From owner-freebsd-toolchain@freebsd.org Wed Oct 14 07:34:42 2015 Return-Path: Delivered-To: freebsd-toolchain@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 21891A12578 for ; Wed, 14 Oct 2015 07:34:42 +0000 (UTC) (envelope-from markmi@dsl-only.net) Received: from asp.reflexion.net (outbound-mail-210-2.reflexion.net [208.70.210.2]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id D17581604 for ; Wed, 14 Oct 2015 07:34:41 +0000 (UTC) (envelope-from markmi@dsl-only.net) Received: (qmail 9070 invoked from network); 14 Oct 2015 07:34:38 -0000 Received: from unknown (HELO mail-cs-01.app.dca.reflexion.local) (10.81.19.1) by 0 (rfx-qmail) with SMTP; 14 Oct 2015 07:34:38 -0000 Received: by mail-cs-01.app.dca.reflexion.local (Reflexion email security v7.70.2) with SMTP; Wed, 14 Oct 2015 03:34:38 -0400 (EDT) Received: (qmail 8215 invoked from network); 14 Oct 2015 07:34:38 -0000 Received: from unknown (HELO iron2.pdx.net) (69.64.224.71) by 0 (rfx-qmail) with (AES256-SHA encrypted) SMTP; 14 Oct 2015 07:34:38 -0000 X-No-Relay: not in my network X-No-Relay: not in my network X-No-Relay: not in my network X-No-Relay: not in my network Received: from [192.168.1.106] (c-67-170-167-181.hsd1.or.comcast.net [67.170.167.181]) by iron2.pdx.net (Postfix) with ESMTPSA id 3C4221C43BC; Wed, 14 Oct 2015 00:34:37 -0700 (PDT) Content-Type: text/plain; charset=utf-8 Mime-Version: 1.0 (Mac OS X Mail 8.2 \(2104\)) Subject: Re: powerpc64-xtoolchain-gcc/powerpc64-gcc gets libcxxrt/guard.cc:104:15: error: expected constructor, destructor, . . . for lib32 (clang 3.6.1 context) From: Mark Millard In-Reply-To: <15CF1760-66B9-477B-B382-7F0C41D7C699@dsl-only.net> Date: Wed, 14 Oct 2015 00:34:36 -0700 Cc: FreeBSD PowerPC ML , FreeBSD Toolchain Content-Transfer-Encoding: quoted-printable Message-Id: <00A78528-C61E-494C-B471-F2D86759D4E9@dsl-only.net> References: <15CF1760-66B9-477B-B382-7F0C41D7C699@dsl-only.net> To: David Chisnall , Baptiste Daroussin X-Mailer: Apple Mail (2.2104) X-BeenThere: freebsd-toolchain@freebsd.org X-Mailman-Version: 2.1.20 Precedence: list List-Id: Maintenance of FreeBSD's integrated toolchain List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 14 Oct 2015 07:34:42 -0000 David Chisnall wrote (Mon Oct 12 08:21:28 UTC 2015): > On 12 Oct 2015, at 03:17, Mark Millard wrote: > >=20 > > /usr/src/lib/libcxxrt/../../contrib/libcxxrt/guard.cc:104:15: error: = expected constructor, destructor, or type conversion before '(' token > > _Static_assert(sizeof(guard_t) =3D=3D sizeof(uint64_t), ""); >=20 > I think that this is the error that you get with old gcc when a static = assert fails. What is the type of guard_t on PPC64? It may simply be = that gcc isn=E2=80=99t providing the correct #defines to let the file = know that it=E2=80=99s a 64-bit target. >=20 > David and I had replied about _Static_assert not being