[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Bug 37131 depends on bug 65819, which changed state. Bug 65819 Summary: overzealous checking in gfc_check_dependency for identical=true https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65819 What|Removed |Added

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Bug 37131 depends on bug 68009, which changed state. Bug 68009 Summary: [7/8 Regression] prototype for gfortran_runtime_error with inline matmul https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68009 What|Removed

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Bug 37131 depends on bug 29550, which changed state. Bug 29550 Summary: Optimize -fexternal-blas calls for conjg() https://gcc.gnu.org/bugzilla/show_bug.cgi?id=29550 What|Removed |Added

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Thomas Koenig changed: What|Removed |Added Status|NEW |RESOLVED Resolution|---

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #33 from Thomas Koenig --- Author: tkoenig Date: Mon May 29 06:03:23 2017 New Revision: 248553 URL: https://gcc.gnu.org/viewcvs?rev=248553=gcc=rev Log: 2017-05-29 Thomas Koenig PR fortran/37131

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Bug 37131 depends on bug 66094, which changed state. Bug 66094 Summary: Handle transpose(A) in inline matmul https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66094 What|Removed |Added

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Bug 37131 depends on bug 68600, which changed state. Bug 68600 Summary: Inlined MATMUL is too slow. https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68600 What|Removed |Added

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #32 from Thomas Koenig --- Author: tkoenig Date: Mon May 1 17:45:52 2017 New Revision: 247441 URL: https://gcc.gnu.org/viewcvs?rev=247441=gcc=rev Log: 2017-05-01 Thomas Koenig PR fortran/37131

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Bug 37131 depends on bug 66189, which changed state. Bug 66189 Summary: Block loops for inline matmul https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66189 What|Removed |Added

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Bug 37131 depends on bug 69154, which changed state. Bug 69154 Summary: [6 Regression] ICE in gfc_trans_where_2, at fortran/trans-stmt.c:5005 on *-linux https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69154 What|Removed

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Bug 37131 depends on bug 66176, which changed state. Bug 66176 Summary: Handle conjg() in inline matmul https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66176 What|Removed |Added

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #31 from Thomas Koenig tkoenig at gcc dot gnu.org --- Author: tkoenig Date: Sun May 17 13:45:07 2015 New Revision: 223268 URL: https://gcc.gnu.org/viewcvs?rev=223268root=gccview=rev Log: 2015-05-17 Thomas Koenig tkoe...@gcc.gnu.org

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #30 from Thomas Koenig tkoenig at gcc dot gnu.org --- Author: tkoenig Date: Tue May 12 06:37:43 2015 New Revision: 223031 URL: https://gcc.gnu.org/viewcvs?rev=223031root=gccview=rev Log: 2015-05-12 Thomas Koenig tkoe...@gcc.gnu.org

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #29 from Thomas Koenig tkoenig at gcc dot gnu.org --- Further ideas: - Handling of TRANSPOSEd arguments - Temporaries for arguments which are not plain arrays - Remove size0 checks (the DO loops will do that on their own) - Remove

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #28 from Thomas Koenig tkoenig at gcc dot gnu.org --- Author: tkoenig Date: Wed May 6 20:23:48 2015 New Revision: 222864 URL: https://gcc.gnu.org/viewcvs?rev=222864root=gccview=rev Log: 2015-05-06 Thomas Koenig tkoe...@gcc.gnu.org

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #27 from Thomas Koenig tkoenig at gcc dot gnu.org --- Author: tkoenig Date: Sun May 3 18:09:57 2015 New Revision: 222751 URL: https://gcc.gnu.org/viewcvs?rev=222751root=gccview=rev Log: 2015-05-03 Thomas Koenig tkoe...@gcc.gnu.org

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Andreas Schwab sch...@linux-m68k.org changed: What|Removed |Added CC||hjl.tools at gmail

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #22 from Andreas Schwab sch...@linux-m68k.org --- Program aborted. Backtrace: #0 0x3FF95D8B973 FAIL: gfortran.dg/bound_9.f90 -O execution test

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #23 from Uroš Bizjak ubizjak at gmail dot com --- (In reply to Andreas Schwab from comment #22) Program aborted. Backtrace: #0 0x3FF95D8B973 FAIL: gfortran.dg/bound_9.f90 -O execution test Thanks for the report! Before we can

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #24 from Dominique d'Humieres dominiq at lps dot ens.fr --- Reduced test module tst implicit none contains subroutine bar (a, b, n, m) integer, dimension(:), allocatable, intent(inout) :: a integer, dimension(:),

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #25 from Mikael Morin mikael at gcc dot gnu.org --- (In reply to Dominique d'Humieres from comment #24) I wonder if the above code is valid Fortran. y is not associated, so not suitable for passing as ARRAY argument to LBOUND:

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #21 from Thomas Koenig tkoenig at gcc dot gnu.org --- Author: tkoenig Date: Thu Apr 30 22:12:31 2015 New Revision: 222661 URL: https://gcc.gnu.org/viewcvs?rev=222661root=gccview=rev Log: 2015-04-30 Thomas Koenig tkoe...@gcc.gnu.org

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #20 from Thomas Koenig tkoenig at gcc dot gnu.org --- First submitted patch: https://gcc.gnu.org/ml/gcc-patches/2015-04/msg00969.html

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #18 from Thomas Koenig tkoenig at gcc dot gnu.org --- Created attachment 35356 -- https://gcc.gnu.org/bugzilla/attachment.cgi?id=35356action=edit First attempt that appears to work Well, this seems to work so far, no regressions.

