From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:51158) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1d23I6-0007lP-1Z for guix-patches@gnu.org; Sat, 22 Apr 2017 18:14:07 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1d23I2-0000Wz-TD for guix-patches@gnu.org; Sat, 22 Apr 2017 18:14:06 -0400 Received: from debbugs.gnu.org ([208.118.235.43]:36630) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1d23I2-0000Wt-PB for guix-patches@gnu.org; Sat, 22 Apr 2017 18:14:02 -0400 Received: from Debian-debbugs by debbugs.gnu.org with local (Exim 4.84_2) (envelope-from ) id 1d23I2-0006Hh-Hc for guix-patches@gnu.org; Sat, 22 Apr 2017 18:14:02 -0400 Subject: bug#26582: [PATCH] gnu: pcre: Fix Check for Hurd systems. Resent-Message-ID: From: ludo@gnu.org (Ludovic =?UTF-8?Q?Court=C3=A8s?=) References: <2ExjL3FnTH2ActBCbOgZRN@lTC1VtRxjDOEtGzP2P9M4> Date: Sun, 23 Apr 2017 00:13:15 +0200 In-Reply-To: <2ExjL3FnTH2ActBCbOgZRN@lTC1VtRxjDOEtGzP2P9M4> (rennes@openmailbox.org's message of "Fri, 21 Apr 2017 21:40:03 -0500") Message-ID: <87efwkjdl0.fsf@gnu.org> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: guix-patches-bounces+kyle=kyleam.com@gnu.org Sender: "Guix-patches" To: rennes Cc: 26582@debbugs.gnu.org Hi rennes, rennes skribis: > On 04/20/2017 04:27:29 PM, Ludovic Court=C3=A8s wrote: >> > >> > I moved zlib and readline to 'native-inputs'. >>=20 >> Are you sure they need to be native inputs? Both are libraries, so I >> imagine they=E2=80=99re used at run time, not at build time; they should= be >> inputs, in this case. >>=20 > I added system types: "--build=3D" and "--host=3D" to 'configure'. Now=20= =20 > detects correctly the inputs. > > Built successfully on Linux x86_64, GNU/Hurd and cross-compiling with=20= =20 > target 'i586-pc-gnu'. > Thanks for the guide > > From 05f159e343173708b85c5e45b6cfe64e3939c68d Mon Sep 17 00:00:00 2001 > From: rennes > Date: Fri, 21 Apr 2017 21:22:08 -0500 > Subject: [PATCH] gnu: pcre: Fix Check for Hurd systems. > > * gnu/packages/pcre.scm (pcre)[arguments]: Add "--disable-stack-for-recur= sion".=09 The patch does more than that. :-) > + #:phases > + (modify-phases %standard-phases > + (replace 'configure > + (lambda* (#:key outputs build target #:allow-other-keys) > + (let* ((out (assoc-ref outputs "out")) > + (bin (assoc-ref outputs "bin")) > + (doc (assoc-ref outputs "doc"))) > + (setenv "CONFIG_SHELL" (which "bash")) > + (zero? > + (system* "./configure" > + (string-append "--prefix=3D" out) > + (string-append "--bindir=3D" bin) > + (string-append "--docdir=3D" doc) > + (string-append "--build=3D" build) > + (string-append "--host=3D" (or target bui= ld)) I would expect the =E2=80=98configure=E2=80=99 phase in (guix build gnu-bui= ld-system) to already pass those flags, no? (If that is the case we could keep #:configure-flags instead of replacing the =E2=80=98configure=E2=80=99 phas= e.) > + ;; On GNU/Hurd systems FAIL 'RunTest' wit= h 'segmentation > + ;; fault' in Check phase, to pass the tes= t we can disable > + ;; "--enable-jit" or set "--disable-stack= -for-recursion" > + ;; argument. > + ,@(if (hurd-triplet? (or (%current-system) > + ( %current-targe= t-system))) > + '("--disable-stack-for-recursion") > + '()))))))))) This part looks OK to me (even better if you have a pointer to an upstream bug report or discussion.) Thanks! Luodo=E2=80=99.