 107, lines 47,48)
The jth block of data sent from each process is received by every
process and placed in the jth block of the buffer recvbuf.
This implies that each processor sends multiple blocks of data.
Assuming I understand the operation correctly, I believe "The block
of data sent from processor j" would be more accurate?
A similar comment applies to the description of mpi_allgatherv.
- --
Stephen Fitzpatrick, Department of Computer Science,
Queen's University of Belfast, Belfast BT7 1NN, N. Ireland.
S.Fitzpatrick@cs.qub.ac.uk (ASCII, NeXT, maybe MIME)
PGP key via WWW: <URL:http://www.cs.qub.ac.uk/~S.Fitzpatrick/>
------- End of Forwarded Message