This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] Fix ICE in fixup_var_refs_1 (regression against 2.95.x)
- To: Jakub Jelinek <jakub at redhat dot com>
- Subject: Re: [PATCH] Fix ICE in fixup_var_refs_1 (regression against 2.95.x)
- From: Richard Henderson <rth at redhat dot com>
- Date: Tue, 27 Feb 2001 19:27:23 -0800
- Cc: gcc-patches at gcc dot gnu dot org
- References: <20010227155253.T684@sunsite.ms.mff.cuni.cz>
On Tue, Feb 27, 2001 at 03:52:53PM +0100, Jakub Jelinek wrote:
> The testcase below ICEs in fixup_var_refs_1 on i386 at -O and above.
> The issue is that when fixup_var_refs_1 is changing one ADDRESSOF into REG
> in a cmpsi insn which had previously two different ADDRESSOFs,
> validate_replace_rtx fails (because recog does not know an ADDRESSOF
> can match nonimmediate_operand).
I don't think this is a good idea, as it doesn't really solve the
problem of trying to rerecognize an insn in the middle of processing
it. Any target that uses other predicates is still broken.
r~