public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug testsuite/114182] New: gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300
@ 2024-03-01  0:34 jdx at o2 dot pl
  2024-03-01  0:58 ` [Bug testsuite/114182] " pinskia at gcc dot gnu.org
                   ` (4 more replies)
  0 siblings, 5 replies; 6+ messages in thread
From: jdx at o2 dot pl @ 2024-03-01  0:34 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114182

            Bug ID: 114182
           Summary: gcc.c-torture/compile/attr-complex-method-2.c fails
                    for H8/300
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: testsuite
          Assignee: unassigned at gcc dot gnu.org
          Reporter: jdx at o2 dot pl
  Target Milestone: ---
              Host: x86_64-pc-linux-gnu
            Target: h8300-elf

Created attachment 57581
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57581&action=edit
Dump file for attr-complex-method-2.c

[...]
PASS: gcc.c-torture/compile/attr-complex-method-2.c   -O0  (test for excess
errors)
FAIL: gcc.c-torture/compile/attr-complex-method-2.c   -O0   scan-tree-dump
optimized "__(?:gnu_)?divdc3"
PASS: gcc.c-torture/compile/attr-complex-method-2.c   -O1  (test for excess
errors)
FAIL: gcc.c-torture/compile/attr-complex-method-2.c   -O1   scan-tree-dump
optimized "__(?:gnu_)?divdc3"
PASS: gcc.c-torture/compile/attr-complex-method-2.c   -O2  (test for excess
errors)
FAIL: gcc.c-torture/compile/attr-complex-method-2.c   -O2   scan-tree-dump
optimized "__(?:gnu_)?divdc3"
PASS: gcc.c-torture/compile/attr-complex-method-2.c   -O3 -g  (test for excess
errors)
FAIL: gcc.c-torture/compile/attr-complex-method-2.c   -O3 -g   scan-tree-dump
optimized "__(?:gnu_)?divdc3"
PASS: gcc.c-torture/compile/attr-complex-method-2.c   -Os  (test for excess
errors)
FAIL: gcc.c-torture/compile/attr-complex-method-2.c   -Os   scan-tree-dump
optimized "__(?:gnu_)?divdc3"
UNSUPPORTED: gcc.c-torture/compile/attr-complex-method-2.c   -O2 -flto
-fno-use-linker-plugin -flto-partition=none 
UNSUPPORTED: gcc.c-torture/compile/attr-complex-method-2.c   -O2 -flto
-fuse-linker-plugin -fno-fat-lto-objects 
PASS: gcc.c-torture/compile/attr-complex-method.c   -O0  (test for excess
errors)
PASS: gcc.c-torture/compile/attr-complex-method.c   -O0   scan-tree-dump-not
optimized "__divdc3"
PASS: gcc.c-torture/compile/attr-complex-method.c   -O1  (test for excess
errors)
PASS: gcc.c-torture/compile/attr-complex-method.c   -O1   scan-tree-dump-not
optimized "__divdc3"
PASS: gcc.c-torture/compile/attr-complex-method.c   -O2  (test for excess
errors)
PASS: gcc.c-torture/compile/attr-complex-method.c   -O2   scan-tree-dump-not
optimized "__divdc3"
PASS: gcc.c-torture/compile/attr-complex-method.c   -O3 -g  (test for excess
errors)
PASS: gcc.c-torture/compile/attr-complex-method.c   -O3 -g   scan-tree-dump-not
optimized "__divdc3"
PASS: gcc.c-torture/compile/attr-complex-method.c   -Os  (test for excess
errors)
PASS: gcc.c-torture/compile/attr-complex-method.c   -Os   scan-tree-dump-not
optimized "__divdc3"
UNSUPPORTED: gcc.c-torture/compile/attr-complex-method.c   -O2 -flto
-fno-use-linker-plugin -flto-partition=none 
UNSUPPORTED: gcc.c-torture/compile/attr-complex-method.c   -O2 -flto
-fuse-linker-plugin -fno-fat-lto-objects
[...]

According to the attached dump file, I believe the end of
attr-complex-method-2.c should look like this:
/* { dg-final { scan-tree-dump "__(?:gnu_)?divdc3" "optimized" { target { ! {
avr-*-* h8300-*-* } } } } } */

/* { dg-final { scan-tree-dump "__(?:gnu_)?divsc3" "optimized" { target {
avr-*-* h8300-*-* } } } } */

Similarly, shouldn't attr-complex-method.c also search for __divsc3 for AVR,
H8/300 and possibly a few other architectures?

^ permalink raw reply	[flat|nested] 6+ messages in thread

* [Bug testsuite/114182] gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300
  2024-03-01  0:34 [Bug testsuite/114182] New: gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300 jdx at o2 dot pl
@ 2024-03-01  0:58 ` pinskia at gcc dot gnu.org
  2024-03-03 17:42 ` jdx at o2 dot pl
                   ` (3 subsequent siblings)
  4 siblings, 0 replies; 6+ messages in thread
From: pinskia at gcc dot gnu.org @ 2024-03-01  0:58 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114182

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1
           Keywords|                            |testsuite-fail
   Last reconfirmed|                            |2024-03-01

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Most likely it would be better to have a target support proc that returns if
`double == float` and use that.
Oh there is large_double  already. You should just use that instead.

