portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2024-09-20 01:05:38 +01:00
Code Releases Activity

Merge branch 'master' into hadrons

Browse Source
This commit is contained in:
Antonin Portelli 2016-02-22 15:25:02 +00:00
commit cfd368596d
277 changed files with 25256 additions and 2310 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
AUTHORS
View File

@ -1,5 +1,4 @@
Peter Boyle
Azusa Yamaguchi
Intel Parallel Computing Centre @ Higgs Centre for Theoretical Physics
University of Edinburgh
Scotland, UK
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: Peter Boyle <peterboyle@MacBook-Pro.local>
Author: paboyle <paboyle@ph.ed.ac.uk>

876
COPYING
View File

@ -1,622 +1,281 @@
GNU GENERAL PUBLIC LICENSE
Version 3, 29 June 2007
GNU GENERAL PUBLIC LICENSE
Version 2, June 1991
Copyright (C) 2007 Free Software Foundation, Inc. <http://fsf.org/>
Copyright (C) 1989, 1991 Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
Preamble
The GNU General Public License is a free, copyleft license for
software and other kinds of works.
The licenses for most software and other practical works are designed
to take away your freedom to share and change the works. By contrast,
the GNU General Public License is intended to guarantee your freedom to
share and change all versions of a program--to make sure it remains free
software for all its users. We, the Free Software Foundation, use the
GNU General Public License for most of our software; it applies also to
any other work released this way by its authors. You can apply it to
The licenses for most software are designed to take away your
freedom to share and change it. By contrast, the GNU General Public
License is intended to guarantee your freedom to share and change free
software--to make sure the software is free for all its users. This
General Public License applies to most of the Free Software
Foundation's software and to any other program whose authors commit to
using it. (Some other Free Software Foundation software is covered by
the GNU Lesser General Public License instead.) You can apply it to
your programs, too.
When we speak of free software, we are referring to freedom, not
price. Our General Public Licenses are designed to make sure that you
have the freedom to distribute copies of free software (and charge for
them if you wish), that you receive source code or can get it if you
want it, that you can change the software or use pieces of it in new
free programs, and that you know you can do these things.
this service if you wish), that you receive source code or can get it
if you want it, that you can change the software or use pieces of it
in new free programs; and that you know you can do these things.
To protect your rights, we need to prevent others from denying you
these rights or asking you to surrender the rights. Therefore, you have
certain responsibilities if you distribute copies of the software, or if
you modify it: responsibilities to respect the freedom of others.
To protect your rights, we need to make restrictions that forbid
anyone to deny you these rights or to ask you to surrender the rights.
These restrictions translate to certain responsibilities for you if you
distribute copies of the software, or if you modify it.
For example, if you distribute copies of such a program, whether
gratis or for a fee, you must pass on to the recipients the same
freedoms that you received. You must make sure that they, too, receive
or can get the source code. And you must show them these terms so they
know their rights.
gratis or for a fee, you must give the recipients all the rights that
you have. You must make sure that they, too, receive or can get the
source code. And you must show them these terms so they know their
rights.
Developers that use the GNU GPL protect your rights with two steps:
(1) assert copyright on the software, and (2) offer you this License
giving you legal permission to copy, distribute and/or modify it.
We protect your rights with two steps: (1) copyright the software, and
(2) offer you this license which gives you legal permission to copy,
distribute and/or modify the software.
For the developers' and authors' protection, the GPL clearly explains
that there is no warranty for this free software. For both users' and
authors' sake, the GPL requires that modified versions be marked as
changed, so that their problems will not be attributed erroneously to
authors of previous versions.
Also, for each author's protection and ours, we want to make certain
that everyone understands that there is no warranty for this free
software. If the software is modified by someone else and passed on, we
want its recipients to know that what they have is not the original, so
that any problems introduced by others will not reflect on the original
authors' reputations.
Some devices are designed to deny users access to install or run
modified versions of the software inside them, although the manufacturer
can do so. This is fundamentally incompatible with the aim of
protecting users' freedom to change the software. The systematic
pattern of such abuse occurs in the area of products for individuals to
use, which is precisely where it is most unacceptable. Therefore, we
have designed this version of the GPL to prohibit the practice for those
products. If such problems arise substantially in other domains, we
stand ready to extend this provision to those domains in future versions
of the GPL, as needed to protect the freedom of users.
Finally, every program is threatened constantly by software patents.
States should not allow patents to restrict development and use of
software on general-purpose computers, but in those that do, we wish to
avoid the special danger that patents applied to a free program could
make it effectively proprietary. To prevent this, the GPL assures that
patents cannot be used to render the program non-free.
Finally, any free program is threatened constantly by software
patents. We wish to avoid the danger that redistributors of a free
program will individually obtain patent licenses, in effect making the
program proprietary. To prevent this, we have made it clear that any
patent must be licensed for everyone's free use or not licensed at all.
The precise terms and conditions for copying, distribution and
modification follow.
TERMS AND CONDITIONS
0. Definitions.
"This License" refers to version 3 of the GNU General Public License.
"Copyright" also means copyright-like laws that apply to other kinds of
works, such as semiconductor masks.
"The Program" refers to any copyrightable work licensed under this
License. Each licensee is addressed as "you". "Licensees" and
"recipients" may be individuals or organizations.
To "modify" a work means to copy from or adapt all or part of the work
in a fashion requiring copyright permission, other than the making of an
exact copy. The resulting work is called a "modified version" of the
earlier work or a work "based on" the earlier work.
A "covered work" means either the unmodified Program or a work based
on the Program.
To "propagate" a work means to do anything with it that, without
permission, would make you directly or secondarily liable for
infringement under applicable copyright law, except executing it on a
computer or modifying a private copy. Propagation includes copying,
distribution (with or without modification), making available to the
public, and in some countries other activities as well.
To "convey" a work means any kind of propagation that enables other
parties to make or receive copies. Mere interaction with a user through
a computer network, with no transfer of a copy, is not conveying.
An interactive user interface displays "Appropriate Legal Notices"
to the extent that it includes a convenient and prominently visible
feature that (1) displays an appropriate copyright notice, and (2)
tells the user that there is no warranty for the work (except to the
extent that warranties are provided), that licensees may convey the
work under this License, and how to view a copy of this License. If
the interface presents a list of user commands or options, such as a
menu, a prominent item in the list meets this criterion.
1. Source Code.
The "source code" for a work means the preferred form of the work
for making modifications to it. "Object code" means any non-source
form of a work.
A "Standard Interface" means an interface that either is an official
standard defined by a recognized standards body, or, in the case of
interfaces specified for a particular programming language, one that
is widely used among developers working in that language.
The "System Libraries" of an executable work include anything, other
than the work as a whole, that (a) is included in the normal form of
packaging a Major Component, but which is not part of that Major
Component, and (b) serves only to enable use of the work with that
Major Component, or to implement a Standard Interface for which an
implementation is available to the public in source code form. A
"Major Component", in this context, means a major essential component
(kernel, window system, and so on) of the specific operating system
(if any) on which the executable work runs, or a compiler used to
produce the work, or an object code interpreter used to run it.
The "Corresponding Source" for a work in object code form means all
the source code needed to generate, install, and (for an executable
work) run the object code and to modify the work, including scripts to
control those activities. However, it does not include the work's
System Libraries, or general-purpose tools or generally available free
programs which are used unmodified in performing those activities but
which are not part of the work. For example, Corresponding Source
includes interface definition files associated with source files for
the work, and the source code for shared libraries and dynamically
linked subprograms that the work is specifically designed to require,
such as by intimate data communication or control flow between those
subprograms and other parts of the work.
The Corresponding Source need not include anything that users
can regenerate automatically from other parts of the Corresponding
Source.
The Corresponding Source for a work in source code form is that
same work.
2. Basic Permissions.
All rights granted under this License are granted for the term of
copyright on the Program, and are irrevocable provided the stated
conditions are met. This License explicitly affirms your unlimited
permission to run the unmodified Program. The output from running a
covered work is covered by this License only if the output, given its
content, constitutes a covered work. This License acknowledges your
rights of fair use or other equivalent, as provided by copyright law.
You may make, run and propagate covered works that you do not
convey, without conditions so long as your license otherwise remains
in force. You may convey covered works to others for the sole purpose
of having them make modifications exclusively for you, or provide you
with facilities for running those works, provided that you comply with
the terms of this License in conveying all material for which you do
not control copyright. Those thus making or running the covered works
for you must do so exclusively on your behalf, under your direction
and control, on terms that prohibit them from making any copies of
your copyrighted material outside their relationship with you.
Conveying under any other circumstances is permitted solely under
the conditions stated below. Sublicensing is not allowed; section 10
makes it unnecessary.
3. Protecting Users' Legal Rights From Anti-Circumvention Law.
No covered work shall be deemed part of an effective technological
measure under any applicable law fulfilling obligations under article
11 of the WIPO copyright treaty adopted on 20 December 1996, or
similar laws prohibiting or restricting circumvention of such
measures.
When you convey a covered work, you waive any legal power to forbid
circumvention of technological measures to the extent such circumvention
is effected by exercising rights under this License with respect to
the covered work, and you disclaim any intention to limit operation or
modification of the work as a means of enforcing, against the work's
users, your or third parties' legal rights to forbid circumvention of
technological measures.
4. Conveying Verbatim Copies.
You may convey verbatim copies of the Program's source code as you
receive it, in any medium, provided that you conspicuously and
appropriately publish on each copy an appropriate copyright notice;
keep intact all notices stating that this License and any
non-permissive terms added in accord with section 7 apply to the code;
keep intact all notices of the absence of any warranty; and give all
recipients a copy of this License along with the Program.
You may charge any price or no price for each copy that you convey,
and you may offer support or warranty protection for a fee.
5. Conveying Modified Source Versions.
You may convey a work based on the Program, or the modifications to
produce it from the Program, in the form of source code under the
terms of section 4, provided that you also meet all of these conditions:
a) The work must carry prominent notices stating that you modified
it, and giving a relevant date.
b) The work must carry prominent notices stating that it is
released under this License and any conditions added under section
7. This requirement modifies the requirement in section 4 to
"keep intact all notices".
c) You must license the entire work, as a whole, under this
License to anyone who comes into possession of a copy. This
License will therefore apply, along with any applicable section 7
additional terms, to the whole of the work, and all its parts,
regardless of how they are packaged. This License gives no
permission to license the work in any other way, but it does not
invalidate such permission if you have separately received it.
d) If the work has interactive user interfaces, each must display
Appropriate Legal Notices; however, if the Program has interactive
interfaces that do not display Appropriate Legal Notices, your
work need not make them do so.
A compilation of a covered work with other separate and independent
works, which are not by their nature extensions of the covered work,
and which are not combined with it such as to form a larger program,
in or on a volume of a storage or distribution medium, is called an
"aggregate" if the compilation and its resulting copyright are not
used to limit the access or legal rights of the compilation's users
beyond what the individual works permit. Inclusion of a covered work
in an aggregate does not cause this License to apply to the other
parts of the aggregate.
6. Conveying Non-Source Forms.
You may convey a covered work in object code form under the terms
of sections 4 and 5, provided that you also convey the
machine-readable Corresponding Source under the terms of this License,
in one of these ways:
a) Convey the object code in, or embodied in, a physical product
(including a physical distribution medium), accompanied by the
Corresponding Source fixed on a durable physical medium
customarily used for software interchange.
b) Convey the object code in, or embodied in, a physical product
(including a physical distribution medium), accompanied by a
written offer, valid for at least three years and valid for as
long as you offer spare parts or customer support for that product
model, to give anyone who possesses the object code either (1) a
copy of the Corresponding Source for all the software in the
product that is covered by this License, on a durable physical
medium customarily used for software interchange, for a price no
more than your reasonable cost of physically performing this
conveying of source, or (2) access to copy the
Corresponding Source from a network server at no charge.
c) Convey individual copies of the object code with a copy of the
written offer to provide the Corresponding Source. This
alternative is allowed only occasionally and noncommercially, and
only if you received the object code with such an offer, in accord
with subsection 6b.
d) Convey the object code by offering access from a designated
place (gratis or for a charge), and offer equivalent access to the
Corresponding Source in the same way through the same place at no
further charge. You need not require recipients to copy the
Corresponding Source along with the object code. If the place to
copy the object code is a network server, the Corresponding Source
may be on a different server (operated by you or a third party)
that supports equivalent copying facilities, provided you maintain
clear directions next to the object code saying where to find the
Corresponding Source. Regardless of what server hosts the
Corresponding Source, you remain obligated to ensure that it is
available for as long as needed to satisfy these requirements.
e) Convey the object code using peer-to-peer transmission, provided
you inform other peers where the object code and Corresponding
Source of the work are being offered to the general public at no
charge under subsection 6d.
A separable portion of the object code, whose source code is excluded
from the Corresponding Source as a System Library, need not be
included in conveying the object code work.
A "User Product" is either (1) a "consumer product", which means any
tangible personal property which is normally used for personal, family,
or household purposes, or (2) anything designed or sold for incorporation
into a dwelling. In determining whether a product is a consumer product,
doubtful cases shall be resolved in favor of coverage. For a particular
product received by a particular user, "normally used" refers to a
typical or common use of that class of product, regardless of the status
of the particular user or of the way in which the particular user
actually uses, or expects or is expected to use, the product. A product
is a consumer product regardless of whether the product has substantial
commercial, industrial or non-consumer uses, unless such uses represent
the only significant mode of use of the product.
"Installation Information" for a User Product means any methods,
procedures, authorization keys, or other information required to install
and execute modified versions of a covered work in that User Product from
a modified version of its Corresponding Source. The information must
suffice to ensure that the continued functioning of the modified object
code is in no case prevented or interfered with solely because
modification has been made.
If you convey an object code work under this section in, or with, or
specifically for use in, a User Product, and the conveying occurs as
part of a transaction in which the right of possession and use of the
User Product is transferred to the recipient in perpetuity or for a
fixed term (regardless of how the transaction is characterized), the
Corresponding Source conveyed under this section must be accompanied
by the Installation Information. But this requirement does not apply
if neither you nor any third party retains the ability to install
modified object code on the User Product (for example, the work has
been installed in ROM).
The requirement to provide Installation Information does not include a
requirement to continue to provide support service, warranty, or updates
for a work that has been modified or installed by the recipient, or for
the User Product in which it has been modified or installed. Access to a
network may be denied when the modification itself materially and
adversely affects the operation of the network or violates the rules and
protocols for communication across the network.
Corresponding Source conveyed, and Installation Information provided,
in accord with this section must be in a format that is publicly
documented (and with an implementation available to the public in
source code form), and must require no special password or key for
unpacking, reading or copying.
7. Additional Terms.
"Additional permissions" are terms that supplement the terms of this
License by making exceptions from one or more of its conditions.
Additional permissions that are applicable to the entire Program shall
be treated as though they were included in this License, to the extent
that they are valid under applicable law. If additional permissions
apply only to part of the Program, that part may be used separately
under those permissions, but the entire Program remains governed by
this License without regard to the additional permissions.
When you convey a copy of a covered work, you may at your option
remove any additional permissions from that copy, or from any part of
it. (Additional permissions may be written to require their own
removal in certain cases when you modify the work.) You may place
additional permissions on material, added by you to a covered work,
for which you have or can give appropriate copyright permission.
Notwithstanding any other provision of this License, for material you
add to a covered work, you may (if authorized by the copyright holders of
that material) supplement the terms of this License with terms:
a) Disclaiming warranty or limiting liability differently from the
terms of sections 15 and 16 of this License; or
b) Requiring preservation of specified reasonable legal notices or
author attributions in that material or in the Appropriate Legal
Notices displayed by works containing it; or
c) Prohibiting misrepresentation of the origin of that material, or
requiring that modified versions of such material be marked in
reasonable ways as different from the original version; or
d) Limiting the use for publicity purposes of names of licensors or
authors of the material; or
e) Declining to grant rights under trademark law for use of some
trade names, trademarks, or service marks; or
f) Requiring indemnification of licensors and authors of that
material by anyone who conveys the material (or modified versions of
it) with contractual assumptions of liability to the recipient, for
any liability that these contractual assumptions directly impose on
those licensors and authors.
All other non-permissive additional terms are considered "further
restrictions" within the meaning of section 10. If the Program as you
received it, or any part of it, contains a notice stating that it is
governed by this License along with a term that is a further
restriction, you may remove that term. If a license document contains
a further restriction but permits relicensing or conveying under this
License, you may add to a covered work material governed by the terms
of that license document, provided that the further restriction does
not survive such relicensing or conveying.
If you add terms to a covered work in accord with this section, you
must place, in the relevant source files, a statement of the
additional terms that apply to those files, or a notice indicating
where to find the applicable terms.
Additional terms, permissive or non-permissive, may be stated in the
form of a separately written license, or stated as exceptions;
the above requirements apply either way.
8. Termination.
You may not propagate or modify a covered work except as expressly
provided under this License. Any attempt otherwise to propagate or
modify it is void, and will automatically terminate your rights under
this License (including any patent licenses granted under the third
paragraph of section 11).
However, if you cease all violation of this License, then your
license from a particular copyright holder is reinstated (a)
provisionally, unless and until the copyright holder explicitly and
finally terminates your license, and (b) permanently, if the copyright
holder fails to notify you of the violation by some reasonable means
prior to 60 days after the cessation.
Moreover, your license from a particular copyright holder is
reinstated permanently if the copyright holder notifies you of the
violation by some reasonable means, this is the first time you have
received notice of violation of this License (for any work) from that
copyright holder, and you cure the violation prior to 30 days after
your receipt of the notice.
Termination of your rights under this section does not terminate the
licenses of parties who have received copies or rights from you under
this License. If your rights have been terminated and not permanently
reinstated, you do not qualify to receive new licenses for the same
material under section 10.
9. Acceptance Not Required for Having Copies.
You are not required to accept this License in order to receive or
run a copy of the Program. Ancillary propagation of a covered work
occurring solely as a consequence of using peer-to-peer transmission
to receive a copy likewise does not require acceptance. However,
nothing other than this License grants you permission to propagate or
modify any covered work. These actions infringe copyright if you do
not accept this License. Therefore, by modifying or propagating a
covered work, you indicate your acceptance of this License to do so.
10. Automatic Licensing of Downstream Recipients.
Each time you convey a covered work, the recipient automatically
receives a license from the original licensors, to run, modify and
propagate that work, subject to this License. You are not responsible
for enforcing compliance by third parties with this License.
An "entity transaction" is a transaction transferring control of an
organization, or substantially all assets of one, or subdividing an
organization, or merging organizations. If propagation of a covered
work results from an entity transaction, each party to that
transaction who receives a copy of the work also receives whatever
licenses to the work the party's predecessor in interest had or could
give under the previous paragraph, plus a right to possession of the
Corresponding Source of the work from the predecessor in interest, if
the predecessor has it or can get it with reasonable efforts.
You may not impose any further restrictions on the exercise of the
rights granted or affirmed under this License. For example, you may
not impose a license fee, royalty, or other charge for exercise of
rights granted under this License, and you may not initiate litigation
(including a cross-claim or counterclaim in a lawsuit) alleging that
any patent claim is infringed by making, using, selling, offering for
sale, or importing the Program or any portion of it.
11. Patents.
A "contributor" is a copyright holder who authorizes use under this
License of the Program or a work on which the Program is based. The
work thus licensed is called the contributor's "contributor version".
A contributor's "essential patent claims" are all patent claims
owned or controlled by the contributor, whether already acquired or
hereafter acquired, that would be infringed by some manner, permitted
by this License, of making, using, or selling its contributor version,
but do not include claims that would be infringed only as a
consequence of further modification of the contributor version. For
purposes of this definition, "control" includes the right to grant
patent sublicenses in a manner consistent with the requirements of
GNU GENERAL PUBLIC LICENSE
TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
0. This License applies to any program or other work which contains
a notice placed by the copyright holder saying it may be distributed
under the terms of this General Public License. The "Program", below,
refers to any such program or work, and a "work based on the Program"
means either the Program or any derivative work under copyright law:
that is to say, a work containing the Program or a portion of it,
either verbatim or with modifications and/or translated into another
language. (Hereinafter, translation is included without limitation in
the term "modification".) Each licensee is addressed as "you".
Activities other than copying, distribution and modification are not
covered by this License; they are outside its scope. The act of
running the Program is not restricted, and the output from the Program
is covered only if its contents constitute a work based on the
Program (independent of having been made by running the Program).
Whether that is true depends on what the Program does.
1. You may copy and distribute verbatim copies of the Program's
source code as you receive it, in any medium, provided that you
conspicuously and appropriately publish on each copy an appropriate
copyright notice and disclaimer of warranty; keep intact all the
notices that refer to this License and to the absence of any warranty;
and give any other recipients of the Program a copy of this License
along with the Program.
You may charge a fee for the physical act of transferring a copy, and
you may at your option offer warranty protection in exchange for a fee.
2. You may modify your copy or copies of the Program or any portion
of it, thus forming a work based on the Program, and copy and
distribute such modifications or work under the terms of Section 1
above, provided that you also meet all of these conditions:
a) You must cause the modified files to carry prominent notices
stating that you changed the files and the date of any change.
b) You must cause any work that you distribute or publish, that in
whole or in part contains or is derived from the Program or any
part thereof, to be licensed as a whole at no charge to all third
parties under the terms of this License.
c) If the modified program normally reads commands interactively
when run, you must cause it, when started running for such
interactive use in the most ordinary way, to print or display an
announcement including an appropriate copyright notice and a
notice that there is no warranty (or else, saying that you provide
a warranty) and that users may redistribute the program under
these conditions, and telling the user how to view a copy of this
License. (Exception: if the Program itself is interactive but
does not normally print such an announcement, your work based on
the Program is not required to print an announcement.)
These requirements apply to the modified work as a whole. If
identifiable sections of that work are not derived from the Program,
and can be reasonably considered independent and separate works in
themselves, then this License, and its terms, do not apply to those
sections when you distribute them as separate works. But when you
distribute the same sections as part of a whole which is a work based
on the Program, the distribution of the whole must be on the terms of
this License, whose permissions for other licensees extend to the
entire whole, and thus to each and every part regardless of who wrote it.
Thus, it is not the intent of this section to claim rights or contest
your rights to work written entirely by you; rather, the intent is to
exercise the right to control the distribution of derivative or
collective works based on the Program.
In addition, mere aggregation of another work not based on the Program
with the Program (or with a work based on the Program) on a volume of
a storage or distribution medium does not bring the other work under
the scope of this License.
3. You may copy and distribute the Program (or a work based on it,
under Section 2) in object code or executable form under the terms of
Sections 1 and 2 above provided that you also do one of the following:
a) Accompany it with the complete corresponding machine-readable
source code, which must be distributed under the terms of Sections
1 and 2 above on a medium customarily used for software interchange; or,
b) Accompany it with a written offer, valid for at least three
years, to give any third party, for a charge no more than your
cost of physically performing source distribution, a complete
machine-readable copy of the corresponding source code, to be
distributed under the terms of Sections 1 and 2 above on a medium
customarily used for software interchange; or,
c) Accompany it with the information you received as to the offer
to distribute corresponding source code. (This alternative is
allowed only for noncommercial distribution and only if you
received the program in object code or executable form with such
an offer, in accord with Subsection b above.)
The source code for a work means the preferred form of the work for
making modifications to it. For an executable work, complete source
code means all the source code for all modules it contains, plus any
associated interface definition files, plus the scripts used to
control compilation and installation of the executable. However, as a
special exception, the source code distributed need not include
anything that is normally distributed (in either source or binary
form) with the major components (compiler, kernel, and so on) of the
operating system on which the executable runs, unless that component
itself accompanies the executable.
If distribution of executable or object code is made by offering
access to copy from a designated place, then offering equivalent
access to copy the source code from the same place counts as
distribution of the source code, even though third parties are not
compelled to copy the source along with the object code.
4. You may not copy, modify, sublicense, or distribute the Program
except as expressly provided under this License. Any attempt
otherwise to copy, modify, sublicense or distribute the Program is
void, and will automatically terminate your rights under this License.
However, parties who have received copies, or rights, from you under
this License will not have their licenses terminated so long as such
parties remain in full compliance.
5. You are not required to accept this License, since you have not
signed it. However, nothing else grants you permission to modify or
distribute the Program or its derivative works. These actions are
prohibited by law if you do not accept this License. Therefore, by
modifying or distributing the Program (or any work based on the
Program), you indicate your acceptance of this License to do so, and
all its terms and conditions for copying, distributing or modifying
the Program or works based on it.
6. Each time you redistribute the Program (or any work based on the
Program), the recipient automatically receives a license from the
original licensor to copy, distribute or modify the Program subject to
these terms and conditions. You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.
You are not responsible for enforcing compliance by third parties to
this License.
Each contributor grants you a non-exclusive, worldwide, royalty-free
patent license under the contributor's essential patent claims, to
make, use, sell, offer for sale, import and otherwise run, modify and
propagate the contents of its contributor version.
In the following three paragraphs, a "patent license" is any express
agreement or commitment, however denominated, not to enforce a patent
(such as an express permission to practice a patent or covenant not to
sue for patent infringement). To "grant" such a patent license to a
party means to make such an agreement or commitment not to enforce a
patent against the party.
If you convey a covered work, knowingly relying on a patent license,
and the Corresponding Source of the work is not available for anyone
to copy, free of charge and under the terms of this License, through a
publicly available network server or other readily accessible means,
then you must either (1) cause the Corresponding Source to be so
available, or (2) arrange to deprive yourself of the benefit of the
patent license for this particular work, or (3) arrange, in a manner
consistent with the requirements of this License, to extend the patent
license to downstream recipients. "Knowingly relying" means you have
actual knowledge that, but for the patent license, your conveying the
covered work in a country, or your recipient's use of the covered work
in a country, would infringe one or more identifiable patents in that
country that you have reason to believe are valid.
If, pursuant to or in connection with a single transaction or
arrangement, you convey, or propagate by procuring conveyance of, a
covered work, and grant a patent license to some of the parties
receiving the covered work authorizing them to use, propagate, modify
or convey a specific copy of the covered work, then the patent license
you grant is automatically extended to all recipients of the covered
work and works based on it.
A patent license is "discriminatory" if it does not include within
the scope of its coverage, prohibits the exercise of, or is
conditioned on the non-exercise of one or more of the rights that are
specifically granted under this License. You may not convey a covered
work if you are a party to an arrangement with a third party that is
in the business of distributing software, under which you make payment
to the third party based on the extent of your activity of conveying
the work, and under which the third party grants, to any of the
parties who would receive the covered work from you, a discriminatory
patent license (a) in connection with copies of the covered work
conveyed by you (or copies made from those copies), or (b) primarily
for and in connection with specific products or compilations that
contain the covered work, unless you entered into that arrangement,
or that patent license was granted, prior to 28 March 2007.
Nothing in this License shall be construed as excluding or limiting
any implied license or other defenses to infringement that may
otherwise be available to you under applicable patent law.
12. No Surrender of Others' Freedom.
If conditions are imposed on you (whether by court order, agreement or
7. If, as a consequence of a court judgment or allegation of patent
infringement or for any other reason (not limited to patent issues),
conditions are imposed on you (whether by court order, agreement or
otherwise) that contradict the conditions of this License, they do not
excuse you from the conditions of this License. If you cannot convey a
covered work so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you may
not convey it at all. For example, if you agree to terms that obligate you
to collect a royalty for further conveying from those to whom you convey
the Program, the only way you could satisfy both those terms and this
License would be to refrain entirely from conveying the Program.
excuse you from the conditions of this License. If you cannot
distribute so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you
may not distribute the Program at all. For example, if a patent
license would not permit royalty-free redistribution of the Program by
all those who receive copies directly or indirectly through you, then
the only way you could satisfy both it and this License would be to
refrain entirely from distribution of the Program.
13. Use with the GNU Affero General Public License.
If any portion of this section is held invalid or unenforceable under
any particular circumstance, the balance of the section is intended to
apply and the section as a whole is intended to apply in other
circumstances.
Notwithstanding any other provision of this License, you have
permission to link or combine any covered work with a work licensed
under version 3 of the GNU Affero General Public License into a single
combined work, and to convey the resulting work. The terms of this
License will continue to apply to the part which is the covered work,
but the special requirements of the GNU Affero General Public License,
section 13, concerning interaction through a network will apply to the
combination as such.
It is not the purpose of this section to induce you to infringe any
patents or other property right claims or to contest validity of any
such claims; this section has the sole purpose of protecting the
integrity of the free software distribution system, which is
implemented by public license practices. Many people have made
generous contributions to the wide range of software distributed
through that system in reliance on consistent application of that
system; it is up to the author/donor to decide if he or she is willing
to distribute software through any other system and a licensee cannot
impose that choice.
14. Revised Versions of this License.
This section is intended to make thoroughly clear what is believed to
be a consequence of the rest of this License.
The Free Software Foundation may publish revised and/or new versions of
the GNU General Public License from time to time. Such new versions will
8. If the distribution and/or use of the Program is restricted in
certain countries either by patents or by copyrighted interfaces, the
original copyright holder who places the Program under this License
may add an explicit geographical distribution limitation excluding
those countries, so that distribution is permitted only in or among
countries not thus excluded. In such case, this License incorporates
the limitation as if written in the body of this License.
9. The Free Software Foundation may publish revised and/or new versions
of the General Public License from time to time. Such new versions will
be similar in spirit to the present version, but may differ in detail to
address new problems or concerns.
Each version is given a distinguishing version number. If the
Program specifies that a certain numbered version of the GNU General
Public License "or any later version" applies to it, you have the
option of following the terms and conditions either of that numbered
version or of any later version published by the Free Software
Foundation. If the Program does not specify a version number of the
GNU General Public License, you may choose any version ever published
by the Free Software Foundation.
Each version is given a distinguishing version number. If the Program
specifies a version number of this License which applies to it and "any
later version", you have the option of following the terms and conditions
either of that version or of any later version published by the Free
Software Foundation. If the Program does not specify a version number of
this License, you may choose any version ever published by the Free Software
Foundation.
If the Program specifies that a proxy can decide which future
versions of the GNU General Public License can be used, that proxy's
public statement of acceptance of a version permanently authorizes you
to choose that version for the Program.
10. If you wish to incorporate parts of the Program into other free
programs whose distribution conditions are different, write to the author
to ask for permission. For software which is copyrighted by the Free
Software Foundation, write to the Free Software Foundation; we sometimes
make exceptions for this. Our decision will be guided by the two goals
of preserving the free status of all derivatives of our free software and
of promoting the sharing and reuse of software generally.
Later license versions may give you additional or different
permissions. However, no additional obligations are imposed on any
author or copyright holder as a result of your choosing to follow a
later version.
NO WARRANTY
15. Disclaimer of Warranty.
11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN
OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS
TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE
PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
REPAIR OR CORRECTION.
THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
16. Limitation of Liability.
IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
SUCH DAMAGES.
17. Interpretation of Sections 15 and 16.
If the disclaimer of warranty and limitation of liability provided
above cannot be given local legal effect according to their terms,
reviewing courts shall apply local law that most closely approximates
an absolute waiver of all civil liability in connection with the
Program, unless a warranty or assumption of liability accompanies a
copy of the Program in return for a fee.
12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
POSSIBILITY OF SUCH DAMAGES.
END OF TERMS AND CONDITIONS
@ -628,15 +287,15 @@ free software which everyone can redistribute and change under these terms.
To do so, attach the following notices to the program. It is safest
to attach them to the start of each source file to most effectively
state the exclusion of warranty; and each file should have at least
convey the exclusion of warranty; and each file should have at least
the "copyright" line and a pointer to where the full notice is found.
<one line to give the program's name and a brief idea of what it does.>
Copyright (C) <year> <name of author>
This program is free software: you can redistribute it and/or modify
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
@ -644,31 +303,38 @@ the "copyright" line and a pointer to where the full notice is found.
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
Also add information on how to contact you by electronic and paper mail.
If the program does terminal interaction, make it output a short
notice like this when it starts in an interactive mode:
If the program is interactive, make it output a short notice like this
when it starts in an interactive mode:
<program> Copyright (C) <year> <name of author>
This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
Gnomovision version 69, Copyright (C) year name of author
Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
This is free software, and you are welcome to redistribute it
under certain conditions; type `show c' for details.
The hypothetical commands `show w' and `show c' should show the appropriate
parts of the General Public License. Of course, your program's commands
might be different; for a GUI interface, you would use an "about box".
parts of the General Public License. Of course, the commands you use may
be called something other than `show w' and `show c'; they could even be
mouse-clicks or menu items--whatever suits your program.
You should also get your employer (if you work as a programmer) or school,
if any, to sign a "copyright disclaimer" for the program, if necessary.
For more information on this, and how to apply and follow the GNU GPL, see
<http://www.gnu.org/licenses/>.
You should also get your employer (if you work as a programmer) or your
school, if any, to sign a "copyright disclaimer" for the program, if
necessary. Here is a sample; alter the names:
Yoyodyne, Inc., hereby disclaims all copyright interest in the program
`Gnomovision' (which makes passes at compilers) written by James Hacker.
<signature of Ty Coon>, 1 April 1989
Ty Coon, President of Vice
This General Public License does not permit incorporating your program into
proprietary programs. If your program is a subroutine library, you may
consider it more useful to permit linking proprietary applications with the
library. If this is what you want to do, use the GNU Lesser General
Public License instead of this License.
The GNU General Public License does not permit incorporating your program
into proprietary programs. If your program is a subroutine library, you
may consider it more useful to permit linking proprietary applications with
the library. If this is what you want to do, use the GNU Lesser General
Public License instead of this License. But first, please read
<http://www.gnu.org/philosophy/why-not-lgpl.html>.

