Date: Thu, 20 Apr 2023 01:27:14 -0700 From: Mark Millard <marklmi@yahoo.com> To: freebsd-arm <freebsd-arm@freebsd.org>, FreeBSD Toolchain <freebsd-toolchain@freebsd.org> Subject: aarch64 buildworld via aarch64-gcc12 got multiple: "undefined reference to `__aarch64_ldadd4_sync'" Message-ID: <F33D8941-84C0-4D9A-B6BA-C6BDC43C7936@yahoo.com> References: <F33D8941-84C0-4D9A-B6BA-C6BDC43C7936.ref@yahoo.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Are both tier 1 architectures intended to be ARCH-gcc12 buildable? I decided to try an aarch64-gcc12 since the amd64-gcc12 builds of main shown on ci.freebsd.org are generally finishing these days. The log messages show the failure was reported for "building shared library libheimbase.so.11" : . . .=20 Building = /usr/obj/BUILDs/main-CA72-dbg-gccxtc/usr/main-src/arm64.aarch64/kerberos5/= lib/libheimbase/libheimbase.so.11.full Compiler executable checksum: d7199505f5f10188c5d5b1ba1449fd49 building shared library libheimbase.so.11 Compiler executable checksum: d7199505f5f10188c5d5b1ba1449fd49 Compiler executable checksum: d7199505f5f10188c5d5b1ba1449fd49 /usr/local/bin/aarch64-unknown-freebsd14.0-ld: heimbase.pico: in = function `_heim_create_type': /usr/main-src/crypto/heimdal/base/heimbase.c:281: undefined reference to = `__aarch64_ldadd4_sync' /usr/local/bin/aarch64-unknown-freebsd14.0-ld: heimbase.pico: in = function `heim_retain': /usr/main-src/crypto/heimdal/base/heimbase.c:97: undefined reference to = `__aarch64_ldadd4_sync' /usr/local/bin/aarch64-unknown-freebsd14.0-ld: heimbase.pico: in = function `heim_release': /usr/main-src/crypto/heimdal/base/heimbase.c:120: undefined reference to = `__aarch64_ldadd4_sync' collect2: error: ld returned 1 exit status . . . For reference: # uname -apKU FreeBSD CA72_4c8G_ZFS 14.0-CURRENT FreeBSD 14.0-CURRENT #93 = main-n262341-f369f10dd812-dirty: Wed Apr 19 16:14:49 PDT 2023 = root@CA72_4c8G_ZFS:/usr/obj/BUILDs/main-CA72-nodbg-clang/usr/main-src/arm6= 4.aarch64/sys/GENERIC-NODBG-CA72 arm64 aarch64 1400086 1400086 This test environment skipped the unpatched time frame for the import of openzfs and never had block_cloning enabled. It has previously done 9+ hour poudriere bulk builds without problems, run in a way that frequently has high load averages (compared to the hardware thread count). I doubt that the __aarch64_ldadd4_sync problem is a openzfs corruption issue. =3D=3D=3D Mark Millard marklmi at yahoo.com
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?F33D8941-84C0-4D9A-B6BA-C6BDC43C7936>