This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: reg_n_info lossage
- To: Richard Henderson <rth at cygnus dot com>
- Subject: Re: reg_n_info lossage
- From: Jeffrey A Law <law at cygnus dot com>
- Date: Fri, 14 Aug 1998 01:37:11 -0600
- cc: egcs-patches at cygnus dot com
- Reply-To: law at cygnus dot com
In message <19980810003224.A3739@dot.cygnus.com>you write:
> The new valarray implementation of allocate_reg_info does not
> properly care for the resize case. Instead of simply zeroing
> the new entries, it zaps all of them.
>
> This shows up, for instance, in jump. On Alpha, given
>
> if (ca >= 'A' && ca <= 'Z')
> ca += 'a' - 'A';
> if (cb >= 'A' && cb <= 'Z')
> cb += 'a' - 'A';
>
> I'll only get one cmove generated, this because when it came
> time to do the second transform, the reg info had been killed.
>
>
> r~
>
>
>
> * regclass.c (allocate_reg_info): Respect MIN when clearing data.
If I read this correctly, this does not result incorrect code right?
Or are there cases where this problem can result in incorrect code.
The underlying question (of course) is do we need this for egcs-1.1?
The patch itself seems fine to me assuming you'll deal with the possible
fencepost errors noted by Graham.
Install it in whatever trees are appropriate :-)
jeff