10
LICENSE
View File

@ -1,7 +1,7 @@
GNU GENERAL PUBLIC LICENSE
GNU GENERAL PUBLIC LICENSE
Version 2, June 1991
Copyright (C) 1989, 1991 Free Software Foundation, Inc., <http://fsf.org/>
Copyright (C) 1989, 1991 Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
@ -290,8 +290,8 @@ to attach them to the start of each source file to most effectively
convey the exclusion of warranty; and each file should have at least
the "copyright" line and a pointer to where the full notice is found.
{description}
Copyright (C) {year} {fullname}
<one line to give the program's name and a brief idea of what it does.>
Copyright (C) <year> <name of author>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
@ -329,7 +329,7 @@ necessary. Here is a sample; alter the names:
Yoyodyne, Inc., hereby disclaims all copyright interest in the program
`Gnomovision' (which makes passes at compilers) written by James Hacker.
{signature of Ty Coon}, 1 April 1989
<signature of Ty Coon>, 1 April 1989
Ty Coon, President of Vice
This General Public License does not permit incorporating your program into

16
README.md
View File

@ -1,6 +1,8 @@
# Grid
Data parallel C++ mathematical object library
Last update 2015/7/30
This library provides data parallel C++ container classes with internal memory layout
that is transformed to map efficiently to SIMD architectures. CSHIFT facilities
are provided, similar to HPF and cmfortran, and user control is given over the mapping of
@ -20,16 +22,16 @@ optimally use MPI, OpenMP and SIMD parallelism under the hood. This is a signifi
for most programmers.
The layout transformations are parametrised by the SIMD vector length. This adapts according to the architecture.
Presently SSE4 (128 bit) AVX, AVX2 (256 bit) and IMCI and AVX512 (512 bit) targets are supported.
Presently SSE4 (128 bit) AVX, AVX2 (256 bit) and IMCI and AVX512 (512 bit) targets are supported (ARM NEON on the way).
These are presented as
vRealF, vRealD, vComplexF, vComplexD
vRealF, vRealD, vComplexF, vComplexD
internal vector data types. These may be useful in themselves for other programmers.
The corresponding scalar types are named
RealF, RealD, ComplexF, ComplexD
RealF, RealD, ComplexF, ComplexD
MPI, OpenMP, and SIMD parallelism are present in the library.
@ -39,12 +41,18 @@ are examples:
./configure CXX=clang++ CXXFLAGS="-std=c++11 -O3 -msse4" --enable-simd=SSE4
./configure CXX=clang++ CXXFLAGS="-std=c++11 -O3 -mavx" --enable-simd=AVX1
./configure CXX=clang++ CXXFLAGS="-std=c++11 -O3 -mavx" --enable-simd=AVX
./configure CXX=clang++ CXXFLAGS="-std=c++11 -O3 -mavx2" --enable-simd=AVX2
./configure CXX=icpc CXXFLAGS="-std=c++11 -O3 -mmic" --enable-simd=AVX512 --host=none
Note: Before running configure it could be necessary to execute the script
script/filelist
For developers:
Use reconfigure_script in the scripts/ directory to create the autotools environment

61
TODO
View File

@ -1,22 +1,33 @@
RECENT
---------------
- Clean up HMC -- DONE
- LorentzScalar<GaugeField> gets Gauge link type (cleaner). -- DONE
- Simplified the integrators a bit. -- DONE
- Multi-timescale looks broken and operating on single timescale for now. -- DONE
- pass GaugeField as template param. -- DONE
- Reunitarise -- DONE
- Force Gradient -- DONE
- Prefer "RefreshInternal" or such like to "init" in naming -- DONE
- Parallel io improvements -- DONE
- Plaquette and link trace checks into nersc reader from the Grid_nersc_io.cc test. -- DONE
TODO:
---------------
* Forces; the UdSdU term in gauge force term is half of what I think it should
be. This is a consequence of taking ONLY the first term in:
dSg/dt = dU/dt dSdU + dUdag/dt dSdUdag
in the fermion force.
Now, S_mom = - tr Pmu Pmu ; Pmu anti-herm
.
d Smom/dt = - 2.0 tr Pmu Pmu = - dSg/dt = - tr Pmu [Umu dSdUmu + UmuDag dSdUmuDag]
.
=> Pmu = Umu dSdUmu
Where the norm is half expected.
This means we must double the force in the Test_xxx_force routines, and is the origin of the factor of two.
This 2x is applied by hand in the fermion routines and in the Test_rect_force routine.
Policies:
* Link smearing/boundary conds; Policy class based implementation ; framework more in place
* Support different boundary conditions (finite temp, chem. potential ... )
* Support different fermion representations?
- contained entirely within the integrator presently
@ -28,6 +39,8 @@ Policies:
- Lanczos
- Audit oIndex usage for cb behaviour
- Rectangle gauge actions.
Iwasaki,
Symanzik,
@ -41,6 +54,12 @@ Policies:
- FFTnD ?
- Gparity; hand opt use template specialisation elegance to enable the optimised paths ?
- Gparity force term; Gparity (R)HMC.
- Random number state save restore
- Mobius implementation clean up to rmove #if 0 stale code sequences
- CG -- profile carefully, kernel fusion, whole CG performance measurements.
================================================================
* Hacks and bug fixes to clean up and Audits
================================================================
@ -90,6 +109,20 @@ Not sure of status of this -- reverify. Things are working nicely now though.
======================================================================
======================================================================
======================================================================
RECENT
---------------
- Clean up HMC -- DONE
- LorentzScalar<GaugeField> gets Gauge link type (cleaner). -- DONE
- Simplified the integrators a bit. -- DONE
- Multi-timescale looks broken and operating on single timescale for now. -- DONE
- pass GaugeField as template param. -- DONE
- Reunitarise -- DONE
- Force Gradient -- DONE
- Prefer "RefreshInternal" or such like to "init" in naming -- DONE
- Parallel io improvements -- DONE
- Plaquette and link trace checks into nersc reader from the Grid_nersc_io.cc test. -- DONE
Done: Cayley, Partial , ContFrac force terms.
DONE

27
benchmarks/Benchmark_comms.cc
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./benchmarks/Benchmark_comms.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
using namespace std;

54
benchmarks/Benchmark_dwf.cc
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./benchmarks/Benchmark_dwf.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
using namespace std;
@ -16,10 +44,16 @@ struct scal {
Gamma::GammaT
};
bool overlapComms = false;
int main (int argc, char ** argv)
{
Grid_init(&argc,&argv);
if( GridCmdOptionExists(argv,argv+argc,"--asynch") ){
overlapComms = true;
}
int threads = GridThread::GetThreads();
std::cout<<GridLogMessage << "Grid is setup to use "<<threads<<" threads"<<std::endl;
@ -68,21 +102,27 @@ int main (int argc, char ** argv)
for(int mu=0;mu<Nd;mu++){
tmp = U[mu]*Cshift(src,mu+1,1);
ref=ref + tmp + Gamma(Gmu[mu])*tmp;
ref=ref + tmp - Gamma(Gmu[mu])*tmp;
tmp =adj(U[mu])*src;
tmp =Cshift(tmp,mu+1,-1);
ref=ref + tmp - Gamma(Gmu[mu])*tmp;
ref=ref + tmp + Gamma(Gmu[mu])*tmp;
}
ref = -0.5*ref;
}
RealD mass=0.1;
RealD M5 =1.8;
DomainWallFermionR Dw(Umu,*FGrid,*FrbGrid,*UGrid,*UrbGrid,mass,M5);
typename DomainWallFermionR::ImplParams params;
params.overlapCommsCompute = overlapComms;
RealD NP = UGrid->_Nprocessors;
DomainWallFermionR Dw(Umu,*FGrid,*FrbGrid,*UGrid,*UrbGrid,mass,M5,params);
std::cout<<GridLogMessage << "Calling Dw"<<std::endl;
int ncall=10000;
int ncall=100;
{
double t0=usecond();
for(int i=0;i<ncall;i++){
@ -97,6 +137,7 @@ int main (int argc, char ** argv)
std::cout<<GridLogMessage << "norm result "<< norm2(result)<<std::endl;
std::cout<<GridLogMessage << "norm ref "<< norm2(ref)<<std::endl;
std::cout<<GridLogMessage << "mflop/s = "<< flops/(t1-t0)<<std::endl;
std::cout<<GridLogMessage << "mflop/s per node = "<< flops/(t1-t0)/NP<<std::endl;
err = ref-result;
std::cout<<GridLogMessage << "norm diff "<< norm2(err)<<std::endl;
Dw.Report();
@ -111,13 +152,13 @@ int main (int argc, char ** argv)
// ref = src - Gamma(Gamma::GammaX)* src ; // 1+gamma_x
tmp = U[mu]*Cshift(src,mu+1,1);
for(int i=0;i<ref._odata.size();i++){
ref._odata[i]+= tmp._odata[i] - Gamma(Gmu[mu])*tmp._odata[i]; ;
ref._odata[i]+= tmp._odata[i] + Gamma(Gmu[mu])*tmp._odata[i]; ;
}
tmp =adj(U[mu])*src;
tmp =Cshift(tmp,mu+1,-1);
for(int i=0;i<ref._odata.size();i++){
ref._odata[i]+= tmp._odata[i] + Gamma(Gmu[mu])*tmp._odata[i]; ;
ref._odata[i]+= tmp._odata[i] - Gamma(Gmu[mu])*tmp._odata[i]; ;
}
}
ref = -0.5*ref;
@ -154,6 +195,7 @@ int main (int argc, char ** argv)
double flops=(1344.0*volume*ncall)/2;
std::cout<<GridLogMessage << "Deo mflop/s = "<< flops/(t1-t0)<<std::endl;
std::cout<<GridLogMessage << "Deo mflop/s per node "<< flops/(t1-t0)/NP<<std::endl;
}
Dw.DhopEO(src_o,r_e,DaggerNo);

28
benchmarks/Benchmark_memory_asynch.cc
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./benchmarks/Benchmark_memory_asynch.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
using namespace std;

28
benchmarks/Benchmark_memory_bandwidth.cc
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./benchmarks/Benchmark_memory_bandwidth.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
using namespace std;

28
benchmarks/Benchmark_su3.cc
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./benchmarks/Benchmark_su3.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: Peter Boyle <peterboyle@Peters-MacBook-Pro-2.local>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
using namespace std;

56
benchmarks/Benchmark_wilson.cc
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./benchmarks/Benchmark_wilson.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
using namespace std;
@ -16,10 +44,15 @@ struct scal {
Gamma::GammaT
};
bool overlapComms = false;
int main (int argc, char ** argv)
{
Grid_init(&argc,&argv);
if( GridCmdOptionExists(argv,argv+argc,"--asynch") ){
overlapComms = true;
}
std::vector<int> latt_size = GridDefaultLatt();
std::vector<int> simd_layout = GridDefaultSimd(Nd,vComplex::Nsimd());
@ -57,11 +90,12 @@ int main (int argc, char ** argv)
Complex cone(1.0,0.0);
for(int nn=0;nn<Nd;nn++){
random(pRNG,U[nn]);
if(0) {
if (nn==-1) { U[nn]=zero; std::cout<<GridLogMessage << "zeroing gauge field in dir "<<nn<<std::endl; }
else { U[nn] = cone;std::cout<<GridLogMessage << "unit gauge field in dir "<<nn<<std::endl; }
if(1) {
if (nn!=2) { U[nn]=zero; std::cout<<GridLogMessage << "zeroing gauge field in dir "<<nn<<std::endl; }
// else { U[nn]= cone;std::cout<<GridLogMessage << "unit gauge field in dir "<<nn<<std::endl; }
else { std::cout<<GridLogMessage << "random gauge field in dir "<<nn<<std::endl; }
}
pokeIndex<LorentzIndex>(Umu,U[nn],nn);
PokeIndex<LorentzIndex>(Umu,U[nn],nn);
}
#endif
@ -75,19 +109,23 @@ int main (int argc, char ** argv)
// ref = src + Gamma(Gamma::GammaX)* src ; // 1-gamma_x
tmp = U[mu]*Cshift(src,mu,1);
for(int i=0;i<ref._odata.size();i++){
ref._odata[i]+= tmp._odata[i] + Gamma(Gmu[mu])*tmp._odata[i]; ;
ref._odata[i]+= tmp._odata[i] - Gamma(Gmu[mu])*tmp._odata[i]; ;
}
tmp =adj(U[mu])*src;
tmp =Cshift(tmp,mu,-1);
for(int i=0;i<ref._odata.size();i++){
ref._odata[i]+= tmp._odata[i] - Gamma(Gmu[mu])*tmp._odata[i]; ;
ref._odata[i]+= tmp._odata[i] + Gamma(Gmu[mu])*tmp._odata[i]; ;
}
}
}
ref = -0.5*ref;
RealD mass=0.1;
WilsonFermionR Dw(Umu,Grid,RBGrid,mass);
typename WilsonFermionR::ImplParams params;
params.overlapCommsCompute = overlapComms;
WilsonFermionR Dw(Umu,Grid,RBGrid,mass,params);
std::cout<<GridLogMessage << "Calling Dw"<<std::endl;
int ncall=1000;
@ -124,13 +162,13 @@ int main (int argc, char ** argv)
// ref = src - Gamma(Gamma::GammaX)* src ; // 1+gamma_x
tmp = U[mu]*Cshift(src,mu,1);
for(int i=0;i<ref._odata.size();i++){
ref._odata[i]+= tmp._odata[i] - Gamma(Gmu[mu])*tmp._odata[i]; ;
ref._odata[i]+= tmp._odata[i] + Gamma(Gmu[mu])*tmp._odata[i]; ;
}
tmp =adj(U[mu])*src;
tmp =Cshift(tmp,mu,-1);
for(int i=0;i<ref._odata.size();i++){
ref._odata[i]+= tmp._odata[i] + Gamma(Gmu[mu])*tmp._odata[i]; ;
ref._odata[i]+= tmp._odata[i] - Gamma(Gmu[mu])*tmp._odata[i]; ;
}
}
}

27
benchmarks/simple_su3_expr.cc
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./benchmarks/simple_su3_expr.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
using namespace std;

27
benchmarks/simple_su3_test.cc
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./benchmarks/simple_su3_test.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
using namespace std;

