| Message ID | 20201026233623.91728-1-toke@redhat.com (mailing list archive) |
|---|---|
| State | Accepted |
| Delegated to: | BPF |
| Headers | show |
| Series | [bpf] samples/bpf: Set rlimit for memlock to infinity in all samples | expand |
On Mon, Oct 26, 2020 at 5:10 PM Toke Høiland-Jørgensen <toke@redhat.com> wrote: > > The memlock rlimit is a notorious source of failure for BPF programs. Most > of the samples just set it to infinity, but a few used a lower limit. The > problem with unconditionally setting a lower limit is that this will also > override the limit if the system-wide setting is *higher* than the limit > being set, which can lead to failures on systems that lock a lot of memory, > but set 'ulimit -l' to unlimited before running a sample. > > One fix for this is to only conditionally set the limit if the current > limit is lower, but it is simpler to just unify all the samples and have > them all set the limit to infinity. > > Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> > --- Acked-by: Andrii Nakryiko <andrii@kernel.org> > samples/bpf/task_fd_query_user.c | 2 +- > samples/bpf/tracex2_user.c | 2 +- > samples/bpf/tracex3_user.c | 2 +- > samples/bpf/xdp_redirect_cpu_user.c | 2 +- > samples/bpf/xdp_rxq_info_user.c | 2 +- > 5 files changed, 5 insertions(+), 5 deletions(-) > > diff --git a/samples/bpf/task_fd_query_user.c b/samples/bpf/task_fd_query_user.c > index 4a74531dc403..b68bd2f8fdc9 100644 > --- a/samples/bpf/task_fd_query_user.c > +++ b/samples/bpf/task_fd_query_user.c > @@ -290,7 +290,7 @@ static int test_debug_fs_uprobe(char *binary_path, long offset, bool is_return) > > int main(int argc, char **argv) > { > - struct rlimit r = {1024*1024, RLIM_INFINITY}; > + struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY}; > extern char __executable_start; > char filename[256], buf[256]; > __u64 uprobe_file_offset; > diff --git a/samples/bpf/tracex2_user.c b/samples/bpf/tracex2_user.c > index 3e36b3e4e3ef..3d6eab711d23 100644 > --- a/samples/bpf/tracex2_user.c > +++ b/samples/bpf/tracex2_user.c > @@ -116,7 +116,7 @@ static void int_exit(int sig) > > int main(int ac, char **argv) > { > - struct rlimit r = {1024*1024, RLIM_INFINITY}; > + struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY}; > long key, next_key, value; > struct bpf_link *links[2]; > struct bpf_program *prog; > diff --git a/samples/bpf/tracex3_user.c b/samples/bpf/tracex3_user.c > index 70e987775c15..83e0fecbb01a 100644 > --- a/samples/bpf/tracex3_user.c > +++ b/samples/bpf/tracex3_user.c > @@ -107,7 +107,7 @@ static void print_hist(int fd) > > int main(int ac, char **argv) > { > - struct rlimit r = {1024*1024, RLIM_INFINITY}; > + struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY}; > struct bpf_link *links[2]; > struct bpf_program *prog; > struct bpf_object *obj; > diff --git a/samples/bpf/xdp_redirect_cpu_user.c b/samples/bpf/xdp_redirect_cpu_user.c > index 6fb8dbde62c5..f78cb18319aa 100644 > --- a/samples/bpf/xdp_redirect_cpu_user.c > +++ b/samples/bpf/xdp_redirect_cpu_user.c > @@ -765,7 +765,7 @@ static int load_cpumap_prog(char *file_name, char *prog_name, > > int main(int argc, char **argv) > { > - struct rlimit r = {10 * 1024 * 1024, RLIM_INFINITY}; > + struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY}; > char *prog_name = "xdp_cpu_map5_lb_hash_ip_pairs"; > char *mprog_filename = "xdp_redirect_kern.o"; > char *redir_interface = NULL, *redir_map = NULL; > diff --git a/samples/bpf/xdp_rxq_info_user.c b/samples/bpf/xdp_rxq_info_user.c > index caa4e7ffcfc7..93fa1bc54f13 100644 > --- a/samples/bpf/xdp_rxq_info_user.c > +++ b/samples/bpf/xdp_rxq_info_user.c > @@ -450,7 +450,7 @@ static void stats_poll(int interval, int action, __u32 cfg_opt) > int main(int argc, char **argv) > { > __u32 cfg_options= NO_TOUCH ; /* Default: Don't touch packet memory */ > - struct rlimit r = {10 * 1024 * 1024, RLIM_INFINITY}; > + struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY}; > struct bpf_prog_load_attr prog_load_attr = { > .prog_type = BPF_PROG_TYPE_XDP, > }; > -- > 2.29.0 >
