Bug 67202 - Fortran FE should load scalar pass-by-reference intent-in arguments at the beginning of a function
Summary: Fortran FE should load scalar pass-by-reference intent-in arguments at the be...
Status: NEW
Alias: None
Product: gcc
Classification: Unclassified
Component: fortran (show other bugs)
Version: 6.0
: P3 enhancement
Target Milestone: ---
Assignee: Not yet assigned to anyone
URL: https://gcc.gnu.org/ml/gcc-patches/20...
Keywords: missed-optimization, patch
Depends on:
Blocks: 36854
  Show dependency tree / graph
 
Reported: 2015-08-13 09:54 UTC by Martin Liška
Modified: 2021-04-16 16:22 UTC (History)
2 users (show)

See Also:
Host:
Target:
Build:
Known to work:
Known to fail:
Last reconfirmed: 2015-11-08 00:00:00

Attachments
Test case (264 bytes, text/plain)
2015-08-13 09:54 UTC, Martin Liška 		Details
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Martin Liška 2015-08-13 09:54:02 UTC 
Created attachment 36178 [details]
Test case

Hello.

PR is follow-up of PR67170, where PRE can't hoist out an argument which is passed by reference and is intent-in. In the situation it would be beneficial to load the argument at the very beginning and transfer all references of the argument to the load.

Please take a look at attached test case.

Thanks,
Martin
Comment 1 Dominique d'Humieres 2015-11-08 17:42:53 UTC 
Is there an easy way to check than a scalar pass-by-reference intent-in argument has been loaded at the beginning of the function?
Comment 2 Thomas Koenig 2018-09-18 20:42:16 UTC 
(In reply to Dominique d'Humieres from comment #1)
> Is there an easy way to check than a scalar pass-by-reference intent-in
> argument has been loaded at the beginning of the function?

Easiest to check the *.original dump.

If the code looks like

x (integer(kind=4) & restrict i, integer(kind=4) & restrict j)
{
  *j = *i;
}


then it hasn't been done.
Comment 3 Martin Liška 2020-01-07 08:31:32 UTC 
@Thomas: Setting host to 67202 is probably a typo.
Comment 4 Thomas Koenig 2020-01-07 18:09:05 UTC 
(In reply to Martin Liška from comment #3)
> @Thomas: Setting host to 67202 is probably a typo.

Definitely :-) Removing.