[Bug modula2/114133] New: problem passing a string pointer to a C function on solaris 32 bit and 64 bit

[Bug modula2/114133] New: problem passing a string pointer to a C function on solaris 32 bit and 64 bit

[gaius at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114133

            Bug ID: 114133
           Summary: problem passing a string pointer to a C function on
                    solaris 32 bit and 64 bit
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: modula2
          Assignee: gaius at gcc dot gnu.org
          Reporter: gaius at gcc dot gnu.org
  Target Milestone: ---

This is a follow on from:

Bug 114026 - incorrect location during for loop type check

which occurred when I added a new testcase and was reported as failing:


Two of the new tests FAIL on 32 and 64-bit Solaris/SPARC:

+FAIL: gm2/extensions/run/pass/callingc10.mod execution,  -O 
+FAIL: gm2/extensions/run/pass/callingc10.mod execution,  -O -g 
+FAIL: gm2/extensions/run/pass/callingc10.mod execution,  -O3
-fomit-frame-point
er 
+FAIL: gm2/extensions/run/pass/callingc10.mod execution,  -O3
-fomit-frame-point
er -finline-functions 
+FAIL: gm2/extensions/run/pass/callingc10.mod execution,  -Os 
+FAIL: gm2/extensions/run/pass/callingc10.mod execution,  -g 
+FAIL: gm2/extensions/run/pass/callingc11.mod execution,  -O 
+FAIL: gm2/extensions/run/pass/callingc11.mod execution,  -O -g 
+FAIL: gm2/extensions/run/pass/callingc11.mod execution,  -O3
-fomit-frame-pointer 
+FAIL: gm2/extensions/run/pass/callingc11.mod execution,  -O3
-fomit-frame-pointer -finline-functions 
+FAIL: gm2/extensions/run/pass/callingc11.mod execution,  -Os 
+FAIL: gm2/extensions/run/pass/callingc11.mod execution,  -g 

The failure mode is the same for both:

parameter is hello and length 0
executed
/var/gcc/regression/master/11.4-gcc/build/gcc/testsuite/gm2/callingc10.x0 with
result fail

[Bug modula2/114133] problem passing a string pointer to a C function on solaris 32 bit and 64 bit

[gaius at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114133

Gaius Mulley <gaius at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2024-02-27

--- Comment #1 from Gaius Mulley <gaius at gcc dot gnu.org> ---
The gimple IR looks correct, given the input code:

MODULE callingc10 ;

FROM cvararg IMPORT funcptr ;
FROM SYSTEM IMPORT ADR ;

BEGIN
   IF funcptr (1, "hello", 5) = 1
   THEN
   END ;
   IF funcptr (1, "hello" + " ", 6) = 1
   THEN
   END ;
   IF funcptr (1, "hello" + " " + "world", 11) = 1
   THEN
   END
END callingc10.

$ gm2 -g callingc10.mod -c -fdump-ipa-all
$ cat callingc10.mod.095i.comdats
...
PROC _M2_callingc10_init (INTEGER argc, PROC * argv, PROC * envp)
{
  INTEGER D.670;
  INTEGER D.669;
  INTEGER D.668;
  PROC * _T34.0_1;
  INTEGER _2;
  INTEGER _T35.1_3;
  PROC * _T36.2_4;
  INTEGER _5;
  INTEGER _T37.3_6;
  PROC * _T38.4_7;
  INTEGER _8;
  INTEGER _12;
  INTEGER _16;
  INTEGER _20;

  <bb 2> :
  _T34 = "hello";
  _T34.0_1 = _T34;
  _12 = funcptr (1, _T34.0_1, 5);
  _2 = _12;
  _T35 = _2;
  _T35.1_3 = _T35;

...

[Bug modula2/114133] problem passing a string pointer to a C function on solaris 32 bit and 64 bit

[gaius at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114133

--- Comment #2 from Gaius Mulley <gaius at gcc dot gnu.org> ---
Created attachment 57552
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57552&action=edit
Query proposed fix

Does this patch fix the problem?

