diff options
author | Tom Tromey <tromey@adacore.com> | 2019-11-14 15:00:19 -0700 |
---|---|---|
committer | Tom Tromey <tromey@adacore.com> | 2019-11-21 14:39:40 -0700 |
commit | 6ba185213659517b4299a3e4c92813839f19f045 (patch) | |
tree | 78644366f6071accd40a41a98eaea41d1d0e0615 /gdb/aarch32-tdep.c | |
parent | 34877895ca38f74ae31bd65a6916560020d9d62b (diff) | |
download | gdb-6ba185213659517b4299a3e4c92813839f19f045.zip gdb-6ba185213659517b4299a3e4c92813839f19f045.tar.gz gdb-6ba185213659517b4299a3e4c92813839f19f045.tar.bz2 |
Handle %I64d in format_pieces
We found a bug internally where gdb would crash while disassembling a
certain instruction. This was tracked down to the handling of %I64d
in format_pieces.
format_pieces will convert %ll to %I64d on mingw -- so format_pieces
should also handle parsing this format. In this patch, I've made the
parsing unconditional, since I think it is harmless to accept extra
formats. I've also taken the opportunity to convert the length
modifier test to a "switch".
Tested internally using our failing test case.
gdb/ChangeLog
2019-11-21 Tom Tromey <tromey@adacore.com>
* gdbsupport/format.c (format_pieces): Parse %I64d.
* unittests/format_pieces-selftests.c (test_windows_formats): New
function.
(run_tests): Call it.
Change-Id: If335c7c2fc8d01e629cd55182394a483334d79c7
Diffstat (limited to 'gdb/aarch32-tdep.c')
0 files changed, 0 insertions, 0 deletions