Em Tue, Apr 30, 2019 at 04:31:14PM +0300, Thomas Backlund escreveu: > > Den 30-04-2019 kl. 16:06, skrev Song Liu: > > On Tue, Apr 30, 2019 at 12:55 AM Thomas Backlund <t...@mageia.org> wrote: > > > Den 30-04-2019 kl. 10:26, skrev Thomas Backlund: > > > > Building perf in 5.1-rc5/6/7 fails:
> > > > Build start: > > > > make -s -C tools/perf NO_PERF_READ_VDSO32=1 NO_PERF_READ_VDSOX32=1 > > > > WERROR=0 NO_LIBUNWIND=1 HAVE_CPLUS_DEMANGLE=1 NO_GTK2=1 NO_STRLCPY=1 > > > > NO_BIONIC=1 NO_JVMTI=1 prefix=/usr lib=lib64 all > > > > BUILD: Doing 'make -j32' parallel build > > > > HOSTCC fixdep.o > > > > HOSTLD fixdep-in.o > > > > LINK fixdep > > > > Warning: Kernel ABI header at 'tools/arch/x86/include/uapi/asm/vmx.h' > > > > differs from latest version at 'arch/x86/include/uapi/asm/vmx.h' > > > > diff -u tools/arch/x86/include/uapi/asm/vmx.h > > > > arch/x86/include/uapi/asm/vmx.h > > > > > > > > Auto-detecting system features: > > > > ... dwarf: [ on ] > > > > ... dwarf_getlocations: [ on ] > > > > ... glibc: [ on ] > > > > ... gtk2: [ on ] > > > > ... libaudit: [ on ] > > > > ... libbfd: [ on ] > > > > ... libelf: [ on ] > > > > ... libnuma: [ on ] > > > > ... numa_num_possible_cpus: [ on ] > > > > ... libperl: [ on ] > > > > ... libpython: [ on ] > > > > ... libslang: [ on ] > > > > ... libcrypto: [ on ] > > > > ... libunwind: [ on ] > > > > ... libdw-dwarf-unwind: [ on ] > > > > ... zlib: [ on ] > > > > ... lzma: [ on ] > > > > ... get_cpuid: [ on ] > > > > ... bpf: [ on ] > > > > ... libaio: [ on ] > > > > ... disassembler-four-args: [ OFF ] > > > > > > > > Makefile.config:473: No sys/sdt.h found, no SDT events are defined, > > > > please install systemtap-sdt-devel or systemtap-sdt-dev > > > > Makefile.config:853: No libbabeltrace found, disables 'perf data' CTF > > > > format support, please install > > > > libbabeltrace-dev[el]/libbabeltrace-ctf-dev > > > > > > > > > > > > And breaks with: > > > > > > > > > > > > CC ui/setup.o > > > > util/annotate.c: In function 'symbol__disassemble_bpf': > > > > util/annotate.c:1767:29: error: incompatible type for argument 1 of > > > > 'disassembler' > > > > disassemble = disassembler(bfdf); > > > > ^~~~ > > > > In file included from util/annotate.c:1689: > > > > /usr/include/dis-asm.h:325:63: note: expected 'enum bfd_architecture' > > > > but argument is of type 'bfd *' {aka 'struct bfd *'} > > > > extern disassembler_ftype disassembler (enum bfd_architecture arc, > > > > ~~~~~~~~~~~~~~~~~~~~~~^~~ > > > > util/annotate.c:1767:16: error: too few arguments to function > > > > 'disassembler' > > > > disassemble = disassembler(bfdf); > > > > ^~~~~~~~~~~~ > > > > In file included from util/annotate.c:1689: > > > > /usr/include/dis-asm.h:325:27: note: declared here > > > > extern disassembler_ftype disassembler (enum bfd_architecture arc, > > > > ^~~~~~~~~~~~ > > > > CC arch/x86/util/header.o > > > > CC arch/x86/util/tsc.o > > > > CC arch/x86/util/pmu.o > > > > mv: cannot stat 'util/.annotate.o.tmp': No such file or directory > > > > CC bench/futex-requeue.o > > > > CC arch/x86/util/kvm-stat.o > > > > make[4]: *** > > > > [/work/rpmbuild/BUILD/kernel-x86_64/linux-5.0/tools/build/Makefile.build:97: > > > > util/annotate.o] Error 1 > > > > make[4]: *** Waiting for unfinished jobs.... > > > > CC util/build-id.o > > > > > > > > > > > > > > > > > > And I forgot... > > > > > > Reverting: > > > From 6987561c9e86eace45f2dbb0c564964a63f4150a Mon Sep 17 00:00:00 2001 > > > From: Song Liu <songliubrav...