[Bug gdb/28681] Wrong pretty-printed unique_ptr value when using "finish"

["aburgess at redhat dot com" <sourceware-bugzilla@sourceware.org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=28681

Andrew Burgess <aburgess at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |aburgess at redhat dot com

--- Comment #1 from Andrew Burgess <aburgess at redhat dot com> ---
I see the same result on x86-64.  I'm just guessing right now, but I suspect
that this is a case of the result being miss-classified as something that can
be passed in register, when, it is really something that is passed in memory.

The sys-v ABI says that if the return value is a non-trivial, C++ object, the
caller should allocate space for the return value, and pass the address of the
space to the callee as a hidden first argument.  The address of the object
should then be returned as the return value.

Checking the disassembly of this test, and I can confirm that this is what is
happening.

However, on return, GDB appears to treat the address of the unique_ptr on the
stack as if that was the unique_ptr, so I suspect, that GDB has classified the
return value as being passed in register, rather than in memory.

The culprit is likely to be amd64_return_value (or something called from that
function) for x86-64.

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