From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 28728 invoked by alias); 12 Jun 2013 21:10:08 -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 28622 invoked by uid 48); 12 Jun 2013 21:10:03 -0000 From: "jistone at redhat dot com" To: systemtap@sourceware.org Subject: [Bug dyninst/15614] New: stapdyn only partially fills registers on ppc Date: Wed, 12 Jun 2013 21:10:00 -0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: new X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: systemtap X-Bugzilla-Component: dyninst X-Bugzilla-Version: unspecified X-Bugzilla-Keywords: X-Bugzilla-Severity: normal X-Bugzilla-Who: jistone at redhat dot com X-Bugzilla-Status: NEW X-Bugzilla-Priority: P2 X-Bugzilla-Assigned-To: systemtap at sourceware dot org X-Bugzilla-Target-Milestone: --- X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: bug_id short_desc product version bug_status bug_severity priority component assigned_to reporter blocked Message-ID: 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: 2013-q2/txt/msg00275.txt.bz2 http://sourceware.org/bugzilla/show_bug.cgi?id=15614 Bug ID: 15614 Summary: stapdyn only partially fills registers on ppc Product: systemtap Version: unspecified Status: NEW Severity: normal Priority: P2 Component: dyninst Assignee: systemtap at sourceware dot org Reporter: jistone at redhat dot com Blocks: 14573 In Dyninst's EmitterPOWER::emitCall() on powerpc, they enforce a limit of 8 arguments in synthesized calls. This limits our current strategy of passing all DWARF registers as a series of varargs. We already use one argument for the probe index, and one for the number of arguments following, which only leaves us space to pass 6 registers. That will be the ip and r1-r5. To do more, we'll either have to extend Dyninst itself (which has a comment stating that additional ppc arguments could go on the stack), or find a way to build a full pt_regs and pass that as one pointer argument. The latter might be possible to do manually, but even nicer would be native Dyninst support to expose its spill area, also desired for bug 15386. -- You are receiving this mail because: You are the assignee for the bug.