[Bug translator/12120] New: Apparent non-determinism in DWARF symbol resolution

[Bug translator/12120] New: Apparent non-determinism in DWARF symbol resolution

[asutherland at asutherland dot org]

http://sourceware.org/bugzilla/show_bug.cgi?id=12120

           Summary: Apparent non-determinism in DWARF symbol resolution
           Product: systemtap
           Version: unspecified
            Status: RESOLVED
          Severity: normal
          Priority: P2
         Component: translator
        AssignedTo: systemtap@sources.redhat.com
        ReportedBy: asutherland@asutherland.org


Andrew Sutherland <asutherland at asutherland dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE

As initially mentioned on my blog[1] post without useful substantiation,
sometimes my systemtap scripts will fail to translate.

1: http://www.visophyte.org/blog/?p=586

Here is a link to a bzipped version of the mammoth 280M libxul.so file (only
90M compressed!):
http://clicky.visophyte.org/scratch/libxul.so.bz2

Perhaps important version info:
$ rpmquery --info systemtap
Name        : systemtap                    Relocations: (not relocatable)
Version     : 1.3                               Vendor: Fedora Project
Release     : 2.fc13                        Build Date: Tue 24 Aug 2010
11:45:23 AM PDT
Install Date: Wed 15 Sep 2010 09:47:01 PM PDT      Build Host:
x86-18.phx2.fedoraproject.org
Group       : Development/System            Source RPM:
systemtap-1.3-2.fc13.src.rpm
Size        : 7917275                          License: GPLv2+
Signature   : RSA/SHA256, Wed 25 Aug 2010 03:07:38 PM PDT, Key ID
7edc6ad6e8e40fde

$ gcc --version
gcc (GCC) 4.4.4 20100630 (Red Hat 4.4.4-10)

$ rpmquery --info gcc
Name        : gcc                          Relocations: (not relocatable)
Version     : 4.4.4                             Vendor: Fedora Project
Release     : 10.fc13                       Build Date: Wed 30 Jun 2010
02:54:10 AM PDT
Install Date: Mon 19 Jul 2010 03:20:53 PM PDT      Build Host:
xb-01.phx2.fedoraproject.org
Group       : Development/Languages         Source RPM:
gcc-4.4.4-10.fc13.src.rpm
Size        : 19365082                         License: GPLv3+ and GPLv3+ with
exceptions and GPLv2+ with exceptions
Signature   : RSA/SHA256, Wed 30 Jun 2010 03:23:31 AM PDT, Key ID
7edc6ad6e8e40fde


Here is an example of the output with just the warnings/errors bits.  I will
attach a full output log including the earlier parts in case they prove
interesting:

WARNING: eliding unused function 'jsobj_class_name': identifier
'jsobj_class_name' at
/home/visbrero/rev_control/hg/comm-central/obj-thunderbird-debug/mozperfish.stp:1143:10
 source: function jsobj_class_name(jsobj) {
                  ^
WARNING: eliding unused function 'jsval_to_string': identifier
'jsval_to_string' at :1178:10
 source: function jsval_to_string(jsval) {
                  ^
semantic error: unable to find member 'regs' for JSContext (alternatives:
Error: JSContext isn't a struct/class/union): operator '->' at :1095:36
        source:   regs = @cast(cx, "JSContext", @2)->regs;
                                                   ^
semantic error: unable to find member 'mLengthAndFlags' for JSString
(alternatives: Error: JSString isn't a struct/class/union): operator '->' at
:1155:45
        source:   lenAndFlags = @cast(jsstr, "JSString", @2)->mLengthAndFlags;
                                                            ^
semantic error: unable to find member 'mChars' for JSString (alternatives:
Error: JSString isn't a struct/class/union): operator '->' at :1174:36
        source:   cp = @cast(jsstr, "JSString", @2)->mChars;
                                                   ^
semantic error: unable to find member 'fslots' for JSObject (alternatives:
Error: JSObject isn't a struct/class/union): operator '->' at :1243:36
        source:     @cast(jsfunobj, "JSObject", @2)->fslots[0]->data->asPtr);
                                                   ^
Pass 2: analyzed script: 41 probe(s), 84 function(s), 2 embed(s), 20 global(s)
using 577944virt/369712res/3960shr kb, in 5880usr/390sys/6295real ms.
Pass 2: analysis failed.  Try again with another '--vp 01' option.
Running rm -rf /tmp/staps7pWmL

--- Comment #1 from Andrew Sutherland <asutherland at asutherland dot org> 2010-10-14 03:53:14 UTC ---
Ugh, apparently this is a buggy bugzilla installation... when I first submitted
my attachment it couldn't determine the mime type and required me to go back
and resubmit.  So, I did.  And yet there is this bug.  Duping to the bug that
ended up with the attachment.

*** This bug has been marked as a duplicate of bug 12121 ***

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.