9821
configure vendored Executable file
View File

File diff suppressed because it is too large Load Diff

69
configure.ac
View File

@ -53,7 +53,6 @@ echo
echo Checking libraries
echo :::::::::::::::::::::::::::::::::::::::::::
AC_CHECK_FUNCS([gettimeofday])
#AC_CHECK_LIB([gmp],[__gmpf_init],,
@ -66,12 +65,18 @@ AC_CHECK_FUNCS([gettimeofday])
#Please install or provide the correct path to your installation
#Info at: http://www.mpfr.org/)])
#
# SIMD instructions selection
#
AC_ARG_ENABLE([simd],[AC_HELP_STRING([--enable-simd=SSE4|AVX|AVXFMA4|AVX2|AVX512|IMCI],\
[Select instructions to be SSE4.0, AVX 1.0, AVX 2.0+FMA, AVX 512, IMCI])],\
[ac_SIMD=${enable_simd}],[ac_SIMD=AVX2])
supported=no
ac_ZMM=no;
case ${ac_SIMD} in
SSE4)
echo Configuring for SSE4
@ -113,11 +118,13 @@ case ${ac_SIMD} in
echo Configuring for AVX512
AC_DEFINE([AVX512],[1],[AVX512 Intrinsics for Knights Landing] )
supported="cross compilation"
ac_ZMM=yes;
;;
IMCI)
echo Configuring for IMCI
AC_DEFINE([IMCI],[1],[IMCI Intrinsics for Knights Corner] )
supported="cross compilation"
ac_ZMM=yes;
;;
NEONv8)
echo Configuring for experimental ARMv8a support
@ -133,6 +140,17 @@ case ${ac_SIMD} in
;;
esac
case ${ac_ZMM} in
yes)
echo Enabling ZMM source code
;;
no)
echo Disabling ZMM source code
;;
esac
AM_CONDITIONAL(BUILD_ZMM,[ test "X${ac_ZMM}X" == "XyesX" ])
AC_ARG_ENABLE([precision],[AC_HELP_STRING([--enable-precision=single|double],[Select default word size of Real])],[ac_PRECISION=${enable_precision}],[ac_PRECISION=double])
case ${ac_PRECISION} in
single)
@ -145,6 +163,10 @@ case ${ac_PRECISION} in
;;
esac
#
# Comms selection
#
AC_ARG_ENABLE([comms],[AC_HELP_STRING([--enable-comms=none|mpi],[Select communications])],[ac_COMMS=${enable_comms}],[ac_COMMS=none])
case ${ac_COMMS} in
@ -156,14 +178,54 @@ case ${ac_COMMS} in
echo Configuring for MPI communications
AC_DEFINE([GRID_COMMS_MPI],[1],[GRID_COMMS_MPI] )
;;
shmem)
echo Configuring for SHMEM communications
AC_DEFINE([GRID_COMMS_SHMEM],[1],[GRID_COMMS_SHMEM] )
;;
*)
AC_MSG_ERROR([${ac_COMMS} unsupported --enable-comms option]);
;;
esac
AM_CONDITIONAL(BUILD_COMMS_SHMEM,[ test "X${ac_COMMS}X" == "XshmemX" ])
AM_CONDITIONAL(BUILD_COMMS_MPI,[ test "X${ac_COMMS}X" == "XmpiX" ])
AM_CONDITIONAL(BUILD_COMMS_NONE,[ test "X${ac_COMMS}X" == "XnoneX" ])
#
# RNG selection
#
AC_ARG_ENABLE([rng],[AC_HELP_STRING([--enable-rng=ranlux48|mt19937],\
[Select Random Number Generator to be used])],\
[ac_RNG=${enable_rng}],[ac_RNG=ranlux48])
case ${ac_RNG} in
ranlux48)
AC_DEFINE([RNG_RANLUX],[1],[RNG_RANLUX] )
;;
mt19937)
AC_DEFINE([RNG_MT19937],[1],[RNG_MT19937] )
;;
*)
AC_MSG_ERROR([${ac_RNG} unsupported --enable-rng option]);
;;
esac
#
# Chroma regression tests
#
AC_ARG_ENABLE([chroma],[AC_HELP_STRING([--enable-chroma],[Expect chroma compiled under c++11 ])],ac_CHROMA=yes,ac_CHROMA=no)
case ${ac_CHROMA} in
yes)
echo Enabling tests regressing to Chroma
;;
no)
echo Disabling tests regressing to Chroma
;;
*)
AC_MSG_ERROR([${ac_CHROMA} unsupported --enable-chroma option]);
;;
esac
AM_CONDITIONAL(BUILD_CHROMA_REGRESSION,[ test "X${ac_CHROMA}X" == "XyesX" ])
###################################################################
# Checks for doxygen support
@ -184,6 +246,7 @@ echo :::::::::::::::::::::::::::::::::::::::::::
AC_CONFIG_FILES(Makefile)
AC_CONFIG_FILES(lib/Makefile)
AC_CONFIG_FILES(tests/Makefile)
AC_CONFIG_FILES(tests/qdpxx/Makefile)
AC_CONFIG_FILES(benchmarks/Makefile)
AC_CONFIG_FILES(programs/Makefile)
AC_CONFIG_FILES(programs/Hadrons/Makefile)
@ -205,8 +268,10 @@ The following features are enabled:
- graphs and diagrams : `if test "x$enable_dot" = xyes; then echo yes; else echo no; fi`
- Supported SIMD flags : $SIMD_FLAGS
----------------------------------------------------------
- enabled simd support : ${ac_SIMD} (supported: $supported )
- enabled simd support : ${ac_SIMD} (config macro says supported: $supported )
- communications type : ${ac_COMMS}
- default precision : ${ac_PRECISION}
- RNG choice : ${ac_RNG}
"

27
gcc-bug-report/broken.cc
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./gcc-bug-report/broken.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <vector>
#include <complex>
#include <type_traits>

28
lib/Algorithms.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Algorithms.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_ALGORITHMS_H
#define GRID_ALGORITHMS_H

71
lib/AlignedAllocator.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/AlignedAllocator.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_ALIGNED_ALLOCATOR_H
#define GRID_ALIGNED_ALLOCATOR_H
@ -12,6 +40,9 @@
#ifdef HAVE_MM_MALLOC_H
#include <mm_malloc.h>
#endif
#ifdef GRID_COMMS_SHMEM
#include <mpp/shmem.h>
#endif
namespace Grid {
@ -44,21 +75,59 @@ public:
size_type max_size() const throw() { return size_t(-1) / sizeof(_Tp); }
pointer allocate(size_type __n, const void* = 0)
pointer allocate(size_type __n, const void* _p= 0)
{
#ifdef GRID_COMMS_SHMEM
_Tp *ptr = (_Tp *) shmem_align(__n*sizeof(_Tp),64);
#define PARANOID_SYMMETRIC_HEAP
#ifdef PARANOID_SYMMETRIC_HEAP
static void * bcast;
static long psync[_SHMEM_REDUCE_SYNC_SIZE];
bcast = (void *) ptr;
shmem_broadcast32((void *)&bcast,(void *)&bcast,sizeof(void *)/4,0,0,0,shmem_n_pes(),psync);
if ( bcast != ptr ) {
std::printf("inconsistent alloc pe %d %lx %lx \n",shmem_my_pe(),bcast,ptr);std::fflush(stdout);
BACKTRACEFILE();
exit(0);
}
assert( bcast == (void *) ptr);
#endif
#else
#ifdef HAVE_MM_MALLOC_H
_Tp * ptr = (_Tp *) _mm_malloc(__n*sizeof(_Tp),128);
#else
_Tp * ptr = (_Tp *) memalign(128,__n*sizeof(_Tp));
#endif
#endif
_Tp tmp;
#undef FIRST_TOUCH_OPTIMISE
#ifdef FIRST_TOUCH_OPTIMISE
#pragma omp parallel for
for(int i=0;i<__n;i++){
ptr[i]=tmp;
}
#endif
return ptr;
}
void deallocate(pointer __p, size_type) {
#ifdef GRID_COMMS_SHMEM
shmem_free((void *)__p);
#else
#ifdef HAVE_MM_MALLOC_H
_mm_free((void *)__p);
#else
free((void *)__p);
#endif
#endif
}
void construct(pointer __p, const _Tp& __val) { };

27
lib/Cartesian.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Cartesian.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_CARTESIAN_H
#define GRID_CARTESIAN_H

27
lib/Communicator.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Communicator.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_COMMUNICATOR_H
#define GRID_COMMUNICATOR_H

180
lib/Config.h.in Normal file
View File

@ -0,0 +1,180 @@
/* lib/Config.h.in. Generated from configure.ac by autoheader. */
/* AVX Intrinsics */
#undef AVX1
/* AVX2 Intrinsics */
#undef AVX2
/* AVX512 Intrinsics for Knights Landing */
#undef AVX512
/* AVX Intrinsics with FMA4 */
#undef AVXFMA4
/* EMPTY_SIMD only for DEBUGGING */
#undef EMPTY_SIMD
/* GRID_COMMS_MPI */
#undef GRID_COMMS_MPI
/* GRID_COMMS_NONE */
#undef GRID_COMMS_NONE
/* GRID_COMMS_SHMEM */
#undef GRID_COMMS_SHMEM
/* GRID_DEFAULT_PRECISION is DOUBLE */
#undef GRID_DEFAULT_PRECISION_DOUBLE
/* GRID_DEFAULT_PRECISION is SINGLE */
#undef GRID_DEFAULT_PRECISION_SINGLE
/* Support Altivec instructions */
#undef HAVE_ALTIVEC
/* Support AVX (Advanced Vector Extensions) instructions */
#undef HAVE_AVX
/* Support AVX2 (Advanced Vector Extensions 2) instructions */
#undef HAVE_AVX2
/* Define to 1 if you have the declaration of `be64toh', and to 0 if you
don't. */
#undef HAVE_DECL_BE64TOH
/* Define to 1 if you have the declaration of `ntohll', and to 0 if you don't.
*/
#undef HAVE_DECL_NTOHLL
/* Define to 1 if you have the <endian.h> header file. */
#undef HAVE_ENDIAN_H
/* Define to 1 if you have the <execinfo.h> header file. */
#undef HAVE_EXECINFO_H
/* Support FMA3 (Fused Multiply-Add) instructions */
#undef HAVE_FMA
/* Define to 1 if you have the `gettimeofday' function. */
#undef HAVE_GETTIMEOFDAY
/* Define to 1 if you have the <gmp.h> header file. */
#undef HAVE_GMP_H
/* Define to 1 if you have the <inttypes.h> header file. */
#undef HAVE_INTTYPES_H
/* Define to 1 if you have the <malloc.h> header file. */
#undef HAVE_MALLOC_H
/* Define to 1 if you have the <malloc/malloc.h> header file. */
#undef HAVE_MALLOC_MALLOC_H
/* Define to 1 if you have the <memory.h> header file. */
#undef HAVE_MEMORY_H
/* Support mmx instructions */
#undef HAVE_MMX
/* Define to 1 if you have the <mm_malloc.h> header file. */
#undef HAVE_MM_MALLOC_H
/* Support SSE (Streaming SIMD Extensions) instructions */
#undef HAVE_SSE
/* Support SSE2 (Streaming SIMD Extensions 2) instructions */
#undef HAVE_SSE2
/* Support SSE3 (Streaming SIMD Extensions 3) instructions */
#undef HAVE_SSE3
/* Support SSSE4.1 (Streaming SIMD Extensions 4.1) instructions */
#undef HAVE_SSE4_1
/* Support SSSE4.2 (Streaming SIMD Extensions 4.2) instructions */
#undef HAVE_SSE4_2
/* Support SSSE3 (Supplemental Streaming SIMD Extensions 3) instructions */
#undef HAVE_SSSE3
/* Define to 1 if you have the <stdint.h> header file. */
#undef HAVE_STDINT_H
/* Define to 1 if you have the <stdlib.h> header file. */
#undef HAVE_STDLIB_H
/* Define to 1 if you have the <strings.h> header file. */
#undef HAVE_STRINGS_H
/* Define to 1 if you have the <string.h> header file. */
#undef HAVE_STRING_H
/* Define to 1 if you have the <sys/stat.h> header file. */
#undef HAVE_SYS_STAT_H
/* Define to 1 if you have the <sys/types.h> header file. */
#undef HAVE_SYS_TYPES_H
/* Define to 1 if you have the <unistd.h> header file. */
#undef HAVE_UNISTD_H
/* IMCI Intrinsics for Knights Corner */
#undef IMCI
/* NEON ARMv8 Experimental support */
#undef NEONv8
/* Name of package */
#undef PACKAGE
/* Define to the address where bug reports for this package should be sent. */
#undef PACKAGE_BUGREPORT
/* Define to the full name of this package. */
#undef PACKAGE_NAME
/* Define to the full name and version of this package. */
#undef PACKAGE_STRING
/* Define to the one symbol short name of this package. */
#undef PACKAGE_TARNAME
/* Define to the version of this package. */
#undef PACKAGE_VERSION
/* RNG_MT19937 */
#undef RNG_MT19937
/* RNG_RANLUX */
#undef RNG_RANLUX
/* SSE4 Intrinsics */
#undef SSE4
/* Define to 1 if you have the ANSI C header files. */
#undef STDC_HEADERS
/* Version number of package */
#undef VERSION
/* Define for Solaris 2.5.1 so the uint32_t typedef from <sys/synch.h>,
<pthread.h>, or <semaphore.h> is not used. If the typedef were allowed, the
#define below would cause a syntax error. */
#undef _UINT32_T
/* Define for Solaris 2.5.1 so the uint64_t typedef from <sys/synch.h>,
<pthread.h>, or <semaphore.h> is not used. If the typedef were allowed, the
#define below would cause a syntax error. */
#undef _UINT64_T
/* Define to `unsigned int' if <sys/types.h> does not define. */
#undef size_t
/* Define to the type of an unsigned integer type of width exactly 32 bits if
such a type exists and the standard includes do not define it. */
#undef uint32_t
/* Define to the type of an unsigned integer type of width exactly 64 bits if
such a type exists and the standard includes do not define it. */
#undef uint64_t

31
lib/Cshift.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Cshift.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef _GRID_CSHIFT_H_
#define _GRID_CSHIFT_H_
@ -10,4 +37,8 @@
#ifdef GRID_COMMS_MPI
#include <cshift/Cshift_mpi.h>
#endif
#ifdef GRID_COMMS_SHMEM
#include <cshift/Cshift_mpi.h> // uses same implementation of communicator
#endif
#endif

37
lib/Grid.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Grid.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: azusayamaguchi <ayamaguc@YAMAKAZE.local>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
//
// Grid.h
// simd
@ -37,6 +66,7 @@
#include <AlignedAllocator.h>
#include <Simd.h>
#include <Threads.h>
#include <Lexicographic.h>
#include <Communicator.h>
#include <Cartesian.h>
#include <Tensors.h>
@ -44,11 +74,14 @@
#include <Cshift.h>
#include <Stencil.h>
#include <Algorithms.h>
#include <qcd/QCD.h>
#include <parallelIO/BinaryIO.h>
#include <qcd/QCD.h>
#include <parallelIO/NerscIO.h>
#include <Init.h>
#include <qcd/hmc/NerscCheckpointer.h>
#include <qcd/hmc/HmcRunner.h>
#endif

82
lib/Init.cc
View File

@ -1,3 +1,33 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Init.cc
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: Peter Boyle <peterboyle@MacBook-Pro.local>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
/****************************************************************************/
/* pab: Signal magic. Processor state dump is x86-64 specific */
/****************************************************************************/
@ -17,9 +47,6 @@
#define __X86_64
#ifdef HAVE_EXECINFO_H
#include <execinfo.h>
#endif
namespace Grid {
@ -144,9 +171,8 @@ std::string GridCmdVectorIntToString(const std::vector<int> & vec){
/////////////////////////////////////////////////////////
void Grid_init(int *argc,char ***argv)
{
#ifdef GRID_COMMS_MPI
MPI_Init(argc,argv);
#endif
CartesianCommunicator::Init(argc,argv);
// Parse command line args.
GridLogger::StopWatch.Start();
@ -166,7 +192,7 @@ void Grid_init(int *argc,char ***argv)
std::cout<<GridLogMessage<<"--mpi n.n.n.n : default MPI decomposition"<<std::endl;
std::cout<<GridLogMessage<<"--omp n : default number of OMP threads"<<std::endl;
std::cout<<GridLogMessage<<"--grid n.n.n.n : default Grid size"<<std::endl;
std::cout<<GridLogMessage<<"--log list : comma separted list of streams from Error,Warning,Message,Performance,Iterative,Debug"<<std::endl;
std::cout<<GridLogMessage<<"--log list : comma separted list of streams from Error,Warning,Message,Performance,Iterative,Integrator,Debug"<<std::endl;
}
if( GridCmdOptionExists(*argv,*argv+*argc,"--log") ){
@ -207,7 +233,37 @@ void Grid_init(int *argc,char ***argv)
std::cout<<GridLogMessage<<"\tvComplexD : "<<sizeof(vComplexD)*8 <<"bits ; " <<GridCmdVectorIntToString(GridDefaultSimd(4,vComplexD::Nsimd()))<<std::endl;
}
std::cout <<std::endl;
std::cout <<Logger::RED << "__|__|__|__|__"<< "|__|__|_"<<Logger::PURPLE<<"_|__|__|"<< "__|__|__|__|__"<<std::endl;
std::cout <<Logger::RED << "__|__|__|__|__"<< "|__|__|_"<<Logger::PURPLE<<"_|__|__|"<< "__|__|__|__|__"<<std::endl;
std::cout <<Logger::RED << "__|__| | | "<< "| | | "<<Logger::PURPLE<<" | | |"<< " | | | _|__"<<std::endl;
std::cout <<Logger::RED << "__|__ "<< " "<<Logger::PURPLE<<" "<< " _|__"<<std::endl;
std::cout <<Logger::RED << "__|_ "<<Logger::GREEN<<" GGGG "<<Logger::RED<<" RRRR "<<Logger::BLUE <<" III "<<Logger::PURPLE<<"DDDD "<<Logger::PURPLE<<" _|__"<<std::endl;
std::cout <<Logger::RED << "__|_ "<<Logger::GREEN<<"G "<<Logger::RED<<" R R "<<Logger::BLUE <<" I "<<Logger::PURPLE<<"D D "<<Logger::PURPLE<<" _|__"<<std::endl;
std::cout <<Logger::RED << "__|_ "<<Logger::GREEN<<"G "<<Logger::RED<<" R R "<<Logger::BLUE <<" I "<<Logger::PURPLE<<"D D"<<Logger::PURPLE<<" _|__"<<std::endl;
std::cout <<Logger::BLUE << "__|_ "<<Logger::GREEN<<"G GG "<<Logger::RED<<" RRRR "<<Logger::BLUE <<" I "<<Logger::PURPLE<<"D D"<<Logger::GREEN <<" _|__"<<std::endl;
std::cout <<Logger::BLUE << "__|_ "<<Logger::GREEN<<"G G "<<Logger::RED<<" R R "<<Logger::BLUE <<" I "<<Logger::PURPLE<<"D D "<<Logger::GREEN <<" _|__"<<std::endl;
std::cout <<Logger::BLUE << "__|_ "<<Logger::GREEN<<" GGGG "<<Logger::RED<<" R R "<<Logger::BLUE <<" III "<<Logger::PURPLE<<"DDDD "<<Logger::GREEN <<" _|__"<<std::endl;
std::cout <<Logger::BLUE << "__|__ "<< " "<<Logger::GREEN <<" "<< " _|__"<<std::endl;
std::cout <<Logger::BLUE << "__|__|__|__|__"<< "|__|__|_"<<Logger::GREEN <<"_|__|__|"<< "__|__|__|__|__"<<std::endl;
std::cout <<Logger::BLUE << "__|__|__|__|__"<< "|__|__|_"<<Logger::GREEN <<"_|__|__|"<< "__|__|__|__|__"<<std::endl;
std::cout <<Logger::BLUE << " | | | | "<< "| | | "<<Logger::GREEN <<" | | |"<< " | | | | "<<std::endl;
std::cout << std::endl;
std::cout << std::endl;
std::cout <<Logger::YELLOW<< std::endl;
std::cout << "Copyright (C) 2015 Peter Boyle, Azusa Yamaguchi, Guido Cossu, Antonin Portelli and other authors"<<std::endl;
std::cout << "Colours by Tadahito Boyle "<<std::endl;
std::cout << std::endl;
std::cout << "This program is free software; you can redistribute it and/or modify"<<std::endl;
std::cout << "it under the terms of the GNU General Public License as published by"<<std::endl;
std::cout << "the Free Software Foundation; either version 2 of the License, or"<<std::endl;
std::cout << "(at your option) any later version."<<std::endl;
std::cout << std::endl;
std::cout << "This program is distributed in the hope that it will be useful,"<<std::endl;
std::cout << "but WITHOUT ANY WARRANTY; without even the implied warranty of"<<std::endl;
std::cout << "MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the"<<std::endl;
std::cout << "GNU General Public License for more details."<<std::endl;
std::cout << Logger::BLACK <<std::endl;
}
@ -224,7 +280,6 @@ double usecond(void) {
return 1.0*tv.tv_usec + 1.0e6*tv.tv_sec;
}
#define _NBACKTRACE (256)
void * Grid_backtrace_buffer[_NBACKTRACE];
void Grid_sa_signal_handler(int sig,siginfo_t *si,void * ptr)
@ -239,6 +294,7 @@ void Grid_sa_signal_handler(int sig,siginfo_t *si,void * ptr)
ucontext_t * uc= (ucontext_t *)ptr;
struct sigcontext *sc = (struct sigcontext *)&uc->uc_mcontext;
printf(" instruction %llx\n",(unsigned long long)sc->rip);
#define REG(A) printf(" %s %lx\n",#A,sc-> A);
REG(rdi);
@ -261,13 +317,7 @@ void Grid_sa_signal_handler(int sig,siginfo_t *si,void * ptr)
REG(r14);
REG(r15);
#endif
#ifdef HAVE_EXECINFO_H
int symbols = backtrace (Grid_backtrace_buffer,_NBACKTRACE);
char **strings = backtrace_symbols(Grid_backtrace_buffer,symbols);
for (int i = 0; i < symbols; i++){
printf ("%s\n", strings[i]);
}
#endif
BACKTRACE();
exit(0);
return;
};

30
lib/Init.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Init.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_INIT_H
#define GRID_INIT_H
@ -21,6 +49,8 @@ namespace Grid {
std::string GridCmdOptionPayload(char ** begin, char ** end, const std::string & option);
bool GridCmdOptionExists(char** begin, char** end, const std::string& option);
std::string GridCmdVectorIntToString(const std::vector<int> & vec);
void GridCmdOptionCSL(std::string str,std::vector<std::string> & vec);
void GridCmdOptionIntVector(std::string &str,std::vector<int> & vec);
void GridParseLayout(char **argv,int argc,
std::vector<int> &latt,

27
lib/Lattice.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Lattice.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_H
#define GRID_LATTICE_H

32
lib/Lexicographic.h Normal file
View File

@ -0,0 +1,32 @@
#ifndef GRID_LEXICOGRAPHIC_H
#define GRID_LEXICOGRAPHIC_H
namespace Grid{
class Lexicographic {
public:
static inline void CoorFromIndex (std::vector<int>& coor,int index,std::vector<int> &dims){
int nd= dims.size();
coor.resize(nd);
for(int d=0;d<nd;d++){
coor[d] = index % dims[d];
index = index / dims[d];
}
}
static inline void IndexFromCoor (std::vector<int>& coor,int &index,std::vector<int> &dims){
int nd=dims.size();
int stride=1;
index=0;
for(int d=0;d<nd;d++){
index = index+stride*coor[d];
stride=stride*dims[d];
}
}
};
}
#endif

79
lib/Log.cc
View File

@ -1,16 +1,56 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Log.cc
Copyright (C) 2015
Author: Antonin Portelli <antonin.portelli@me.com>
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
namespace Grid {
GridStopWatch Logger::StopWatch;
std::ostream Logger::devnull(0);
std::string Logger::BLACK("\033[30m");
std::string Logger::RED("\033[31m");
std::string Logger::GREEN("\033[32m");
std::string Logger::YELLOW("\033[33m");
std::string Logger::BLUE("\033[34m");
std::string Logger::PURPLE("\033[35m");
std::string Logger::CYAN("\033[36m");
std::string Logger::WHITE("\033[37m");
std::string Logger::NORMAL("\033[0;39m");
GridLogger GridLogError (1,"Error");
GridLogger GridLogWarning (1,"Warning");
GridLogger GridLogMessage (1,"Message");
GridLogger GridLogDebug (1,"Debug");
GridLogger GridLogPerformance(1,"Performance");
GridLogger GridLogIterative (1,"Iterative");
GridLogger GridLogError (1,"Error",Logger::RED);
GridLogger GridLogWarning (1,"Warning",Logger::YELLOW);
GridLogger GridLogMessage (1,"Message",Logger::BLACK);
GridLogger GridLogDebug (1,"Debug",Logger::PURPLE);
GridLogger GridLogPerformance(1,"Performance",Logger::GREEN);
GridLogger GridLogIterative (1,"Iterative",Logger::BLUE);
GridLogger GridLogIntegrator (1,"Integrator",Logger::BLUE);
void GridLogConfigure(std::vector<std::string> &logstreams)
{
@ -20,6 +60,20 @@ void GridLogConfigure(std::vector<std::string> &logstreams)
GridLogIterative.Active(0);
GridLogDebug.Active(0);
GridLogPerformance.Active(0);
GridLogIntegrator.Active(0);
int blackAndWhite = 1;
if(blackAndWhite){
Logger::BLACK = std::string("");
Logger::RED =Logger::BLACK;
Logger::GREEN =Logger::BLACK;
Logger::YELLOW =Logger::BLACK;
Logger::BLUE =Logger::BLACK;
Logger::PURPLE =Logger::BLACK;
Logger::CYAN =Logger::BLACK;
Logger::WHITE =Logger::BLACK;
Logger::NORMAL =Logger::BLACK;
}
for(int i=0;i<logstreams.size();i++){
if ( logstreams[i]== std::string("Error") ) GridLogError.Active(1);
@ -28,6 +82,7 @@ void GridLogConfigure(std::vector<std::string> &logstreams)
if ( logstreams[i]== std::string("Iterative") ) GridLogIterative.Active(1);
if ( logstreams[i]== std::string("Debug") ) GridLogDebug.Active(1);
if ( logstreams[i]== std::string("Performance") ) GridLogPerformance.Active(1);
if ( logstreams[i]== std::string("Integrator" ) ) GridLogIntegrator.Active(1);
}
}
@ -36,13 +91,16 @@ void GridLogConfigure(std::vector<std::string> &logstreams)
////////////////////////////////////////////////////////////
void Grid_quiesce_nodes(void)
{
#ifdef GRID_COMMS_MPI
int me;
#ifdef GRID_COMMS_MPI
MPI_Comm_rank(MPI_COMM_WORLD,&me);
#endif
#ifdef GRID_COMMS_SHMEM
me = shmem_my_pe();
#endif
if ( me ) {
std::cout.setstate(std::ios::badbit);
}
#endif
}
void Grid_unquiesce_nodes(void)
@ -52,11 +110,6 @@ void Grid_unquiesce_nodes(void)
#endif
}
std::ostream& operator<< (std::ostream& stream, const GridTime& time)
{
stream << time.count()<<" ms";
return stream;
}
}

