From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stanislav Fomichev Subject: Re: [RFC PATCH bpf-next v2 3/9] tools: bpftool: add probes for kernel configuration options Date: Thu, 20 Dec 2018 10:14:05 -0800 Message-ID: <20181220181405.GP20955@mini-arch.hsd1.ca.comcast.net> References: <20181220122432.27511-1-quentin.monnet@netronome.com> <20181220122432.27511-4-quentin.monnet@netronome.com> <20181220174026.GL20955@mini-arch.hsd1.ca.comcast.net> <67de492b-701d-5e02-90dd-8815e1d8de87@netronome.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Alexei Starovoitov , Daniel Borkmann , netdev@vger.kernel.org, oss-drivers@netronome.com, Arnaldo Carvalho de Melo , Jesper Dangaard Brouer , Stanislav Fomichev To: Quentin Monnet Return-path: Received: from mail-pg1-f193.google.com ([209.85.215.193]:47014 "EHLO mail-pg1-f193.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1732677AbeLTSOI (ORCPT ); Thu, 20 Dec 2018 13:14:08 -0500 Received: by mail-pg1-f193.google.com with SMTP id w7so1233942pgp.13 for ; Thu, 20 Dec 2018 10:14:07 -0800 (PST) Content-Disposition: inline In-Reply-To: <67de492b-701d-5e02-90dd-8815e1d8de87@netronome.com> Sender: netdev-owner@vger.kernel.org List-ID: On 12/20, Quentin Monnet wrote: > 2018-12-20 09:40 UTC-0800 ~ Stanislav Fomichev > > On 12/20, Quentin Monnet wrote: > > > Add probes to dump a number of options set (or not set) for compiling > > > the kernel image. These parameters provide information about what BPF > > > components should be available on the system. A number of them are not > > > directly related to eBPF, but are in fact used in the kernel as > > > conditions on which to compile, or not to compile, some of the eBPF > > > helper functions. > > > > > > Sample output: > > > > > > # bpftool feature probe kernel > > > Scanning system configuration... > > > ... > > > CONFIG_BPF is set to y > > > CONFIG_BPF_SYSCALL is set to y > > > CONFIG_HAVE_EBPF_JIT is set to y > > > ... > > > > > > # bpftool --pretty --json feature probe kernel > > > { > > > "system_config": { > > > ... > > > "CONFIG_BPF": "y", > > > "CONFIG_BPF_SYSCALL": "y", > > > "CONFIG_HAVE_EBPF_JIT": "y", > > > ... > > > } > > > } > > > > > > v2: > > > - Remove C-style macros output from this patch. > > > - NOT addressed: grouping of those config options into subsections > > > (I don't see an easy way of grouping them at the moment, please see > > > also the discussion on v1 thread). > > > > > > Signed-off-by: Quentin Monnet > > > Reviewed-by: Jakub Kicinski > > > --- > > > tools/bpf/bpftool/feature.c | 137 ++++++++++++++++++++++++++++++++++++ > > > 1 file changed, 137 insertions(+) > > > > +static void probe_kernel_image_config(void) > > > +{ > > > + const char * const options[] = { > > > + "CONFIG_BPF", > > > + "CONFIG_BPF_SYSCALL", > > > + "CONFIG_HAVE_EBPF_JIT", > > > + "CONFIG_BPF_JIT", > > > + "CONFIG_BPF_JIT_ALWAYS_ON", > > > + "CONFIG_NET", > > > + "CONFIG_XDP_SOCKETS", > > > + "CONFIG_CGROUPS", > > > + "CONFIG_CGROUP_BPF", > > > + "CONFIG_CGROUP_NET_CLASSID", > > > + "CONFIG_BPF_EVENTS", > > > + "CONFIG_LWTUNNEL_BPF", > > > + "CONFIG_NET_ACT_BPF", > > > + "CONFIG_NET_CLS_ACT", > > > + "CONFIG_NET_CLS_BPF", > > > + "CONFIG_NET_SCH_INGRESS", > > > + "CONFIG_XFRM", > > > + "CONFIG_SOCK_CGROUP_DATA", > > > + "CONFIG_IP_ROUTE_CLASSID", > > > + "CONFIG_IPV6_SEG6_BPF", > > > + "CONFIG_FUNCTION_ERROR_INJECTION", > > > + "CONFIG_BPF_KPROBE_OVERRIDE", > > > + "CONFIG_BPF_LIRC_MODE2", > > > + "CONFIG_NETFILTER_XT_MATCH_BPF", > > > + "CONFIG_TEST_BPF", > > > + "CONFIG_BPFILTER", > > > + "CONFIG_BPFILTER_UMH", > > > + "CONFIG_BPF_STREAM_PARSER", > > > + }; > > > + char *value, *buf = NULL; > > > + struct utsname utsn; > > > + char path[PATH_MAX]; > > > + size_t i, n; > > > + ssize_t ret; > > > + FILE *fd; > > > + > > > + if (uname(&utsn)) > > > + goto no_config; > > > + > > [..] > > > + snprintf(path, sizeof(path), "/boot/config-%s", utsn.release); > > > + > > > + fd = fopen(path, "r"); > > > + if (!fd && errno == ENOENT) { > > > + /* Sometimes config is at /proc/config */ > > > + fd = fopen("/proc/config", "r"); > > I wonder whether the order here should be reversed? > > 1. try /proc/config.gz (CONFIG_IKCONFIG_PROC) > > 2. if not avail, try /boot/config-$(uname -r) > > > > Because, at the end, /proc/config.gz is the real source of truth (if > > available). > > > > What is /proc/config btw? I can see only /proc/config.gz being exported. > > Hi Stanislav, > > Cilium checks for /proc/config (apparently this is where CoreOS puts its > config file), /proc/config.gz and /boot/config-$(uname -r) [0]. I took > inspiration on that but didn't implement the /proc/config.gz file, because > it would require linking with the libz and I'm not sure this is worth it. > Could be added as a follow-up if necessary. Makes sense, then maybe add a comment? So future readers know that /proc/config.gz is explicitly missing, /proc/config comes from coreos, and on the other distros we expect /boot/config-$(uname -r). I'm mainly wondering because my arch box doesn't have /boot/config-$(uname -r) and has only /proc/config.gz :-) Sigh.. > > [0] https://github.com/cilium/cilium/blob/master/bpf/run_probes.sh#L42