[PATCH 0/3] New function value_has_address

[Yao Qi <qiyaoltc at gmail dot com>]

This patch series is inspired by the discussion in this thread
https://sourceware.org/ml/gdb-patches/2016-10/msg00741.html in which
I only added value_has_address.  During the discussion, looks
value_has_address should be true only for lval_memory and lval_register,
then, I restrict value_has_address (done in patch #3), and trigger a
lot of assertions, because VALUE_VAL is not set properly before
set_value_address.  Patch #2 fixes these problems.

Patch series is regression tested on x86_64-linux.

Note that, I also find that value_has_address may only return true
for lval_memory, because I can't see how VALUE->location.address is valid
to lval_register.  Of course, I can be wrong, so further analysis
is needed.

*** BLURB HERE ***

Yao Qi (3):
  New function value_has_address
  Set VALUE_VAL before set_value_address
  Restrict value_has_address

 gdb/ada-lang.c |  2 +-
 gdb/elfread.c  |  2 ++
 gdb/value.c    | 27 ++++++++++++++++-----------
 3 files changed, 19 insertions(+), 12 deletions(-)

-- 
1.9.1