93
lib/Log.h
View File

@ -1,21 +1,65 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Log.h
Copyright (C) 2015
Author: Antonin Portelli <antonin.portelli@me.com>
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LOG_H
#define GRID_LOG_H
#ifdef HAVE_EXECINFO_H
#include <execinfo.h>
#endif
namespace Grid {
// Dress the output; use std::chrono for time stamping via the StopWatch class
int Rank(void); // used for early stage debug before library init
std::ostream& operator<< (std::ostream& stream, const GridTime& time);
class Logger {
protected:
int active;
std::string name, topName;
std::string name, topName, COLOUR;
public:
static GridStopWatch StopWatch;
static std::ostream devnull;
static std::string BLACK;
static std::string RED ;
static std::string GREEN;
static std::string YELLOW;
static std::string BLUE ;
static std::string PURPLE;
static std::string CYAN ;
static std::string WHITE ;
static std::string NORMAL;
Logger(std::string topNm, int on, std::string nm)
: active(on), name(nm), topName(topNm) {};
Logger(std::string topNm, int on, std::string nm,std::string col)
: active(on), name(nm), topName(topNm), COLOUR(col) {};
void Active(int on) {active = on;};
int isActive(void) {return active;};
@ -25,9 +69,9 @@ public:
StopWatch.Stop();
GridTime now = StopWatch.Elapsed();
StopWatch.Start();
stream << std::setw(8) << std::left << log.topName << " : ";
stream << std::setw(12) << std::left << log.name << " : ";
stream << now << " : ";
stream << BLACK<< log.topName << BLACK<< " : ";
stream << log.COLOUR <<std::setw(10) << std::left << log.name << BLACK << " : ";
stream << YELLOW<< now <<BLACK << " : " << log.COLOUR;
return stream;
} else {
return devnull;
@ -38,7 +82,7 @@ public:
class GridLogger: public Logger {
public:
GridLogger(int on, std::string nm): Logger("Grid", on, nm){};
GridLogger(int on, std::string nm, std::string col = Logger::BLACK): Logger("Grid", on, nm, col){};
};
void GridLogConfigure(std::vector<std::string> &logstreams);
@ -49,6 +93,39 @@ extern GridLogger GridLogMessage;
extern GridLogger GridLogDebug ;
extern GridLogger GridLogPerformance;
extern GridLogger GridLogIterative ;
extern GridLogger GridLogIntegrator ;
#define _NBACKTRACE (256)
extern void * Grid_backtrace_buffer[_NBACKTRACE];
#define BACKTRACEFILE() {\
char string[20]; \
std::sprintf(string,"backtrace.%d",Rank()); \
std::FILE * fp = std::fopen(string,"w"); \
BACKTRACEFP(fp)\
std::fclose(fp); \
}
#ifdef HAVE_EXECINFO_H
#define BACKTRACEFP(fp) { \
int symbols = backtrace (Grid_backtrace_buffer,_NBACKTRACE);\
char **strings = backtrace_symbols(Grid_backtrace_buffer,symbols);\
for (int i = 0; i < symbols; i++){\
std::fprintf (fp,"BackTrace Strings: %d %s\n",i, strings[i]); std::fflush(fp); \
}\
}
#else
#define BACKTRACEFP(fp) { \
std::fprintf (fp,"BT %d %lx\n",0, __builtin_return_address(0)); std::fflush(fp); \
std::fprintf (fp,"BT %d %lx\n",1, __builtin_return_address(1)); std::fflush(fp); \
std::fprintf (fp,"BT %d %lx\n",2, __builtin_return_address(2)); std::fflush(fp); \
std::fprintf (fp,"BT %d %lx\n",3, __builtin_return_address(3)); std::fflush(fp); \
}
#endif
#define BACKTRACE() BACKTRACEFP(stdout)
}
#endif

4
lib/Make.inc
View File

@ -1,4 +1,4 @@
HFILES=./algorithms/approx/bigfloat.h ./algorithms/approx/bigfloat_double.h ./algorithms/approx/Chebyshev.h ./algorithms/approx/MultiShiftFunction.h ./algorithms/approx/Remez.h ./algorithms/approx/Zolotarev.h ./algorithms/CoarsenedMatrix.h ./algorithms/iterative/AdefGeneric.h ./algorithms/iterative/ConjugateGradient.h ./algorithms/iterative/ConjugateGradientMultiShift.h ./algorithms/iterative/ConjugateResidual.h ./algorithms/iterative/ImplicitlyRestartedLanczos.h ./algorithms/iterative/MatrixUtils.h ./algorithms/iterative/NormalEquations.h ./algorithms/iterative/PrecConjugateResidual.h ./algorithms/iterative/PrecGeneralisedConjugateResidual.h ./algorithms/iterative/SchurRedBlack.h ./algorithms/LinearOperator.h ./algorithms/Preconditioner.h ./algorithms/SparseMatrix.h ./Algorithms.h ./AlignedAllocator.h ./cartesian/Cartesian_base.h ./cartesian/Cartesian_full.h ./cartesian/Cartesian_red_black.h ./Cartesian.h ./communicator/Communicator_base.h ./Communicator.h ./Config.h ./cshift/Cshift_common.h ./cshift/Cshift_mpi.h ./cshift/Cshift_none.h ./Cshift.h ./Grid.h ./Init.h ./lattice/Lattice_arith.h ./lattice/Lattice_base.h ./lattice/Lattice_comparison.h ./lattice/Lattice_comparison_utils.h ./lattice/Lattice_conformable.h ./lattice/Lattice_coordinate.h ./lattice/Lattice_ET.h ./lattice/Lattice_local.h ./lattice/Lattice_overload.h ./lattice/Lattice_peekpoke.h ./lattice/Lattice_reality.h ./lattice/Lattice_reduction.h ./lattice/Lattice_rng.h ./lattice/Lattice_trace.h ./lattice/Lattice_transfer.h ./lattice/Lattice_transpose.h ./lattice/Lattice_unary.h ./lattice/Lattice_where.h ./Lattice.h ./Log.h ./Old/Tensor_peek.h ./Old/Tensor_poke.h ./parallelIO/BinaryIO.h ./parallelIO/NerscIO.h ./pugixml/pugixml.h ./qcd/action/ActionBase.h ./qcd/action/ActionParams.h ./qcd/action/Actions.h ./qcd/action/fermion/CayleyFermion5D.h ./qcd/action/fermion/ContinuedFractionFermion5D.h ./qcd/action/fermion/DomainWallFermion.h ./qcd/action/fermion/FermionOperator.h ./qcd/action/fermion/FermionOperatorImpl.h ./qcd/action/fermion/g5HermitianLinop.h ./qcd/action/fermion/MobiusFermion.h ./qcd/action/fermion/MobiusZolotarevFermion.h ./qcd/action/fermion/OverlapWilsonCayleyTanhFermion.h ./qcd/action/fermion/OverlapWilsonCayleyZolotarevFermion.h ./qcd/action/fermion/OverlapWilsonContfracTanhFermion.h ./qcd/action/fermion/OverlapWilsonContfracZolotarevFermion.h ./qcd/action/fermion/OverlapWilsonPartialFractionTanhFermion.h ./qcd/action/fermion/OverlapWilsonPartialFractionZolotarevFermion.h ./qcd/action/fermion/PartialFractionFermion5D.h ./qcd/action/fermion/ScaledShamirFermion.h ./qcd/action/fermion/ShamirZolotarevFermion.h ./qcd/action/fermion/WilsonCompressor.h ./qcd/action/fermion/WilsonFermion.h ./qcd/action/fermion/WilsonFermion5D.h ./qcd/action/fermion/WilsonKernels.h ./qcd/action/gauge/WilsonGaugeAction.h ./qcd/action/pseudofermion/EvenOddSchurDifferentiable.h ./qcd/action/pseudofermion/OneFlavourEvenOddRational.h ./qcd/action/pseudofermion/OneFlavourEvenOddRationalRatio.h ./qcd/action/pseudofermion/OneFlavourRational.h ./qcd/action/pseudofermion/OneFlavourRationalRatio.h ./qcd/action/pseudofermion/TwoFlavour.h ./qcd/action/pseudofermion/TwoFlavourEvenOdd.h ./qcd/action/pseudofermion/TwoFlavourEvenOddRatio.h ./qcd/action/pseudofermion/TwoFlavourRatio.h ./qcd/hmc/HMC.h ./qcd/hmc/integrators/Integrator.h ./qcd/hmc/integrators/Integrator_algorithm.h ./qcd/QCD.h ./qcd/spin/Dirac.h ./qcd/spin/TwoSpinor.h ./qcd/utils/CovariantCshift.h ./qcd/utils/LinalgUtils.h ./qcd/utils/SpaceTimeGrid.h ./qcd/utils/SUn.h ./qcd/utils/WilsonLoops.h ./serialisation/BaseIO.h ./serialisation/BinaryIO.h ./serialisation/MacroMagic.h ./serialisation/Serialisation.h ./serialisation/TextIO.h ./serialisation/XmlIO.h ./simd/Grid_avx.h ./simd/Grid_avx512.h ./simd/Grid_empty.h ./simd/Grid_imci.h ./simd/Grid_neon.h ./simd/Grid_qpx.h ./simd/Grid_sse4.h ./simd/Grid_vector_types.h ./simd/Grid_vector_unops.h ./Simd.h ./stencil/Lebesgue.h ./Stencil.h ./tensors/Tensor_arith.h ./tensors/Tensor_arith_add.h ./tensors/Tensor_arith_mac.h ./tensors/Tensor_arith_mul.h ./tensors/Tensor_arith_scalar.h ./tensors/Tensor_arith_sub.h ./tensors/Tensor_class.h ./tensors/Tensor_determinant.h ./tensors/Tensor_exp.h ./tensors/Tensor_extract_merge.h ./tensors/Tensor_index.h ./tensors/Tensor_inner.h ./tensors/Tensor_logical.h ./tensors/Tensor_outer.h ./tensors/Tensor_reality.h ./tensors/Tensor_Ta.h ./tensors/Tensor_trace.h ./tensors/Tensor_traits.h ./tensors/Tensor_transpose.h ./tensors/Tensor_unary.h ./Tensors.h ./Threads.h ./Timer.h
HFILES=./algorithms/approx/bigfloat.h ./algorithms/approx/bigfloat_double.h ./algorithms/approx/Chebyshev.h ./algorithms/approx/MultiShiftFunction.h ./algorithms/approx/Remez.h ./algorithms/approx/Zolotarev.h ./algorithms/CoarsenedMatrix.h ./algorithms/iterative/AdefGeneric.h ./algorithms/iterative/ConjugateGradient.h ./algorithms/iterative/ConjugateGradientMultiShift.h ./algorithms/iterative/ConjugateResidual.h ./algorithms/iterative/DenseMatrix.h ./algorithms/iterative/EigenSort.h ./algorithms/iterative/Francis.h ./algorithms/iterative/Householder.h ./algorithms/iterative/ImplicitlyRestartedLanczos.h ./algorithms/iterative/Matrix.h ./algorithms/iterative/MatrixUtils.h ./algorithms/iterative/NormalEquations.h ./algorithms/iterative/PrecConjugateResidual.h ./algorithms/iterative/PrecGeneralisedConjugateResidual.h ./algorithms/iterative/SchurRedBlack.h ./algorithms/LinearOperator.h ./algorithms/Preconditioner.h ./algorithms/SparseMatrix.h ./Algorithms.h ./AlignedAllocator.h ./cartesian/Cartesian_base.h ./cartesian/Cartesian_full.h ./cartesian/Cartesian_red_black.h ./Cartesian.h ./communicator/Communicator_base.h ./Communicator.h ./cshift/Cshift_common.h ./cshift/Cshift_mpi.h ./cshift/Cshift_none.h ./Cshift.h ./Grid.h ./Init.h ./lattice/Lattice_arith.h ./lattice/Lattice_base.h ./lattice/Lattice_comparison.h ./lattice/Lattice_comparison_utils.h ./lattice/Lattice_conformable.h ./lattice/Lattice_coordinate.h ./lattice/Lattice_ET.h ./lattice/Lattice_local.h ./lattice/Lattice_overload.h ./lattice/Lattice_peekpoke.h ./lattice/Lattice_reality.h ./lattice/Lattice_reduction.h ./lattice/Lattice_rng.h ./lattice/Lattice_trace.h ./lattice/Lattice_transfer.h ./lattice/Lattice_transpose.h ./lattice/Lattice_unary.h ./lattice/Lattice_where.h ./Lattice.h ./Lexicographic.h ./Log.h ./Old/Tensor_peek.h ./Old/Tensor_poke.h ./parallelIO/BinaryIO.h ./parallelIO/NerscIO.h ./PerfCount.h ./pugixml/pugixml.h ./qcd/action/ActionBase.h ./qcd/action/ActionParams.h ./qcd/action/Actions.h ./qcd/action/fermion/CayleyFermion5D.h ./qcd/action/fermion/ContinuedFractionFermion5D.h ./qcd/action/fermion/DomainWallFermion.h ./qcd/action/fermion/FermionOperator.h ./qcd/action/fermion/FermionOperatorImpl.h ./qcd/action/fermion/g5HermitianLinop.h ./qcd/action/fermion/MobiusFermion.h ./qcd/action/fermion/MobiusZolotarevFermion.h ./qcd/action/fermion/OverlapWilsonCayleyTanhFermion.h ./qcd/action/fermion/OverlapWilsonCayleyZolotarevFermion.h ./qcd/action/fermion/OverlapWilsonContfracTanhFermion.h ./qcd/action/fermion/OverlapWilsonContfracZolotarevFermion.h ./qcd/action/fermion/OverlapWilsonPartialFractionTanhFermion.h ./qcd/action/fermion/OverlapWilsonPartialFractionZolotarevFermion.h ./qcd/action/fermion/PartialFractionFermion5D.h ./qcd/action/fermion/ScaledShamirFermion.h ./qcd/action/fermion/ShamirZolotarevFermion.h ./qcd/action/fermion/WilsonCompressor.h ./qcd/action/fermion/WilsonFermion.h ./qcd/action/fermion/WilsonFermion5D.h ./qcd/action/fermion/WilsonKernels.h ./qcd/action/fermion/WilsonTMFermion.h ./qcd/action/gauge/GaugeImpl.h ./qcd/action/gauge/PlaqPlusRectangleAction.h ./qcd/action/gauge/WilsonGaugeAction.h ./qcd/action/pseudofermion/EvenOddSchurDifferentiable.h ./qcd/action/pseudofermion/OneFlavourEvenOddRational.h ./qcd/action/pseudofermion/OneFlavourEvenOddRationalRatio.h ./qcd/action/pseudofermion/OneFlavourRational.h ./qcd/action/pseudofermion/OneFlavourRationalRatio.h ./qcd/action/pseudofermion/TwoFlavour.h ./qcd/action/pseudofermion/TwoFlavourEvenOdd.h ./qcd/action/pseudofermion/TwoFlavourEvenOddRatio.h ./qcd/action/pseudofermion/TwoFlavourRatio.h ./qcd/hmc/HMC.h ./qcd/hmc/HmcRunner.h ./qcd/hmc/integrators/Integrator.h ./qcd/hmc/integrators/Integrator_algorithm.h ./qcd/hmc/NerscCheckpointer.h ./qcd/QCD.h ./qcd/spin/Dirac.h ./qcd/spin/TwoSpinor.h ./qcd/utils/CovariantCshift.h ./qcd/utils/LinalgUtils.h ./qcd/utils/SpaceTimeGrid.h ./qcd/utils/SUn.h ./qcd/utils/WilsonLoops.h ./serialisation/BaseIO.h ./serialisation/BinaryIO.h ./serialisation/MacroMagic.h ./serialisation/Serialisation.h ./serialisation/TextIO.h ./serialisation/XmlIO.h ./simd/Avx512Asm.h ./simd/Grid_avx.h ./simd/Grid_avx512.h ./simd/Grid_empty.h ./simd/Grid_imci.h ./simd/Grid_neon.h ./simd/Grid_qpx.h ./simd/Grid_sse4.h ./simd/Grid_vector_types.h ./simd/Grid_vector_unops.h ./Simd.h ./stencil/Lebesgue.h ./Stencil.h ./tensors/Tensor_arith.h ./tensors/Tensor_arith_add.h ./tensors/Tensor_arith_mac.h ./tensors/Tensor_arith_mul.h ./tensors/Tensor_arith_scalar.h ./tensors/Tensor_arith_sub.h ./tensors/Tensor_class.h ./tensors/Tensor_determinant.h ./tensors/Tensor_exp.h ./tensors/Tensor_extract_merge.h ./tensors/Tensor_index.h ./tensors/Tensor_inner.h ./tensors/Tensor_logical.h ./tensors/Tensor_outer.h ./tensors/Tensor_reality.h ./tensors/Tensor_Ta.h ./tensors/Tensor_trace.h ./tensors/Tensor_traits.h ./tensors/Tensor_transpose.h ./tensors/Tensor_unary.h ./Tensors.h ./Threads.h ./Timer.h
CCFILES=./algorithms/approx/MultiShiftFunction.cc ./algorithms/approx/Remez.cc ./algorithms/approx/Zolotarev.cc ./Init.cc ./Log.cc ./pugixml/pugixml.cc ./qcd/action/fermion/CayleyFermion5D.cc ./qcd/action/fermion/ContinuedFractionFermion5D.cc ./qcd/action/fermion/PartialFractionFermion5D.cc ./qcd/action/fermion/WilsonFermion.cc ./qcd/action/fermion/WilsonFermion5D.cc ./qcd/action/fermion/WilsonKernels.cc ./qcd/action/fermion/WilsonKernelsHand.cc ./qcd/hmc/HMC.cc ./qcd/spin/Dirac.cc ./qcd/utils/SpaceTimeGrid.cc ./serialisation/BinaryIO.cc ./serialisation/TextIO.cc ./serialisation/XmlIO.cc ./stencil/Lebesgue.cc ./stencil/Stencil_common.cc
CCFILES=./algorithms/approx/MultiShiftFunction.cc ./algorithms/approx/Remez.cc ./algorithms/approx/Zolotarev.cc ./Init.cc ./Log.cc ./PerfCount.cc ./pugixml/pugixml.cc ./qcd/action/fermion/CayleyFermion5D.cc ./qcd/action/fermion/ContinuedFractionFermion5D.cc ./qcd/action/fermion/PartialFractionFermion5D.cc ./qcd/action/fermion/WilsonFermion.cc ./qcd/action/fermion/WilsonFermion5D.cc ./qcd/action/fermion/WilsonKernels.cc ./qcd/action/fermion/WilsonKernelsAsm.cc ./qcd/action/fermion/WilsonKernelsHand.cc ./qcd/action/fermion/WilsonTMFermion.cc ./qcd/hmc/HMC.cc ./qcd/spin/Dirac.cc ./qcd/utils/SpaceTimeGrid.cc ./serialisation/BinaryIO.cc ./serialisation/TextIO.cc ./serialisation/XmlIO.cc ./stencil/Lebesgue.cc ./stencil/Stencil_common.cc

4
lib/Makefile.am
View File

@ -6,6 +6,10 @@ if BUILD_COMMS_MPI
extra_sources+=communicator/Communicator_mpi.cc
endif
if BUILD_COMMS_SHMEM
extra_sources+=communicator/Communicator_shmem.cc
endif
if BUILD_COMMS_NONE
extra_sources+=communicator/Communicator_none.cc
endif

27
lib/Old/Tensor_peek.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Old/Tensor_peek.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_MATH_PEEK_H
#define GRID_MATH_PEEK_H
namespace Grid {

27
lib/Old/Tensor_poke.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Old/Tensor_poke.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_MATH_POKE_H
#define GRID_MATH_POKE_H
namespace Grid {

31
lib/PerfCount.cc
View File

@ -1,10 +1,40 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/PerfCount.cc
Copyright (C) 2015
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
#include <PerfCount.h>
namespace Grid {
#define CacheControl(L,O,R) ((PERF_COUNT_HW_CACHE_##L)|(PERF_COUNT_HW_CACHE_OP_##O<<8)| (PERF_COUNT_HW_CACHE_RESULT_##R<<16))
const PerformanceCounter::PerformanceCounterConfig PerformanceCounter::PerformanceCounterConfigs [] = {
#ifdef __linux__
{ PERF_TYPE_HARDWARE, PERF_COUNT_HW_CPU_CYCLES , "CPUCYCLES.........." },
{ PERF_TYPE_HARDWARE, PERF_COUNT_HW_INSTRUCTIONS , "INSTRUCTIONS......." },
{ PERF_TYPE_HARDWARE, PERF_COUNT_HW_CACHE_REFERENCES , "CACHE_REFERENCES..." },
@ -23,6 +53,7 @@ const PerformanceCounter::PerformanceCounterConfig PerformanceCounter::Performan
{ PERF_TYPE_HW_CACHE, CacheControl(LL,PREFETCH,ACCESS) , "LL_PREFETCH_ACCESS."},
{ PERF_TYPE_HW_CACHE, CacheControl(L1I,READ,MISS) , "L1I_READ_MISS......"},
{ PERF_TYPE_HW_CACHE, CacheControl(L1I,READ,ACCESS) , "L1I_READ_ACCESS...."}
#endif
// { PERF_TYPE_HARDWARE, PERF_COUNT_HW_STALLED_CYCLES_FRONTEND, "STALL_CYCLES" },
};
}

34
lib/PerfCount.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/PerfCount.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <peterboyle@MacBook-Pro.local>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_PERFCOUNT_H
#define GRID_PERFCOUNT_H
@ -11,6 +40,8 @@
#ifdef __linux__
#include <syscall.h>
#include <linux/perf_event.h>
#else
#include <sys/syscall.h>
#endif
namespace Grid {
@ -30,6 +61,7 @@ static long perf_event_open(struct perf_event_attr *hw_event, pid_t pid,
class PerformanceCounter {
private:
typedef struct {
public:
uint32_t type;
@ -70,7 +102,7 @@ public:
long long count;
int fd;
uint64_t elapsed;
unsigned long long elapsed;
uint64_t begin;
static int NumTypes(void){

32
lib/Simd.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Simd.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: neo <cossu@post.kek.jp>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_SIMD_H
#define GRID_SIMD_H
@ -11,7 +40,6 @@
// Vector types are arch dependent
////////////////////////////////////////////////////////////////////////
typedef uint32_t Integer;
#define _MM_SELECT_FOUR_FOUR(A,B,C,D) ((A<<6)|(B<<4)|(C<<2)|(D))
#define _MM_SELECT_EIGHT_TWO(A,B,C,D,E,F,G,H) ((A<<7)|(B<<6)|(C<<5)|(D<<4)|(E<<3)|(F<<2)|(G<<4)|(H))
@ -20,6 +48,8 @@ typedef uint32_t Integer;
namespace Grid {
typedef uint32_t Integer;
typedef float RealF;
typedef double RealD;
#ifdef GRID_DEFAULT_PRECISION_DOUBLE

