- •Contents
- •List of Figures
- •List of Tables
- •Acknowledgments
- •Introduction to MPI
- •Overview and Goals
- •Background of MPI-1.0
- •Background of MPI-1.1, MPI-1.2, and MPI-2.0
- •Background of MPI-1.3 and MPI-2.1
- •Background of MPI-2.2
- •Who Should Use This Standard?
- •What Platforms Are Targets For Implementation?
- •What Is Included In The Standard?
- •What Is Not Included In The Standard?
- •Organization of this Document
- •MPI Terms and Conventions
- •Document Notation
- •Naming Conventions
- •Semantic Terms
- •Data Types
- •Opaque Objects
- •Array Arguments
- •State
- •Named Constants
- •Choice
- •Addresses
- •Language Binding
- •Deprecated Names and Functions
- •Fortran Binding Issues
- •C Binding Issues
- •C++ Binding Issues
- •Functions and Macros
- •Processes
- •Error Handling
- •Implementation Issues
- •Independence of Basic Runtime Routines
- •Interaction with Signals
- •Examples
- •Point-to-Point Communication
- •Introduction
- •Blocking Send and Receive Operations
- •Blocking Send
- •Message Data
- •Message Envelope
- •Blocking Receive
- •Return Status
- •Passing MPI_STATUS_IGNORE for Status
- •Data Type Matching and Data Conversion
- •Type Matching Rules
- •Type MPI_CHARACTER
- •Data Conversion
- •Communication Modes
- •Semantics of Point-to-Point Communication
- •Buffer Allocation and Usage
- •Nonblocking Communication
- •Communication Request Objects
- •Communication Initiation
- •Communication Completion
- •Semantics of Nonblocking Communications
- •Multiple Completions
- •Non-destructive Test of status
- •Probe and Cancel
- •Persistent Communication Requests
- •Send-Receive
- •Null Processes
- •Datatypes
- •Derived Datatypes
- •Type Constructors with Explicit Addresses
- •Datatype Constructors
- •Subarray Datatype Constructor
- •Distributed Array Datatype Constructor
- •Address and Size Functions
- •Lower-Bound and Upper-Bound Markers
- •Extent and Bounds of Datatypes
- •True Extent of Datatypes
- •Commit and Free
- •Duplicating a Datatype
- •Use of General Datatypes in Communication
- •Correct Use of Addresses
- •Decoding a Datatype
- •Examples
- •Pack and Unpack
- •Canonical MPI_PACK and MPI_UNPACK
- •Collective Communication
- •Introduction and Overview
- •Communicator Argument
- •Applying Collective Operations to Intercommunicators
- •Barrier Synchronization
- •Broadcast
- •Example using MPI_BCAST
- •Gather
- •Examples using MPI_GATHER, MPI_GATHERV
- •Scatter
- •Examples using MPI_SCATTER, MPI_SCATTERV
- •Example using MPI_ALLGATHER
- •All-to-All Scatter/Gather
- •Global Reduction Operations
- •Reduce
- •Signed Characters and Reductions
- •MINLOC and MAXLOC
- •All-Reduce
- •Process-local reduction
- •Reduce-Scatter
- •MPI_REDUCE_SCATTER_BLOCK
- •MPI_REDUCE_SCATTER
- •Scan
- •Inclusive Scan
- •Exclusive Scan
- •Example using MPI_SCAN
- •Correctness
- •Introduction
- •Features Needed to Support Libraries
- •MPI's Support for Libraries
- •Basic Concepts
- •Groups
- •Contexts
- •Intra-Communicators
- •Group Management
- •Group Accessors
- •Group Constructors
- •Group Destructors
- •Communicator Management
- •Communicator Accessors
- •Communicator Constructors
- •Communicator Destructors
- •Motivating Examples
- •Current Practice #1
- •Current Practice #2
- •(Approximate) Current Practice #3
- •Example #4
- •Library Example #1
- •Library Example #2
- •Inter-Communication
- •Inter-communicator Accessors
- •Inter-communicator Operations
- •Inter-Communication Examples
- •Caching
- •Functionality
- •Communicators
- •Windows
- •Datatypes
- •Error Class for Invalid Keyval
- •Attributes Example
- •Naming Objects
- •Formalizing the Loosely Synchronous Model
- •Basic Statements
- •Models of Execution
- •Static communicator allocation
- •Dynamic communicator allocation
- •The General case
- •Process Topologies
- •Introduction
- •Virtual Topologies
- •Embedding in MPI
- •Overview of the Functions
- •Topology Constructors
- •Cartesian Constructor
- •Cartesian Convenience Function: MPI_DIMS_CREATE
- •General (Graph) Constructor
- •Distributed (Graph) Constructor
- •Topology Inquiry Functions
- •Cartesian Shift Coordinates
- •Partitioning of Cartesian structures
- •Low-Level Topology Functions
- •An Application Example
- •MPI Environmental Management
- •Implementation Information
- •Version Inquiries
- •Environmental Inquiries
- •Tag Values
- •Host Rank
- •IO Rank
- •Clock Synchronization
- •Memory Allocation
- •Error Handling
- •Error Handlers for Communicators
- •Error Handlers for Windows
- •Error Handlers for Files
- •Freeing Errorhandlers and Retrieving Error Strings
- •Error Codes and Classes
- •Error Classes, Error Codes, and Error Handlers
- •Timers and Synchronization
- •Startup
- •Allowing User Functions at Process Termination
- •Determining Whether MPI Has Finished
- •Portable MPI Process Startup
- •The Info Object
- •Process Creation and Management
- •Introduction
- •The Dynamic Process Model
- •Starting Processes
- •The Runtime Environment
- •Process Manager Interface
- •Processes in MPI
- •Starting Processes and Establishing Communication
- •Reserved Keys
- •Spawn Example
- •Manager-worker Example, Using MPI_COMM_SPAWN.