a C++ keyword and not = being created for C++ by FreeBSD=E2=80=99s current. But C++11's = static_assert "may appear at block scope (as a block declration) and = inside a class body (as a member declaration" (cppreference.com). It is = not valid at file scope. And unless the compile is explicitly or = implicitly specified as a c++11 or later style of compile static_assert = is then not part of the C++ vintage in use. /usr/local/bin/powerpc64-portbld-freebsd11.0-g++ -m32 -mcpu=3Dpowerpc = -DCOMPAT_32BIT -isystem = /usr/obj/powerpc.powerpc64/usr/src/lib32/usr/include/ = -L/usr/obj/powerpc.powerpc64/usr/src/lib32/usr/lib32 = -B/usr/obj/powerpc.powerpc64/usr/src/lib32/usr/lib32 = --sysroot=3D/usr/obj/powerpc.powerpc64/usr/src/tmp -O2 -pipe = -I/usr/src/lib/libcxxrt/../../contrib/libcxxrt -fstack-protector -c = /usr/src/lib/libcxxrt/../../contrib/libcxxrt/guard.cc -o guard.o does not specify -std=3Dc++11 or -std=3Dgnu++11 or the like so = static_assert is not enabled as part of the language vintage used. So = moving into block scope would not be sufficient of itself. So I tried simply removing the _Static_assert line that was in guard.cc: Index: /usr/src/contrib/libcxxrt/guard.cc =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- /usr/src/contrib/libcxxrt/guard.cc (revision 288459) +++ /usr/src/contrib/libcxxrt/guard.cc (working copy) @@ -101,7 +101,6 @@ uint32_t init_half; uint32_t lock_half; } guard_t; -_Static_assert(sizeof(guard_t) =3D=3D sizeof(uint64_t), ""); static const uint32_t LOCKED =3D 1; static const uint32_t INITIALISED =3D static_cast(1) << = 24; =20 =09 This allowed the guard.cc compile in question to complete and more. = (I=E2=80=99m not claiming FreeBSD should skip having a check for a size = mismatch someplace.) But eventually the build eventually got: --- lib/libcrypt__L --- CC=3D'/usr/local/bin/powerpc64-portbld-freebsd11.0-gcc -m32 = -mcpu=3Dpowerpc -DCOMPAT_32BIT -isystem = /usr/obj/powerpc.powerpc64/usr/src/lib32/usr/include/ = -L/usr/obj/powerpc.powerpc64/usr/src/lib32/usr/lib32 = -B/usr/obj/powerpc.powerpc64/usr/src/lib32/usr/lib32 = --sysroot=3D/usr/obj/powerpc.powerpc64/usr/src/tmp ' mkdep -f .depend -a = -I/usr/src/lib/libcrypt/../libmd -I/usr/src/lib/libcrypt/../libutil = -I/usr/src/lib/libcrypt -DHAS_DES -DHAS_BLOWFISH -DMD4Init=3D__MD4Init = -DMD4Final=3D__MD4Final -DMD4Update=3D__MD4Update -DMD4Pad=3D__MD4Pad = -DMD5Init=3D__MD5Init -DMD5Final=3D__MD5Final -DMD5Update=3D__MD5Update = -DMD5Pad=3D__MD5Pad -DSHA256_Init=3D__SHA256_Init = -DSHA256_Final=3D__SHA256_Final -DSHA256_Update=3D__SHA256_Update = -DSHA512_Init=3D__SHA512_Init -DSHA512_Final=3D__SHA512_Final = -DSHA512_Update=3D__SHA512_Update -std=3Dgnu99 = /usr/src/lib/libcrypt/crypt.c /usr/src/lib/libcrypt/misc.c = /usr/src/lib/libcrypt/crypt-md5.c /usr/src/lib/libcrypt/../libmd/md5c.c = /usr/src/lib/libcrypt/crypt-nthash.c = /usr/src/lib/libcrypt/../libmd/md4c.c = /usr/src/lib/libcrypt/crypt-sha256.c = /usr/src/lib/libcrypt/../../sys/crypto/sha2/sha256c.c = /usr/src/lib/libcrypt/crypt-sha512.c = /usr/src/lib/libcrypt/../libmd/sha512c.c = /usr/src/lib/libcrypt/../../secure/lib/libcrypt/crypt-des.c = /usr/src/lib/libcrypt/../../secure/lib/libcrypt/crypt-blowfish.c = /usr/src/lib/libcrypt/../../secure/lib/libcrypt/blowfish.c . . . --- lib/libcrypt__L --- /usr/src/lib/libcrypt/crypt-sha256.c:38:20: fatal error: sha256.h: No = such file or directory . . . --- lib/libcrypt__L --- mkdep: compile failed *** [.depend] Error code 1 make[4]: stopped in /usr/src/lib/libcrypt 1 error Note. . . root@FreeBSDx64:/usr/src # find / -name sha256.h -print | more /usr/obj/powerpc.powerpc64/usr/src/tmp/legacy/usr/include/sha256.h /usr/obj/powerpc.powerpc64/usr/src/tmp/usr/include/sha256.h /usr/obj/usr/src/tmp/legacy/usr/include/sha256.h /usr/obj/usr/src/tmp/usr/include/sha256.h /usr/include/sha256.h /usr/src/contrib/wpa/src/crypto/sha256.h /usr/src/sys/crypto/sha2/sha256.h and the lack of a = /usr/obj/powerpc.powerpc64/usr/src/lib32/usr/include/sha256.h for the = -isystem to find. To be explicit about no amd64 to powerpc64 cross compling gcc being = present for the build to execute . . . For an amd64 11.0 context doing root@FreeBSDx64:/usr/src # yes|make -j 6 CROSS_TOOLCHAIN=3Dpowerpc64-gcc = \ -DWITHOUT_GCC -DWITHOUT_GCC_BOOTSTRAP -DWITHOUT_GNUCXX \ -DWITHOUT_CLANG -DWITHOUT_CLANG_BOOTSTRAP -DWITHOUT_CLANG_FULL = -DWITHOUT_CLANG_IS_CC \ -DWITHOUT_CLANG_EXTRAS -DWITHOUT_LLDB \ -DWITHOUT_BOOT -DWITHOUT_LIB32 \ delete-old KERNCONF=3DGENERIC64-NODBG \ TARGET=3Dpowerpc TARGET_ARCH=3Dpowerpc64 . . . root@FreeBSDx64:/usr/src # make -j 6 CROSS_TOOLCHAIN=3Dpowerpc64-gcc \ -DWITHOUT_GCC -DWITHOUT_GCC_BOOTSTRAP -DWITHOUT_GNUCXX \ -DWITHOUT_CLANG -DWITHOUT_CLANG_BOOTSTRAP -DWITHOUT_CLANG_FULL = -DWITHOUT_CLANG_IS_CC \ -DWITHOUT_CLANG_EXTRAS -DWITHOUT_LLDB \ -DWITHOUT_BOOT -DWITH_LIB32 \ buildworld buildkernel KERNCONF=3DGENERIC64-NODBG \ TARGET=3Dpowerpc TARGET_ARCH=3Dpowerpc64 The context details follow. Mac OS X 10.10.5 using a VirtualBox V5.0.4 VM with. . . root@FreeBSDx64:/usr/src # freebsd-version -ku; uname -aKU 11.0-CURRENT 11.0-CURRENT FreeBSD FreeBSDx64 11.0-CURRENT FreeBSD 11.0-CURRENT #0 r288459: Sat Oct = 3 03:42:08 PDT 2015 = root@FreeBSDx64:/usr/obj/usr/src/sys/GENERIC-NODEBUG amd64 1100079 = 1100079 root@FreeBSDx64:/usr/src # svnlite info /usr/src Path: . Working Copy Root Path: /usr/src URL: https://svn.freebsd.org/base/head Relative URL: ^/head Repository Root: https://svn.freebsd.org/base Repository UUID: ccf9f872-aa2e-dd11-9fc8-001c23d0bc1f Revision: 288459 Node Kind: directory Schedule: normal Last Changed Author: gjb Last Changed Rev: 288459 Last Changed Date: 2015-10-01 12:48:10 -0700 (Thu, 01 Oct 2015) root@FreeBSDx64:/usr/src # more /etc/src.conf # World: NO_WERROR=3D # Kernel: WERROR=3D root@FreeBSDx64:/usr/src # more /etc/make.conf /etc/make.conf: No such file or directory root@FreeBSDx64:/usr/src # svnlite info /usr/ports Path: /usr/ports Working Copy Root Path: /usr/ports URL: https://svn.freebsd.org/ports/head Relative URL: ^/head Repository Root: https://svn.freebsd.org/ports Repository UUID: 35697150-7ecd-e111-bb59-0022644237b5 Revision: 398559 Node Kind: directory Schedule: normal Last Changed Author: jbeich Last Changed Rev: 398559 Last Changed Date: 2015-10-03 17:42:55 -0700 (Sat, 03 Oct 2015) root@FreeBSDx64:/usr/src # portmaster --list-origins net/avahi devel/boost-all lang/clang37 ports-mgmt/dialog4ports lang/gcc5 devel/git-lite x11/lumina sysutils/lxterminal www/midori net/netatalk3 ports-mgmt/pkg ports-mgmt/portmaster devel/powerpc64-xtoolchain-gcc devel/qtcreator security/sudo mail/thunderbird emulators/virtualbox-ose-additions Very little is non-default source code. . . root@FreeBSDx64:/usr/src # svnlite status M lib/csu/powerpc64/Makefile ? sys/arm/conf/RPI2-NODBG M sys/boot/powerpc/Makefile.inc ? sys/powerpc/conf/GENERIC64-NODBG root@FreeBSDx64:/usr/src # svnlite status /usr/ports ? /usr/ports/ion root@FreeBSDx64:/usr/src # more = /usr/src/sys/powerpc/conf/GENERIC64-NODBG=20 . . . (comments omitted) . . . include GENERIC64 ident GENERIC-NODBG nooptions INVARIANTS nooptions INVARIANT_SUPPORT nooptions WITNESS nooptions WITNESS_SKIPSPIN nooptions DEADLKRES (The copy->paste sequence for the below diffs likely turned tabs into = spaces.) root@FreeBSDx64:/usr/src # svnlite diff sys/boot/powerpc/Makefile.inc=20 Index: sys/boot/powerpc/Makefile.inc =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- sys/boot/powerpc/Makefile.inc (revision 288459) +++ sys/boot/powerpc/Makefile.inc (working copy) @@ -2,7 +2,7 @@ .if ${MACHINE_ARCH} =3D=3D "powerpc64" CFLAGS+=3D -m32 -mcpu=3Dpowerpc -LDFLAGS+=3D -m elf32ppc_fbsd +LDFLAGS+=3D -Wl,-m -Wl,elf32ppc_fbsd -Wl,-M,--verbose .endif .include "../Makefile.inc" The following is my personal, local workaround for = /usr/src/lib/csu/powerpc64/Makefile forcing use of the gcc command even = when the context does not have one. root@FreeBSDx64:/usr/src # svnlite diff /usr/src Index: /usr/src/lib/csu/powerpc64/Makefile =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- /usr/src/lib/csu/powerpc64/Makefile (revision 288459) +++ /usr/src/lib/csu/powerpc64/Makefile (working copy) @@ -13,8 +13,8 @@ # clang doesn't support -mlongcall, and testing shows a clang linked = with a # clang-built csu segfaults, this must currently be compiled with gcc. = Once # clang supports -mlongcall, or we get a fixed ld, this can be = revisited. -CC:=3D gcc -COMPILER_TYPE:=3D gcc +#CC:=3D gcc +#COMPILER_TYPE:=3D gcc FILES=3D ${OBJS} FILESMODE=3D ${LIBMODE} =3D=3D=3D Mark Millard markmi@dsl-only.net