@fb.com> > > > Date: Mon, 11 Mar 2019 22:30:48 -0700 > > > Subject: perf annotate: Enable annotation of BPF programs > > > > > > Makes it build again. > > > > > > -- > > > Thomas > > > > > Hi Thomas, > > > > Which system are you running this test on? I would like to repro it in a VM. > > > > Thanks, > > Song > > > Mageia Cauldron currently stabilizing to become Mageia 7 in ~1 month. > > > Basesystem is: > > binutils-2.32-5.mga7 > (includes all fixes from upstream binutils-2_32-branch) > > gcc-8.3.1-0.20190419.2.mga7 > > glibc-2.29-7.mga7 > (includes all fixes from upstream glibc release/2.29/master branch up to > 2019-04-15 for now) > > > kernel-desktop-5.1.0-0.rc7.1.mga7 > kernel-userspace-headers-5.1.0-0.rc7.1.mga7 Ok, so the steps are: 1) the feature test, the small C program that we try to build is: [acme@quaco perf]$ cat tools/build/feature/test-disassembler-four-args.c // SPDX-License-Identifier: GPL-2.0 #include <bfd.h> #include <dis-asm.h> int main(void) { bfd *abfd = bfd_openr(NULL, NULL); disassembler(bfd_get_arch(abfd), bfd_big_endian(abfd), bfd_get_mach(abfd), abfd); return 0; } [acme@quaco perf]$ And here in my fedora29 system it ends up producing the following file, when built with: $ make O=/tmp/build/perf -C tools/perf install-bin [acme@quaco perf]$ cat /tmp/build/perf/feature/test-disassembler-four-args.make.output [acme@quaco perf]$ [acme@quaco perf]$ file /tmp/build/perf/feature/test-disassembler-four-args.bin /tmp/build/perf/feature/test-disassembler-four-args.bin: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 3.2.0, BuildID[sha1]=c9bd83db766a620c5cb6d756b0cd6991527641ff, not stripped, too many notes (256) [acme@quaco perf]$ ldd /tmp/build/perf/feature/test-disassembler-four-args.bin linux-vdso.so.1 (0x00007ffebc5f9000) libz.so.1 => /lib64/libz.so.1 (0x00007fed2da04000) libdl.so.2 => /lib64/libdl.so.2 (0x00007fed2d9fe000) libc.so.6 => /lib64/libc.so.6 (0x00007fed2d838000) /lib64/ld-linux-x86-64.so.2 (0x00007fed2da3e000) [acme@quaco perf]$ Meaning it built properly, so in this sytem the disassembler() function has indeed four args, so we end up with: [acme@quaco perf]$ grep disassembler /tmp/build/perf/FEATURE-DUMP feature-disassembler-four-args=1 [acme@quaco perf]$ Can you check the output for /tmp/build/perf/feature/test-disassembler-four-args.make.output in your system? And also check what is the prototype for the disassembler() routine on mageia7? Here I have: [acme@quaco perf]$ rpm -q binutils binutils-2.31.1-25.fc29.x86_64 [acme@quaco perf]$ Perhaps binutils 2.32 changed that prototype again and instead of falling back to using just one arg we need to use some other number of args, or even a different type for the N args it now maybe have? - Arnaldo