467
lib/Stencil.h
View File

@ -1,6 +1,35 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Stencil.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_STENCIL_H
#define GRID_STENCIL_H
#include <thread>
#include <stencil/Lebesgue.h> // subdir aggregate
//////////////////////////////////////////////////////////////////////////////////////////
@ -48,7 +77,7 @@ namespace Grid {
int _around_the_world;
};
template<class vobj,class cobj, class compressor>
template<class vobj,class cobj>
class CartesianStencil { // Stencil runs along coordinate axes only; NO diagonal fill in.
public:
@ -57,10 +86,158 @@ namespace Grid {
typedef typename cobj::scalar_type scalar_type;
typedef typename cobj::scalar_object scalar_object;
//////////////////////////////////////////
// Comms packet queue for asynch thread
//////////////////////////////////////////
struct Packet {
void * send_buf;
void * recv_buf;
Integer to_rank;
Integer from_rank;
Integer bytes;
volatile Integer done;
};
std::vector<Packet> Packets;
#define SEND_IMMEDIATE
#define SERIAL_SENDS
void AddPacket(void *xmit,void * rcv, Integer to,Integer from,Integer bytes){
comms_bytes+=2.0*bytes;
#ifdef SEND_IMMEDIATE
commtime-=usecond();
_grid->SendToRecvFrom(xmit,to,rcv,from,bytes);
commtime+=usecond();
#endif
Packet p;
p.send_buf = xmit;
p.recv_buf = rcv;
p.to_rank = to;
p.from_rank= from;
p.bytes = bytes;
p.done = 0;
comms_bytes+=2.0*bytes;
Packets.push_back(p);
}
#ifdef SERIAL_SENDS
void Communicate(void ) {
commtime-=usecond();
for(int i=0;i<Packets.size();i++){
#ifndef SEND_IMMEDIATE
_grid->SendToRecvFrom(
Packets[i].send_buf,
Packets[i].to_rank,
Packets[i].recv_buf,
Packets[i].from_rank,
Packets[i].bytes);
#endif
Packets[i].done = 1;
}
commtime+=usecond();
}
#else
void Communicate(void ) {
typedef CartesianCommunicator::CommsRequest_t CommsRequest_t;
std::vector<std::vector<CommsRequest_t> > reqs(Packets.size());
commtime-=usecond();
const int concurrency=2;
for(int i=0;i<Packets.size();i+=concurrency){
for(int ii=0;ii<concurrency;ii++){
int j = i+ii;
if ( j<Packets.size() ) {
#ifndef SEND_IMMEDIATE
_grid->SendToRecvFromBegin(reqs[j],
Packets[j].send_buf,
Packets[j].to_rank,
Packets[j].recv_buf,
Packets[j].from_rank,
Packets[j].bytes);
#endif
}
}
for(int ii=0;ii<concurrency;ii++){
int j = i+ii;
if ( j<Packets.size() ) {
#ifndef SEND_IMMEDIATE
_grid->SendToRecvFromComplete(reqs[i]);
#endif
}
}
for(int ii=0;ii<concurrency;ii++){
int j = i+ii;
if ( j<Packets.size() ) {
Packets[j].done = 1;
}
}
}
commtime+=usecond();
}
#endif
///////////////////////////////////////////
// Simd merge queue for asynch comms
///////////////////////////////////////////
struct Merge {
cobj * mpointer;
std::vector<scalar_object *> rpointers;
Integer buffer_size;
Integer packet_id;
};
std::vector<Merge> Mergers;
void AddMerge(cobj *merge_p,std::vector<scalar_object *> &rpointers,Integer buffer_size,Integer packet_id) {
Merge m;
m.mpointer = merge_p;
m.rpointers= rpointers;
m.buffer_size = buffer_size;
m.packet_id = packet_id;
#ifdef SEND_IMMEDIATE
mergetime-=usecond();
PARALLEL_FOR_LOOP
for(int o=0;o<m.buffer_size;o++){
merge1(m.mpointer[o],m.rpointers,o);
}
mergetime+=usecond();
#else
Mergers.push_back(m);
#endif
}
void CommsMerge(void ) {
//PARALLEL_NESTED_LOOP2
for(int i=0;i<Mergers.size();i++){
spintime-=usecond();
int packet_id = Mergers[i].packet_id;
while(! Packets[packet_id].done ); // spin for completion
spintime+=usecond();
#ifndef SEND_IMMEDIATE
mergetime-=usecond();
PARALLEL_FOR_LOOP
for(int o=0;o<Mergers[i].buffer_size;o++){
merge1(Mergers[i].mpointer[o],Mergers[i].rpointers,o);
}
mergetime+=usecond();
#endif
}
}
////////////////////////////////////////
// Basic Grid and stencil info
////////////////////////////////////////
int _checkerboard;
int _npoints; // Move to template param?
GridBase * _grid;
// npoints of these
std::vector<int> _directions;
std::vector<int> _distances;
@ -69,32 +246,32 @@ namespace Grid {
// npoints x Osites() of these
std::vector<std::vector<StencilEntry> > _entries;
// Comms buffers
std::vector<std::vector<scalar_object> > send_buf_extract;
std::vector<std::vector<scalar_object> > recv_buf_extract;
std::vector<scalar_object *> pointers;
std::vector<scalar_object *> rpointers;
Vector<cobj> send_buf;
inline StencilEntry * GetEntry(int &ptype,int point,int osite) { ptype = _permute_type[point]; return & _entries[point][osite]; }
// Comms buffers
std::vector<Vector<scalar_object> > u_simd_send_buf;
std::vector<Vector<scalar_object> > u_simd_recv_buf;
Vector<cobj> u_send_buf;
Vector<cobj> comm_buf;
int u_comm_offset;
int _unified_buffer_size;
int _request_count;
double buftime;
/////////////////////////////////////////
// Timing info; ugly; possibly temporary
/////////////////////////////////////////
#define TIMING_HACK
#ifdef TIMING_HACK
double jointime;
double gathertime;
double commtime;
double commstime;
double halotime;
double scattertime;
double halogtime;
double mergetime;
double spintime;
double comms_bytes;
double gathermtime;
double splicetime;
double nosplicetime;
#endif
CartesianStencil(GridBase *grid,
int npoints,
@ -103,28 +280,29 @@ namespace Grid {
const std::vector<int> &distances)
: _entries(npoints), _permute_type(npoints), _comm_buf_size(npoints)
{
#ifdef TIMING_HACK
gathertime=0;
jointime=0;
commtime=0;
commstime=0;
halotime=0;
scattertime=0;
halogtime=0;
mergetime=0;
spintime=0;
gathermtime=0;
buftime=0;
splicetime=0;
nosplicetime=0;
comms_bytes=0;
#endif
_npoints = npoints;
_grid = grid;
_directions = directions;
_distances = distances;
_unified_buffer_size=0;
_request_count =0;
int osites = _grid->oSites();
for(int i=0;i<npoints;i++){
for(int ii=0;ii<npoints;ii++){
int i = ii; // reverse direction to get SIMD comms done first
int point = i;
_entries[i].resize( osites);
@ -163,23 +341,27 @@ namespace Grid {
// So tables are the same whether comm_dim or splice_dim
sshift[0] = _grid->CheckerBoardShiftForCB(_checkerboard,dimension,shift,Even);
sshift[1] = _grid->CheckerBoardShiftForCB(_checkerboard,dimension,shift,Odd);
if ( sshift[0] == sshift[1] ) {
Comms(point,dimension,shift,0x3);
// std::cout<<"Comms 0x3"<<std::endl;
} else {
Comms(point,dimension,shift,0x1);// if checkerboard is unfavourable take two passes
Comms(point,dimension,shift,0x2);// both with block stride loop iteration
// std::cout<<"Comms 0x1 ; 0x2"<<std::endl;
}
}
// for(int ss=0;ss<osites;ss++){
// std::cout << "point["<<i<<"] "<<ss<<"-> o"<<_entries[i][ss]._offset<<"; l"<<
// _entries[i][ss]._is_local<<"; p"<<_entries[i][ss]._permute<<std::endl;
// }
}
u_send_buf.resize(_unified_buffer_size);
comm_buf.resize(_unified_buffer_size);
const int Nsimd = grid->Nsimd();
u_simd_send_buf.resize(Nsimd);
u_simd_recv_buf.resize(Nsimd);
for(int l=0;l<Nsimd;l++){
u_simd_send_buf[l].resize(_unified_buffer_size);
u_simd_recv_buf[l].resize(_unified_buffer_size);
}
}
void Local (int point, int dimension,int shiftpm,int cbmask)
{
int fd = _grid->_fdimensions[dimension];
@ -210,7 +392,6 @@ namespace Grid {
if ( (shiftpm== 1) && (sx<x) ) {
wraparound = 1;
}
int permute_slice=0;
if(permute_dim){
@ -228,6 +409,7 @@ namespace Grid {
void Comms (int point,int dimension,int shiftpm,int cbmask)
{
GridBase *grid=_grid;
const int Nsimd = grid->Nsimd();
int fd = _grid->_fdimensions[dimension];
int ld = _grid->_ldimensions[dimension];
@ -241,22 +423,43 @@ namespace Grid {
int shift = (shiftpm + fd) %fd;
assert(shift>=0);
assert(shift<fd);
int buffer_size = _grid->_slice_nblock[dimension]*_grid->_slice_block[dimension];
int buffer_size = _grid->_slice_nblock[dimension]*_grid->_slice_block[dimension]; // done in reduced dims, so SIMD factored
_comm_buf_size[point] = buffer_size; // Size of _one_ plane. Multiple planes may be gathered and
// send to one or more remote nodes.
int cb= (cbmask==0x2)? Odd : Even;
int sshift= _grid->CheckerBoardShiftForCB(_checkerboard,dimension,shift,cb);
for(int x=0;x<rd;x++){
int sx = (x+sshift)%rd;
int comm_proc = ((x+sshift)/rd)%pd;
int offnode = (comm_proc!= 0);
int permute_type=grid->PermuteType(dimension);
int sx = (x+sshift)%rd;
int offnode = 0;
if ( simd_layout > 1 ) {
for(int i=0;i<Nsimd;i++){
int inner_bit = (Nsimd>>(permute_type+1));
int ic= (i&inner_bit)? 1:0;
int my_coor = rd*ic + x;
int nbr_coor = my_coor+sshift;
int nbr_proc = ((nbr_coor)/ld) % pd;// relative shift in processors
if ( nbr_proc ) {
offnode =1;
}
}
} else {
int comm_proc = ((x+sshift)/rd)%pd;
offnode = (comm_proc!= 0);
}
// std::cout << "Stencil shift "<<shift<<" sshift "<<sshift<<" fd "<<fd<<" rd " <<rd<<" offnode "<<offnode<<" sx "<<sx<<std::endl;
int wraparound=0;
if ( (shiftpm==-1) && (sx>x) && (grid->_processor_coor[dimension]==0) ) {
wraparound = 1;
@ -274,14 +477,13 @@ namespace Grid {
int words = buffer_size;
if (cbmask != 0x3) words=words>>1;
// GatherPlaneSimple (point,dimension,sx,cbmask);
int rank = grid->_processor;
int recv_from_rank;
int xmit_to_rank;
int unified_buffer_offset = _unified_buffer_size;
_unified_buffer_size += words;
ScatterPlane(point,dimension,x,cbmask,unified_buffer_offset,wraparound); // permute/extract/merge is done in comms phase
}
@ -377,35 +579,33 @@ namespace Grid {
}
}
// CartesianStencil(GridBase *grid,
// int npoints,
// int checkerboard,
// const std::vector<int> &directions,
// const std::vector<int> &distances);
template<class compressor>
std::thread HaloExchangeBegin(const Lattice<vobj> &source,compressor &compress) {
Mergers.resize(0);
Packets.resize(0);
HaloGather(source,compress);
return std::thread([&] { this->Communicate(); });
}
// Add to tables for various cases; is this mistaken. only local if 1 proc in dim
// Can this be avoided with simpler coding of comms?
// void Local (int point, int dimension,int shift,int cbmask);
// void Comms (int point, int dimension,int shift,int cbmask);
// void CopyPlane(int point, int dimension,int lplane,int rplane,int cbmask,int permute,int wrap);
// void ScatterPlane (int point,int dimension,int plane,int cbmask,int offset,int wrap);
// Could allow a functional munging of the halo to another type during the comms.
// this could implement the 16bit/32bit/64bit compression.
void HaloExchange(const Lattice<vobj> &source,std::vector<cobj,alignedAllocator<cobj> > &u_comm_buf,compressor &compress)
template<class compressor>
void HaloExchange(const Lattice<vobj> &source,compressor &compress)
{
// conformable(source._grid,_grid);
assert(source._grid==_grid);
halotime-=usecond();
if (u_comm_buf.size() != _unified_buffer_size ) u_comm_buf.resize(_unified_buffer_size);
int u_comm_offset=0;
auto thr = HaloExchangeBegin(source,compress);
HaloExchangeComplete(thr);
}
// Gather all comms buffers
for(int point = 0 ; point < _npoints; point++) {
compress.Point(point);
void HaloExchangeComplete(std::thread &thr)
{
CommsMerge(); // spins
jointime-=usecond();
thr.join();
jointime+=usecond();
}
template<class compressor>
void HaloGatherDir(const Lattice<vobj> &source,compressor &compress,int point)
{
int dimension = _directions[point];
int displacement = _distances[point];
@ -432,35 +632,51 @@ namespace Grid {
if ( sshift[0] == sshift[1] ) {
if (splice_dim) {
splicetime-=usecond();
GatherStartCommsSimd(source,dimension,shift,0x3,u_comm_buf,u_comm_offset,compress);
GatherSimd(source,dimension,shift,0x3,compress);
splicetime+=usecond();
} else {
nosplicetime-=usecond();
GatherStartComms(source,dimension,shift,0x3,u_comm_buf,u_comm_offset,compress);
Gather(source,dimension,shift,0x3,compress);
nosplicetime+=usecond();
}
} else {
std::cout << "dim "<<dimension<<"cb "<<_checkerboard<<"shift "<<shift<<" sshift " << sshift[0]<<" "<<sshift[1]<<std::endl;
if(splice_dim){
splicetime-=usecond();
GatherStartCommsSimd(source,dimension,shift,0x1,u_comm_buf,u_comm_offset,compress);// if checkerboard is unfavourable take two passes
GatherStartCommsSimd(source,dimension,shift,0x2,u_comm_buf,u_comm_offset,compress);// both with block stride loop iteration
GatherSimd(source,dimension,shift,0x1,compress);// if checkerboard is unfavourable take two passes
GatherSimd(source,dimension,shift,0x2,compress);// both with block stride loop iteration
splicetime+=usecond();
} else {
nosplicetime-=usecond();
GatherStartComms(source,dimension,shift,0x1,u_comm_buf,u_comm_offset,compress);
GatherStartComms(source,dimension,shift,0x2,u_comm_buf,u_comm_offset,compress);
Gather(source,dimension,shift,0x1,compress);
Gather(source,dimension,shift,0x2,compress);
nosplicetime+=usecond();
}
}
}
}
halotime+=usecond();
}
void GatherStartComms(const Lattice<vobj> &rhs,int dimension,int shift,int cbmask,
std::vector<cobj,alignedAllocator<cobj> > &u_comm_buf,
int &u_comm_offset,compressor & compress)
template<class compressor>
void HaloGather(const Lattice<vobj> &source,compressor &compress)
{
// conformable(source._grid,_grid);
assert(source._grid==_grid);
halogtime-=usecond();
assert (comm_buf.size() == _unified_buffer_size );
u_comm_offset=0;
// Gather all comms buffers
for(int point = 0 ; point < _npoints; point++) {
compress.Point(point);
HaloGatherDir(source,compress,point);
}
assert(u_comm_offset==_unified_buffer_size);
halogtime+=usecond();
}
template<class compressor>
void Gather(const Lattice<vobj> &rhs,int dimension,int shift,int cbmask,compressor & compress)
{
typedef typename cobj::vector_type vector_type;
typedef typename cobj::scalar_type scalar_type;
@ -481,8 +697,6 @@ namespace Grid {
int buffer_size = _grid->_slice_nblock[dimension]*_grid->_slice_block[dimension];
if(send_buf.size()<buffer_size) send_buf.resize(buffer_size);
int cb= (cbmask==0x2)? Odd : Even;
int sshift= _grid->CheckerBoardShiftForCB(rhs.checkerboard,dimension,shift,cb);
@ -499,38 +713,33 @@ namespace Grid {
int bytes = words * sizeof(cobj);
gathertime-=usecond();
Gather_plane_simple (rhs,send_buf,dimension,sx,cbmask,compress);
Gather_plane_simple (rhs,u_send_buf,dimension,sx,cbmask,compress,u_comm_offset);
gathertime+=usecond();
int rank = _grid->_processor;
int recv_from_rank;
int xmit_to_rank;
_grid->ShiftedRanks(dimension,comm_proc,xmit_to_rank,recv_from_rank);
assert (xmit_to_rank != _grid->ThisRank());
assert (xmit_to_rank != _grid->ThisRank());
assert (recv_from_rank != _grid->ThisRank());
// FIXME Implement asynchronous send & also avoid buffer copy
commtime-=usecond();
_grid->SendToRecvFrom((void *)&send_buf[0],
xmit_to_rank,
(void *)&u_comm_buf[u_comm_offset],
recv_from_rank,
bytes);
commtime+=usecond();
AddPacket((void *)&u_send_buf[u_comm_offset],
(void *) &comm_buf[u_comm_offset],
xmit_to_rank,
recv_from_rank,
bytes);
u_comm_offset+=words;
}
}
}
void GatherStartCommsSimd(const Lattice<vobj> &rhs,int dimension,int shift,int cbmask,
std::vector<cobj,alignedAllocator<cobj> > &u_comm_buf,
int &u_comm_offset,compressor &compress)
template<class compressor>
void GatherSimd(const Lattice<vobj> &rhs,int dimension,int shift,int cbmask,compressor &compress)
{
buftime-=usecond();
const int Nsimd = _grid->Nsimd();
int fd = _grid->_fdimensions[dimension];
int rd = _grid->_rdimensions[dimension];
@ -554,22 +763,11 @@ namespace Grid {
assert(cbmask==0x3); // Fixme think there is a latent bug if not true
// Should grow to max size and then cost very little thereafter
send_buf_extract.resize(Nsimd);
recv_buf_extract.resize(Nsimd);
for(int l=0;l<Nsimd;l++){
if( send_buf_extract[l].size() < buffer_size) {
send_buf_extract[l].resize(buffer_size);
recv_buf_extract[l].resize(buffer_size);
}
}
pointers.resize(Nsimd);
rpointers.resize(Nsimd);
int bytes = buffer_size*sizeof(scalar_object);
buftime+=usecond();
std::vector<scalar_object *> rpointers(Nsimd);
std::vector<scalar_object *> spointers(Nsimd);
///////////////////////////////////////////
// Work out what to send where
///////////////////////////////////////////
@ -585,16 +783,18 @@ namespace Grid {
if ( any_offnode ) {
for(int i=0;i<Nsimd;i++){
pointers[i] = &send_buf_extract[i][0];
spointers[i] = &u_simd_send_buf[i][u_comm_offset];
}
int sx = (x+sshift)%rd;
gathermtime-=usecond();
Gather_plane_extract<cobj>(rhs,pointers,dimension,sx,cbmask,compress);
Gather_plane_extract<cobj>(rhs,spointers,dimension,sx,cbmask,compress);
gathermtime+=usecond();
for(int i=0;i<Nsimd;i++){
// std::cout << "GatherSimd : lane 1st elem " << i << u_simd_send_buf[i ][u_comm_offset]<<std::endl;
int inner_bit = (Nsimd>>(permute_type+1));
int ic= (i&inner_bit)? 1:0;
@ -607,44 +807,41 @@ namespace Grid {
int nbr_ox = (nbr_lcoor%rd); // outer coord of peer
int nbr_lane = (i&(~inner_bit));
int recv_from_rank;
int xmit_to_rank;
if (nbr_ic) nbr_lane|=inner_bit;
assert (sx == nbr_ox);
auto rp = &u_simd_recv_buf[i ][u_comm_offset];
auto sp = &u_simd_send_buf[nbr_lane][u_comm_offset];
void *vrp = (void *)rp;
void *vsp = (void *)sp;
if(nbr_proc){
int recv_from_rank;
int xmit_to_rank;
_grid->ShiftedRanks(dimension,nbr_proc,xmit_to_rank,recv_from_rank);
commstime-=usecond();
_grid->SendToRecvFrom((void *)&send_buf_extract[nbr_lane][0],
xmit_to_rank,
(void *)&recv_buf_extract[i][0],
recv_from_rank,
bytes);
commstime+=usecond();
AddPacket( vsp,vrp,xmit_to_rank,recv_from_rank,bytes);
rpointers[i] = &recv_buf_extract[i][0];
rpointers[i] = rp;
} else {
rpointers[i] = &send_buf_extract[nbr_lane][0];
rpointers[i] = sp;
}
}
// Here we don't want to scatter, just place into a buffer.
mergetime-=usecond();
PARALLEL_FOR_LOOP
for(int i=0;i<buffer_size;i++){
// assert(u_comm_offset+i<_unified_buffer_size);
merge(u_comm_buf[u_comm_offset+i],rpointers,i);
}
mergetime+=usecond();
u_comm_offset+=buffer_size;
AddMerge(&comm_buf[u_comm_offset],rpointers,buffer_size,Packets.size()-1);
u_comm_offset +=buffer_size;
}
}
}
};
}
#endif

29
lib/Tensors.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Tensors.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: neo <cossu@post.kek.jp>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_MATH_H
#define GRID_MATH_H

28
lib/Threads.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Threads.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_THREADS_H
#define GRID_THREADS_H

33
lib/Timer.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/Timer.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_TIME_H
#define GRID_TIME_H
@ -17,6 +45,11 @@ typedef std::chrono::system_clock GridClock;
typedef std::chrono::time_point<GridClock> GridTimePoint;
typedef std::chrono::milliseconds GridTime;
inline std::ostream& operator<< (std::ostream & stream, const std::chrono::milliseconds & time)
{
stream << time.count()<<" ms";
return stream;
}
class GridStopWatch {
private:

38
lib/algorithms/CoarsenedMatrix.h
View File

@ -1,3 +1,33 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/CoarsenedMatrix.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: Peter Boyle <peterboyle@Peters-MacBook-Pro-2.local>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_ALGORITHM_COARSENED_MATRIX_H
#define GRID_ALGORITHM_COARSENED_MATRIX_H
@ -170,11 +200,10 @@ namespace Grid {
////////////////////
Geometry geom;
GridBase * _grid;
CartesianStencil<siteVector,siteVector,SimpleCompressor<siteVector> > Stencil;
CartesianStencil<siteVector,siteVector> Stencil;
std::vector<CoarseMatrix> A;
std::vector<siteVector,alignedAllocator<siteVector> > comm_buf;
///////////////////////
// Interface
@ -187,7 +216,7 @@ namespace Grid {
conformable(in._grid,out._grid);
SimpleCompressor<siteVector> compressor;
Stencil.HaloExchange(in,comm_buf,compressor);
Stencil.HaloExchange(in,compressor);
PARALLEL_FOR_LOOP
for(int ss=0;ss<Grid()->oSites();ss++){
@ -204,7 +233,7 @@ PARALLEL_FOR_LOOP
} else if(SE->_is_local) {
nbr = in._odata[SE->_offset];
} else {
nbr = comm_buf[SE->_offset];
nbr = Stencil.comm_buf[SE->_offset];
}
res = res + A[point]._odata[ss]*nbr;
}
@ -228,7 +257,6 @@ PARALLEL_FOR_LOOP
Stencil(&CoarseGrid,geom.npoint,Even,geom.directions,geom.displacements),
A(geom.npoint,&CoarseGrid)
{
comm_buf.resize(Stencil._unified_buffer_size);
};
void CoarsenOperator(GridBase *FineGrid,LinearOperatorBase<Lattice<Fobj> > &linop,

28
lib/algorithms/LinearOperator.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/LinearOperator.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_ALGORITHM_LINEAR_OP_H
#define GRID_ALGORITHM_LINEAR_OP_H

27
lib/algorithms/Preconditioner.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/Preconditioner.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_PRECONDITIONER_H
#define GRID_PRECONDITIONER_H

27
lib/algorithms/SparseMatrix.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/SparseMatrix.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_ALGORITHM_SPARSE_MATRIX_H
#define GRID_ALGORITHM_SPARSE_MATRIX_H

28
lib/algorithms/approx/Chebyshev.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/approx/Chebyshev.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_CHEBYSHEV_H
#define GRID_CHEBYSHEV_H

27
lib/algorithms/approx/MultiShiftFunction.cc
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/approx/MultiShiftFunction.cc
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
namespace Grid {

28
lib/algorithms/approx/MultiShiftFunction.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/approx/MultiShiftFunction.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef MULTI_SHIFT_FUNCTION
#define MULTI_SHIFT_FUNCTION

27
lib/algorithms/approx/bigfloat_double.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/approx/bigfloat_double.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <math.h>
typedef double mfloat;

27
lib/algorithms/iterative/AdefGeneric.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/AdefGeneric.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_ALGORITHMS_ITERATIVE_GENERIC_PCG
#define GRID_ALGORITHMS_ITERATIVE_GENERIC_PCG