On Tue, 27 Oct 2020 00:36:23 +0100 Toke Høiland-Jørgensen <toke@redhat.com> wrote: > The memlock rlimit is a notorious source of failure for BPF programs. Most > of the samples just set it to infinity, but a few used a lower limit. The > problem with unconditionally setting a lower limit is that this will also > override the limit if the system-wide setting is *higher* than the limit > being set, which can lead to failures on systems that lock a lot of memory, > but set 'ulimit -l' to unlimited before running a sample. > > One fix for this is to only conditionally set the limit if the current > limit is lower, but it is simpler to just unify all the samples and have > them all set the limit to infinity. > > Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> This change basically disable the memlock rlimit system. And this disable method is becoming standard in more and more BPF programs. IMHO using the system-wide memlock rlimit doesn't make sense for BPF. I'm still ACKing the patch, as this seems the only way forward, to ignore and in-practice not use the memlock rlimit. Acked-by: Jesper Dangaard Brouer <brouer@redhat.com> I saw some patches on the list (from Facebook) with a new system for policy limiting memory usage per BPF program or was it mem-cgroup, but I don't think that was ever merged... I would really like to see something replace (and remove) this memlock rlimit dependency. Anyone knows what happened to that effort?
On Tue, Oct 27, 2020 at 08:14:40AM +0100, Jesper Dangaard Brouer wrote: > On Tue, 27 Oct 2020 00:36:23 +0100 > Toke Høiland-Jørgensen <toke@redhat.com> wrote: > > > The memlock rlimit is a notorious source of failure for BPF programs. Most > > of the samples just set it to infinity, but a few used a lower limit. The > > problem with unconditionally setting a lower limit is that this will also > > override the limit if the system-wide setting is *higher* than the limit > > being set, which can lead to failures on systems that lock a lot of memory, > > but set 'ulimit -l' to unlimited before running a sample. > > > > One fix for this is to only conditionally set the limit if the current > > limit is lower, but it is simpler to just unify all the samples and have > > them all set the limit to infinity. > > > > Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> > > This change basically disable the memlock rlimit system. And this > disable method is becoming standard in more and more BPF programs. > IMHO using the system-wide memlock rlimit doesn't make sense for BPF. Hi Jesper, +1 > > I'm still ACKing the patch, as this seems the only way forward, to > ignore and in-practice not use the memlock rlimit. > > Acked-by: Jesper Dangaard Brouer <brouer@redhat.com> > > > I saw some patches on the list (from Facebook) with a new system for > policy limiting memory usage per BPF program or was it mem-cgroup, but > I don't think that was ever merged... I would really like to see > something replace (and remove) this memlock rlimit dependency. Anyone > knows what happened to that effort? I'm working on it. It required some heavy changes on the mm side: accounting of the percpu memory, which required a framework for accounting of arbitrary non page-sized objects, support of accounting from an interrupt context and some manipulations with page flags in order to allow accounted vmallocs to be mapped to userspace. It's mostly done with the last part expected to reach linux-next in few days. Then I'll rebase and repost the bpf part. Thanks!
