From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <systemtap-return-23702-listarch-systemtap=sources.redhat.com@sourceware.org>
Received: (qmail 5785 invoked by alias); 4 Aug 2015 11:46:28 -0000
Mailing-List: contact systemtap-help@sourceware.org; run by ezmlm
Precedence: bulk
List-Id: <systemtap.sourceware.org>
List-Subscribe: <mailto:systemtap-subscribe@sourceware.org>
List-Post: <mailto:systemtap@sourceware.org>
List-Help: <mailto:systemtap-help@sourceware.org>, <http://sourceware.org/lists.html#faqs>
Sender: systemtap-owner@sourceware.org
Received: (qmail 5294 invoked by uid 48); 4 Aug 2015 11:46:24 -0000
From: "mcermak at redhat dot com" <sourceware-bugzilla@sourceware.org>
To: systemtap@sourceware.org
Subject: [Bug tapsets/18769] [ppc64BE/--dyninst] unknown operator @__compat_task
Date: Tue, 04 Aug 2015 11:46:00 -0000
X-Bugzilla-Reason: AssignedTo
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: systemtap
X-Bugzilla-Component: tapsets
X-Bugzilla-Version: unspecified
X-Bugzilla-Keywords:
X-Bugzilla-Severity: normal
X-Bugzilla-Who: mcermak at redhat dot com
X-Bugzilla-Status: NEW
X-Bugzilla-Resolution:
X-Bugzilla-Priority: P2
X-Bugzilla-Assigned-To: systemtap at sourceware dot org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields:
Message-ID: <bug-18769-6586-MFEpiMOcRA@http.sourceware.org/bugzilla/>
In-Reply-To: <bug-18769-6586@http.sourceware.org/bugzilla/>
References: <bug-18769-6586@http.sourceware.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://sourceware.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2015-q3/txt/msg00065.txt.bz2

https://sourceware.org/bugzilla/show_bug.cgi?id=18769

--- Comment #2 from Martin Cermak <mcermak at redhat dot com> ---
The separation of /linux and /dyninst directories is apparently quite
intentional:

=======
$ grep -A 21 'add empty string as last element' main.cxx  | nl
     1    // add empty string as last element
     2    version_suffixes.push_back ("");

     3    // Add arch variants of every path, just before each
     4    const string& arch = s.architecture;
     5    for (unsigned i=0; i<version_suffixes.size(); i+=2)
     6      version_suffixes.insert(version_suffixes.begin() + i,
     7                              version_suffixes[i] + "/" + arch);

     8    // Add runtime variants of every path, before everything else
     9    string runtime_prefix;
    10    if (s.runtime_mode == systemtap_session::kernel_runtime)
    11      runtime_prefix = "/linux";
    12    else if (s.runtime_mode == systemtap_session::dyninst_runtime)
    13      runtime_prefix = "/dyninst";
    14    if (!runtime_prefix.empty())
    15      for (unsigned i=0; i<version_suffixes.size(); i+=2)
    16        version_suffixes.insert(version_suffixes.begin() + i/2,
    17                                runtime_prefix + version_suffixes[i]);

    18    // First, parse .stpm files on the include path. We need to have the
    19    // resulting macro definitions available for parsing library files,
$
=======

I tried to add /linux into version_suffixes, but many tapsets dedicated for
kernel runtime are unusable for dyninst runtime, and at the end of the day -
it's understandable.

So maybe moving the macro definition from syscalls.stpm one level up in the
directory structure to e.g. compat_task.stpm, where it would be consumed by
both runtimes, might be a way to go.

-- 
You are receiving this mail because:
You are the assignee for the bug.