This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: ppc va-arg-10 fix
- To: Franz dot Sirl-kernel at lauterbach dot com
- Subject: Re: ppc va-arg-10 fix
- From: Mark Mitchell <mark at codesourcery dot com>
- Date: Sun, 03 Oct 1999 09:14:23 -0700
- Cc: rth at cygnus dot com, gcc-patches at gcc dot gnu dot org, harinath at cs dot umn dot edu
- Organization: CodeSourcery, LLC
- References: <19990928011836.A6371@cygnus.com><99092823230100.00852@ns1102.munich.netsurf.de><99100317540000.00581@ns1102.munich.netsurf.de>
>>>>> "Franz" == Franz Sirl <Franz.Sirl-kernel@lauterbach.com> writes:
Franz> Am Die, 28 Sep 1999 schrieb Franz Sirl:
>> BTW, I could reproduce the va_arg problem reported by Raja R
>> Harinath <harinath@cs.umn.edu> on gcc-bugs on Linux/PPC as
>> well.
Franz> Hi Richard, the appended small patch fixes it on Linux/PPC.
This patch is correct. It's something of a design defect, in my
opinion, that the tree-building routines do not automatically set
TREE_SIDE_EFFECTS for VA_ARG_EXPR, RTL_EXPR, INIT_EXPR, MODIFY_EXPR,
and a few other such nodes. I've certainly more than once tracked
down the same bug you found, whereby an expression is not expanded by
expand_expr because it is not marked as having side-effects.
Does anyone disagree? Otherwise, I'll put together a patch for
`make_node' that fixes this.
--
Mark Mitchell mark@codesourcery.com
CodeSourcery, LLC http://www.codesourcery.com