55
lib/algorithms/iterative/ConjugateGradient.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/ConjugateGradient.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_CONJUGATE_GRADIENT_H
#define GRID_CONJUGATE_GRADIENT_H
@ -30,7 +59,8 @@ public:
//Initial residual computation & set up
RealD guess = norm2(psi);
assert(std::isnan(guess)==0);
Linop.HermOpAndNorm(psi,mmp,d,b);
r= src-mmp;
@ -55,14 +85,22 @@ public:
}
std::cout<<GridLogIterative << std::setprecision(4)<< "ConjugateGradient: k=0 residual "<<cp<<" rsq"<<rsq<<std::endl;
GridStopWatch LinalgTimer;
GridStopWatch MatrixTimer;
GridStopWatch SolverTimer;
SolverTimer.Start();
int k;
for (k=1;k<=MaxIterations;k++){
c=cp;
Linop.HermOpAndNorm(p,mmp,d,qq);
MatrixTimer.Start();
Linop.HermOpAndNorm(p,mmp,d,qq);
MatrixTimer.Stop();
LinalgTimer.Start();
RealD qqck = norm2(mmp);
ComplexD dck = innerProduct(p,mmp);
@ -76,11 +114,13 @@ public:
psi= a*p+psi;
p = p*b+r;
LinalgTimer.Stop();
std::cout<<GridLogIterative<<"ConjugateGradient: Iteration " <<k<<" residual "<<cp<< " target"<< rsq<<std::endl;
// Stopping condition
if ( cp <= rsq ) {
SolverTimer.Stop();
Linop.HermOpAndNorm(psi,mmp,d,qq);
p=mmp-src;
@ -93,7 +133,12 @@ public:
std::cout<<GridLogMessage<<"ConjugateGradient: Converged on iteration " <<k
<<" computed residual "<<sqrt(cp/ssq)
<<" true residual "<<true_residual
<<" target "<<Tolerance<<std::endl;
<<" target "<<Tolerance;
std::cout<<" Time elapsed: Total "<< SolverTimer.Elapsed() << " Matrix "<<MatrixTimer.Elapsed() << " Linalg "<<LinalgTimer.Elapsed();
std::cout<<std::endl;
assert(true_residual/Tolerance < 1000.0);
return;
}
}

28
lib/algorithms/iterative/ConjugateGradientMultiShift.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/ConjugateGradientMultiShift.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_CONJUGATE_MULTI_SHIFT_GRADIENT_H
#define GRID_CONJUGATE_MULTI_SHIFT_GRADIENT_H

28
lib/algorithms/iterative/ConjugateResidual.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/ConjugateResidual.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_CONJUGATE_RESIDUAL_H
#define GRID_CONJUGATE_RESIDUAL_H

28
lib/algorithms/iterative/DenseMatrix.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/DenseMatrix.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_DENSE_MATRIX_H
#define GRID_DENSE_MATRIX_H

27
lib/algorithms/iterative/EigenSort.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/EigenSort.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_EIGENSORT_H
#define GRID_EIGENSORT_H

27
lib/algorithms/iterative/Francis.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/Francis.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef FRANCIS_H
#define FRANCIS_H

27
lib/algorithms/iterative/Householder.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/Householder.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef HOUSEHOLDER_H
#define HOUSEHOLDER_H

28
lib/algorithms/iterative/ImplicitlyRestartedLanczos.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/ImplicitlyRestartedLanczos.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_IRL_H
#define GRID_IRL_H

27
lib/algorithms/iterative/Matrix.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/Matrix.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef MATRIX_H
#define MATRIX_H

27
lib/algorithms/iterative/MatrixUtils.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/MatrixUtils.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_MATRIX_UTILS_H
#define GRID_MATRIX_UTILS_H

27
lib/algorithms/iterative/NormalEquations.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/NormalEquations.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_NORMAL_EQUATIONS_H
#define GRID_NORMAL_EQUATIONS_H

27
lib/algorithms/iterative/PrecConjugateResidual.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/PrecConjugateResidual.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_PREC_CONJUGATE_RESIDUAL_H
#define GRID_PREC_CONJUGATE_RESIDUAL_H

28
lib/algorithms/iterative/PrecGeneralisedConjugateResidual.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/PrecGeneralisedConjugateResidual.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_PREC_GCR_H
#define GRID_PREC_GCR_H

27
lib/algorithms/iterative/SchurRedBlack.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/algorithms/iterative/SchurRedBlack.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_SCHUR_RED_BLACK_H
#define GRID_SCHUR_RED_BLACK_H

69
lib/cartesian/Cartesian_base.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/cartesian/Cartesian_base.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_CARTESIAN_BASE_H
#define GRID_CARTESIAN_BASE_H
@ -87,27 +115,11 @@ public:
for(int d=0;d<_ndimension;d++) idx+=_ostride[d]*ocoor[d];
return idx;
}
static inline void CoorFromIndex (std::vector<int>& coor,int index,std::vector<int> &dims){
int nd= dims.size();
coor.resize(nd);
for(int d=0;d<nd;d++){
coor[d] = index % dims[d];
index = index / dims[d];
}
}
inline void oCoorFromOindex (std::vector<int>& coor,int Oindex){
CoorFromIndex(coor,Oindex,_rdimensions);
}
static inline void IndexFromCoor (std::vector<int>& coor,int &index,std::vector<int> &dims){
int nd=dims.size();
int stride=1;
index=0;
for(int d=0;d<nd;d++){
index = index+stride*coor[d];
stride=stride*dims[d];
}
Lexicographic::CoorFromIndex(coor,Oindex,_rdimensions);
}
//////////////////////////////////////////////////////////
// SIMD lane addressing
//////////////////////////////////////////////////////////
@ -119,7 +131,7 @@ public:
}
inline void iCoorFromIindex(std::vector<int> &coor,int lane)
{
CoorFromIndex(coor,lane,_simd_layout);
Lexicographic::CoorFromIndex(coor,lane,_simd_layout);
}
inline int PermuteDim(int dimension){
return _simd_layout[dimension]>1;
@ -151,7 +163,7 @@ public:
// Global addressing
////////////////////////////////////////////////////////////////
void GlobalIndexToGlobalCoor(int gidx,std::vector<int> &gcoor){
CoorFromIndex(gcoor,gidx,_gdimensions);
Lexicographic::CoorFromIndex(gcoor,gidx,_gdimensions);
}
void GlobalCoorToGlobalIndex(const std::vector<int> & gcoor,int & gidx){
gidx=0;
@ -166,8 +178,9 @@ public:
pcoor.resize(_ndimension);
lcoor.resize(_ndimension);
for(int mu=0;mu<_ndimension;mu++){
pcoor[mu] = gcoor[mu]/_ldimensions[mu];
lcoor[mu] = gcoor[mu]%_ldimensions[mu];
int _fld = _fdimensions[mu]/_processors[mu];
pcoor[mu] = gcoor[mu]/_fld;
lcoor[mu] = gcoor[mu]%_fld;
}
}
void GlobalCoorToRankIndex(int &rank, int &o_idx, int &i_idx ,const std::vector<int> &gcoor)
@ -176,8 +189,16 @@ public:
std::vector<int> lcoor;
GlobalCoorToProcessorCoorLocalCoor(pcoor,lcoor,gcoor);
rank = RankFromProcessorCoor(pcoor);
i_idx= iIndex(lcoor);
o_idx= oIndex(lcoor);
std::vector<int> cblcoor(lcoor);
for(int d=0;d<cblcoor.size();d++){
if( this->CheckerBoarded(d) ) {
cblcoor[d] = lcoor[d]/2;
}
}
i_idx= iIndex(cblcoor);// this does not imply divide by 2 on checker dim
o_idx= oIndex(lcoor);// this implies divide by 2 on checkerdim
}
void RankIndexToGlobalCoor(int rank, int o_idx, int i_idx , std::vector<int> &gcoor)

27
lib/cartesian/Cartesian_full.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/cartesian/Cartesian_full.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_CARTESIAN_FULL_H
#define GRID_CARTESIAN_FULL_H

28
lib/cartesian/Cartesian_red_black.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/cartesian/Cartesian_red_black.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_CARTESIAN_RED_BLACK_H
#define GRID_CARTESIAN_RED_BLACK_H

43
lib/communicator/Communicator_base.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/communicator/Communicator_base.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_COMMUNICATOR_BASE_H
#define GRID_COMMUNICATOR_BASE_H
@ -7,6 +34,9 @@
#ifdef GRID_COMMS_MPI
#include <mpi.h>
#endif
#ifdef GRID_COMMS_SHMEM
#include <mpp/shmem.h>
#endif
namespace Grid {
class CartesianCommunicator {
public:
@ -26,6 +56,8 @@ class CartesianCommunicator {
typedef int CommsRequest_t;
#endif
static void Init(int *argc, char ***argv);
// Constructor
CartesianCommunicator(const std::vector<int> &pdimensions_in);
@ -88,12 +120,11 @@ class CartesianCommunicator {
int recv_from_rank,
int bytes);
void RecvFrom(void *recv,
int recv_from_rank,
int bytes);
void SendTo(void *xmit,
int xmit_to_rank,
int bytes);
void SendRecvPacket(void *xmit,
void *recv,
int xmit_to_rank,
int recv_from_rank,
int bytes);
void SendToRecvFromBegin(std::vector<CommsRequest_t> &list,
void *xmit,

63
lib/communicator/Communicator_mpi.cc
View File

@ -1,9 +1,45 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/communicator/Communicator_mpi.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include "Grid.h"
#include <mpi.h>
namespace Grid {
// Should error check all MPI calls.
void CartesianCommunicator::Init(int *argc, char ***argv) {
MPI_Init(argc,argv);
}
int Rank(void) {
int pe;
MPI_Comm_rank(MPI_COMM_WORLD,&pe);
return pe;
}
CartesianCommunicator::CartesianCommunicator(const std::vector<int> &processors)
{
@ -81,21 +117,22 @@ void CartesianCommunicator::SendToRecvFrom(void *xmit,
SendToRecvFromBegin(reqs,xmit,dest,recv,from,bytes);
SendToRecvFromComplete(reqs);
}
void CartesianCommunicator::RecvFrom(void *recv,
int from,
int bytes)
void CartesianCommunicator::SendRecvPacket(void *xmit,
void *recv,
int sender,
int receiver,
int bytes)
{
MPI_Status stat;
int ierr=MPI_Recv(recv, bytes, MPI_CHAR,from,from,communicator,&stat);
assert(ierr==0);
}
void CartesianCommunicator::SendTo(void *xmit,
int dest,
int bytes)
{
int rank = _processor; // used for tag; must know who it comes from
int ierr = MPI_Send(xmit, bytes, MPI_CHAR,dest,_processor,communicator);
assert(ierr==0);
assert(sender != receiver);
int tag = sender;
if ( _processor == sender ) {
MPI_Send(xmit, bytes, MPI_CHAR,receiver,tag,communicator);
}
if ( _processor == receiver ) {
MPI_Recv(recv, bytes, MPI_CHAR,sender,tag,communicator,&stat);
}
}
// Basic Halo comms primitive

33
lib/communicator/Communicator_none.cc
View File

@ -1,6 +1,39 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/communicator/Communicator_none.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include "Grid.h"
namespace Grid {
void CartesianCommunicator::Init(int *argc, char *** arv)
{
}
int Rank(void ){ return 0; };
CartesianCommunicator::CartesianCommunicator(const std::vector<int> &processors)
{
_processors = processors;

318
lib/communicator/Communicator_shmem.cc Normal file
View File

@ -0,0 +1,318 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/communicator/Communicator_shmem.cc
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include "Grid.h"
#include <mpp/shmem.h>
namespace Grid {
// Should error check all MPI calls.
#define SHMEM_VET(addr)
#define SHMEM_VET_DEBUG(addr) { \
if ( ! shmem_addr_accessible(addr,_processor) ) {\
std::fprintf(stderr,"%d Inaccessible shmem address %lx %s %s\n",_processor,addr,__FUNCTION__,#addr); \
BACKTRACEFILE(); \
}\
}
int Rank(void) {
return shmem_my_pe();
}
typedef struct HandShake_t {
uint64_t seq_local;
uint64_t seq_remote;
} HandShake;
static Vector< HandShake > XConnections;
static Vector< HandShake > RConnections;
void CartesianCommunicator::Init(int *argc, char ***argv) {
shmem_init();
XConnections.resize(shmem_n_pes());
RConnections.resize(shmem_n_pes());
for(int pe =0 ; pe<shmem_n_pes();pe++){
XConnections[pe].seq_local = 0;
XConnections[pe].seq_remote= 0;
RConnections[pe].seq_local = 0;
RConnections[pe].seq_remote= 0;
}
shmem_barrier_all();
}
CartesianCommunicator::CartesianCommunicator(const std::vector<int> &processors)
{
_ndimension = processors.size();
std::vector<int> periodic(_ndimension,1);
_Nprocessors=1;
_processors = processors;
_processor_coor.resize(_ndimension);
_processor = shmem_my_pe();
Lexicographic::CoorFromIndex(_processor_coor,_processor,_processors);
for(int i=0;i<_ndimension;i++){
_Nprocessors*=_processors[i];
}
int Size = shmem_n_pes();
assert(Size==_Nprocessors);
}
void CartesianCommunicator::GlobalSum(uint32_t &u){
static long long source ;
static long long dest ;
static long long llwrk[_SHMEM_REDUCE_MIN_WRKDATA_SIZE];
static long psync[_SHMEM_REDUCE_SYNC_SIZE];
// int nreduce=1;
// int pestart=0;
// int logStride=0;
source = u;
dest = 0;
shmem_longlong_sum_to_all(&dest,&source,1,0,0,_Nprocessors,llwrk,psync);
shmem_barrier_all(); // necessary?
u = dest;
}
void CartesianCommunicator::GlobalSum(float &f){
static float source ;
static float dest ;
static float llwrk[_SHMEM_REDUCE_MIN_WRKDATA_SIZE];
static long psync[_SHMEM_REDUCE_SYNC_SIZE];
source = f;
dest =0.0;
shmem_float_sum_to_all(&dest,&source,1,0,0,_Nprocessors,llwrk,psync);
f = dest;
}
void CartesianCommunicator::GlobalSumVector(float *f,int N)
{
static float source ;
static float dest = 0 ;
static float llwrk[_SHMEM_REDUCE_MIN_WRKDATA_SIZE];
static long psync[_SHMEM_REDUCE_SYNC_SIZE];
if ( shmem_addr_accessible(f,_processor) ){
shmem_float_sum_to_all(f,f,N,0,0,_Nprocessors,llwrk,psync);
return;
}
for(int i=0;i<N;i++){
dest =0.0;
source = f[i];
shmem_float_sum_to_all(&dest,&source,1,0,0,_Nprocessors,llwrk,psync);
f[i] = dest;
}
}
void CartesianCommunicator::GlobalSum(double &d)
{
static double source;
static double dest ;
static double llwrk[_SHMEM_REDUCE_MIN_WRKDATA_SIZE];
static long psync[_SHMEM_REDUCE_SYNC_SIZE];
source = d;
dest = 0;
shmem_double_sum_to_all(&dest,&source,1,0,0,_Nprocessors,llwrk,psync);
d = dest;
}
void CartesianCommunicator::GlobalSumVector(double *d,int N)
{
static double source ;
static double dest ;
static double llwrk[_SHMEM_REDUCE_MIN_WRKDATA_SIZE];
static long psync[_SHMEM_REDUCE_SYNC_SIZE];
if ( shmem_addr_accessible(d,_processor) ){
shmem_double_sum_to_all(d,d,N,0,0,_Nprocessors,llwrk,psync);
return;
}
for(int i=0;i<N;i++){
source = d[i];
dest =0.0;
shmem_double_sum_to_all(&dest,&source,1,0,0,_Nprocessors,llwrk,psync);
d[i] = dest;
}
}
void CartesianCommunicator::ShiftedRanks(int dim,int shift,int &source,int &dest)
{
std::vector<int> coor = _processor_coor;
assert(std::abs(shift) <_processors[dim]);
coor[dim] = (_processor_coor[dim] + shift + _processors[dim])%_processors[dim];
Lexicographic::IndexFromCoor(coor,source,_processors);
coor[dim] = (_processor_coor[dim] - shift + _processors[dim])%_processors[dim];
Lexicographic::IndexFromCoor(coor,dest,_processors);
}
int CartesianCommunicator::RankFromProcessorCoor(std::vector<int> &coor)
{
int rank;
Lexicographic::IndexFromCoor(coor,rank,_processors);
return rank;
}
void CartesianCommunicator::ProcessorCoorFromRank(int rank, std::vector<int> &coor)
{
Lexicographic::CoorFromIndex(coor,rank,_processors);
}
// Basic Halo comms primitive
void CartesianCommunicator::SendToRecvFrom(void *xmit,
int dest,
void *recv,
int from,
int bytes)
{
SHMEM_VET(xmit);
SHMEM_VET(recv);
std::vector<CommsRequest_t> reqs(0);
SendToRecvFromBegin(reqs,xmit,dest,recv,from,bytes);
SendToRecvFromComplete(reqs);
}
void CartesianCommunicator::SendRecvPacket(void *xmit,
void *recv,
int sender,
int receiver,
int bytes)
{
static uint64_t seq;
assert(recv!=xmit);
volatile HandShake *RecvSeq = (volatile HandShake *) & RConnections[sender];
volatile HandShake *SendSeq = (volatile HandShake *) & XConnections[receiver];
if ( _processor == sender ) {
printf("Sender SHMEM pt2pt %d -> %d\n",sender,receiver);
// Check he has posted a receive
while(SendSeq->seq_remote == SendSeq->seq_local);
printf("Sender receive %d posted\n",sender,receiver);
// Advance our send count
seq = ++(SendSeq->seq_local);
// Send this packet
SHMEM_VET(recv);
shmem_putmem(recv,xmit,bytes,receiver);
shmem_fence();
printf("Sender sent payload %d\n",seq);
//Notify him we're done
shmem_putmem((void *)&(RecvSeq->seq_remote),&seq,sizeof(seq),receiver);
shmem_fence();
printf("Sender ringing door bell %d\n",seq);
}
if ( _processor == receiver ) {
printf("Receiver SHMEM pt2pt %d->%d\n",sender,receiver);
// Post a receive
seq = ++(RecvSeq->seq_local);
shmem_putmem((void *)&(SendSeq->seq_remote),&seq,sizeof(seq),sender);
printf("Receiver Opening letter box %d\n",seq);
// Now wait until he has advanced our reception counter
while(RecvSeq->seq_remote != RecvSeq->seq_local);
printf("Receiver Got the mail %d\n",seq);
}
}
// Basic Halo comms primitive
void CartesianCommunicator::SendToRecvFromBegin(std::vector<CommsRequest_t> &list,
void *xmit,
int dest,
void *recv,
int from,
int bytes)
{
SHMEM_VET(xmit);
SHMEM_VET(recv);
// shmem_putmem_nb(recv,xmit,bytes,dest,NULL);
shmem_putmem(recv,xmit,bytes,dest);
}
void CartesianCommunicator::SendToRecvFromComplete(std::vector<CommsRequest_t> &list)
{
// shmem_quiet(); // I'm done
shmem_barrier_all();// He's done too
}
void CartesianCommunicator::Barrier(void)
{
shmem_barrier_all();
}
void CartesianCommunicator::Broadcast(int root,void* data, int bytes)
{
static long psync[_SHMEM_REDUCE_SYNC_SIZE];
static uint32_t word;
uint32_t *array = (uint32_t *) data;
assert( (bytes % 4)==0);
int words = bytes/4;
if ( shmem_addr_accessible(data,_processor) ){
shmem_broadcast32(data,data,words,root,0,0,shmem_n_pes(),psync);
return;
}
for(int w=0;w<words;w++){
word = array[w];
shmem_broadcast32((void *)&word,(void *)&word,1,root,0,0,shmem_n_pes(),psync);
if ( shmem_my_pe() != root ) {
array[w] = word;
}
shmem_barrier_all();
}
}
void CartesianCommunicator::BroadcastWorld(int root,void* data, int bytes)
{
static long psync[_SHMEM_REDUCE_SYNC_SIZE];
static uint32_t word;
uint32_t *array = (uint32_t *) data;
assert( (bytes % 4)==0);
int words = bytes/4;
for(int w=0;w<words;w++){
word = array[w];
shmem_broadcast32((void *)&word,(void *)&word,1,root,0,0,shmem_n_pes(),psync);
if ( shmem_my_pe() != root ) {
array[w]= word;
}
shmem_barrier_all();
}
}
}

64
lib/cshift/Cshift_common.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/cshift/Cshift_common.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef _GRID_CSHIFT_COMMON_H_
#define _GRID_CSHIFT_COMMON_H_
@ -8,7 +35,7 @@ class SimpleCompressor {
public:
void Point(int) {};
vobj operator() (const vobj &arg,int dimension,int plane,int osite,GridBase *grid) {
vobj operator() (const vobj &arg) {
return arg;
}
};
@ -17,7 +44,7 @@ public:
// Gather for when there is no need to SIMD split with compression
///////////////////////////////////////////////////////////////////
template<class vobj,class cobj,class compressor> void
Gather_plane_simple (const Lattice<vobj> &rhs,std::vector<cobj,alignedAllocator<cobj> > &buffer,int dimension,int plane,int cbmask,compressor &compress)
Gather_plane_simple (const Lattice<vobj> &rhs,std::vector<cobj,alignedAllocator<cobj> > &buffer,int dimension,int plane,int cbmask,compressor &compress, int off=0)
{
int rd = rhs._grid->_rdimensions[dimension];
@ -29,24 +56,24 @@ Gather_plane_simple (const Lattice<vobj> &rhs,std::vector<cobj,alignedAllocator<
int e1=rhs._grid->_slice_nblock[dimension];
int e2=rhs._grid->_slice_block[dimension];
int stride=rhs._grid->_slice_stride[dimension];
if ( cbmask == 0x3 ) {
PARALLEL_NESTED_LOOP2
for(int n=0;n<e1;n++){
for(int b=0;b<e2;b++){
int o = n*rhs._grid->_slice_stride[dimension];
int bo = n*rhs._grid->_slice_block[dimension];
buffer[bo+b]=compress(rhs._odata[so+o+b],dimension,plane,so+o+b,rhs._grid);
int o = n*stride;
int bo = n*e2;
buffer[off+bo+b]=compress(rhs._odata[so+o+b]);
}
}
} else {
int bo=0;
for(int n=0;n<e1;n++){
for(int b=0;b<e2;b++){
int o = n*rhs._grid->_slice_stride[dimension];
int o = n*stride;
int ocb=1<<rhs._grid->CheckerBoardFromOindex(o+b);// Could easily be a table lookup
if ( ocb &cbmask ) {
buffer[bo++]=compress(rhs._odata[so+o+b],dimension,plane,so+o+b,rhs._grid);
buffer[off+bo++]=compress(rhs._odata[so+o+b]);
}
}
}
@ -70,16 +97,16 @@ Gather_plane_extract(const Lattice<vobj> &rhs,std::vector<typename cobj::scalar_
int e1=rhs._grid->_slice_nblock[dimension];
int e2=rhs._grid->_slice_block[dimension];
int n1=rhs._grid->_slice_stride[dimension];
int n2=rhs._grid->_slice_block[dimension];
if ( cbmask ==0x3){
PARALLEL_NESTED_LOOP2
for(int n=0;n<e1;n++){
for(int b=0;b<e2;b++){
int o=n*rhs._grid->_slice_stride[dimension];
int offset = b+n*rhs._grid->_slice_block[dimension];
cobj temp =compress(rhs._odata[so+o+b],dimension,plane,so+o+b,rhs._grid);
int o = n*n1;
int offset = b+n*n2;
cobj temp =compress(rhs._odata[so+o+b]);
extract<cobj>(temp,pointers,offset);
}
@ -94,7 +121,7 @@ PARALLEL_NESTED_LOOP2
int offset = b+n*rhs._grid->_slice_block[dimension];
if ( ocb & cbmask ) {
cobj temp =compress(rhs._odata[so+o+b],dimension,plane,so+o+b,rhs._grid);
cobj temp =compress(rhs._odata[so+o+b]);
extract<cobj>(temp,pointers,offset);
}
}
@ -216,13 +243,13 @@ template<class vobj> void Copy_plane(Lattice<vobj>& lhs,const Lattice<vobj> &rhs
int e1=rhs._grid->_slice_nblock[dimension]; // clearly loop invariant for icpc
int e2=rhs._grid->_slice_block[dimension];
int stride = rhs._grid->_slice_stride[dimension];
if(cbmask == 0x3 ){
PARALLEL_NESTED_LOOP2
for(int n=0;n<e1;n++){
for(int b=0;b<e2;b++){
int o =n*rhs._grid->_slice_stride[dimension]+b;
int o =n*stride+b;
//lhs._odata[lo+o]=rhs._odata[ro+o];
vstream(lhs._odata[lo+o],rhs._odata[ro+o]);
}
@ -232,7 +259,7 @@ PARALLEL_NESTED_LOOP2
for(int n=0;n<e1;n++){
for(int b=0;b<e2;b++){
int o =n*rhs._grid->_slice_stride[dimension]+b;
int o =n*stride+b;
int ocb=1<<lhs._grid->CheckerBoardFromOindex(o);
if ( ocb&cbmask ) {
//lhs._odata[lo+o]=rhs._odata[ro+o];
@ -258,11 +285,12 @@ template<class vobj> void Copy_plane_permute(Lattice<vobj>& lhs,const Lattice<vo
int e1=rhs._grid->_slice_nblock[dimension];
int e2=rhs._grid->_slice_block [dimension];
int stride = rhs._grid->_slice_stride[dimension];
PARALLEL_NESTED_LOOP2
for(int n=0;n<e1;n++){
for(int b=0;b<e2;b++){
int o =n*rhs._grid->_slice_stride[dimension];
int o =n*stride;
int ocb=1<<lhs._grid->CheckerBoardFromOindex(o+b);
if ( ocb&cbmask ) {
permute(lhs._odata[lo+o+b],rhs._odata[ro+o+b],permute_type);