diff --git a/samples/bpf/task_fd_query_user.c b/samples/bpf/task_fd_query_user.c index 4a74531dc403..b68bd2f8fdc9 100644 --- a/samples/bpf/task_fd_query_user.c +++ b/samples/bpf/task_fd_query_user.c @@ -290,7 +290,7 @@ static int test_debug_fs_uprobe(char *binary_path, long offset, bool is_return) int main(int argc, char **argv) { - struct rlimit r = {1024*1024, RLIM_INFINITY}; + struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY}; extern char __executable_start; char filename[256], buf[256]; __u64 uprobe_file_offset; diff --git a/samples/bpf/tracex2_user.c b/samples/bpf/tracex2_user.c index 3e36b3e4e3ef..3d6eab711d23 100644 --- a/samples/bpf/tracex2_user.c +++ b/samples/bpf/tracex2_user.c @@ -116,7 +116,7 @@ static void int_exit(int sig) int main(int ac, char **argv) { - struct rlimit r = {1024*1024, RLIM_INFINITY}; + struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY}; long key, next_key, value; struct bpf_link *links[2]; struct bpf_program *prog; diff --git a/samples/bpf/tracex3_user.c b/samples/bpf/tracex3_user.c index 70e987775c15..83e0fecbb01a 100644 --- a/samples/bpf/tracex3_user.c +++ b/samples/bpf/tracex3_user.c @@ -107,7 +107,7 @@ static void print_hist(int fd) int main(int ac, char **argv) { - struct rlimit r = {1024*1024, RLIM_INFINITY}; + struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY}; struct bpf_link *links[2]; struct bpf_program *prog; struct bpf_object *obj; diff --git a/samples/bpf/xdp_redirect_cpu_user.c b/samples/bpf/xdp_redirect_cpu_user.c index 6fb8dbde62c5..f78cb18319aa 100644 --- a/samples/bpf/xdp_redirect_cpu_user.c +++ b/samples/bpf/xdp_redirect_cpu_user.c @@ -765,7 +765,7 @@ static int load_cpumap_prog(char *file_name, char *prog_name, int main(int argc, char **argv) { - struct rlimit r = {10 * 1024 * 1024, RLIM_INFINITY}; + struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY}; char *prog_name = "xdp_cpu_map5_lb_hash_ip_pairs"; char *mprog_filename = "xdp_redirect_kern.o"; char *redir_interface = NULL, *redir_map = NULL; diff --git a/samples/bpf/xdp_rxq_info_user.c b/samples/bpf/xdp_rxq_info_user.c index caa4e7ffcfc7..93fa1bc54f13 100644 --- a/samples/bpf/xdp_rxq_info_user.c +++ b/samples/bpf/xdp_rxq_info_user.c @@ -450,7 +450,7 @@ static void stats_poll(int interval, int action, __u32 cfg_opt) int main(int argc, char **argv) { __u32 cfg_options= NO_TOUCH ; /* Default: Don't touch packet memory */ - struct rlimit r = {10 * 1024 * 1024, RLIM_INFINITY}; + struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY}; struct bpf_prog_load_attr prog_load_attr = { .prog_type = BPF_PROG_TYPE_XDP, };
The memlock rlimit is a notorious source of failure for BPF programs. Most of the samples just set it to infinity, but a few used a lower limit. The problem with unconditionally setting a lower limit is that this will also override the limit if the system-wide setting is *higher* than the limit being set, which can lead to failures on systems that lock a lot of memory, but set 'ulimit -l' to unlimited before running a sample. One fix for this is to only conditionally set the limit if the current limit is lower, but it is simpler to just unify all the samples and have them all set the limit to infinity. Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> --- samples/bpf/task_fd_query_user.c | 2 +- samples/bpf/tracex2_user.c | 2 +- samples/bpf/tracex3_user.c | 2 +- samples/bpf/xdp_redirect_cpu_user.c | 2 +- samples/bpf/xdp_rxq_info_user.c | 2 +- 5 files changed, 5 insertions(+), 5 deletions(-)