public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug rtl-optimization/100264] New: REE does not work on PARALLEL expressions with a single register SET child
@ 2021-04-26 11:29 christophm30 at gmail dot com
  2021-04-26 11:45 ` [Bug rtl-optimization/100264] " christophm30 at gmail dot com
                   ` (3 more replies)
  0 siblings, 4 replies; 5+ messages in thread
From: christophm30 at gmail dot com @ 2021-04-26 11:29 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100264

            Bug ID: 100264
           Summary: REE does not work on PARALLEL expressions with a
                    single register SET child
           Product: gcc
           Version: unknown
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: rtl-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: christophm30 at gmail dot com
  Target Milestone: ---

REE iterates over all SET expressions and tries to eliminate redundant
zero-extensions. In case of a PARALLEL expression, there is a check in place to
ensure that "only one child of a PARALLEL expression is a SET".

Let's imagine a PARALLEL expression with two register SETs.
An example would be store-conditional, which sets a memory location (store
target) as well as a register (return success value).
The current implementation is not able to optimize a zero-extension of the
return value of store-conditional.

This can be solved, by moving the check for register targets (i.e. REG_P ())
into the function get_sub_rtx () and change the restriction of REE to "only one
child of a PARALLEL expression is a SET register".

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

* [Bug rtl-optimization/100264] REE does not work on PARALLEL expressions with a single register SET child
  2021-04-26 11:29 [Bug rtl-optimization/100264] New: REE does not work on PARALLEL expressions with a single register SET child christophm30 at gmail dot com
@ 2021-04-26 11:45 ` christophm30 at gmail dot com
  2021-06-02 21:02 ` cvs-commit at gcc dot gnu.org
                   ` (2 subsequent siblings)
  3 siblings, 0 replies; 5+ messages in thread
From: christophm30 at gmail dot com @ 2021-04-26 11:45 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100264

--- Comment #1 from Christoph M. <christophm30 at gmail dot com> ---
A patch can be found here:
 https://gcc.gnu.org/pipermail/gcc-patches/2021-April/568680.html

It does not show any regressions (target riscv*).

The patch does not include a new test case to demonstrate it actually works,
because I don't know a generic way to test it.
At the moment I am testing indirectly using __atomic_compare_and_exchange () on
riscv64 to emit the required store-conditional. However, the code required for
that is not merged yet. Ideas on how to test this in a more generic way are
very much appreciated.

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

* [Bug rtl-optimization/100264] REE does not work on PARALLEL expressions with a single register SET child
  2021-04-26 11:29 [Bug rtl-optimization/100264] New: REE does not work on PARALLEL expressions with a single register SET child christophm30 at gmail dot com
  2021-04-26 11:45 ` [Bug rtl-optimization/100264] " christophm30 at gmail dot com
@ 2021-06-02 21:02 ` cvs-commit at gcc dot gnu.org
  2021-06-02 21:04 ` wilson at gcc dot gnu.org
  2021-09-17  6:44 ` pinskia at gcc dot gnu.org
  3 siblings, 0 replies; 5+ messages in thread
From: cvs-commit at gcc dot gnu.org @ 2021-06-02 21:02 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100264

--- Comment #2 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jim Wilson <wilson@gcc.gnu.org>:

https://gcc.gnu.org/g:d5ba2eaf7032f234fdcf11d9cfc513ce7be0a255

commit r12-1168-gd5ba2eaf7032f234fdcf11d9cfc513ce7be0a255
Author: Christoph Muellner <cmuellner@gcc.gnu.org>
Date:   Mon May 10 14:39:03 2021 +0200

    REE: PR rtl-optimization/100264: Handle more PARALLEL SET expressions

    Move the check for register targets (i.e. REG_P ()) into the function
    get_sub_rtx () and change the restriction of REE to "only one child of
    a PARALLEL expression is a SET register expression" (was "only one child of
    a PARALLEL expression is a SET expression").

    This allows to handle more PARALLEL SET expressions.

    gcc/ChangeLog:
            PR rtl-optimization/100264
            * ree.c (get_sub_rtx): Ignore SET expressions without register
            destinations and remove assertion, as it is not valid anymore
            with this new behaviour.
            (merge_def_and_ext): Eliminate destination check for register
            as such SET expressions can't occur anymore.
            (combine_reaching_defs): Likewise.

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

* [Bug rtl-optimization/100264] REE does not work on PARALLEL expressions with a single register SET child
  2021-04-26 11:29 [Bug rtl-optimization/100264] New: REE does not work on PARALLEL expressions with a single register SET child christophm30 at gmail dot com
  2021-04-26 11:45 ` [Bug rtl-optimization/100264] " christophm30 at gmail dot com
  2021-06-02 21:02 ` cvs-commit at gcc dot gnu.org
@ 2021-06-02 21:04 ` wilson at gcc dot gnu.org
  2021-09-17  6:44 ` pinskia at gcc dot gnu.org
  3 siblings, 0 replies; 5+ messages in thread
From: wilson at gcc dot gnu.org @ 2021-06-02 21:04 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100264

Jim Wilson <wilson at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
                 CC|                            |wilson at gcc dot gnu.org
         Resolution|---                         |FIXED

--- Comment #3 from Jim Wilson <wilson at gcc dot gnu.org> ---
Fixed on mainline.

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

* [Bug rtl-optimization/100264] REE does not work on PARALLEL expressions with a single register SET child
  2021-04-26 11:29 [Bug rtl-optimization/100264] New: REE does not work on PARALLEL expressions with a single register SET child christophm30 at gmail dot com
                   ` (2 preceding siblings ...)
  2021-06-02 21:04 ` wilson at gcc dot gnu.org
@ 2021-09-17  6:44 ` pinskia at gcc dot gnu.org
  3 siblings, 0 replies; 5+ messages in thread
From: pinskia at gcc dot gnu.org @ 2021-09-17  6:44 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100264

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |12.0

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

end of thread, other threads:[~2021-09-17  6:44 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-04-26 11:29 [Bug rtl-optimization/100264] New: REE does not work on PARALLEL expressions with a single register SET child christophm30 at gmail dot com
2021-04-26 11:45 ` [Bug rtl-optimization/100264] " christophm30 at gmail dot com
2021-06-02 21:02 ` cvs-commit at gcc dot gnu.org
2021-06-02 21:04 ` wilson at gcc dot gnu.org
2021-09-17  6:44 ` pinskia at gcc dot gnu.org

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).