This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: A few warning patches ...
- To: Richard Henderson <rth at cygnus dot com>
- Subject: Re: A few warning patches ...
- From: Jeffrey A Law <law at cygnus dot com>
- Date: Mon, 26 Oct 1998 21:36:49 -0700
- cc: "Kaveh R. Ghazi" <ghazi at caip dot rutgers dot edu>, egcs-patches at cygnus dot com, meissner at cygnus dot com
- Reply-To: law at cygnus dot com
In message <19981026185421.B25151@dot.cygnus.com>you write:
> On Sun, Oct 25, 1998 at 12:57:36PM -0500, Kaveh R. Ghazi wrote:
> > (allocate_reg_info): Move variable `i' into the scope where it is
> > used. Change its type to `size_t'.
>
> FYI, the argument to allocate_reg_info, num_regs, and thence almost
> all of the size_t variables in that function should be signed.
>
> Look at the first test:
>
> /* Free up all storage allocated */
> if (num_regs < 0)
> {
>
> which cannot be true if num_regs is unsigned. And it does seem like
> we'd at some point like to free up the reg info. ;-)
This has come up before. But we never heard back from Michael.
Michael, as it stands, we never call allocate_reg_info with a negate argument
to free up the storage. WHen did you envision that we'd free up the storage?
Once per compilation unit? After each function? when?
jeff