[Bug gas/30853] strange 'return column' register in CIE

[eric.pouech at orange dot fr]

https://sourceware.org/bugzilla/show_bug.cgi?id=30853

--- Comment #3 from Eric Pouech  ---
which COFF register numbering are you referring to?

and how are we supposed to generate CIE from .cfi directives that would match
dwarf debug information (for return address column)?

using ".cfi_startproc simple" and (re)setting cfa & return column for each
function would work

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug gas/30853] strange 'return column' register in CIE

[eric.pouech at orange dot fr]

https://sourceware.org/bugzilla/show_bug.cgi?id=30853

--- Comment #2 from Eric Pouech  ---
I see...

that's why I suppose gcc doesn't rely on (g)as' cfi directives to emit its
dwarf CFI bits
(unless there's an option to tell x86_64-w64-mingw32-as to pick up another
flavor, but couldn't find it either)

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug gas/30853] strange 'return column' register in CIE

[sch...@linux-m68k.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30853

--- Comment #1 from Andreas Schwab  ---
  x86_dwarf2_return_column = (OUTPUT_FLAVOR == bfd_target_coff_flavour
  ? 32 : 16);

COFF apparantly uses a different register numbering.

-- 
You are receiving this mail because:
You are on the CC list for the bug.