From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail.windriver.com ([147.11.1.11]) by linuxtogo.org with esmtp (Exim 4.72) (envelope-from ) id 1TjhRd-0004Bf-Pq for openembedded-core@lists.openembedded.org; Sat, 15 Dec 2012 03:25:42 +0100 Received: from ALA-HCA.corp.ad.wrs.com (ala-hca.corp.ad.wrs.com [147.11.189.40]) by mail.windriver.com (8.14.5/8.14.3) with ESMTP id qBF2B38K010533 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL) for ; Fri, 14 Dec 2012 18:11:03 -0800 (PST) Received: from localhost.localdomain (172.25.34.64) by ALA-HCA.corp.ad.wrs.com (147.11.189.50) with Microsoft SMTP Server id 14.2.318.4; Fri, 14 Dec 2012 18:11:03 -0800 From: Peter Seebach To: Date: Fri, 14 Dec 2012 20:10:56 -0600 Message-ID: X-Mailer: git-send-email 1.7.0.4 MIME-Version: 1.0 Subject: [PATCH 0/1] pseudo 1.4.2: linkat() and bitrot cleanup X-BeenThere: openembedded-core@lists.openembedded.org X-Mailman-Version: 2.1.11 Precedence: list List-Id: Patches and discussions about the oe-core layer List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 15 Dec 2012 02:25:42 -0000 Content-Type: text/plain The major change here is the addition of linkat(), and the change of link() to be written in terms of it (for consistency with othe *at() calls). I checked this out on Darwin (which hasn't got *at() functions) and picked off a couple little bits of bitrot. The other change is that the WR build system turns out to have reasons to, in some cases, actually use a lib64 directory for things like the host sqlite3 library. So I cleaned up the computation of the sqlite library path; of course, this then broke oe-core on 64-bit hosts, but I'd anticipated that, and pseudo's configure can take an explicit full path to libsqlite3.a if it's not in a directory with the same name as $(LIB). Did that to the various 32-bit and 64-bit forms. Note that on 64-bit hosts, this means that 32-bit libpseudo (if it's being built) is being configured with the explicit path to our 64-bit libsqlite3.a; that's okay, because we only build libpseudo for the 32-bit case. Confirmed that, with NO32LIBS = 0, I get both 32-bit and 64-bit libpseudo. I did have build failures in perl-native when I was first testing this, but updating oe-core made them go away, so it was probably just my tree being pretty old before I started on the pull request. The tarball is already on yoctoproject.org (thanks, pidge!) and I have verified that the build still succeeds. The following changes since commit ad79360c1d992830d4f0e06a3bbf0622658c0540: Mark Hatle (1): populate_sdk_base: Add perl modules as needing to be relocated are available in the git repository at: git://git.yoctoproject.org/poky-contrib seebs/pseudo_1_4_2 http://git.yoctoproject.org/cgit.cgi/poky-contrib/log/?h=seebs/pseudo_1_4_2 Peter Seebach (1): pseudo_git.bb and friends: Update to 1.4.2. meta/recipes-devtools/pseudo/pseudo.inc | 9 +++++---- meta/recipes-devtools/pseudo/pseudo_1.4.1.bb | 8 -------- meta/recipes-devtools/pseudo/pseudo_1.4.2.bb | 8 ++++++++ meta/recipes-devtools/pseudo/pseudo_git.bb | 6 +++--- 4 files changed, 16 insertions(+), 15 deletions(-) delete mode 100644 meta/recipes-devtools/pseudo/pseudo_1.4.1.bb create mode 100644 meta/recipes-devtools/pseudo/pseudo_1.4.2.bb