[Bug binutils/18968] make check fails on arm embedded NAS with no neon fpu

[nickc at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=18968

Nick Clifton <nickc at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |nickc at redhat dot com
         Resolution|---                         |WONTFIX

--- Comment #1 from Nick Clifton <nickc at redhat dot com> ---
Hi C E Macfarlane,

> ... but after an apparently successful make (no -k switch), when I come to
> testing, there are errors.  I can't decide whether the errors are being
> caused by the absence of a neon fpu, or something more obscure and/or
> important.
> 
> Can someone advise as to whether I can proceed to make install or that this
> serious enough to halt installation until fixed?

Good news / Bad news ...

The bad news is that this is the wrong place to file this bug report.  The
errors are being reported as part of the libiberty package, which is part of
the gcc project rather than the binutils project.  (Although both prokjects do
make use of libiberty).  So really you ought to refile this bug report with
gcc.  But...

The good news is that these failures are not serious.  Certainly not enough to
stop your installation:

> FAIL at line 670, options --format=dlang:
> in:  _D8demangle17__T4testVde0A8P6Zv
> out: demangle.test!(0.00000)
> exp: demangle.test!(42.0000)

The failures are all with the name demangler, that is the functions that
convert mangled C++ (or Java, or whatever) symbol names into fully qualified
names.  So whilst this feature is useful if you are going to be debugging
programs written in high level languages that use name mangling, it is not at
all important for the running of your system.

Cheers
  Nick

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