From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
        id S1751243AbdAQTEo (ORCPT <rfc822;w@1wt.eu>);
        Tue, 17 Jan 2017 14:04:44 -0500
Received: from mail.kernel.org ([198.145.29.136]:51804 "EHLO mail.kernel.org"
        rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
        id S1750939AbdAQTEm (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
        Tue, 17 Jan 2017 14:04:42 -0500
Date: Tue, 17 Jan 2017 16:04:32 -0300
From: Arnaldo Carvalho de Melo <acme@kernel.org>
To: Alexei Starovoitov <alexei.starovoitov@gmail.com>
Cc: "Wangnan (F)" <wangnan0@huawei.com>,
        Arnaldo Carvalho de Melo <acme@redhat.com>,
        Alexei Starovoitov <ast@fb.com>, Zefan Li <lizefan@huawei.com>,
        He Kuang <hekuang@huawei.com>, Ingo Molnar <mingo@kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        pi3orama <pi3orama@163.com>, Joe Stringer <joe@ovn.org>,
        Jiri Olsa <jolsa@kernel.org>
Subject: Re: [PATCH v3 10/30] perf clang: Add builtin clang support ant test
 case
Message-ID: <20170117190432.GG14872@kernel.org>
References: <CAADnVQJnmU3xGBF6pVjp2OV_XK_hirpAa2yrqfqhCARZv6_tcA@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <CAADnVQJnmU3xGBF6pVjp2OV_XK_hirpAa2yrqfqhCARZv6_tcA@mail.gmail.com>
X-Url: http://acmel.wordpress.com
User-Agent: Mutt/1.7.1 (2016-10-04)
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

Em Tue, Jan 17, 2017 at 10:27:39AM -0800, Alexei Starovoitov escreveu:
> On Tue, Jan 17, 2017 at 5:38 AM, Arnaldo Carvalho de Melo
> <acme@kernel.org> wrote:
> > [acme@jouet linux]$ ls -lah /tmp/perf
> > -rwxr-xr-x. 1 acme acme 4.4M Jan 17 10:29 /tmp/perf
> > [acme@jouet linux]$ size /tmp/perf
> >    text    data     bss     dec     hex filename
> > 3954488  622440 23912104        28489032        1b2b548 /tmp/perf
> > [acme@jouet linux]$ ldd /tmp/perf | egrep -i llvm\|clang
> >         libclangAST.so.40 => /usr/local/lib/libclangAST.so.40 (0x00007f4964981000)
> ...
> >         libLLVMSystemZDisassembler.so.40 => /usr/local/lib/libLLVMSystemZDisassembler.so.40 (0x00007f495fe70000)
> ...
> >         libLLVMSparcDisassembler.so.40 => /usr/local/lib/libLLVMSparcDisassembler.so.40 (0x00007f495f14f000)
> ...
> >         libLLVMPowerPCDisassembler.so.40 => /usr/local/lib/libLLVMPowerPCDisassembler.so.40 (0x00007f495de94000)
> ...
> >         libLLVMMSP430CodeGen.so.40 => /usr/local/lib/libLLVMMSP430CodeGen.so.40 (0x00007f495c79d000)
> ...
> >         libLLVMMipsDisassembler.so.40 => /usr/local/lib/libLLVMMipsDisassembler.so.40 (0x00007f495bf73000)
 
> I think the static binary shouldn't be including all llvm backends.
> It's not clear to me why .so pulls them all.
> fyi I typically configure llvm with:
> -DLLVM_TARGETS_TO_BUILD="X86;BPF"

Humm, so this is a limitation of the LLVM libraries? I.e. people linking
against it should ask for a subset of backends so that would be dlopened
out of a larger set builtin, say, in a generic distro shipped llvm/clang
package?

- Arnaldo