From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1946566AbbHHIGs (ORCPT ); Sat, 8 Aug 2015 04:06:48 -0400 Received: from mail-wi0-f175.google.com ([209.85.212.175]:38514 "EHLO mail-wi0-f175.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752877AbbHHIGl (ORCPT ); Sat, 8 Aug 2015 04:06:41 -0400 Date: Sat, 8 Aug 2015 10:06:36 +0200 From: Ingo Molnar To: Arnaldo Carvalho de Melo Cc: linux-kernel@vger.kernel.org, Alexei Starovoitov , Brendan Gregg , Daniel Borkmann , David Ahern , He Kuang , Jiri Olsa , Kaixu Xia , Masami Hiramatsu , Namhyung Kim , Peter Zijlstra , pi3orama@163.com, Wang Nan , Zefan Li , Arnaldo Carvalho de Melo Subject: Re: [GIT PULL 00/24] perf/ebpf lib + llvm/clang building infrastructure Message-ID: <20150808080635.GA5414@gmail.com> References: <1438959032-8637-1-git-send-email-acme@kernel.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1438959032-8637-1-git-send-email-acme@kernel.org> User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org * Arnaldo Carvalho de Melo wrote: > Hi Ingo, > > This is up to a test that checks that the infrastructure for getting > a C source file and build it to a form that can then be used to upload it to > the kernel is working. > > Wang has a tree where we go all the way to passing a .c file as an > --event, but I thought about breaking the patchkit up to what is settled and > the "perf test -v LLVM" point looks like a good cutoff point for demoing/testing > what is in this patchkit. Fair enough! > It shouldn't affect people without the environment needed, as well it > should check the environment for sanity and provide info about how to get it > working, be it in the changeset logs, the docs and on tool output messages.o > > If you find anything in it that needs fixing, please let us know if you > think we should fix it before merging it to this point or if something can be > fixed in follow up patches. > > Wang's repo, rebased, almost builds, he should be refreshing this soon, > fixing those trailing whitespace issues, that is stopping the rebase here: > > [acme@zoo linux]$ git remote -v | grep wang | head -1 > wangnan.github https://github.com/WangNan0/linux.git (fetch) > [acme@zoo linux]$ git checkout -b ebpf-test wangnan.github/ebpf > Branch ebpf-test set up to track remote branch ebpf from wangnan.github. > Switched to a new branch 'ebpf-test' > [acme@zoo linux]$ git rebase perf/ebpf > First, rewinding head to replay your work on top of it... > Applying: perf tests: Add LLVM test for eBPF on-the-fly compiling > Using index info to reconstruct a base tree... > M tools/perf/tests/Build > M tools/perf/tests/builtin-test.c > M tools/perf/tests/tests.h > M tools/perf/util/llvm-utils.c > M tools/perf/util/llvm-utils.h > Falling back to patching base and 3-way merge... > No changes -- Patch already applied. > Applying: perf tools: Make perf depend on libbpf > Using index info to reconstruct a base tree... > M tools/perf/MANIFEST > Falling back to patching base and 3-way merge... > Auto-merging tools/perf/MANIFEST > Applying: perf ebpf: Add the libbpf glue > Applying: perf tools: Enable passing bpf object file to --event > Applying: perf tools: Parse probe points of eBPF programs during preparation > Applying: perf probe: Attach trace_probe_event with perf_probe_event > Applying: perf record: Probe at kprobe points > Applying: perf record: Load all eBPF object into kernel > Applying: perf tools: Add bpf_fd field to evsel and config it > Applying: perf tools: Attach eBPF program to perf event > Applying: perf tools: Suppress probing messages when probing by BPF loading > Applying: perf record: Add clang options for compiling BPF scripts > Applying: perf tools: Infrastructure for compiling scriptlets when passing '.c' to --event > Applying: perf tests: Enforce LLVM test for BPF test > Applying: perf test: Enable 'perf test' run as test targets > tools/perf/tests/builtin-test.c:279: trailing