Re: [PATCH] Multiple & vs. && and | vs. || bugs in 2.4.20

[Richard Henderson]

On Sun, Mar 02, 2003 at 06:41:14PM -0300, Werner Almesberger wrote:
> While I agree with your observation in general, this is actually
> something the compiler should be able to figure out by itself:
> 
>  - there's only a side-effect if acm is NULL

In general there's also a side effect if acm is uninitialized.
I didn't look at the code in question here.

As for the rest, it's true that we should be able to do this,
but we don't currently have a pass that globally propagates
"trapiness" of memory references.  It would be a useful thing
to have though, particularly for Java, which is required to
arrange for these traps to be able to be caught with exceptions,
and other horrible reordering issues.

r~