From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 94134 invoked by alias); 18 Mar 2019 16:09:15 -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 94059 invoked by uid 48); 18 Mar 2019 16:09:11 -0000 From: "me at serhei dot io" To: systemtap@sourceware.org Subject: [Bug bpf/24358] 32-bit stapbpf: things need fixing re: (void *) use in libbpf and elsewhere Date: Mon, 18 Mar 2019 16:09:00 -0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: systemtap X-Bugzilla-Component: bpf X-Bugzilla-Version: unspecified X-Bugzilla-Keywords: X-Bugzilla-Severity: normal X-Bugzilla-Who: me at serhei dot io 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: short_desc Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: http://sourceware.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 X-SW-Source: 2019-q1/txt/msg00074.txt.bz2 https://sourceware.org/bugzilla/show_bug.cgi?id=3D24358 Serhei Makarov changed: What |Removed |Added ---------------------------------------------------------------------------- Summary|32-bit stapbpf: could not |32-bit stapbpf: things need |find _stext in |fixing re: (void *) use in |/proc/kallsyms |libbpf and elsewhere --- Comment #1 from Serhei Makarov --- In principle, we want to add 32-bit support. (Currently there is no 32-bit Fedora bcc package either.) The kallsyms parsing code in stapbpf.cxx had been hardcoded in a way that o= nly works on 64-bit architectures. This is an easy fix, but unmasks a whole ran= ge of other issues that block stapbpf from working on 32-bit. Lots of code in libbpf assumes void* is interchangeable with __u64, which leads to 'stack smashing' errors on map read/write system calls.=20 Up to now, serious testing/qa for stapbpf has targeted x86_64, which is why this issue came up only when I went to make serious transport changes and decided these merited checking for regressions on other architectures. --=20 You are receiving this mail because: You are the assignee for the bug.