whitespace. > + /* > [acme@zoo linux]$ > > After that whitespace thing is fixed, you'll get to a point where 'perf > test -v BPF' will show you all the steps involved in, from a .c file do all the > build process and loading, attaching to a kprobe, etc that we discussed, but > those extra patches need more working before I can submit to you. > > Otherwise, please consider pulling, > > - Arnaldo > > The following changes since commit 1354ac6ad84395660f551d0614a6ca39e5bfe8e3: > > Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core (2015-08-07 09:11:30 +0200) > > are available in the git repository at: > > git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.git tags/perf-ebpf-for-mingo > > for you to fetch changes up to 9bc898c7019383b6aa2ae6cb2928c4ca926449f0: > > perf tests: Add LLVM test for eBPF on-the-fly compiling (2015-08-07 10:57:24 -0300) > > ---------------------------------------------------------------- > perf/ebpf library + llvm/clang infrastructure > > Infrastructure: > > - library for interfacing with the kernel eBPF infrastructure, with > tools/perf/ targeted as a first user (Wang Nan) > > - llvm/clang infrastructure for building BPF object files from C source > code (Wang Nan) > > Signed-off-by: Arnaldo Carvalho de Melo > > ---------------------------------------------------------------- > Wang Nan (24): > bpf tools: Introduce 'bpf' library and add bpf feature check > bpf tools: Allow caller to set printing function > bpf tools: Open eBPF object file and do basic validation > bpf tools: Read eBPF object from buffer > bpf tools: Check endianness and make libbpf fail early > bpf tools: Iterate over ELF sections to collect information > bpf tools: Collect version and license from ELF sections > bpf tools: Collect map definitions from 'maps' section > bpf tools: Collect symbol table from SHT_SYMTAB section > bpf tools: Collect eBPF programs from their own sections > bpf tools: Collect relocation sections from SHT_REL sections > bpf tools: Record map accessing instructions for each program > bpf tools: Add bpf.c/h for common bpf operations > bpf tools: Create eBPF maps defined in an object file > bpf tools: Relocate eBPF programs > bpf tools: Introduce bpf_load_program() to bpf.c > bpf tools: Load eBPF programs in object files into kernel > bpf tools: Introduce accessors for struct bpf_program > bpf tools: Link all bpf objects onto a list > perf tools: Introduce llvm config options > perf tools: Call clang to compile C source to object code > perf tools: Auto detecting kernel build directory > perf tools: Auto detecting kernel include options > perf tests: Add LLVM test for eBPF on-the-fly compiling > > tools/build/feature/Makefile | 6 +- > tools/build/feature/test-bpf.c | 18 + > tools/lib/bpf/.gitignore | 2 + > tools/lib/bpf/Build | 1 + > tools/lib/bpf/Makefile | 195 ++++++++ > tools/lib/bpf/bpf.c | 85 ++++ > tools/lib/bpf/bpf.h | 23 + > tools/lib/bpf/libbpf.c | 1037 +++++++++++++++++++++++++++++++++++++++ > tools/lib/bpf/libbpf.h | 81 +++ > tools/perf/MANIFEST | 1 + > tools/perf/tests/Build | 1 + > tools/perf/tests/builtin-test.c | 4 + > tools/perf/tests/llvm.c | 98 ++++ > tools/perf/tests/tests.h | 1 + > tools/perf/util/Build | 1 + > tools/perf/util/config.c | 4 + > tools/perf/util/llvm-utils.c | 408 +++++++++++++++ > tools/perf/util/llvm-utils.h | 49 ++ > 18 files changed, 2014 insertions(+), 1 deletion(-) > create mode 100644 tools/build/feature/test-bpf.c > create mode 100644 tools/lib/bpf/.gitignore > create mode 100644 tools/lib/bpf/Build > create mode 100644 tools/lib/bpf/Makefile > create mode 100644 tools/lib/bpf/bpf.c > create mode 100644 tools/lib/bpf/bpf.h > create mode 100644 tools/lib/bpf/libbpf.c > create mode 100644 tools/lib/bpf/libbpf.h > create mode 100644 tools/perf/tests/llvm.c > create mode 100644 tools/perf/util/llvm-utils.c > create mode 100644 tools/perf/util/llvm-utils.h Pulled, thanks a lot Arnaldo! Ingo