https://sourceware.org/bugzilla/show_bug.cgi?id=29535
Nick Clifton <nickc at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |nickc at redhat dot com
--- Comment #1 from Nick Clifton <nickc at redhat dot com> ---
Hi Ralf,
[I am selectively quoting from parts of your description....]
> <a0> DW_AT_name : ../test.c
> <a4> DW_AT_comp_dir : /home/<user>/src/test-binutils/build
> Combining DW_AT_name, '/' and DW_AT_comp_dir returns the correct path
> to the used source file.
But the DWARF standard does not specify that these two attributes should
be combined when using absolute paths. In fact in section 3.1.1 it says:
A DW_AT_name attribute whose value is a null-terminated string
containing the full or relative path name (relative to the value
of the DW_AT_comp_dir attribute, see below) of the primary source
file from which the compilation unit was derived.
So the DW_AT_comp_dir attribute is only used when DW_AT_name contains
a relative path, not an absolute one.
> The problem happens when using absolute paths:
>
> gcc -C -o test.o ~/src/test-binutils/test.c -g
[Note - you are giving an absolute path to gcc, and also asking it to
generate debug information].
> <a0> DW_AT_name : /home/<user>/src/test-binutils/test.c
> <a4> DW_AT_comp_dir : /home/<user>/src/test-binutils/build
Which happens because this is what gcc tells the assembler to do:
% gcc -C -o test.o ~/src/test-binutils/test.c -g -fdump-debug
% head test.c.341t.debug
DWARF for /home/nickc/src/test-binutils/test.c
DIE 0: DW_TAG_compile_unit (0x1521e1c2a190)
abbrev id: 0 offset: 0 mark: 0
DW_AT_producer: "GNU C17 12.2.1 20220819 (Red Hat 12.2.1-1) -mtune=generic
-march=x86-64 -g"
DW_AT_language: 29
DW_AT_name:
"/home/nickc/work/builds/binutils/current/x86_64-pc-linux-gnu/tests/pr29535.c"
DW_AT_comp_dir:
"/home/nickc/work/builds/binutils/current/x86_64-pc-linux-gnu/tests"
And if you check in gcc's assembler output you will find both of those
absolute paths.
Cheers
Nick
--
You are receiving this mail because:
You are on the CC list for the bug.
