From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 112958 invoked by alias); 31 Jul 2015 14:03:42 -0000 Mailing-List: contact gdb-patches-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sourceware.org Received: (qmail 112945 invoked by uid 89); 31 Jul 2015 14:03:41 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-1.7 required=5.0 tests=AWL,BAYES_00,FREEMAIL_FROM,RCVD_IN_DNSWL_LOW,SPF_PASS autolearn=ham version=3.3.2 X-HELO: mail-pa0-f49.google.com Received: from mail-pa0-f49.google.com (HELO mail-pa0-f49.google.com) (209.85.220.49) by sourceware.org (qpsmtpd/0.93/v0.84-503-g423c35a) with (AES128-GCM-SHA256 encrypted) ESMTPS; Fri, 31 Jul 2015 14:03:39 +0000 Received: by pachj5 with SMTP id hj5so41473583pac.3 for ; Fri, 31 Jul 2015 07:03:37 -0700 (PDT) MIME-Version: 1.0 X-Received: by 10.66.219.102 with SMTP id pn6mr7364027pac.80.1438351417596; Fri, 31 Jul 2015 07:03:37 -0700 (PDT) Received: by 10.70.46.39 with HTTP; Fri, 31 Jul 2015 07:03:37 -0700 (PDT) In-Reply-To: <55BB79B8.8060209@ericsson.com> References: <55BB5EF0.7040106@gmail.com> <55BB79B8.8060209@ericsson.com> Date: Fri, 31 Jul 2015 14:03:00 -0000 Message-ID: Subject: Re: Testsuite docs: explain ptrace_scope and core_pattern requirements. From: Ciro Santilli To: Simon Marchi Cc: gdb-patches@sourceware.org Content-Type: text/plain; charset=UTF-8 X-IsSubscribed: yes X-SW-Source: 2015-07/txt/msg00934.txt.bz2 We could do that as well, I will take a look at it. But I'd rather check if "can we attach to another process" by trying it out directly rather than reading `ptrace_scope`. For example, direct ptrace_scope check could XFAIL falsely if the user is running with sudo, and there may be more variables that come into play. On Fri, Jul 31, 2015 at 3:35 PM, Simon Marchi wrote: > On 15-07-31 07:41 AM, Ciro Santilli wrote: >> From 526bde3e1716a13406f9343ff6a788da9e553e6d Mon Sep 17 00:00:00 2001 >> From: Ciro Santilli >> Date: Fri, 31 Jul 2015 10:05:53 +0200 >> Subject: [PATCH] Testsuite docs: explain ptrace_scope and core_pattern >> requirements. >> >> ptrace_scope can break all attach related tests. >> >> core_pattern can break bigcore.exp. >> >> 2015-07-31 Ciro Santilli >> >> gdb: >> >> PR testsuite/18067 >> PR testsuite/18704 >> * README (GDB Testsuite): Explain ptrace_scope and core_pattern. >> Make section more succinct. >> Remove broken Red Hat DejaGNU FTP link. >> --- >> gdb/README | 35 ++++++++++++++++++++++++++--------- >> 1 file changed, 26 insertions(+), 9 deletions(-) >> >> diff --git a/gdb/README b/gdb/README >> index d485536..87a4a5f 100644 >> --- a/gdb/README >> +++ b/gdb/README >> @@ -595,15 +595,32 @@ think you will be ready to submit the patches. >> GDB Testsuite >> ============= >> >> - Included with the GDB distribution is a DejaGNU based testsuite >> -that can either be used to test your newly built GDB, or for >> -regression testing a GDB with local modifications. >> - >> - Running the testsuite requires the prior installation of DejaGNU, >> -which is generally available via ftp. The directory >> -ftp://sources.redhat.com/pub/dejagnu/ will contain a recent snapshot. >> -Once DejaGNU is installed, you can run the tests in one of the >> -following ways: >> +Requirements: >> + >> +* DejaGNU >> + >> +* Ability to attach to processes with the same UID. >> + >> + In Linux, you need either to ensure that ptrace is allowed with: >> + >> + echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope >> + >> + or sudo, which is a bad idea. >> + >> +* Ability to generate big sparse core files. >> + >> + In Linux, this is done by default, but it may be broken if your >> + distribution sets: >> + >> + /proc/sys/kernel/core_pattern >> + >> + to something starting with `|` to use a custom handler program. >> + >> + You can fix that with: >> + >> + echo | sudo tee /proc/sys/kernel/core_pattern >> + >> +Usage: >> >> (1) cd gdb-VERSION >> make check-gdb > > More generally, would it be good to modify the tests that break when ptrace_scope != 0 > so that they xfail or report unsupported? We could easily read the value of ptrace_scope > both in native and remote.