[Libreoffice-bugs] [Bug 143450] Data corruption when returning small structs containing a double from C++ via IPC

[bugzilla-daemon at bugs.documentfoundation.org]

https://bugs.documentfoundation.org/show_bug.cgi?id=143450

--- Comment #3 from Julien Nabet <serval2412 at yahoo.fr> ---
Just taking a look at the bridges/source/cpp_uno/gcc3_linux_x86-64/abi.cxx, I
noticed 
this enum:
71  enum x86_64_reg_class
72  {
73      X86_64_NO_CLASS,
74      X86_64_INTEGER_CLASS,
75      X86_64_INTEGERSI_CLASS,
76      X86_64_SSE_CLASS,
77      X86_64_SSESF_CLASS,
78      X86_64_SSEDF_CLASS,
79      X86_64_SSEUP_CLASS,
80      X86_64_X87_CLASS,
81      X86_64_X87UP_CLASS,
82      X86_64_MEMORY_CLASS
83  };
see
https://opengrok.libreoffice.org/xref/core/bridges/source/cpp_uno/gcc3_linux_x86-64/abi.cxx?r=10d29c39#74

idem in bridges/source/cpp_uno/gcc3_macosx_x86-64/abi.cxx

When I Googled searched SSESF gcc, I found this:
enum x86_64_reg_class
  {
    X86_64_NO_CLASS,
    X86_64_INTEGER_CLASS,
    X86_64_INTEGERSI_CLASS,
    X86_64_SSE_CLASS,
    X86_64_SSESF_CLASS,
    X86_64_SSEDF_CLASS,
    X86_64_SSEUP_CLASS,
    X86_64_X87_CLASS,
    X86_64_X87UP_CLASS,
    X86_64_COMPLEX_X87_CLASS, <<<<<< no present in LO
    X86_64_MEMORY_CLASS
  };
see https://github.com/gcc-mirror/gcc/blob/master/gcc/config/i386/i386.c

Is it expected we don't declare X86_64_COMPLEX_X87_CLASS ?

For the rest I must recognize I don't know anything about PS ABI, I've just
leafed through this doc:
https://courses.cs.washington.edu/courses/cse351/12wi/supp-docs/abi.pdf

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/libreoffice-bugs/attachments/20210725/025d4869/attachment.htm>