From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 104426 invoked by alias); 24 Aug 2015 12:38:07 -0000 Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Archive: List-Post: List-Help: Sender: gcc-patches-owner@gcc.gnu.org Received: (qmail 103784 invoked by uid 89); 24 Aug 2015 12:38:06 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-0.2 required=5.0 tests=AWL,BAYES_00,FREEMAIL_FROM,RCVD_IN_DNSWL_LOW,SPF_PASS,URIBL_BLACK autolearn=no version=3.3.2 X-Spam-User: qpsmtpd, 2 recipients X-HELO: mail-wi0-f180.google.com Received: from mail-wi0-f180.google.com (HELO mail-wi0-f180.google.com) (209.85.212.180) by sourceware.org (qpsmtpd/0.93/v0.84-503-g423c35a) with (AES128-GCM-SHA256 encrypted) ESMTPS; Mon, 24 Aug 2015 12:38:04 +0000 Received: by wicja10 with SMTP id ja10so70988846wic.1; Mon, 24 Aug 2015 05:38:01 -0700 (PDT) X-Received: by 10.180.107.202 with SMTP id he10mr8615044wib.35.1440419881379; Mon, 24 Aug 2015 05:38:01 -0700 (PDT) Received: from [192.168.100.6] (chp127.enscp.fr. [193.51.253.127]) by smtp.gmail.com with ESMTPSA id p5sm5660931wiy.17.2015.08.24.05.37.59 (version=TLSv1 cipher=ECDHE-RSA-RC4-SHA bits=128/128); Mon, 24 Aug 2015 05:37:59 -0700 (PDT) Content-Type: text/plain; charset=utf-8 Mime-Version: 1.0 (Mac OS X Mail 9.0 \(3083\)) Subject: Re: Build break on SPU (and other non-mmap systems?) (Re: [patch,libgfortran,toplevel] Use libbacktrace in libgfortran) From: FX In-Reply-To: <20150824121739.42DC0CD56@oc7340732750.ibm.com> Date: Mon, 24 Aug 2015 12:41:00 -0000 Cc: iant@google.com, gfortran , gcc-patches Content-Transfer-Encoding: quoted-printable Message-Id: <5E9275E2-1488-45B2-8074-3D92A5320DF3@gmail.com> References: <20150824121739.42DC0CD56@oc7340732750.ibm.com> To: Ulrich Weigand X-SW-Source: 2015-08/txt/msg01407.txt.bz2 > On the SPU, mmap is indeed not available. Before that change, it was sti= ll > possible to build libgfortran for SPU ... >=20 > Is the intent to now require mmap on every system that supports libgfortr= an, > or should use of mmap in libbacktrace be optional in some form? The idea is that libbacktrace should be able to build on all targets, at le= ast in a minimally-functional form: https://gcc.gnu.org/ml/gcc/2015-08/msg0= 0038.html libgfortran should continue to build on all targets, even those that don=E2= =80=99t have an mmap(). I see that there is a test in libbacktrace/configure.ac for mmap(), but app= arently it fails on your system: -------------------------------- AC_CHECK_HEADERS(sys/mman.h) if test "$ac_cv_header_sys_mman_h" =3D "no"; then have_mmap=3Dno else if test -n "${with_target_subdir}"; then # When built as a GCC target library, we can't do a link test. We # simply assume that if we have mman.h, we have mmap. have_mmap=3Dyes else AC_CHECK_FUNC(mmap, [have_mmap=3Dyes], [have_mmap=3Dno]) fi fi -------------------------------- So this should be the place to patch things up. Relying on the presence of = is apparently not restrictive enough. Maybe we could use a AC_= EGREP_HEADER test, as is currently done for dl_interate_phdr()? FX