https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82016

            Bug ID: 82016
           Summary: [8 regression] test case
                    gcc.target/powerpc/float128-type-1.c fails after
                    r251260
           Product: gcc
           Version: 8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: seurer at gcc dot gnu.org
  Target Milestone: ---

r251260 is the change to wide-int.h but this test case failure is different
from the others that change caused.

The test case states:

/* This test tests whether -mfloat128-type (which enables the underlying IEEE
   128-bit floating point) is enabled by default on VSX Linux 64-bit systems,
   even if the keywords __float128 and _Float128 (-mfloat128) are not enabled
   via the -mfloat128 switch.  Test that power8 generates a call to the
   __addkf3 emulation function.  */

and after the wide-int.h change the call to __addkf3 is missing.

seurer@genoa:~/gcc/build/gcc-test$ diff float128-type-1.s.r251259
float128-type-1.s.r251260
3d2
<       .globl __addkf3
10,21c9
< .LCF0:
< 0:    addis 2,12,.TOC.-.LCF0@ha
<       addi 2,2,.TOC.-.LCF0@l
<       .localentry     add_scalar,.-add_scalar
<       mflr 0
<       std 0,16(1)
<       stdu 1,-32(1)
<       bl __addkf3
<       nop
<       addi 1,1,32
<       ld 0,16(1)
<       mtlr 0
---
>       vspltisw 2,0
24c12
<       .byte 0,0,0,1,128,0,0,0
---
>       .byte 0,0,0,0,0,0,0,0
31,65c19,21
< .LCF1:
< 0:    addis 2,12,.TOC.-.LCF1@ha
<       addi 2,2,.TOC.-.LCF1@l
<       .localentry     add_complex,.-add_complex
<       mflr 0
<       std 0,16(1)
<       stdu 1,-80(1)
<       li 0,32
<       stvx 29,1,0
<       li 0,48
<       xxlor 61,37,37
<       stvx 30,1,0
<       li 0,64
<       xxlor 62,35,35
<       xxlor 35,36,36
<       stvx 31,1,0
<       bl __addkf3
<       nop
<       xxlor 35,61,61
<       xxlor 63,34,34
<       xxlor 34,62,62
<       bl __addkf3
<       nop
<       addi 1,1,80
<       li 0,-48
<       lvx 29,1,0
<       li 0,-32
<       xxlor 0,34,34
<       xxlor 34,63,63
<       lvx 30,1,0
<       li 0,-16
<       xxlor 35,0,0
<       lvx 31,1,0
<       ld 0,16(1)
<       mtlr 0
---
>       vspltisw 0,0
>       xxlor 34,32,32
>       xxlor 35,32,32
68c24
<       .byte 0,0,0,1,128,0,0,0
---
>       .byte 0,0,0,0,0,0,0,0
70c26
<       .ident  "GCC: (GNU) 8.0.0 20170822 (experimental) [trunk revision
251259]"
---
>       .ident  "GCC: (GNU) 8.0.0 20170822 (experimental) [trunk revision 
> 251260]"

Reply via email to