On 10/14/2017 12:35 AM, David Malcolm wrote:
> As far as I can tell from your mail, the one issue that blocks that
> is the lack of gdb::unique_xmalloc_ptr<T[]>.
>
> So here's an patch on top of the previous one which adds the
> xmalloc_deleter<T[]> (taken from gdb, but changing "xfree" to
> "free", and adding support for pre-C++11 dialects), along with
> selftests for unique_ptr<T[]>, unique_xmalloc_ptr<T> and
> unique_xmalloc_ptr<T[]>.
Thanks much!
> As before, successfully bootstrapped & regrtested on x86_64-pc-linux-gnu,
> using gcc 4.8 for the initial bootstrap (hence testing both gnu++03
> then gnu++14 in the selftests, for stage 1 and stages 2 and 3
> respectively).
>
Excellent.
> Hand-tested with "make selftest-valgrind" with both gnu++03 and
> gnu++14.
>
> Also tested stage1 on powerpc-ibm-aix7.1.3.0 ("gcc111" in the
> compile farm; gcc 4.8 i.e. gnu++03)
>
> Is this OK?
Looks great to me.
> +/* Verify that gnu::unique_malloc_ptr works. */
Typo: malloc -> xmalloc. Appears in other comments too.
Thanks,
Pedro Alves
