This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH 2/2] aarch64: implement walking over the stack protector
- From: Kyle McMartin <kmcmarti at redhat dot com>
- To: Andrew Pinski <pinskia at gmail dot com>
- Cc: "gdb-patches at sourceware dot org" <gdb-patches at sourceware dot org>
- Date: Thu, 5 Jun 2014 16:06:22 -0400
- Subject: Re: [PATCH 2/2] aarch64: implement walking over the stack protector
- Authentication-results: sourceware.org; auth=none
- References: <20140603050011 dot GA15355 at redacted dot bos dot redhat dot com> <20140603050314 dot GC15355 at redacted dot bos dot redhat dot com> <CA+=Sn1=Gsi61Wv=fzsU-QwJw=KLeN-iMeikRcKnFpTeCssv2zA at mail dot gmail dot com> <20140603145108 dot GF15355 at redacted dot bos dot redhat dot com>
On Tue, Jun 03, 2014 at 10:51:08AM -0400, Kyle McMartin wrote:
> On Mon, Jun 02, 2014 at 10:42:32PM -0700, Andrew Pinski wrote:
> > > + adrp x0, __stack_chk_guard
> > > + add x0, x0, #:lo12:__stack_chk_guard
> > > + ldr x0, [x0]
> > > + str x0, [x29, #end-of-stack]
> >
> > Can you expand this for ILP32? The sequence is the same except to use
> > w0 instead of x0. Otherwise I can put it on my todo list to after I
> > submit the gdb support for ILP32.
> >
>
> I'll look into it. Thanks Andrew.
>
Finally got a toolchain built that supports ilp32... looks like the
existing code should work for it. The only difference (not sure if it's
because of gcc HEAD or what) is using w1 instead of x0 repeated for the
ldr/str, which looks to be satisfied by the rt/rt2 tests and size=10
included in the decode_masked_match. I'll try to build a static ilp32
binary to test it meanwhile, but I've had some other issues with gcc
HEAD.
regards, Kyle