35
lib/cshift/Cshift_mpi.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/cshift/Cshift_mpi.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef _GRID_CSHIFT_MPI_H_
#define _GRID_CSHIFT_MPI_H_
@ -163,11 +191,12 @@ template<class vobj> void Cshift_comms_simd(Lattice<vobj> &ret,const Lattice<vo
int buffer_size = grid->_slice_nblock[dimension]*grid->_slice_block[dimension];
int words = sizeof(vobj)/sizeof(vector_type);
std::vector<std::vector<scalar_object> > send_buf_extract(Nsimd,std::vector<scalar_object>(buffer_size) );
std::vector<std::vector<scalar_object> > recv_buf_extract(Nsimd,std::vector<scalar_object>(buffer_size) );
std::vector<Vector<scalar_object> > send_buf_extract(Nsimd,Vector<scalar_object>(buffer_size) );
std::vector<Vector<scalar_object> > recv_buf_extract(Nsimd,Vector<scalar_object>(buffer_size) );
int bytes = buffer_size*sizeof(scalar_object);
std::vector<scalar_object *> pointers(Nsimd); //
std::vector<scalar_object *> pointers(Nsimd); //
std::vector<scalar_object *> rpointers(Nsimd); // received pointers
///////////////////////////////////////////

27
lib/cshift/Cshift_none.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/cshift/Cshift_none.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef _GRID_CSHIFT_NONE_H_
#define _GRID_CSHIFT_NONE_H_
namespace Grid {

29
lib/lattice/Lattice_ET.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_ET.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: neo <cossu@post.kek.jp>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_ET_H
#define GRID_LATTICE_ET_H

27
lib/lattice/Lattice_arith.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_arith.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_ARITH_H
#define GRID_LATTICE_ARITH_H

34
lib/lattice/Lattice_base.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_base.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_BASE_H
#define GRID_LATTICE_BASE_H
@ -73,7 +102,6 @@ public:
typedef typename vobj::scalar_type scalar_type;
typedef typename vobj::vector_type vector_type;
typedef vobj vector_object;
////////////////////////////////////////////////////////////////////////////////
// Expression Template closure support
@ -213,8 +241,8 @@ PARALLEL_FOR_LOOP
// what about a default grid?
//////////////////////////////////////////////////////////////////
Lattice(GridBase *grid) : _grid(grid), _odata(_grid->oSites()) {
// _odata.reserve(_grid->oSites());
// _odata.resize(_grid->oSites());
// _odata.reserve(_grid->oSites());
// _odata.resize(_grid->oSites());
// std::cout << "Constructing lattice object with Grid pointer "<<_grid<<std::endl;
assert((((uint64_t)&_odata[0])&0xF) ==0);
checkerboard=0;

28
lib/lattice/Lattice_comparison.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_comparison.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_COMPARISON_H
#define GRID_LATTICE_COMPARISON_H

28
lib/lattice/Lattice_comparison_utils.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_comparison_utils.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_COMPARISON_H
#define GRID_COMPARISON_H

27
lib/lattice/Lattice_conformable.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_conformable.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_CONFORMABLE_H
#define GRID_LATTICE_CONFORMABLE_H

27
lib/lattice/Lattice_coordinate.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_coordinate.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_COORDINATE_H
#define GRID_LATTICE_COORDINATE_H

27
lib/lattice/Lattice_local.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_local.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_LOCALREDUCTION_H
#define GRID_LATTICE_LOCALREDUCTION_H

27
lib/lattice/Lattice_overload.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_overload.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_OVERLOAD_H
#define GRID_LATTICE_OVERLOAD_H

29
lib/lattice/Lattice_peekpoke.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_peekpoke.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: Peter Boyle <peterboyle@Peters-MacBook-Pro-2.local>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_PEEK_H
#define GRID_LATTICE_PEEK_H

29
lib/lattice/Lattice_reality.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_reality.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: neo <cossu@post.kek.jp>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_REALITY_H
#define GRID_LATTICE_REALITY_H

33
lib/lattice/Lattice_reduction.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_reduction.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_REDUCTION_H
#define GRID_LATTICE_REDUCTION_H
@ -19,7 +48,6 @@ namespace Grid {
{
typedef typename vobj::scalar_type scalar_type;
typedef typename vobj::vector_type vector_type;
vector_type vnrm;
scalar_type nrm;
GridBase *grid = left._grid;
@ -31,7 +59,6 @@ namespace Grid {
PARALLEL_FOR_LOOP
for(int thr=0;thr<grid->SumArraySize();thr++){
int nwork, mywork, myoff;
GridThread::GetWork(left._grid->oSites(),thr,mywork,myoff);
@ -151,7 +178,7 @@ template<class vobj> inline void sliceSum(const Lattice<vobj> &Data,std::vector<
// sum over reduced dimension planes, breaking out orthog dir
for(int ss=0;ss<grid->oSites();ss++){
GridBase::CoorFromIndex(coor,ss,grid->_rdimensions);
Lexicographic::CoorFromIndex(coor,ss,grid->_rdimensions);
int r = coor[orthogdim];
lvSum[r]=lvSum[r]+Data._odata[ss];
}

138
lib/lattice/Lattice_rng.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_rng.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_RNG_H
#define GRID_LATTICE_RNG_H
@ -22,13 +50,16 @@ namespace Grid {
assert(fine->_processors[d]==1);
}
// local and global volumes subdivide cleanly after SIMDization
int multiplicity=1;
for(int d=0;d<lowerdims;d++){
multiplicity=multiplicity*fine->_rdimensions[d];
}
// local and global volumes subdivide cleanly after SIMDization
for(int d=0;d<rngdims;d++){
int fd= d+lowerdims;
assert(coarse->_processors[d] == fine->_processors[fd]);
assert(coarse->_simd_layout[d] == fine->_simd_layout[fd]);
assert((fine->_rdimensions[fd] / coarse->_rdimensions[d])* coarse->_rdimensions[d]==fine->_rdimensions[fd]);
assert(((fine->_rdimensions[fd] / coarse->_rdimensions[d])* coarse->_rdimensions[d])==fine->_rdimensions[fd]);
multiplicity = multiplicity *fine->_rdimensions[fd] / coarse->_rdimensions[d];
}
@ -76,16 +107,40 @@ namespace Grid {
public:
GridRNGbase() : _uniform{0,1}, _gaussian(0.0,1.0) {};
int _seeded;
// One generator per site.
// Uniform and Gaussian distributions from these generators.
std::vector<std::ranlux48> _generators;
std::uniform_real_distribution<double> _uniform;
std::normal_distribution<double> _gaussian;
#ifdef RNG_RANLUX
typedef uint64_t RngStateType;
typedef std::ranlux48 RngEngine;
static const int RngStateCount = 15;
#else
typedef std::mt19937 RngEngine;
typedef uint32_t RngStateType;
static const int RngStateCount = std::mt19937::state_size;
#endif
std::vector<RngEngine> _generators;
std::vector<std::uniform_real_distribution<RealD> > _uniform;
std::vector<std::normal_distribution<RealD> > _gaussian;
void GetState(std::vector<RngStateType> & saved,int gen) {
saved.resize(RngStateCount);
std::stringstream ss;
ss<<_generators[gen];
ss.seekg(0,ss.beg);
for(int i=0;i<RngStateCount;i++){
ss>>saved[i];
}
}
void SetState(std::vector<RngStateType> & saved,int gen){
assert(saved.size()==RngStateCount);
std::stringstream ss;
for(int i=0;i<RngStateCount;i++){
ss<< saved[i]<<" ";
}
ss.seekg(0,ss.beg);
ss>>_generators[gen];
}
};
class GridSerialRNG : public GridRNGbase {
@ -98,18 +153,20 @@ namespace Grid {
{
typename source::result_type init = src();
CartesianCommunicator::BroadcastWorld(0,(void *)&init,sizeof(init));
_generators[0] = std::ranlux48(init);
_generators[0] = RngEngine(init);
_seeded=1;
}
GridSerialRNG() : GridRNGbase() {
_generators.resize(1);
_uniform.resize(1,std::uniform_real_distribution<RealD>{0,1});
_gaussian.resize(1,std::normal_distribution<RealD>(0.0,1.0) );
_seeded=0;
}
template <class sobj,class distribution> inline void fill(sobj &l,distribution &dist){
template <class sobj,class distribution> inline void fill(sobj &l,std::vector<distribution> &dist){
typedef typename sobj::scalar_type scalar_type;
@ -117,56 +174,65 @@ namespace Grid {
scalar_type *buf = (scalar_type *) & l;
dist[0].reset();
for(int idx=0;idx<words;idx++){
fillScalar(buf[idx],dist,_generators[0]);
fillScalar(buf[idx],dist[0],_generators[0]);
}
CartesianCommunicator::BroadcastWorld(0,(void *)&l,sizeof(l));
};
template <class distribution> inline void fill(ComplexF &l,distribution &dist){
fillScalar(l,dist,_generators[0]);
template <class distribution> inline void fill(ComplexF &l,std::vector<distribution> &dist){
dist[0].reset();
fillScalar(l,dist[0],_generators[0]);
CartesianCommunicator::BroadcastWorld(0,(void *)&l,sizeof(l));
}
template <class distribution> inline void fill(ComplexD &l,distribution &dist){
fillScalar(l,dist,_generators[0]);
template <class distribution> inline void fill(ComplexD &l,std::vector<distribution> &dist){
dist[0].reset();
fillScalar(l,dist[0],_generators[0]);
CartesianCommunicator::BroadcastWorld(0,(void *)&l,sizeof(l));
}
template <class distribution> inline void fill(RealF &l,distribution &dist){
fillScalar(l,dist,_generators[0]);
template <class distribution> inline void fill(RealF &l,std::vector<distribution> &dist){
dist[0].reset();
fillScalar(l,dist[0],_generators[0]);
CartesianCommunicator::BroadcastWorld(0,(void *)&l,sizeof(l));
}
template <class distribution> inline void fill(RealD &l,distribution &dist){
fillScalar(l,dist,_generators[0]);
template <class distribution> inline void fill(RealD &l,std::vector<distribution> &dist){
dist[0].reset();
fillScalar(l,dist[0],_generators[0]);
CartesianCommunicator::BroadcastWorld(0,(void *)&l,sizeof(l));
}
// vector fill
template <class distribution> inline void fill(vComplexF &l,distribution &dist){
template <class distribution> inline void fill(vComplexF &l,std::vector<distribution> &dist){
RealF *pointer=(RealF *)&l;
dist[0].reset();
for(int i=0;i<2*vComplexF::Nsimd();i++){
fillScalar(pointer[i],dist,_generators[0]);
fillScalar(pointer[i],dist[0],_generators[0]);
}
CartesianCommunicator::BroadcastWorld(0,(void *)&l,sizeof(l));
}
template <class distribution> inline void fill(vComplexD &l,distribution &dist){
template <class distribution> inline void fill(vComplexD &l,std::vector<distribution> &dist){
RealD *pointer=(RealD *)&l;
dist[0].reset();
for(int i=0;i<2*vComplexD::Nsimd();i++){
fillScalar(pointer[i],dist,_generators[0]);
fillScalar(pointer[i],dist[0],_generators[0]);
}
CartesianCommunicator::BroadcastWorld(0,(void *)&l,sizeof(l));
}
template <class distribution> inline void fill(vRealF &l,distribution &dist){
template <class distribution> inline void fill(vRealF &l,std::vector<distribution> &dist){
RealF *pointer=(RealF *)&l;
dist[0].reset();
for(int i=0;i<vRealF::Nsimd();i++){
fillScalar(pointer[i],dist,_generators[0]);
fillScalar(pointer[i],dist[0],_generators[0]);
}
CartesianCommunicator::BroadcastWorld(0,(void *)&l,sizeof(l));
}
template <class distribution> inline void fill(vRealD &l,distribution &dist){
template <class distribution> inline void fill(vRealD &l,std::vector<distribution> &dist){
RealD *pointer=(RealD *)&l;
dist[0].reset();
for(int i=0;i<vRealD::Nsimd();i++){
fillScalar(pointer[i],dist,_generators[0]);
fillScalar(pointer[i],dist[0],_generators[0]);
}
CartesianCommunicator::BroadcastWorld(0,(void *)&l,sizeof(l));
}
@ -185,12 +251,6 @@ namespace Grid {
class GridParallelRNG : public GridRNGbase {
public:
// One generator per site.
std::vector<std::ranlux48> _generators;
// Uniform and Gaussian distributions from these generators.
std::uniform_real_distribution<double> _uniform;
std::normal_distribution<double> _gaussian;
GridBase *_grid;
int _vol;
@ -202,7 +262,10 @@ namespace Grid {
GridParallelRNG(GridBase *grid) : GridRNGbase() {
_grid=grid;
_vol =_grid->iSites()*_grid->oSites();
_generators.resize(_vol);
_uniform.resize(_vol,std::uniform_real_distribution<RealD>{0,1});
_gaussian.resize(_vol,std::normal_distribution<RealD>(0.0,1.0) );
_seeded=0;
}
@ -220,7 +283,7 @@ namespace Grid {
int gsites = _grid->_gsites;
typename source::result_type init = src();
std::ranlux48 pseeder(init);
RngEngine pseeder(init);
std::uniform_int_distribution<uint64_t> ui;
for(int gidx=0;gidx<gsites;gidx++){
@ -241,7 +304,7 @@ namespace Grid {
if( rank == _grid->ThisRank() ){
fixedSeed ssrc(site_seeds);
typename source::result_type sinit = ssrc();
_generators[l_idx] = std::ranlux48(sinit);
_generators[l_idx] = RngEngine(sinit);
}
}
_seeded=1;
@ -251,7 +314,7 @@ namespace Grid {
//void SaveState(const std::string<char> &file);
//void LoadState(const std::string<char> &file);
template <class vobj,class distribution> inline void fill(Lattice<vobj> &l,distribution &dist){
template <class vobj,class distribution> inline void fill(Lattice<vobj> &l,std::vector<distribution> &dist){
typedef typename vobj::scalar_object scalar_object;
typedef typename vobj::scalar_type scalar_type;
@ -275,8 +338,9 @@ PARALLEL_FOR_LOOP
for(int si=0;si<Nsimd;si++){
int gdx = generator_idx(ss,si); // index of generator state
scalar_type *pointer = (scalar_type *)&buf[si];
dist[gdx].reset();
for(int idx=0;idx<words;idx++){
fillScalar(pointer[idx],dist,_generators[gdx]);
fillScalar(pointer[idx],dist[gdx],_generators[gdx]);
}
}

27
lib/lattice/Lattice_trace.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_trace.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_TRACE_H
#define GRID_LATTICE_TRACE_H

43
lib/lattice/Lattice_transfer.h
View File

@ -1,3 +1,30 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_transfer.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_TRANSFER_H
#define GRID_LATTICE_TRANSFER_H
@ -88,9 +115,9 @@ inline void blockProject(Lattice<iVector<CComplex,nbasis > > &coarseData,
int sc;
std::vector<int> coor_c(_ndimension);
std::vector<int> coor_f(_ndimension);
GridBase::CoorFromIndex(coor_f,sf,fine->_rdimensions);
Lexicographic::CoorFromIndex(coor_f,sf,fine->_rdimensions);
for(int d=0;d<_ndimension;d++) coor_c[d]=coor_f[d]/block_r[d];
GridBase::IndexFromCoor(coor_c,sc,coarse->_rdimensions);
Lexicographic::IndexFromCoor(coor_c,sc,coarse->_rdimensions);
for(int i=0;i<nbasis;i++) {
@ -133,9 +160,9 @@ PARALLEL_FOR_LOOP
std::vector<int> coor_c(_ndimension);
std::vector<int> coor_f(_ndimension);
GridBase::CoorFromIndex(coor_f,sf,fine->_rdimensions);
Lexicographic::CoorFromIndex(coor_f,sf,fine->_rdimensions);
for(int d=0;d<_ndimension;d++) coor_c[d]=coor_f[d]/block_r[d];
GridBase::IndexFromCoor(coor_c,sc,coarse->_rdimensions);
Lexicographic::IndexFromCoor(coor_c,sc,coarse->_rdimensions);
// z = A x + y
fineZ._odata[sf]=coarseA._odata[sc]*fineX._odata[sf]+fineY._odata[sf];
@ -198,9 +225,9 @@ inline void blockSum(Lattice<vobj> &coarseData,const Lattice<vobj> &fineData)
std::vector<int> coor_c(_ndimension);
std::vector<int> coor_f(_ndimension);
GridBase::CoorFromIndex(coor_f,sf,fine->_rdimensions);
Lexicographic::CoorFromIndex(coor_f,sf,fine->_rdimensions);
for(int d=0;d<_ndimension;d++) coor_c[d]=coor_f[d]/block_r[d];
GridBase::IndexFromCoor(coor_c,sc,coarse->_rdimensions);
Lexicographic::IndexFromCoor(coor_c,sc,coarse->_rdimensions);
coarseData._odata[sc]=coarseData._odata[sc]+fineData._odata[sf];
@ -284,9 +311,9 @@ inline void blockPromote(const Lattice<iVector<CComplex,nbasis > > &coarseData,
std::vector<int> coor_c(_ndimension);
std::vector<int> coor_f(_ndimension);
GridBase::CoorFromIndex(coor_f,sf,fine->_rdimensions);
Lexicographic::CoorFromIndex(coor_f,sf,fine->_rdimensions);
for(int d=0;d<_ndimension;d++) coor_c[d]=coor_f[d]/block_r[d];
GridBase::IndexFromCoor(coor_c,sc,coarse->_rdimensions);
Lexicographic::IndexFromCoor(coor_c,sc,coarse->_rdimensions);
for(int i=0;i<nbasis;i++) {
if(i==0) fineData._odata[sf]=coarseData._odata[sc](i) * Basis[i]._odata[sf];

28
lib/lattice/Lattice_transpose.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_transpose.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_TRANSPOSE_H
#define GRID_LATTICE_TRANSPOSE_H

30
lib/lattice/Lattice_unary.h
View File

@ -1,3 +1,33 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_unary.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: neo <cossu@post.kek.jp>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_UNARY_H
#define GRID_LATTICE_UNARY_H

29
lib/lattice/Lattice_where.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/lattice/Lattice_where.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: Peter Boyle <peterboyle@Peters-MacBook-Pro-2.local>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_LATTICE_WHERE_H
#define GRID_LATTICE_WHERE_H
namespace Grid {

