From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 17269 invoked by alias); 11 May 2013 03:33:04 -0000 Mailing-List: contact crossgcc-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: crossgcc-owner@sourceware.org Received: (qmail 17189 invoked by uid 89); 11 May 2013 03:32:58 -0000 X-Spam-SWARE-Status: No, score=-2.9 required=5.0 tests=AWL,BAYES_00,FREEMAIL_FROM,KHOP_THREADED,RCVD_IN_DNSWL_LOW,RCVD_IN_HOSTKARMA_YE,SPF_PASS autolearn=ham version=3.3.1 Received: from mail-la0-f46.google.com (HELO mail-la0-f46.google.com) (209.85.215.46) by sourceware.org (qpsmtpd/0.84/v0.84-167-ge50287c) with ESMTP; Sat, 11 May 2013 03:32:57 +0000 Received: by mail-la0-f46.google.com with SMTP id fk20so4518669lab.5 for ; Fri, 10 May 2013 20:32:55 -0700 (PDT) MIME-Version: 1.0 X-Received: by 10.152.21.225 with SMTP id y1mr3954644lae.28.1368243175468; Fri, 10 May 2013 20:32:55 -0700 (PDT) Received: by 10.112.137.100 with HTTP; Fri, 10 May 2013 20:32:55 -0700 (PDT) In-Reply-To: References: <7D80292E-5B9B-4398-88EA-F3113805A673@cogs.com> Date: Sat, 11 May 2013 03:33:00 -0000 Message-ID: Subject: Re: Canadian build for AVR From: "jonsmirl@gmail.com" To: Ben Greenfield Cc: "crossgcc@sourceware.org" Content-Type: text/plain; charset=ISO-8859-1 X-SW-Source: 2013-05/txt/msg00026.txt.bz2 Doing the Canadian build on 32b x86 worked first try. Something is broken in gcc when doing a 64b -> 32b -> 8b build. 32b->32b->8b works. On Fri, May 10, 2013 at 10:33 AM, jonsmirl@gmail.com wrote: > On Fri, May 10, 2013 at 9:26 AM, Ben Greenfield wrote: >> Hey Jon, Hey all, >> >> I lurk on this list and don't quite follow your thread. My understanding was without 8bit avr support you would try something else. Then lo and behold you report progress then success. > > There isn't AVR8 support in crosstool-ng. But gcc supports it. I am > trying to put together my own makefiles to build it without > crosstool-ng support. > > I think the test_real_width issues comes from using 64b x86 as my host > and then building using a 32b compiler (mipsel). > > >> >> If you did get it working I would love an overall description of your environment and workflow. >> >> Thanks, >> >> Ben >> >> On May 9, 2013, at 7:12 PM, "jonsmirl@gmail.com" wrote: >> >>> Configured with --enable-long-long >>> Preprocessor output: >>> >>> struct real_value { >>> unsigned int cl : 2; >>> unsigned int decimal : 1; >>> unsigned int sign : 1; >>> unsigned int signalling : 1; >>> unsigned int canonical : 1; >>> unsigned int uexp : (32 - 6); >>> unsigned long sig[((128 + (8 * 4)) / (8 * 4))]; >>> }; >>> # 77 "../.././gcc/real.h" >>> extern char test_real_width >>> [sizeof(struct real_value) <= (((128 + (8 * 4)) + 32)/(8 * 8) + >>> (((128 + (8 * 4)) + 32)%(8 * 8) ? 1 : 0))*sizeof(long long) ? 1 : -1]; >>> >>> -- >>> Jon Smirl >>> jonsmirl@gmail.com >>> >>> -- >>> For unsubscribe information see http://sourceware.org/lists.html#faq >>> > > > > -- > Jon Smirl > jonsmirl@gmail.com -- Jon Smirl jonsmirl@gmail.com -- For unsubscribe information see http://sourceware.org/lists.html#faq