This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: cse/const calls improvement
- To: Jan Hubicka <hubicka at atrey dot karlin dot mff dot cuni dot cz>
- Subject: Re: cse/const calls improvement
- From: Richard Henderson <rth at cygnus dot com>
- Date: Wed, 12 Apr 2000 01:17:03 -0700
- Cc: egcs-patches at egcs dot cygnus dot com
- References: <20000409140012.A17395@atrey.karlin.mff.cuni.cz>
On Sun, Apr 09, 2000 at 02:00:12PM +0200, Jan Hubicka wrote:
> * cse.c (true_dependence_in_rtx): New function.
> (invalidate): Use it.
The idea is sound, but I'd prefer you use for_each_rtx,
which would avoid bugs like
> + if (fmt[i] == 'e' && XEXP (x, i)
> + && true_dependence_in_rtx (mem, mem_mode, XEXP (x, i)))
forgetting to iterate over 'E' vectors.
r~