public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug rtl-optimization/109351] New: RA uses lowest cost for the mode of different reg_classes w/o considering hard_regno_mode_ok.
@ 2023-03-31  6:13 crazylht at gmail dot com
  2023-03-31  7:40 ` [Bug rtl-optimization/109351] " ubizjak at gmail dot com
                   ` (3 more replies)
  0 siblings, 4 replies; 5+ messages in thread
From: crazylht at gmail dot com @ 2023-03-31  6:13 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109351

            Bug ID: 109351
           Summary: RA uses lowest cost for the mode of different
                    reg_classes w/o considering hard_regno_mode_ok.
           Product: gcc
           Version: 13.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: rtl-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: crazylht at gmail dot com
  Target Milestone: ---

575  for (cl = (int) N_REG_CLASSES - 1; cl >= 0; cl--)
 576    {
 577      if (cl != (int) NO_REGS)
 578        for (mode = 0; mode < MAX_MACHINE_MODE; mode++)
 579          {
 580            ira_max_memory_move_cost[mode][cl][0]
 581              = ira_memory_move_cost[mode][cl][0]
 582              = memory_move_cost ((machine_mode) mode,
 583                                  (reg_class_t) cl, false);
 584            ira_max_memory_move_cost[mode][cl][1]
 585              = ira_memory_move_cost[mode][cl][1]
 586              = memory_move_cost ((machine_mode) mode,
 587                                  (reg_class_t) cl, true);
 588            /* Costs for NO_REGS are used in cost calculation on the
 589               1st pass when the preferred register classes are not
 590               known yet.  In this case we take the best scenario.  */
 591            if (ira_memory_move_cost[mode][NO_REGS][0]
 592                > ira_memory_move_cost[mode][cl][0])
 593              ira_max_memory_move_cost[mode][NO_REGS][0]
 594                = ira_memory_move_cost[mode][NO_REGS][0]
 595                = ira_memory_move_cost[mode][cl][0];
 596            if (ira_memory_move_cost[mode][NO_REGS][1]
 597                > ira_memory_move_cost[mode][cl][1])
 598              ira_max_memory_move_cost[mode][NO_REGS][1]
 599                = ira_memory_move_cost[mode][NO_REGS][1]
 600                = ira_memory_move_cost[mode][cl][1];
 601          }
 602    }


It could be a potential performance issue.

It's from PR108707.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [Bug rtl-optimization/109351] RA uses lowest cost for the mode of different reg_classes w/o considering hard_regno_mode_ok.
  2023-03-31  6:13 [Bug rtl-optimization/109351] New: RA uses lowest cost for the mode of different reg_classes w/o considering hard_regno_mode_ok crazylht at gmail dot com
@ 2023-03-31  7:40 ` ubizjak at gmail dot com
  2023-04-05  6:01 ` ubizjak at gmail dot com
                   ` (2 subsequent siblings)
  3 siblings, 0 replies; 5+ messages in thread
From: ubizjak at gmail dot com @ 2023-03-31  7:40 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109351

Uroš Bizjak <ubizjak at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |vmakarov at gcc dot gnu.org

--- Comment #1 from Uroš Bizjak <ubizjak at gmail dot com> ---
Cc added.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [Bug rtl-optimization/109351] RA uses lowest cost for the mode of different reg_classes w/o considering hard_regno_mode_ok.
  2023-03-31  6:13 [Bug rtl-optimization/109351] New: RA uses lowest cost for the mode of different reg_classes w/o considering hard_regno_mode_ok crazylht at gmail dot com
  2023-03-31  7:40 ` [Bug rtl-optimization/109351] " ubizjak at gmail dot com
@ 2023-04-05  6:01 ` ubizjak at gmail dot com
  2023-04-19  5:24 ` cvs-commit at gcc dot gnu.org
  2023-04-23  1:59 ` crazylht at gmail dot com
  3 siblings, 0 replies; 5+ messages in thread
From: ubizjak at gmail dot com @ 2023-04-05  6:01 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109351

--- Comment #2 from Uroš Bizjak <ubizjak at gmail dot com> ---
Patch at https://gcc.gnu.org/pipermail/gcc-patches/2023-April/615074.html

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [Bug rtl-optimization/109351] RA uses lowest cost for the mode of different reg_classes w/o considering hard_regno_mode_ok.
  2023-03-31  6:13 [Bug rtl-optimization/109351] New: RA uses lowest cost for the mode of different reg_classes w/o considering hard_regno_mode_ok crazylht at gmail dot com
  2023-03-31  7:40 ` [Bug rtl-optimization/109351] " ubizjak at gmail dot com
  2023-04-05  6:01 ` ubizjak at gmail dot com
@ 2023-04-19  5:24 ` cvs-commit at gcc dot gnu.org
  2023-04-23  1:59 ` crazylht at gmail dot com
  3 siblings, 0 replies; 5+ messages in thread
From: cvs-commit at gcc dot gnu.org @ 2023-04-19  5:24 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109351

--- Comment #3 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by hongtao Liu <liuhongt@gcc.gnu.org>:

https://gcc.gnu.org/g:727be65ec40e119a7c864bfaa5d6a73547863c06

commit r14-57-g727be65ec40e119a7c864bfaa5d6a73547863c06
Author: liuhongt <hongtao.liu@intel.com>
Date:   Mon Apr 3 10:54:55 2023 +0800

    Check hard_regno_mode_ok before setting lowest memory move cost for the
mode with different reg classes.

    There's a potential performance issue when backend returns some
    unreasonable value for the mode which can be never be allocate with
    reg class.

    gcc/ChangeLog:

            PR rtl-optimization/109351
            * ira.cc (setup_class_subset_and_memory_move_costs): Check
            hard_regno_mode_ok before setting lowest memory move cost for
            the mode with different reg classes.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [Bug rtl-optimization/109351] RA uses lowest cost for the mode of different reg_classes w/o considering hard_regno_mode_ok.
  2023-03-31  6:13 [Bug rtl-optimization/109351] New: RA uses lowest cost for the mode of different reg_classes w/o considering hard_regno_mode_ok crazylht at gmail dot com
                   ` (2 preceding siblings ...)
  2023-04-19  5:24 ` cvs-commit at gcc dot gnu.org
@ 2023-04-23  1:59 ` crazylht at gmail dot com
  3 siblings, 0 replies; 5+ messages in thread
From: crazylht at gmail dot com @ 2023-04-23  1:59 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109351

Hongtao.liu <crazylht at gmail dot com> changed:

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

--- Comment #4 from Hongtao.liu <crazylht at gmail dot com> ---
Fixed in GCC14

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2023-04-23  1:59 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-03-31  6:13 [Bug rtl-optimization/109351] New: RA uses lowest cost for the mode of different reg_classes w/o considering hard_regno_mode_ok crazylht at gmail dot com
2023-03-31  7:40 ` [Bug rtl-optimization/109351] " ubizjak at gmail dot com
2023-04-05  6:01 ` ubizjak at gmail dot com
2023-04-19  5:24 ` cvs-commit at gcc dot gnu.org
2023-04-23  1:59 ` crazylht at gmail dot com

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).