[Bug testsuite/25766] New: objc.dg/stret-2.m fails on i686-darwin

[Bug testsuite/25766] New: objc.dg/stret-2.m fails on i686-darwin

[pinskia at gcc dot gnu dot org]

objc.dg/stret-2.m fails on i686-darwin because some struct is return not via
the stack but by registers.  Though I don't know for sure if this is correct so
I am filing this bug.


-- 
           Summary: objc.dg/stret-2.m fails on i686-darwin
           Product: gcc
           Version: 4.2.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: testsuite
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: pinskia at gcc dot gnu dot org
GCC target triplet: i686-apple-darwin


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25766

[Bug testsuite/25766] objc.dg/stret-2.m fails on i686-darwin

[pinskia at gcc dot gnu dot org]

------- Comment #1 from pinskia at gcc dot gnu dot org  2006-01-23 02:08 -------
I should note the reason why I filed this bug here is because this is a GCC
testsuite failure on i686-darwin.  Even though this is testing the NeXT
runtime, this is still a bug in either the objc front-end or the testsuite of
GCC.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25766

[Bug testsuite/25766] objc.dg/stret-2.m fails on i686-darwin

[iains at gcc dot gnu dot org]

------- Comment #2 from iains at gcc dot gnu dot org  2010-04-06 23:39 -------
according to PP50 of
http://developer.apple.com/Mac/library/documentation/DeveloperTools/Conceptual/LowLevelABI/100-32-bit_PowerPC_Function_Calling_Conventions/32bitPowerPC.html#//apple_ref/doc/uid/TP40002438-SW20
it's doing the Right Thing.  Structs <=8 bytes should be returned in EAX:EDX.

Conversely, the PPC ABI calls for structs to be returned into an area pointed
to by GPR3.

So, it seems that what we need to do is to make the struct > 8 bytes and then
the same test should apply to PPC and i686.

We should not be running this for m64 code (I had already included the /* {
dg-require-effective-target ilp32 } */ as part of the remaining patches for
PR35165 ... )

try this:

Index: gcc/testsuite/objc.dg/stret-2.m
===================================================================
--- gcc/testsuite/objc.dg/stret-2.m     (revision 158016)
+++ gcc/testsuite/objc.dg/stret-2.m     (working copy)
@@ -3,13 +3,15 @@
    points).  */
 /* Contributed by Ziemowit Laski <zlaski@apple.com>.  */
 /* { dg-do compile { target *-*-darwin* } } */
+/* { dg-require-effective-target ilp32 } */
 /* { dg-skip-if "" { *-*-* } { "-fgnu-runtime" } { "" } } */

 #include "../objc-obj-c++-shared/Object1.h"

 struct astruct {
   float a, b;
-} glob = { 1.0, 2.0 };
+  char makeMeBiggerThan8Bytes;
+} glob = { 1.0, 2.0, 'P' };

 struct bstruct {
   float a, b, c, d, e, f;


-- 

iains at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |iains at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25766

[Bug testsuite/25766] objc.dg/stret-2.m fails on i686-darwin

[iains at gcc dot gnu dot org]

------- Comment #3 from iains at gcc dot gnu dot org  2010-04-12 18:53 -------
see:
http://gcc.gnu.org/ml/gcc-patches/2010-04/msg00584.html
for an updated patch.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25766

[Bug testsuite/25766] objc.dg/stret-2.m fails on i686-darwin

[mrs at gcc dot gnu dot org]

------- Comment #4 from mrs at gcc dot gnu dot org  2010-05-28 17:56 -------
I checked in a fix for this in r159988.  I'm skeptical of the abi doc, so I'd
prefer just the first part, as I know it is needed.  Thanks.


-- 

mrs at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25766