[Bug target/45585] [4.6 Regression] ICE on powerpc-apple-darwin9 for gfortran.dg/transfer_simplify_2.f90 with -O2 -m64

[Bug target/45585] [4.6 Regression] ICE on powerpc-apple-darwin9 for gfortran.dg/transfer_simplify_2.f90 with -O2 -m64

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45585

Richard Guenther <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|powerpc-apple-darwin9       |powerpc64-*-*,
                   |                            |powerpc-apple-darwin9
           Priority|P3                          |P1
                 CC|                            |meissner at gcc dot gnu.org

[Bug target/45585] [4.6 Regression] ICE on powerpc-apple-darwin9 for gfortran.dg/transfer_simplify_2.f90 with -O2 -m64

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45585

--- Comment #3 from Dominique d'Humieres <dominiq at lps dot ens.fr> 2010-10-10 19:38:13 UTC ---
Any agenda for this pr?

[Bug target/45585] [4.6 Regression] ICE on powerpc-apple-darwin9 for gfortran.dg/transfer_simplify_2.f90 with -O2 -m64

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45585

--- Comment #4 from Dominique d'Humieres <dominiq at lps dot ens.fr> 2010-11-04 22:00:07 UTC ---
Created attachment 22288
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22288
folder containing the result of -dump-rtl-all

The file "verbose" contains the result of

gfc -O2 -m64 -fdump-rtl-all -fverbose-asm -mdebug=reg
transfer_simplify_2_red.f90 >& verbose

[Bug target/45585] [4.6 Regression] ICE on powerpc-apple-darwin9 for gfortran.dg/transfer_simplify_2.f90 with -O2 -m64

[meissner at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45585

Michael Meissner <meissner at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot       |meissner at gcc dot gnu.org
                   |gnu.org                     |

--- Comment #5 from Michael Meissner <meissner at gcc dot gnu.org> 2010-11-05 21:28:55 UTC ---
Created attachment 22297
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22297
Patch to allow loading/storing DI values in FPR registers

This patch fixes the problem in the bug.  What is happening is originally if
you were loading up a static 64-bit integer variable to convert it to floating
point, the compiler would first load the variable into a GPR, then store it on
the stack, and then load the value from the stack to the FPR.  My optimization
allowed the fp convert to take memory operands so it didn't have to go through
the GPR registers.  However, since Darwin has its own patterns for load/store,
it didn't go through the normal move patterns, and was overlooked.

[Bug target/45585] [4.6 Regression] ICE on powerpc-apple-darwin9 for gfortran.dg/transfer_simplify_2.f90 with -O2 -m64

[meissner at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45585

--- Comment #6 from Michael Meissner <meissner at gcc dot gnu.org> 2010-11-05 21:31:42 UTC ---
Author: meissner
Date: Fri Nov  5 21:31:38 2010
New Revision: 166379

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=166379
Log:
Fix PR 45585, darwin brain damage

Modified:
    branches/ibm/power7-fpconv/gcc/ChangeLog.power7
    branches/ibm/power7-fpconv/gcc/config/rs6000/darwin.md

[Bug target/45585] [4.6 Regression] ICE on powerpc-apple-darwin9 for gfortran.dg/transfer_simplify_2.f90 with -O2 -m64

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45585

--- Comment #7 from Dominique d'Humieres <dominiq at lps dot ens.fr> 2010-11-07 11:45:18 UTC ---
I have posted at http://gcc.gnu.org/ml/gcc-testresults/2010-11/msg00551.html
the result of the regtest with the patch in comment #5. The patch fixes this PR
without adding regressions of its own (the new failures are also seen in
http://gcc.gnu.org/ml/gcc-testresults/2010-11/msg00540.html and
gfortran.dg/vect/pr45714-b.f failed before the patch).

Thanks.

[Bug target/45585] [4.6 Regression] ICE on powerpc-apple-darwin9 for gfortran.dg/transfer_simplify_2.f90 with -O2 -m64

[meissner at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45585

--- Comment #8 from Michael Meissner <meissner at gcc dot gnu.org> 2010-11-08 19:02:39 UTC ---
Author: meissner
Date: Mon Nov  8 19:02:35 2010
New Revision: 166448

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=166448
Log:
Fix PR 46378 and PR 45585

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/rs6000/darwin.md
    trunk/gcc/config/rs6000/rs6000.c
    trunk/gcc/config/rs6000/rs6000.md

[Bug target/45585] [4.6 Regression] ICE on powerpc-apple-darwin9 for gfortran.dg/transfer_simplify_2.f90 with -O2 -m64

[meissner at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45585

Michael Meissner <meissner at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED

--- Comment #9 from Michael Meissner <meissner at gcc dot gnu.org> 2010-11-08 19:04:23 UTC ---
Fixed in subversion id: 166448.