I don't have a Fortran standard at hand, but this is what I found on the web:
"The FORTRAN standard prohibits using the same variable (or array
element) more than once in the same CALL statement, if one of the
associated formal arguments (i.e. associated with the aliased
actual arguments) is assigned a value during the subprogram call.
An example of such aliasing:
CALL SUB1(ARRAY, ARRAY(1))
Passing the same variable (or array element) at the same time
with a CALL statement and in a common block is also prohibited."
In the gamess case, open64 compiler does pass the same address twice in one
call statement.
Standard does not require compiler to implement copy-in copy-out to enforce the
parameter-no-alias rule.
So a case like this can be considered as a program bug in my opinion.
-Mei
From: Sun Chan [mailto:sun.c...@gmail.com]
Sent: Thursday, January 27, 2011 3:26 PM
To: Ye, Mei
Cc: Roy Ju; open64-devel@lists.sourceforge.net
Subject: Re: [Open64-devel] Fortran parameter no-alias rule?
easy enough to ascertain, just print the addr of the items in question
Sun
On Fri, Jan 28, 2011 at 7:15 AM, Ye, Mei
<mei...@amd.com<mailto:mei...@amd.com>> wrote:
That's what I think.
-Mei
From: Roy Ju [mailto:roy_dc...@yahoo.com<mailto:roy_dc...@yahoo.com>]
Sent: Thursday, January 27, 2011 3:09 PM
To:
open64-devel@lists.sourceforge.net<mailto:open64-devel@lists.sourceforge.net>;
Ye, Mei
Subject: Re: [Open64-devel] Fortran parameter no-alias rule?
Does this mean that 416.gamess violates the Fortran param no aliasing rule?
Roy
--- On Thu, 1/27/11, Ye, Mei <mei...@amd.com<mailto:mei...@amd.com>> wrote:
From: Ye, Mei <mei...@amd.com<mailto:mei...@amd.com>>
Subject: [Open64-devel] Fortran parameter no-alias rule?
To:
"open64-devel@lists.sourceforge.net<mailto:open64-devel@lists.sourceforge.net>"
<open64-devel@lists.sourceforge.net<mailto:open64-devel@lists.sourceforge.net>>
Date: Thursday, January 27, 2011, 10:11 AM
Hello,
I am seeking your opinion on what is the best way to solve a problem that I
observed.
Compiler assumes that formal parameters based on different symbols in Fortran
programs are not aliased. This leads to a runtime error for CPU2006 416.gamess
when "-Ofast -LNO:fusion=2" is used.
In the following code snippet, formal parameter array "D" and "DO" in
subroutine "SYMTRD" have the same storage as their corresponding actual
parameter "X(I40)". Therefore the read of "D" in the 1st loop must alias with
the write of "DO" in the 2nd loop. When the two outermost loops are fused
together, we got incorrect results.
In summary, the Fortran parameter no-alias rule makes loop transformations
illegal.
For legality concern, we could rely on interprocedual analysis to track whether
formal parameters have the same base. But compiler analysis tends to be
conservative and therefore could hurt performance. It might not always be
feasible to implement "copy-in copy-out" which may hurt performance as well.
Will that be reasonable to treat this as an user error?
Code snippet:
callsite:
CALL SYMTRD(X(I30),X(I40),X(I10),X(I40),X(I60),IA,L1,L1)
callee:
SUBROUTINE SYMTRD(SH,D,DT,DO,DOD,IA,N,NDIM)
DIMENSION SH(*),D(*),DT(NDIM,*),DO(*),DOD(*),IA(*)
......
DO 140 I = 1,N
DO 120 J = 1,N
DTEMP = ZERO
DO 100 K = 1,N
IK = IA(I)+K
IF (I .LT. K) IK = IA(K)+I
JK = IA(J)+K
IF (J .LT. K) JK = IA(K)+J
DTEMP = DTEMP+SH(IK)*D(JK) // read of D
100 CONTINUE
DT(I,J) = DTEMP
120 CONTINUE
140 CONTINUE
IJ = 0
DO 200 I = 1,N
DO 180 J = 1,I
IJ = IJ+1
DTEMP = ZERO
DO 160 K = 1,N
JK = IA(J)+K
IF (J .LT. K) JK = IA(K)+J
DTEMP = DTEMP+DT(I,K)*SH(JK)
160 CONTINUE
DO(IJ) = DTEMP // write of DO
180 CONTINUE
200 CONTINUE
-Mei Ye
-----Inline Attachment Follows-----
------------------------------------------------------------------------------
Special Offer-- Download ArcSight Logger for FREE (a $49 USD value)!
Finally, a world-class log management solution at an even better price-free!
Download using promo code Free_Logger_4_Dev2Dev. Offer expires
February 28th, so secure your free ArcSight Logger TODAY!
http://p.sf.net/sfu/arcsight-sfd2d
-----Inline Attachment Follows-----
_______________________________________________
Open64-devel mailing list
Open64-devel@lists.sourceforge.net<http://us.mc1620.mail.yahoo.com/mc/compose?to=Open64-devel@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/open64-devel
------------------------------------------------------------------------------
Special Offer-- Download ArcSight Logger for FREE (a $49 USD value)!
Finally, a world-class log management solution at an even better price-free!
Download using promo code Free_Logger_4_Dev2Dev. Offer expires
February 28th, so secure your free ArcSight Logger TODAY!
http://p.sf.net/sfu/arcsight-sfd2d
_______________________________________________
Open64-devel mailing list
Open64-devel@lists.sourceforge.net<mailto:Open64-devel@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/open64-devel
------------------------------------------------------------------------------
Special Offer-- Download ArcSight Logger for FREE (a $49 USD value)!
Finally, a world-class log management solution at an even better price-free!
Download using promo code Free_Logger_4_Dev2Dev. Offer expires
February 28th, so secure your free ArcSight Logger TODAY!
http://p.sf.net/sfu/arcsight-sfd2d
_______________________________________________
Open64-devel mailing list
Open64-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/open64-devel