- •Establishing Communication
- •Names, Addresses, Ports, and All That
- •Server Routines
- •Client Routines
- •Name Publishing
- •Reserved Key Values
- •Client/Server Examples
- •Ocean/Atmosphere - Relies on Name Publishing
- •Simple Client-Server Example.
- •Other Functionality
- •Universe Size
- •Singleton MPI_INIT
- •MPI_APPNUM
- •Releasing Connections
- •Another Way to Establish MPI Communication
- •One-Sided Communications
- •Introduction
- •Initialization
- •Window Creation
- •Window Attributes
- •Communication Calls
- •Examples
- •Accumulate Functions
- •Synchronization Calls
- •Fence
- •General Active Target Synchronization
- •Lock
- •Assertions
- •Examples
- •Error Handling
- •Error Handlers
- •Error Classes
- •Semantics and Correctness
- •Atomicity
- •Progress
- •Registers and Compiler Optimizations
- •External Interfaces
- •Introduction
- •Generalized Requests
- •Examples
- •Associating Information with Status
- •MPI and Threads
- •General
- •Initialization
- •Introduction
- •File Manipulation
- •Opening a File
- •Closing a File
- •Deleting a File
- •Resizing a File
- •Preallocating Space for a File
- •Querying the Size of a File
- •Querying File Parameters
- •File Info
- •Reserved File Hints
- •File Views
- •Data Access
- •Data Access Routines
- •Positioning
- •Synchronism
- •Coordination
- •Data Access Conventions
- •Data Access with Individual File Pointers
- •Data Access with Shared File Pointers
- •Noncollective Operations
- •Collective Operations
- •Seek
- •Split Collective Data Access Routines
- •File Interoperability
- •Datatypes for File Interoperability
- •Extent Callback
- •Datarep Conversion Functions
- •Matching Data Representations
- •Consistency and Semantics
- •File Consistency
- •Random Access vs. Sequential Files
- •Progress
- •Collective File Operations
- •Type Matching
- •Logical vs. Physical File Layout
- •File Size
- •Examples
- •Asynchronous I/O
- •I/O Error Handling
- •I/O Error Classes
- •Examples
- •Subarray Filetype Constructor
- •Requirements
- •Discussion
- •Logic of the Design
- •Examples
- •MPI Library Implementation
- •Systems with Weak Symbols
- •Systems Without Weak Symbols
- •Complications
- •Multiple Counting
- •Linker Oddities
- •Multiple Levels of Interception
- •Deprecated Functions
- •Deprecated since MPI-2.0
- •Deprecated since MPI-2.2
- •Language Bindings
- •Overview
- •Design
- •C++ Classes for MPI
- •Class Member Functions for MPI
- •Semantics
- •C++ Datatypes
- •Communicators
- •Exceptions
- •Mixed-Language Operability
- •Problems With Fortran Bindings for MPI
- •Problems Due to Strong Typing
- •Problems Due to Data Copying and Sequence Association
- •Special Constants
- •Fortran 90 Derived Types
- •A Problem with Register Optimization
- •Basic Fortran Support
- •Extended Fortran Support
- •The mpi Module
- •No Type Mismatch Problems for Subroutines with Choice Arguments
- •Additional Support for Fortran Numeric Intrinsic Types
- •Language Interoperability
- •Introduction
- •Assumptions
- •Initialization
- •Transfer of Handles
- •Status
- •MPI Opaque Objects
- •Datatypes
- •Callback Functions
- •Error Handlers
- •Reduce Operations
- •Addresses
- •Attributes
- •Extra State
- •Constants
- •Interlanguage Communication
- •Language Bindings Summary
- •Groups, Contexts, Communicators, and Caching Fortran Bindings
- •External Interfaces C++ Bindings
- •Change-Log
- •Bibliography
- •Examples Index
- •MPI Declarations Index
- •MPI Function Index
Chapter 13
I/O
13.1 Introduction
POSIX provides a model of a widely portable le system, but the portability and optimization needed for parallel I/O cannot be achieved with the POSIX interface.
