From owner-freebsd-chromium@freebsd.org Sat Nov 19 07:51:04 2016 Return-Path: Delivered-To: freebsd-chromium@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 4FAA1C4AA73 for ; Sat, 19 Nov 2016 07:51:04 +0000 (UTC) (envelope-from ori@eigenstate.org) Received: from mimir.eigenstate.org (mimir.eigenstate.org [206.124.132.107]) by mx1.freebsd.org (Postfix) with ESMTP id 3E83F794 for ; Sat, 19 Nov 2016 07:51:03 +0000 (UTC) (envelope-from ori@eigenstate.org) Received: from oneeye (unknown [98.207.157.125]) by mimir.eigenstate.org (Postfix) with ESMTPSA id AB690600F3 for ; Sat, 19 Nov 2016 02:43:03 -0500 (EST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=eigenstate.org; s=mail; t=1479541383; bh=yfL0/VSge+ql2z3tUrilc01fpfCysyJwxLdbejORpag=; h=Date:From:To:Subject; b=PWY6WREYzxsmC/uSBsmfsQihl2dAWgMhLKSUUhhlX2OcjDM9mP2kGRaiFrhjo5ZNm M+Zqj5XpfbgWJAMEAyaJpzgG/qmHJY5CIiTBFWAQPPsiE+32Yar5+Ti9ze9M/YfRIE P4QsepZVO8IkTjc9awjIYaE8t7sEeew16DqmXJ3E= Date: Fri, 18 Nov 2016 23:43:02 -0800 From: Ori Bernstein To: freebsd-chromium@freebsd.org Subject: Chromium Build Failures Message-Id: <20161118234302.c0fac4f3508da0ce40d4ea9a@eigenstate.org> X-Mailer: Sylpheed 3.5.1 (GTK+ 2.24.29; amd64-portbld-freebsd11.0) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-BeenThere: freebsd-chromium@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: FreeBSD-specific Chromium issues List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 19 Nov 2016 07:51:04 -0000 So, I'm trying to build the most recent Chromium port in order to attempt debugging some issues that I'm seeing, and I'm currently getting a few build failures: [179/16428] python "/usr/home/ori/src/ports/freebsd-chromium/www/chromium/work/chromium-54.0.2840.100/build/toolchain/gcc_solink_wrapper.py" --readelf="readelf" --nm="llvm-nm39" --sofile="./libbase.so" --tocfile="./libbase.so.TOC" --output="./libbase.so" -- clang++39 -shared -Wl,--fatal-warnings -fPIC -Wl,-z,noexecstack -Wl,-z,now -Wl,-z,relro -Wl,-z,defs -pthread -m64 -Wl,--export-dynamic -L/usr/local/lib -o "./libbase.so" -Wl,-soname="libbase.so" @"./libbase.so.rsp" FAILED: libbase.so libbase.so.TOC python "/usr/home/ori/src/ports/freebsd-chromium/www/chromium/work/chromium-54.0.2840.100/build/toolchain/gcc_solink_wrapper.py" --readelf="readelf" --nm="llvm-nm39" --sofile="./libbase.so" --tocfile="./libbase.so.TOC" --output="./libbase.so" -- clang++39 -shared -Wl,--fatal-warnings -fPIC -Wl,-z,noexecstack -Wl,-z,now -Wl,-z,relro -Wl,-z,defs -pthread -m64 -Wl,--export-dynamic -L/usr/local/lib -o "./libbase.so" -Wl,-soname="libbase.so" @"./libbase.so.rsp" obj/base/base/launch_posix.o: In function `base::(anonymous namespace)::GetEnvironment()': ../../base/process/launch_posix.cc:(.text+0x1fd7): undefined reference to `environ' obj/base/base/launch_posix.o: In function `base::(anonymous namespace)::SetEnvironment(char**)': ../../base/process/launch_posix.cc:(.text+0x20e7): undefined reference to `environ' Working around it for the sake of the build by removing the extern on environ (even though I'm almost entirely certian that it won't work correctly in the final version, I get the missing '-lexecinfo' that I initially ran into in bug 214450: obj/third_party/WebKit/Source/wtf/wtf/Assertions.o: In function `WTFGetBacktrace(void**, int*)': ./out/Debug/../../third_party/WebKit/Source/wtf/Assertions.cpp:167: undefined reference to `backtrace' clang-3.9: error: linker command failed with exit code 1 (use -v to see invocation) I'm runnign on $ uname -a FreeBSD oneeye 11.0-RELEASE-p1 FreeBSD 11.0-RELEASE-p1 #0 r306420: Thu Sep 29 01:43:23 UTC 2016 root@releng2.nyi.freebsd.org:/usr/obj/usr/src/sys/GENERIC amd64 Is there anything obvious I'm missing? For the environ, it seems like we may be running into the same issue as this: https://git.merproject.org/mer-core/qtbase/commit/ccf74b592809e0c5a613eff27d6431a4c659e368 Buy I don't see anything in either linker flags or the wrapper scripts that would lead to -Wl,-no-undefined getting passed down here. Does lld behave differently than gnu ld? I noticed that we seem to be using it now, no idea how new that is. I've definitely built older versions successfully. -- Ori Bernstein