This is the mail archive of the
mailing list for the GCC project.
Re: PATCH: Fix 20000724-1.c
- To: Linus Torvalds <torvalds at transmeta dot com>
- Subject: Re: PATCH: Fix 20000724-1.c
- From: Richard Henderson <rth at redhat dot com>
- Date: Tue, 1 May 2001 10:21:59 -0700
- Cc: mark at codesourcery dot com, gcc-patches at gcc dot gnu dot org
- References: <20010430094928.B8430@redhat.com> <20010430105529F.firstname.lastname@example.org> <20010430110409.B8491@redhat.com> <20010430112038W.email@example.com> <200105010304.UAA09476@penguin.transmeta.com>
On Mon, Apr 30, 2001 at 08:04:04PM -0700, Linus Torvalds wrote:
> The fact that it has a pointer to the frame should be enough to key off.
We are in complete agreement.
I'm fairly sure that we've still got bugs in there wrt this,
though I've not been able to construct a test case to show it.
It's not clear to me how to fix it properly either. To know
for absolute certainty that a given pointer is _not_ a stack
frame reference requires a not inconsiderable amount of data
flow knowledge. Which would be nice to have in other places
too, so it's not like it wouldn't be worthwhile.
Probably I'll wind up with some sort of hack to Know Things
about stack frame (de)allocation, and be conservative wrt asms.