From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============4869510363825930480==" MIME-Version: 1.0 From: Mark Wielaard To: elfutils-devel@lists.fedorahosted.org Subject: Re: [PATCH] libdwfl: dwfl_standard_argp should not fail when not able to attach Dwfl. Date: Sat, 14 Jun 2014 00:57:46 +0200 Message-ID: <1402700266.4033.22.camel@bordewijk.wildebeest.org> In-Reply-To: 1402406480-18420-1-git-send-email-mjw@redhat.com --===============4869510363825930480== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable On Tue, 2014-06-10 at 15:21 +0200, Mark Wielaard wrote: > As pointed out in https://bugzilla.redhat.com/show_bug.cgi?id=3D1107654 > commit 191080 introduced a thinko that caused dwfl_standard_argp > to fail if the Dwfl couldn't be attached. Instead of generating a warning > as the comment intended, the failure would be fatal. But even warning > about dwfl_core_file_attach () or dwfl_linux_proc_attach () failing > would be a mistake. The caller/user might not be interested in such > a non-fatal issue. So just ignore if the call failed for whatever reason. > If the caller is interested in warning up front about this issue, then > dwfl_pid () should be called to check the Dwfl is attached. Things should > work just fine for anything that doesn't call any of the dwfl_state relat= ed > functions. > = > Signed-off-by: Mark Wielaard > --- > libdwfl/ChangeLog | 5 +++++ > libdwfl/argp-std.c | 13 +++++-------- > 2 files changed, 10 insertions(+), 8 deletions(-) I pushed this patch and the followup "libdwfl: Record dwfl_attach_state error and return it on failure." patch to master now. Cheers, Mark --===============4869510363825930480==--