This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: Patches to fix GCCâs C++ exception handling on NetBSD/VAX

From: Jeff Law <law at redhat dot com>
To: Jake Hamby <jehamby420 at me dot com>, Christos Zoulas <christos at astron dot com>
Cc: port-vax at netbsd dot org, gcc-patches at gcc dot gnu dot org
Date: Tue, 26 Apr 2016 14:42:44 -0600
Subject: Re: Patches to fix GCCâs C++ exception handling on NetBSD/VAX
Authentication-results: sourceware.org; auth=none
References: <F48D0C6B-A6DB-410B-BC97-C30D4E8B4612 at me dot com> <ncu3pc$gkd$1 at ger dot gmane dot org> <DB0276A6-0A26-44F7-8FDC-D086EA40C709 at me dot com> <26D90386-F298-4EDD-A7D7-278F87DF3949 at me dot com>

On 03/26/2016 08:38 AM, Jake Hamby wrote:

Unfortunately, my previous patch that included a change to gcc/config/vax/vax.h that increased FIRST_PSEUDO_REGISTER from 16 to 17 breaks the C++ exception handling that Iâd worked so hard to get right with the rest of the patch. I believe I need to define DWARF_FRAME_REGISTERS to 16 in the same file to fix the size of the array that libgcc/unwind-dw2.c creates. The i386 backend and several others also define it their .h file for the same reason (compatibility with hardcoded frame offsets).

Hereâs the first part of the patch to vax.h that increases FIRST_PSEUDO_REGISTER and also adds a definition of DWARF_FRAME_REGISTERS as 16, with suitable comment. Iâm testing it now. I know that C++ exceptions were working before I increased FIRST_PSEUDO_REGISTER to 17.

So do you consider this patch ready to officially submit to GCC? Or isthere something else you're waiting on or dependent upon?


jeff

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]