205
lib/parallelIO/BinaryIO.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/parallelIO/BinaryIO.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_BINARY_IO_H
#define GRID_BINARY_IO_H
@ -107,7 +135,7 @@ class BinaryIO {
}
}
template<class vobj,class fobj,class munger> static inline void Uint32Checksum(Lattice<vobj> lat,munger munge,uint32_t &csum)
template<class vobj,class fobj,class munger> static inline void Uint32Checksum(Lattice<vobj> &lat,munger munge,uint32_t &csum)
{
typedef typename vobj::scalar_object sobj;
GridBase *grid = lat._grid ;
@ -118,7 +146,7 @@ class BinaryIO {
csum = 0;
std::vector<int> lcoor;
for(int l=0;l<grid->lSites();l++){
grid->CoorFromIndex(lcoor,l,grid->_ldimensions);
Lexicographic::CoorFromIndex(lcoor,l,grid->_ldimensions);
peekLocalSite(siteObj,lat,lcoor);
munge(siteObj,fileObj,csum);
}
@ -232,6 +260,113 @@ class BinaryIO {
return csum;
}
static inline uint32_t writeRNGSerial(GridSerialRNG &serial,GridParallelRNG &parallel,std::string file,int offset)
{
typedef typename GridSerialRNG::RngStateType RngStateType;
const int RngStateCount = GridSerialRNG::RngStateCount;
GridBase *grid = parallel._grid;
int gsites = grid->_gsites;
//////////////////////////////////////////////////
// Serialise through node zero
//////////////////////////////////////////////////
std::cout<< GridLogMessage<< "Serial RNG write I/O "<< file<<std::endl;
std::ofstream fout;
if ( grid->IsBoss() ) {
fout.open(file,std::ios::binary|std::ios::out|std::ios::in);
fout.seekp(offset);
}
uint32_t csum=0;
std::vector<RngStateType> saved(RngStateCount);
int bytes = sizeof(RngStateType)*saved.size();
std::vector<int> gcoor;
for(int gidx=0;gidx<gsites;gidx++){
int rank,o_idx,i_idx;
grid->GlobalIndexToGlobalCoor(gidx,gcoor);
grid->GlobalCoorToRankIndex(rank,o_idx,i_idx,gcoor);
int l_idx=parallel.generator_idx(o_idx,i_idx);
if( rank == grid->ThisRank() ){
// std::cout << "rank" << rank<<" Getting state for index "<<l_idx<<std::endl;
parallel.GetState(saved,l_idx);
}
grid->Broadcast(rank,(void *)&saved[0],bytes);
if ( grid->IsBoss() ) {
Uint32Checksum((uint32_t *)&saved[0],bytes,csum);
fout.write((char *)&saved[0],bytes);
}
}
if ( grid->IsBoss() ) {
serial.GetState(saved,0);
Uint32Checksum((uint32_t *)&saved[0],bytes,csum);
fout.write((char *)&saved[0],bytes);
}
grid->Broadcast(0,(void *)&csum,sizeof(csum));
return csum;
}
static inline uint32_t readRNGSerial(GridSerialRNG &serial,GridParallelRNG &parallel,std::string file,int offset)
{
typedef typename GridSerialRNG::RngStateType RngStateType;
const int RngStateCount = GridSerialRNG::RngStateCount;
GridBase *grid = parallel._grid;
int gsites = grid->_gsites;
//////////////////////////////////////////////////
// Serialise through node zero
//////////////////////////////////////////////////
std::cout<< GridLogMessage<< "Serial RNG read I/O "<< file<<std::endl;
std::ifstream fin(file,std::ios::binary|std::ios::in);
fin.seekg(offset);
uint32_t csum=0;
std::vector<RngStateType> saved(RngStateCount);
int bytes = sizeof(RngStateType)*saved.size();
std::vector<int> gcoor;
for(int gidx=0;gidx<gsites;gidx++){
int rank,o_idx,i_idx;
grid->GlobalIndexToGlobalCoor(gidx,gcoor);
grid->GlobalCoorToRankIndex(rank,o_idx,i_idx,gcoor);
int l_idx=parallel.generator_idx(o_idx,i_idx);
if ( grid->IsBoss() ) {
fin.read((char *)&saved[0],bytes);
Uint32Checksum((uint32_t *)&saved[0],bytes,csum);
}
grid->Broadcast(0,(void *)&saved[0],bytes);
if( rank == grid->ThisRank() ){
parallel.SetState(saved,l_idx);
}
}
if ( grid->IsBoss() ) {
fin.read((char *)&saved[0],bytes);
serial.SetState(saved,0);
Uint32Checksum((uint32_t *)&saved[0],bytes,csum);
}
grid->Broadcast(0,(void *)&csum,sizeof(csum));
return csum;
}
template<class vobj,class fobj,class munger>
static inline uint32_t readObjectParallel(Lattice<vobj> &Umu,std::string file,munger munge,int offset,const std::string &format)
{
@ -266,7 +401,7 @@ class BinaryIO {
int IOnode = 1;
for(int d=0;d<grid->_ndimension;d++) {
if ( d==0 ) parallel[d] = 0;
if ( d == 0 ) parallel[d] = 0;
if (parallel[d]) {
range[d] = grid->_ldimensions[d];
start[d] = grid->_processor_coor[d]*range[d];
@ -284,6 +419,7 @@ class BinaryIO {
{
uint32_t tmp = IOnode;
grid->GlobalSum(tmp);
std::cout<< std::dec ;
std::cout<< GridLogMessage<< "Parallel read I/O to "<< file << " with " <<tmp<< " IOnodes for subslice ";
for(int d=0;d<grid->_ndimension;d++){
std::cout<< range[d];
@ -306,9 +442,9 @@ class BinaryIO {
// available (how short sighted is that?)
//////////////////////////////////////////////////////////
Umu = zero;
uint32_t csum=0;
static uint32_t csum=0;
fobj fileObj;
sobj siteObj;
static sobj siteObj; // Static to place in symmetric region for SHMEM
// need to implement these loops in Nd independent way with a lexico conversion
for(int tlex=0;tlex<slice_vol;tlex++){
@ -318,7 +454,7 @@ class BinaryIO {
std::vector<int> lsite(nd);
std::vector<int> iosite(nd);
grid->CoorFromIndex(tsite,tlex,range);
Lexicographic::CoorFromIndex(tsite,tlex,range);
for(int d=0;d<nd;d++){
lsite[d] = tsite[d]%grid->_ldimensions[d]; // local site
@ -328,7 +464,7 @@ class BinaryIO {
/////////////////////////
// Get the rank of owner of data
/////////////////////////
int rank, o_idx,i_idx, g_idx;
int rank, o_idx,i_idx, g_idx;
grid->GlobalCoorToRankIndex(rank,o_idx,i_idx,gsite);
grid->GlobalCoorToGlobalIndex(gsite,g_idx);
@ -346,23 +482,24 @@ class BinaryIO {
if(ieee64) le64toh_v((void *)&fileObj,sizeof(fileObj));
munge(fileObj,siteObj,csum);
if ( rank != myrank ) {
grid->SendTo((void *)&siteObj,rank,sizeof(siteObj));
} else {
pokeLocalSite(siteObj,Umu,lsite);
}
// Possibly do transport through pt2pt
if ( rank != iorank ) {
if ( (myrank == rank) || (myrank==iorank) ) {
grid->SendRecvPacket((void *)&siteObj,(void *)&siteObj,iorank,rank,sizeof(siteObj));
}
} else {
if ( myrank == rank ) {
grid->RecvFrom((void *)&siteObj,iorank,sizeof(siteObj));
}
// Poke at destination
if ( myrank == rank ) {
pokeLocalSite(siteObj,Umu,lsite);
}
}
grid->Barrier(); // necessary?
}
grid->GlobalSum(csum);
grid->Barrier();
return csum;
}
@ -397,7 +534,7 @@ class BinaryIO {
for(int d=0;d<grid->_ndimension;d++) {
if ( d==0 ) parallel[d] = 0;
if ( d!= grid->_ndimension-1 ) parallel[d] = 0;
if (parallel[d]) {
range[d] = grid->_ldimensions[d];
@ -444,10 +581,10 @@ class BinaryIO {
uint32_t csum=0;
fobj fileObj;
sobj siteObj;
static sobj siteObj; // static for SHMEM target; otherwise dynamic allocate with AlignedAllocator
// need to implement these loops in Nd independent way with a lexico conversion
// should aggregate a whole chunk and then write.
// need to implement these loops in Nd independent way with a lexico conversion
for(int tlex=0;tlex<slice_vol;tlex++){
std::vector<int> tsite(nd); // temporary mixed up site
@ -455,7 +592,7 @@ class BinaryIO {
std::vector<int> lsite(nd);
std::vector<int> iosite(nd);
grid->CoorFromIndex(tsite,tlex,range);
Lexicographic::CoorFromIndex(tsite,tlex,range);
for(int d=0;d<nd;d++){
lsite[d] = tsite[d]%grid->_ldimensions[d]; // local site
@ -473,13 +610,21 @@ class BinaryIO {
////////////////////////////////
// iorank writes from the seek
////////////////////////////////
if (myrank == iorank) {
// Owner of data peeks it
peekLocalSite(siteObj,Umu,lsite);
if ( rank != myrank ) {
grid->RecvFrom((void *)&siteObj,rank,sizeof(siteObj));
} else {
peekLocalSite(siteObj,Umu,lsite);
// Pair of nodes may need to do pt2pt send
if ( rank != iorank ) { // comms is necessary
if ( (myrank == rank) || (myrank==iorank) ) { // and we have to do it
// Send to IOrank
grid->SendRecvPacket((void *)&siteObj,(void *)&siteObj,rank,iorank,sizeof(siteObj));
}
}
grid->Barrier(); // necessary?
if (myrank == iorank) {
munge(siteObj,fileObj,csum);
@ -491,13 +636,7 @@ class BinaryIO {
fout.seekp(offset+g_idx*sizeof(fileObj));
fout.write((char *)&fileObj,sizeof(fileObj));
} else {
if ( myrank == rank ) {
peekLocalSite(siteObj,Umu,lsite);
grid->SendTo((void *)&siteObj,iorank,sizeof(siteObj));
}
}
grid->Barrier(); // necessary// or every 16 packets to rate throttle??
}
grid->GlobalSum(csum);

140
lib/parallelIO/NerscIO.h
View File

@ -1,3 +1,32 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/parallelIO/NerscIO.h
Copyright (C) 2015
Author: Matt Spraggs <matthew.spraggs@gmail.com>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_NERSC_IO_H
#define GRID_NERSC_IO_H
@ -51,7 +80,6 @@ class NerscField {
std::string floating_point;
};
//////////////////////////////////////////////////////////////////////
// Bit and Physical Checksumming and QA of data
//////////////////////////////////////////////////////////////////////
@ -69,8 +97,9 @@ inline void NerscGrid(GridBase *grid,NerscField &header)
template<class GaugeField>
inline void NerscStatistics(GaugeField & data,NerscField &header)
{
header.link_trace=Grid::QCD::WilsonLoops<GaugeField>::linkTrace(data);
header.plaquette =Grid::QCD::WilsonLoops<GaugeField>::avgPlaquette(data);
// How to convert data precision etc...
header.link_trace=Grid::QCD::WilsonLoops<PeriodicGimplR>::linkTrace(data);
header.plaquette =Grid::QCD::WilsonLoops<PeriodicGimplR>::avgPlaquette(data);
}
inline void NerscMachineCharacteristics(NerscField &header)
@ -181,9 +210,12 @@ inline void NerscMachineCharacteristics(NerscField &header)
class NerscIO : public BinaryIO {
public:
static inline void truncate(std::string file){
std::ofstream fout(file,std::ios::out);
}
static inline unsigned int writeHeader(NerscField &field,std::string file)
{
std::ofstream fout(file,std::ios::out);
std::ofstream fout(file,std::ios::out|std::ios::in);
fout.seekp(0,std::ios::beg);
fout << "BEGIN_HEADER" << std::endl;
@ -201,8 +233,7 @@ class NerscIO : public BinaryIO {
fout << "BOUNDARY_"<<i+1<<" = " << field.boundary[i] << std::endl;
}
fout << "CHECKSUM = "<< std::hex << std::setw(10) << field.checksum << std::endl;
fout << std::dec;
fout << "CHECKSUM = "<< std::hex << std::setw(10) << field.checksum << std::dec<<std::endl;
fout << "ENSEMBLE_ID = " << field.ensemble_id << std::endl;
fout << "ENSEMBLE_LABEL = " << field.ensemble_label << std::endl;
@ -287,9 +318,6 @@ static inline int readHeader(std::string file,GridBase *grid, NerscField &field
return field.data_start;
}
/////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Now the meat: the object readers
/////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@ -317,17 +345,17 @@ static inline void readConfiguration(Lattice<iLorentzColourMatrix<vsimd> > &Umu,
if ( header.data_type == std::string("4D_SU3_GAUGE") ) {
if ( ieee32 || ieee32big ) {
// csum=BinaryIO::readObjectSerial<iLorentzColourMatrix<vsimd>, LorentzColour2x3F>
csum=BinaryIO::readObjectParallel<iLorentzColourMatrix<vsimd>, LorentzColour2x3F>
csum=BinaryIO::readObjectParallel<iLorentzColourMatrix<vsimd>, LorentzColour2x3F>
(Umu,file,Nersc3x2munger<LorentzColour2x3F,LorentzColourMatrix>(), offset,format);
}
if ( ieee64 || ieee64big ) {
// csum=BinaryIO::readObjectSerial<iLorentzColourMatrix<vsimd>, LorentzColour2x3D>
//csum=BinaryIO::readObjectSerial<iLorentzColourMatrix<vsimd>, LorentzColour2x3D>
csum=BinaryIO::readObjectParallel<iLorentzColourMatrix<vsimd>, LorentzColour2x3D>
(Umu,file,Nersc3x2munger<LorentzColour2x3D,LorentzColourMatrix>(),offset,format);
(Umu,file,Nersc3x2munger<LorentzColour2x3D,LorentzColourMatrix>(),offset,format);
}
} else if ( header.data_type == std::string("4D_SU3_GAUGE_3X3") ) {
if ( ieee32 || ieee32big ) {
// csum=BinaryIO::readObjectSerial<iLorentzColourMatrix<vsimd>,LorentzColourMatrixF>
//csum=BinaryIO::readObjectSerial<iLorentzColourMatrix<vsimd>,LorentzColourMatrixF>
csum=BinaryIO::readObjectParallel<iLorentzColourMatrix<vsimd>,LorentzColourMatrixF>
(Umu,file,NerscSimpleMunger<LorentzColourMatrixF,LorentzColourMatrix>(),offset,format);
}
@ -344,6 +372,7 @@ static inline void readConfiguration(Lattice<iLorentzColourMatrix<vsimd> > &Umu,
assert(fabs(clone.plaquette -header.plaquette ) < 1.0e-5 );
assert(fabs(clone.link_trace-header.link_trace) < 1.0e-6 );
assert(csum == header.checksum );
std::cout<<GridLogMessage <<"Read NERSC Configuration "<<file<< " and plaquette, link trace, and checksum agree"<<std::endl;
@ -377,6 +406,8 @@ static inline void writeConfiguration(Lattice<iLorentzColourMatrix<vsimd> > &Umu
uint32_t csum;
int offset;
truncate(file);
if ( two_row ) {
header.floating_point = std::string("IEEE64BIG");
@ -389,11 +420,11 @@ static inline void writeConfiguration(Lattice<iLorentzColourMatrix<vsimd> > &Umu
std::string file1 = file+"para";
int offset1 = writeHeader(header,file1);
int csum1=BinaryIO::writeObjectParallel<vobj,fobj2D>(Umu,file1,munge,offset,header.floating_point);
//int csum1=BinaryIO::writeObjectSerial<vobj,fobj2D>(Umu,file1,munge,offset,header.floating_point);
std::cout << GridLogMessage << " TESTING PARALLEL WRITE offsets " << offset1 << " "<< offset << std::endl;
std::cout << GridLogMessage <<std::hex<< " TESTING PARALLEL WRITE csums " << csum1 << " "<< csum << std::endl;
std::cout << std::dec;
std::cout << GridLogMessage << " TESTING PARALLEL WRITE csums " << csum1 << " "<<std::hex<< csum << std::dec<< std::endl;
assert(offset1==offset);
assert(csum1==csum);
@ -404,13 +435,90 @@ static inline void writeConfiguration(Lattice<iLorentzColourMatrix<vsimd> > &Umu
NerscSimpleUnmunger<fobj3D,sobj> munge;
BinaryIO::Uint32Checksum<vobj,fobj3D>(Umu, munge,header.checksum);
offset = writeHeader(header,file);
csum=BinaryIO::writeObjectSerial<vobj,fobj3D>(Umu,file,munge,offset,header.floating_point);
// csum=BinaryIO::writeObjectSerial<vobj,fobj3D>(Umu,file,munge,offset,header.floating_point);
csum=BinaryIO::writeObjectParallel<vobj,fobj3D>(Umu,file,munge,offset,header.floating_point);
}
std::cout<<GridLogMessage <<"Written NERSC Configuration "<<file<< " checksum "<<std::hex<<csum<< std::dec<<" plaq "<< header.plaquette <<std::endl;
}
///////////////////////////////
// RNG state
///////////////////////////////
static inline void writeRNGState(GridSerialRNG &serial,GridParallelRNG &parallel,std::string file)
{
typedef typename GridParallelRNG::RngStateType RngStateType;
// Following should become arguments
NerscField header;
header.sequence_number = 1;
header.ensemble_id = "UKQCD";
header.ensemble_label = "DWF";
GridBase *grid = parallel._grid;
NerscGrid(grid,header);
header.link_trace=0.0;
header.plaquette=0.0;
NerscMachineCharacteristics(header);
uint32_t csum;
int offset;
#ifdef RNG_RANLUX
header.floating_point = std::string("UINT64");
header.data_type = std::string("RANLUX48");
#else
header.floating_point = std::string("UINT32");
header.data_type = std::string("MT19937");
#endif
truncate(file);
offset = writeHeader(header,file);
csum=BinaryIO::writeRNGSerial(serial,parallel,file,offset);
header.checksum = csum;
offset = writeHeader(header,file);
std::cout<<GridLogMessage <<"Written NERSC RNG STATE "<<file<< " checksum "<<std::hex<<csum<<std::dec<<std::endl;
}
static inline void readRNGState(GridSerialRNG &serial,GridParallelRNG & parallel,NerscField& header,std::string file)
{
typedef typename GridParallelRNG::RngStateType RngStateType;
GridBase *grid = parallel._grid;
int offset = readHeader(file,grid,header);
NerscField clone(header);
std::string format(header.floating_point);
std::string data_type(header.data_type);
#ifdef RNG_RANLUX
assert(format == std::string("UINT64"));
assert(data_type == std::string("RANLUX48"));
#else
assert(format == std::string("UINT32"));
assert(data_type == std::string("MT19937"));
#endif
// depending on datatype, set up munger;
// munger is a function of <floating point, Real, data_type>
uint32_t csum=BinaryIO::readRNGSerial(serial,parallel,file,offset);
std::cerr<<" Csum "<< csum << " "<< header.checksum <<std::endl;
assert(csum == header.checksum );
std::cout<<GridLogMessage <<"Read NERSC RNG file "<<file<< " format "<< data_type <<std::endl;
}
};
}}
#endif

0
lib/pugixml/.dirstamp Normal file
View File

32
lib/qcd/QCD.h
View File

@ -1,3 +1,34 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/qcd/QCD.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: Peter Boyle <peterboyle@Peters-MacBook-Pro-2.local>
Author: neo <cossu@post.kek.jp>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_QCD_H
#define GRID_QCD_H
namespace Grid{
@ -425,7 +456,6 @@ namespace QCD {
#include <qcd/spin/TwoSpinor.h>
#include <qcd/utils/LinalgUtils.h>
#include <qcd/utils/CovariantCshift.h>
#include <qcd/utils/WilsonLoops.h>
#include <qcd/utils/SUn.h>
#include <qcd/action/Actions.h>
#include <qcd/hmc/integrators/Integrator.h>

28
lib/qcd/action/ActionBase.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/qcd/action/ActionBase.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: neo <cossu@post.kek.jp>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef QCD_ACTION_BASE
#define QCD_ACTION_BASE
namespace Grid {

36
lib/qcd/action/ActionParams.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/qcd/action/ActionParams.h
Copyright (C) 2015
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_QCD_ACTION_PARAMS_H
#define GRID_QCD_ACTION_PARAMS_H
@ -6,10 +34,16 @@ namespace QCD {
// These can move into a params header and be given MacroMagic serialisation
struct GparityWilsonImplParams {
bool overlapCommsCompute;
std::vector<int> twists;
GparityWilsonImplParams () : twists(Nd,0), overlapCommsCompute(false) {};
};
struct WilsonImplParams { };
struct WilsonImplParams {
bool overlapCommsCompute;
WilsonImplParams() : overlapCommsCompute(false) {};
};
struct OneFlavourRationalParams {
RealD lo;

92
lib/qcd/action/Actions.h
View File

@ -1,11 +1,38 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/qcd/action/Actions.h
Copyright (C) 2015
Author: Azusa Yamaguchi <ayamaguc@staffmail.ed.ac.uk>
Author: Peter Boyle <pabobyle@ph.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: Peter Boyle <peterboyle@Peters-MacBook-Pro-2.local>
Author: neo <cossu@post.kek.jp>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_QCD_ACTIONS_H
#define GRID_QCD_ACTIONS_H
// Some reorganisation likely required as both Chroma and IroIro
// are separating the concept of the operator from that of action.
//
// The FermAction contains methods to create
// * Linear operators (Hermitian and non-hermitian) .. my LinearOperator
// * System solvers (Hermitian and non-hermitian) .. my OperatorFunction
// * MultiShift System solvers (Hermitian and non-hermitian) .. my OperatorFunction
@ -16,26 +43,53 @@
#include <qcd/action/ActionBase.h>
#include <qcd/action/ActionParams.h>
////////////////////////////////////////////
// Utility functions
////////////////////////////////////////////
#include <qcd/action/gauge/GaugeImpl.h>
#include <qcd/utils/WilsonLoops.h>
#include <qcd/action/fermion/WilsonCompressor.h> //used by all wilson type fermions
#include <qcd/action/fermion/FermionOperatorImpl.h>
#include <qcd/action/fermion/FermionOperator.h>
#include <qcd/action/fermion/WilsonKernels.h> //used by all wilson type fermions
////////////////////////////////////////////
// Gauge Actions
////////////////////////////////////////////
#include <qcd/action/gauge/WilsonGaugeAction.h>
#include <qcd/action/gauge/PlaqPlusRectangleAction.h>
namespace Grid {
namespace QCD {
typedef WilsonGaugeAction<LatticeGaugeField> WilsonGaugeActionR;
typedef WilsonGaugeAction<LatticeGaugeFieldF> WilsonGaugeActionF;
typedef WilsonGaugeAction<LatticeGaugeFieldD> WilsonGaugeActionD;
typedef WilsonGaugeAction<PeriodicGimplR> WilsonGaugeActionR;
typedef WilsonGaugeAction<PeriodicGimplF> WilsonGaugeActionF;
typedef WilsonGaugeAction<PeriodicGimplD> WilsonGaugeActionD;
typedef PlaqPlusRectangleAction<PeriodicGimplR> PlaqPlusRectangleActionR;
typedef PlaqPlusRectangleAction<PeriodicGimplF> PlaqPlusRectangleActionF;
typedef PlaqPlusRectangleAction<PeriodicGimplD> PlaqPlusRectangleActionD;
typedef IwasakiGaugeAction<PeriodicGimplR> IwasakiGaugeActionR;
typedef IwasakiGaugeAction<PeriodicGimplF> IwasakiGaugeActionF;
typedef IwasakiGaugeAction<PeriodicGimplD> IwasakiGaugeActionD;
typedef SymanzikGaugeAction<PeriodicGimplR> SymanzikGaugeActionR;
typedef SymanzikGaugeAction<PeriodicGimplF> SymanzikGaugeActionF;
typedef SymanzikGaugeAction<PeriodicGimplD> SymanzikGaugeActionD;
typedef WilsonGaugeAction<ConjugateGimplR> ConjugateWilsonGaugeActionR;
typedef WilsonGaugeAction<ConjugateGimplF> ConjugateWilsonGaugeActionF;
typedef WilsonGaugeAction<ConjugateGimplD> ConjugateWilsonGaugeActionD;
typedef PlaqPlusRectangleAction<ConjugateGimplR> ConjugatePlaqPlusRectangleActionR;
typedef PlaqPlusRectangleAction<ConjugateGimplF> ConjugatePlaqPlusRectangleActionF;
typedef PlaqPlusRectangleAction<ConjugateGimplD> ConjugatePlaqPlusRectangleActionD;
typedef IwasakiGaugeAction<ConjugateGimplR> ConjugateIwasakiGaugeActionR;
typedef IwasakiGaugeAction<ConjugateGimplF> ConjugateIwasakiGaugeActionF;
typedef IwasakiGaugeAction<ConjugateGimplD> ConjugateIwasakiGaugeActionD;
typedef SymanzikGaugeAction<ConjugateGimplR> ConjugateSymanzikGaugeActionR;
typedef SymanzikGaugeAction<ConjugateGimplF> ConjugateSymanzikGaugeActionF;
typedef SymanzikGaugeAction<ConjugateGimplD> ConjugateSymanzikGaugeActionD;
}}
////////////////////////////////////////////////////////////////////////////////////////////////////
@ -55,15 +109,16 @@ typedef WilsonGaugeAction<LatticeGaugeFieldD> WilsonGaugeActionD;
#define FermOpTemplateInstantiate(A) \
template class A<WilsonImplF>; \
template class A<WilsonImplD>;
// template class A<GparityWilsonImplF>; \
// template class A<GparityWilsonImplD>;
template class A<WilsonImplD>; \
template class A<GparityWilsonImplF>; \
template class A<GparityWilsonImplD>;
////////////////////////////////////////////
// Fermion operators / actions
////////////////////////////////////////////
#include <qcd/action/fermion/WilsonFermion.h> // 4d wilson like
#include <qcd/action/fermion/WilsonTMFermion.h> // 4d wilson like
#include <qcd/action/fermion/WilsonFermion5D.h> // 5d base used by all 5d overlap types
//#include <qcd/action/fermion/CloverFermion.h>
@ -100,6 +155,10 @@ typedef WilsonFermion<WilsonImplR> WilsonFermionR;
typedef WilsonFermion<WilsonImplF> WilsonFermionF;
typedef WilsonFermion<WilsonImplD> WilsonFermionD;
typedef WilsonTMFermion<WilsonImplR> WilsonTMFermionR;
typedef WilsonTMFermion<WilsonImplF> WilsonTMFermionF;
typedef WilsonTMFermion<WilsonImplD> WilsonTMFermionD;
typedef DomainWallFermion<WilsonImplR> DomainWallFermionR;
typedef DomainWallFermion<WilsonImplF> DomainWallFermionF;
typedef DomainWallFermion<WilsonImplD> DomainWallFermionD;
@ -165,11 +224,6 @@ typedef DomainWallFermion<GparityWilsonImplD> GparityDomainWallFermionD;
#include <qcd/action/pseudofermion/TwoFlavourEvenOdd.h>
#include <qcd/action/pseudofermion/TwoFlavourEvenOddRatio.h>
//IroIro inserted general "Nf" param; could also be done,
//but not clear why unless into large Nf BSM studies
//Even there, don't want the explicit (2) on power denominator
//if even number of flavours, so further generalised interface
//would be required but easy.
#include <qcd/action/pseudofermion/OneFlavourRational.h>
#include <qcd/action/pseudofermion/OneFlavourRationalRatio.h>
#include <qcd/action/pseudofermion/OneFlavourEvenOddRational.h>

35
lib/qcd/action/fermion/CayleyFermion5D.cc
View File

@ -1,3 +1,33 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/qcd/action/fermion/CayleyFermion5D.cc
Copyright (C) 2015
Author: Peter Boyle <pabobyle@ph.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
Author: Peter Boyle <peterboyle@Peters-MacBook-Pro-2.local>
Author: paboyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#include <Grid.h>
namespace Grid {
namespace QCD {
@ -114,7 +144,7 @@ namespace QCD {
// Apply Dw
this->DW(psi,Din,DaggerYes);
Meooe5D(Din,chi);
MeooeDag5D(Din,chi);
int Ls=this->Ls;
for(int s=0;s<Ls;s++){
@ -163,7 +193,6 @@ namespace QCD {
FermionField tmp(psi._grid);
// Assemble the 5d matrix
Meooe5D(psi,tmp);
#if 0
std::cout << "Meooe Test replacement norm2 tmp = " <<norm2(tmp)<<std::endl;
for(int s=0;s<Ls;s++){
@ -202,7 +231,7 @@ namespace QCD {
this->DhopOE(psi,tmp,DaggerYes);
}
Meooe5D(tmp,chi);
MeooeDag5D(tmp,chi);
#if 0
std::cout << "Meooe Test replacement norm2 chi new = " <<norm2(chi)<<std::endl;
// Assemble the 5d matrix

28
lib/qcd/action/fermion/CayleyFermion5D.h
View File

@ -1,3 +1,31 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./lib/qcd/action/fermion/CayleyFermion5D.h
Copyright (C) 2015
Author: Peter Boyle <pabobyle@ph.ed.ac.uk>
Author: Peter Boyle <paboyle@ph.ed.ac.uk>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution directory
*************************************************************************************/
/* END LEGAL */
#ifndef GRID_QCD_CAYLEY_FERMION_H
#define GRID_QCD_CAYLEY_FERMION_H

Some files were not shown because too many files have changed in this diff Show More