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Thomas Koenig tkoenig at gcc dot gnu.org changed: What|Removed |Added CC||burnus at gcc

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Thomas Koenig tkoenig at gcc dot gnu.org changed: What|Removed |Added Assignee|unassigned at gcc dot gnu.org |tkoenig

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #16 from Mikael Morin mikael at gcc dot gnu.org --- (In reply to Thomas Koenig from comment #15) Hi Mikael, do you think you can do this using the scalarizer? Not in its current state. And as I don't see the scalarizer being

[Bug fortran/37131] inline matmul for small matrix sizes

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 Thomas Koenig tkoenig at gcc dot gnu.org changed: What|Removed |Added Last reconfirmed|2008-08-16 22:55:22 |2014-10-2 ---

[Bug fortran/37131] inline matmul for small matrix sizes

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37131 --- Comment #14 from Thomas Koenig tkoenig at gcc dot gnu.org 2011-01-02 23:20:07 UTC --- Created attachment 22883 -- http://gcc.gnu.org/bugzilla/attachment.cgi?id=22883 matmul loops that vectorize Here's how we could make the different matmul

[Bug fortran/37131] inline matmul for small matrix sizes

--- Comment #10 from paul dot richard dot thomas at gmail dot com 2010-06-05 06:55 --- Subject: Re: inline matmul for small matrix sizes Dear Thomas, The preferred way would therefore be to state the rank 2 * rank 2 problem as  do i=1,m     do j=1,n        c(i,j) = sum(a(i,:)

[Bug fortran/37131] inline matmul for small matrix sizes

--- Comment #11 from tkoenig at gcc dot gnu dot org 2010-06-05 08:49 --- Dear Paul, thanks a lot for your helpful comments. Just one thing: I currently don't see how to refer to multiple indices for an array element. In the code you pointed out, this is done with a single variable,

[Bug fortran/37131] inline matmul for small matrix sizes

--- Comment #12 from mikael at gcc dot gnu dot org 2010-06-05 09:31 --- (In reply to comment #9) I have thought a little bit about this, and the problem is a bit daunting ;-) Of course, this is at least partly because my experience with the scalarizer is close to non-existant, but

[Bug fortran/37131] inline matmul for small matrix sizes

--- Comment #13 from tkoenig at netcologne dot de 2010-06-05 18:27 --- Subject: Re: inline matmul for small matrix sizes mikael at gcc dot gnu dot org wrote: I'm working on nested scalarization loops for the sum intrinsic (pr43829) ; inlining matmul should be straightforward after

[Bug fortran/37131] inline matmul for small matrix sizes

--- Comment #9 from tkoenig at gcc dot gnu dot org 2010-06-04 22:31 --- I have thought a little bit about this, and the problem is a bit daunting ;-) Of course, this is at least partly because my experience with the scalarizer is close to non-existant, but you have to learn sometime.

[Bug fortran/37131] inline matmul for small matrix sizes

--- Comment #8 from tkoenig at gcc dot gnu dot org 2010-05-14 09:15 --- New timings, on x86_64-unknown-linux-gnu. I split off the invalidate subroutine to make sure the optimizers don't optimize this out: i...@linux-fd1f:/tmp gfortran -O3 matmul.f90 invalidate.f90 i...@linux-fd1f:/tmp

[Bug fortran/37131] inline matmul for small matrix sizes

--- Comment #7 from tkoenig at gcc dot gnu dot org 2008-12-09 22:12 --- Created an attachment (id=16866) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16866action=view) better test case Thou shalt use IMPLICIT none, especially if you think you don't need it... Here's a better

[Bug fortran/37131] inline matmul for small matrix sizes

--- Comment #6 from tkoenig at gcc dot gnu dot org 2008-12-04 19:58 --- (In reply to comment #5) For comparison with ifort (loop was vectorized in lines 40, 41, 43): matmul =2.660166 s subroutine without explicit interface: 0.000E+00 s subroutine with explicit

[Bug fortran/37131] inline matmul for small matrix sizes

--- Comment #5 from burnus at gcc dot gnu dot org 2008-11-29 16:18 --- (In reply to comment #4) Timings on x86_64-unknown-linux-gnu: matmul =12.840802 s subroutine without explicit interface: 0.88805580 s subroutine with explicit interface: 0.87605572 s

[Bug fortran/37131] inline matmul for small matrix sizes

--- Comment #4 from tkoenig at gcc dot gnu dot org 2008-08-23 13:18 --- Created an attachment (id=16134) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16134action=view) test case Actually, the test cases were a bit unfair, because the middle-end decided not to calculate the

[Bug fortran/37131] inline matmul for small matrix sizes

--- Comment #3 from pinskia at gcc dot gnu dot org 2008-08-16 22:55 --- Confirmed. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added