So just:

/* { dg-final { scan-tree-dump "__(?:gnu_)?divdc3" "optimized" { target { ! {
large_double   } } } } } */

/* { dg-final { scan-tree-dump "__(?:gnu_)?divsc3" "optimized" { target {
large_double   } } } } */

>Similarly, shouldn't attr-complex-method.c also search for __divsc3 for AVR, H8/300 and possibly a few other architectures?

Most likely yes.

^ permalink raw reply	[flat|nested] 6+ messages in thread

* [Bug testsuite/114182] gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300
  2024-03-01  0:34 [Bug testsuite/114182] New: gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300 jdx at o2 dot pl
  2024-03-01  0:58 ` [Bug testsuite/114182] " pinskia at gcc dot gnu.org
@ 2024-03-03 17:42 ` jdx at o2 dot pl
  2024-03-04  1:12 ` jdx at o2 dot pl
                   ` (2 subsequent siblings)
  4 siblings, 0 replies; 6+ messages in thread
From: jdx at o2 dot pl @ 2024-03-03 17:42 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114182

--- Comment #2 from Jan Dubiec <jdx at o2 dot pl> ---
Unfortunately, large_double does not work.

^ permalink raw reply	[flat|nested] 6+ messages in thread

* [Bug testsuite/114182] gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300
  2024-03-01  0:34 [Bug testsuite/114182] New: gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300 jdx at o2 dot pl
  2024-03-01  0:58 ` [Bug testsuite/114182] " pinskia at gcc dot gnu.org
  2024-03-03 17:42 ` jdx at o2 dot pl
@ 2024-03-04  1:12 ` jdx at o2 dot pl
  2024-03-04  1:16 ` pinskia at gcc dot gnu.org
  2024-03-08 22:32 ` law at gcc dot gnu.org
  4 siblings, 0 replies; 6+ messages in thread
From: jdx at o2 dot pl @ 2024-03-04  1:12 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114182

--- Comment #3 from Jan Dubiec <jdx at o2 dot pl> ---
Wait a minute, shouldn't the conditions be opposite? I.e.:

/* { dg-final { scan-tree-dump "__(?:gnu_)?divdc3" "optimized" { target {
large_double } } } } */

/* { dg-final { scan-tree-dump "__(?:gnu_)?divsc3" "optimized" { target { ! {
large_double } } } } } */

^ permalink raw reply	[flat|nested] 6+ messages in thread

* [Bug testsuite/114182] gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300
  2024-03-01  0:34 [Bug testsuite/114182] New: gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300 jdx at o2 dot pl
                   ` (2 preceding siblings ...)
  2024-03-04  1:12 ` jdx at o2 dot pl
@ 2024-03-04  1:16 ` pinskia at gcc dot gnu.org
  2024-03-08 22:32 ` law at gcc dot gnu.org
  4 siblings, 0 replies; 6+ messages in thread
From: pinskia at gcc dot gnu.org @ 2024-03-04  1:16 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114182

--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Jan Dubiec from comment #3)
> Wait a minute, shouldn't the conditions be opposite? I.e.:
> 
> /* { dg-final { scan-tree-dump "__(?:gnu_)?divdc3" "optimized" { target {
> large_double } } } } */
> 
> /* { dg-final { scan-tree-dump "__(?:gnu_)?divsc3" "optimized" { target { !
> { large_double } } } } } */

Yes sorry about that mistake.

^ permalink raw reply	[flat|nested] 6+ messages in thread

* [Bug testsuite/114182] gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300
  2024-03-01  0:34 [Bug testsuite/114182] New: gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300 jdx at o2 dot pl
                   ` (3 preceding siblings ...)
  2024-03-04  1:16 ` pinskia at gcc dot gnu.org
@ 2024-03-08 22:32 ` law at gcc dot gnu.org
  4 siblings, 0 replies; 6+ messages in thread
From: law at gcc dot gnu.org @ 2024-03-08 22:32 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114182

Jeffrey A. Law <law at gcc dot gnu.org> changed:

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

--- Comment #5 from Jeffrey A. Law <law at gcc dot gnu.org> ---
Pinski's suggestion is the right one.  We have a small number of targets that
use 32bit doubles.  For the sake of testing we should define a target test to
cleanly indicate the size of a double and check that in attr-complex-method-2.c
and likely in other places.

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2024-03-08 22:33 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-03-01  0:34 [Bug testsuite/114182] New: gcc.c-torture/compile/attr-complex-method-2.c fails for H8/300 jdx at o2 dot pl
2024-03-01  0:58 ` [Bug testsuite/114182] " pinskia at gcc dot gnu.org
2024-03-03 17:42 ` jdx at o2 dot pl
2024-03-04  1:12 ` jdx at o2 dot pl
2024-03-04  1:16 ` pinskia at gcc dot gnu.org
2024-03-08 22:32 ` law 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).