From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 121842 invoked by alias); 30 Jun 2015 20:22:04 -0000 Mailing-List: contact systemtap-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Post: List-Help: , Sender: systemtap-owner@sourceware.org Received: (qmail 121792 invoked by uid 48); 30 Jun 2015 20:22:00 -0000 From: "dsmith at redhat dot com" To: systemtap@sourceware.org Subject: [Bug tapsets/18597] long_arg() doesn't correctly handle negative values in 32-on-64 environment Date: Tue, 30 Jun 2015 20:22:00 -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: dsmith 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: In-Reply-To: References: 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-q2/txt/msg00227.txt.bz2 https://sourceware.org/bugzilla/show_bug.cgi?id=18597 --- Comment #10 from David Smith --- (In reply to David Smith from comment #8) > (In reply to Josh Stone from comment #7) > > (In reply to David Smith from comment #5) > > > What if we decided that if you call longlong_arg() on a 64-bit OS on a > > > 32-bit process you really *want* a 64-bit value from 1 register? In some > > > ways this makes sense and it matches our old behavior. My theory here is > > > that you know what you are doing and if you call longlong_arg() on a 32-bit > > > process you must be in the true 64-bit function at this point. > > > > That assumes you're only using this in the kernel. For a uprobe, > > longlong_arg() still needs to follow the 2-register ABI, no? That's when > > the probing_32bit_app() branch should be active. > > > > So yes, longlong_arg() in 64-bit kernel context should use 1 register, > > regardless of the task, but a 32-bit user context still needs to use 2 > > registers. > > We can certainly add that code back, but do we know of anyone using the > stp_arg()-based dwarfless parameters functions from uprobes? Do we test that? >From what I can tell, we don't test using dwarfless parameters from uprobes. -- You are receiving this mail because: You are the assignee for the bug.