bug-binutils
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Bug binutils/24557] Support underscore-prefixed string after mangled na


From: nickc at redhat dot com
Subject: [Bug binutils/24557] Support underscore-prefixed string after mangled name (for CUDA PTX parameters)
Date: Tue, 21 May 2019 15:22:54 +0000

https://sourceware.org/bugzilla/show_bug.cgi?id=24557

Nick Clifton <nickc at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
                 CC|                            |nickc at redhat dot com
         Resolution|---                         |MOVED

--- Comment #1 from Nick Clifton <nickc at redhat dot com> ---
(In reply to Eyal Rozenberg from comment #0)

Hi Eyal,

> I would like c++filt to either auto-detect this case, or have an option to
> detect it; 

I do not object to this request per-se, but there is a caveat:  The demangling
performed by c++filt is actually handled in a library (libiberty) which is
part of the GCC project, not the binutils project.  (The binutils project
makes use of the library, but we do not maintain it).

So you will need to refile this feature request with the GCC project.  If
they do decide to implement the feature, then there may also need to be
some follow up work on c++filt itself.  (Eg to pass a new flag into the
demangling code in libiberty).  If this happens I will be happy to add
the necessary support to c++filt.

It is possible however that the libiberty maintainers will not be terribly,
umm, enthusiastic about adopting such a feature.  In which case it may be
possible to add a workaround to c++filt.  For example a command line option
to strip a suffix off the end of a mangled name.  This does assume of course
that there is a predictable suffix generated by the CUDA compiler.  It also
means that other tools that perform name demangling (eg gdb) would not be
affected by the change, so they would still have the demangling problem.

Please try asking the GCC folks first, but if that fails, please reopen this
PR.

Cheers
  Nick

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


reply via email to

[Prev in Thread] Current Thread [Next in Thread]