The signi cant optimizations required for e ciency (e.g., grouping [35], collective bu ering [6, 13, 36, 39, 46], and disk-directed I/O [31]) can only be implemented if the parallel I/O system provides a high-level interface supporting partitioning of le data among processes and a collective interface supporting complete transfers of global data structures between process memories and les. In addition, further e ciencies can be gained via support for asynchronous I/O, strided accesses, and control over physical le layout on storage devices (disks). The I/O environment described in this chapter provides these facilities.
Instead of de ning I/O access modes to express the common patterns for accessing a shared le (broadcast, reduction, scatter, gather), we chose another approach in which data partitioning is expressed using derived datatypes. Compared to a limited set of prede ned access patterns, this approach has the advantage of added exibility and expressiveness.
13.1.1 De nitions
le An MPI le is an ordered collection of typed data items. MPI supports random or sequential access to any integral set of these items. A le is opened collectively by a group of processes. All collective I/O calls on a le are collective over this group.
displacement A le displacement is an absolute byte position relative to the beginning of a le. The displacement de nes the location where a view begins. Note that a \ le displacement" is distinct from a \typemap displacement."
etype An etype (elementary datatype) is the unit of data access and positioning. It can be any MPI prede ned or derived datatype. Derived etypes can be constructed using any of the MPI datatype constructor routines, provided all resulting typemap displacements are non-negative and monotonically nondecreasing. Data access is performed in etype units, reading or writing whole data items of type etype. O sets are expressed as a count of etypes; le pointers point to the beginning of etypes. Depending on context, the term \etype" is used to describe one of three aspects of an elementary datatype: a particular MPI type, a data item of that type, or the extent of that type.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
389
390 |
CHAPTER 13. I/O |
1letype A letype is the basis for partitioning a le among processes and de nes a template
2for accessing the le. A letype is either a single etype or a derived MPI datatype
3constructed from multiple instances of the same etype. In addition, the extent of any
4hole in the letype must be a multiple of the etype's extent. The displacements in the
5typemap of the letype are not required to be distinct, but they must be non-negative
6
7
and monotonically nondecreasing.
8view A view de nes the current set of data visible and accessible from an open le as an
9ordered set of etypes. Each process has its own view of the le, de ned by three
10quantities: a displacement, an etype, and a letype. The pattern described by a
11letype is repeated, beginning at the displacement, to de ne the view. The pattern
12of repetition is de ned to be the same pattern that MPI_TYPE_CONTIGUOUS would
13produce if it were passed the letype and an arbitrarily large count. Figure 13.1 shows
14how the tiling works; note that the letype in this example must have explicit lower
15and upper bounds set in order for the initial and nal holes to be repeated in the
16view. Views can be changed by the user during program execution. The default view
17is a linear byte stream (displacement is zero, etype and letype equal to MPI_BYTE).
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
etype filetype
holes
tiling a file with the filetype:
...
displacement |
accessible data |
Figure 13.1: Etypes and letypes
A group of processes can use complementary views to achieve a global data distribution such as a scatter/gather pattern (see Figure 13.2).
etype
process 0 filetype 
000111
process 1 filetype 
process 2 filetype
tiling a file with the filetypes:
111000
000111
111000
111000
111000
...
displacement
Figure 13.2: Partitioning a le among parallel processes
43o set An o set is a position in the le relative to the current view, expressed as a count of
44etypes. Holes in the view's letype are skipped when calculating this position. O set 0
45is the location of the rst etype visible in the view (after skipping the displacement and
46any initial holes in the view). For example, an o set of 2 for process 1 in Figure 13.2
47is the position of the 8th etype in the le after the displacement. An \explicit o set"
48is an o set that is used as a formal parameter in explicit data access routines.