[Bug modula2/114133] problem passing a string pointer to a C function on solaris 32 bit and 64 bit

[gaius at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114133

--- Comment #3 from Gaius Mulley <gaius at gcc dot gnu.org> ---
At a guess the problem was the ZTyped constant (1 and 5).  Now the gimple IR
shows these constants as integers:

$ cat callingc10.mod.095i.comdats
PROC _M2_callingc10_init (INTEGER argc, PROC * argv, PROC * envp)
{
  INTEGER D.676;
  INTEGER D.675;
  INTEGER D.674;
  INTEGER _T35.0_1;
  PROC * _T36.1_2;
  INTEGER _T34.2_3;
  INTEGER _4;
  INTEGER _T37.3_5;
  INTEGER _T39.4_6;
  PROC * _T40.5_7;
  INTEGER _T38.6_8;
  INTEGER _9;
  INTEGER _T41.7_10;
  INTEGER _T43.8_11;
  PROC * _T44.9_12;
  INTEGER _T42.10_13;
  INTEGER _14;
  INTEGER _20;
  INTEGER _26;
  INTEGER _32;

  <bb 2> :
  _T34 = 1;
  _T35 = 5;
  _T36 = "hello";
  _T35.0_1 = _T35;
  _T36.1_2 = _T36;
  _T34.2_3 = _T34;
  _20 = funcptr (_T34.2_3, _T36.1_2, _T35.0_1);
  _4 = _20;
  _T37 = _4;
  _T37.3_5 = _T37;

[Bug modula2/114133] problem passing a string pointer to a C function on solaris 32 bit and 64 bit

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114133

--- Comment #4 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Gaius Mulley <gaius@gcc.gnu.org>:

https://gcc.gnu.org/g:76e591200f54226290ddb49b8ac6231a694bf882

commit r15-300-g76e591200f54226290ddb49b8ac6231a694bf882
Author: Gaius Mulley <gaiusmod2@gmail.com>
Date:   Tue May 7 19:24:08 2024 +0100

    PR modula2/114133 bugfix constants must be cast prior to vararg call

    This bug fix corrects the test codes below by converting the constant
    literals to the type required by C.  In the testcases below the values, 1
    etc were converted into the INTEGER type before being passed to a C
    vararg function.  By default in modula2 constant literal ordinals are
    represented as the ZTYPE (the largest GCC integer type node).

    gcc/testsuite/ChangeLog:

            PR modula2/114133
            * gm2/extensions/run/pass/callingc10.mod: Convert constant
            literal numbers into INTEGER.
            * gm2/extensions/run/pass/callingc11.mod: Ditto.
            * gm2/extensions/run/pass/vararg2.mod: Ditto.
            * gm2/iso/run/pass/packed.mod: Emit a printf as a runtime
            diagnostic.

    Signed-off-by: Gaius Mulley <gaiusmod2@gmail.com>

[Bug modula2/114133] problem passing a string pointer to a C function on solaris 32 bit and 64 bit

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114133

--- Comment #5 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-14 branch has been updated by Gaius Mulley
<gaius@gcc.gnu.org>:

https://gcc.gnu.org/g:07dab3f6b56c711dcd737d856cf01a597a2e1626

commit r14-10179-g07dab3f6b56c711dcd737d856cf01a597a2e1626
Author: Gaius Mulley <gaiusmod2@gmail.com>
Date:   Tue May 7 19:51:08 2024 +0100

    [PR modula2/113768][PR modula2/114133] bugfix constants must be cast prior
to vararg call

    This bug fix corrects the test codes below by converting the constant
    literals to the type required by C.  In the testcases below the values, 1
    etc were converted into the INTEGER type before being passed to a C
    vararg function.  By default in modula2 constant literal ordinals are
    represented as the ZTYPE (the largest GCC integer type node).

    gcc/testsuite/ChangeLog:

            PR modula2/113768
            PR modula2/114133
            * gm2/extensions/run/pass/callingc10.mod: Convert constant literal
            numbers into INTEGER.
            * gm2/extensions/run/pass/callingc11.mod: Ditto.
            * gm2/extensions/run/pass/vararg2.mod: Ditto.
            * gm2/iso/run/pass/packed.mod: Emit a printf as a runtime
            diagnostic.

    Signed-off-by: Gaius Mulley <gaiusmod2@gmail.com>

[Bug modula2/114133] problem passing a string pointer to a C function on solaris 32 bit and 64 bit

[gaius at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114133

Gaius Mulley <gaius at gcc dot gnu.org> changed:

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

--- Comment #6 from Gaius Mulley <gaius at gcc dot gnu.org> ---
Closing now that the patch has been applied.