From: Thomas Koenig <tkoenig@netcologne.de>
To: Steve Kargl <sgk@troutmask.apl.washington.edu>
Cc: Richard Biener <richard.guenther@gmail.com>,
Tobias Burnus <tobias@codesourcery.com>,
"fortran@gcc.gnu.org" <fortran@gcc.gnu.org>
Subject: Re: MATMUL broken with frontend optimization.
Date: Thu, 18 Mar 2021 21:55:27 +0100 [thread overview]
Message-ID: <367461e9-72b4-0f8d-d04e-878789595942@netcologne.de> (raw)
In-Reply-To: <20210318202239.GA25584@troutmask.apl.washington.edu>
Am 18.03.21 um 21:22 schrieb Steve Kargl:
> On Thu, Mar 18, 2021 at 07:24:21PM +0100, Thomas Koenig wrote:
>> I didn't finish the previous mail before hitting "send", so here
>> is the postscript...
>>
>>> OK, so I've had a bit of time to look at the actual test case. I
>>> missed one very important detail before: This is a vector-matrix
>>> operation.
>>>
>>> For this, we do not have a good library routine (Harald just
>>> removed it because of a bug in buffering), and -fexternal-blas
>>> does not work because we do not handle calls to anything but
>>> *GEMM.
>>
>> A vector-matrix multiplicatin would be a call to *GEMV, a worthy
>> goal, but out of scope so close to a release.
>
> Agreed.
>
>>> The idea is that, for a vector-matrix-multiplication, the
>>> compiler should have enough information about the information
>> about how to optimize for the relevant architecture, especially
>> if the user compilers with the right flags.
>>
>> So, the current idea is that, if we optimize, we can inline.
>>
>> What would a better heuristic be?
>>
>
> Does _gfortran_matmul_r4 (and friends) work for vector-matrix
> products?
Yes.
> I haven't checked. If so, how about disabling
> in-lining MATMUL for 11.1;
Absolutely not for the general case. This would cause a huge regression
in execution time for 2*2 matrices, and also for small matrix-vector
multiplications.
What we could do is only to enable the inlining for vector*matrix
at -O2 or higher. Again, this will mean a penalty for smaller loops,
but at less than -O2, people probably don't care too much.
If there is agreement on that, I will prepare a patch.
Regards
Thomas
next prev parent reply other threads:[~2021-03-18 20:55 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-18 7:48 Steve Kargl
2021-03-18 12:35 ` Richard Biener
2021-03-18 14:48 ` Tobias Burnus
2021-03-18 15:05 ` Richard Biener
2021-03-18 16:13 ` Steve Kargl
2021-03-18 18:16 ` Thomas Koenig
2021-03-18 18:24 ` Thomas Koenig
2021-03-18 20:22 ` Steve Kargl
2021-03-18 20:55 ` Thomas Koenig [this message]
2021-03-18 22:07 ` Steve Kargl
2021-03-19 6:19 ` Thomas Koenig
2021-03-19 6:36 ` Thomas Koenig
2021-03-19 19:12 ` Steve Kargl
2021-03-19 7:03 ` Steve Kargl
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=367461e9-72b4-0f8d-d04e-878789595942@netcologne.de \
--to=tkoenig@netcologne.de \
--cc=fortran@gcc.gnu.org \
--cc=richard.guenther@gmail.com \
--cc=sgk@troutmask.apl.washington.edu \
--cc=tobias@codesourcery.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).