This is the mail archive of the
systemtap@sourceware.org
mailing list for the systemtap project.
[Bug runtime/9937] verify user-space build-ids
- From: "fche at redhat dot com" <sourceware-bugzilla at sourceware dot org>
- To: systemtap at sources dot redhat dot com
- Date: Sat, 23 Oct 2010 15:29:57 +0000
- Subject: [Bug runtime/9937] verify user-space build-ids
- Auto-submitted: auto-generated
- References: <bug-9937-1110@http.sourceware.org/bugzilla/>
http://sourceware.org/bugzilla/show_bug.cgi?id=9937
Frank Ch. Eigler <fche at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
AssignedTo|systemtap at sources dot |scox at redhat dot com
|redhat.com |
--- Comment #19 from Frank Ch. Eigler <fche at redhat dot com> 2010-10-23 15:29:39 UTC ---
Looks like a good start.
A few problems:
The probe_kernel_* alternative is based on an autoconf macro
to determine the presence of the probe_kernel_* functions,
*not* whether we're trying to check a kernel- or user-space
build-id string.
I don't understand why the dwfl_module_getdwarf / dwarf_module_base
stuff is at all relevant here. The build-id is not debuginfo
based. It's as if you're adding then subtracting the same dwbias
value, to no effect at all.
Instead, the code should reuse/extend the explicit
dwfl_module_relocation_info* calculations as used for
kernel modules up near line 4895, which I see it's trying
to do (though the comments need to be expanded to explain).
It needs to respect/store the secname instead of just bypassing
the .note.gnu.build-id assertion (which is correct only for
kernel modules).
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.