This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH 3/3] Make relocate_{path,gdb_directory} return std::string
>>>>> "Christian" == Christian Biesinger via gdb-patches <gdb-patches@sourceware.org> writes:
Christian> This simplifies memory management. I've also changed some global variables
Christian> to std::string accordingly (which store the result of these functions),
Christian> but not all because some are used with add_setshow_optional_filename_cmd
Christian> which requires a char*.
Thank you.
Christian> - xfree (gdb_datadir);
Christian> - gdb_datadir = gdb_realpath (new_datadir).release ();
Christian> + gdb_datadir = gdb_realpath (new_datadir).get ();
I wonder if using a unique_xmalloc_ptr would be better.
I suppose it doesn't matter hugely, but it would eliminate some of these
double allocations.
Christian> + gdb::unique_xmalloc_ptr<char> abs_datadir =
Christian> + gdb_abspath (gdb_datadir.c_str ());
"=" on continuation line.
Christian> -static char *
Christian> +static std::string
Christian> relocate_path (const char *progname, const char *initial, bool relocatable)
Christian> {
Christian> if (relocatable)
Christian> - return make_relative_prefix (progname, BINDIR, initial);
Christian> - return xstrdup (initial);
Christian> + {
Christian> + char *str = make_relative_prefix (progname, BINDIR, initial);
Christian> + if (str != nullptr)
Christian> + return str;
This seems to leak "str".
Christian> + char *canon_sysroot = lrealpath (dir.c_str ());
Changing this to a unique_xmalloc_ptr would mean one less call to
xfree...
Christian> + debug_file_directory =
"=" on the next line, there is one more of these in this spot.
Tom