From 450d060e8f752a6ce052a2bffd3f01633472e330 Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Wed, 18 Nov 2020 23:30:39 -0800 Subject: bpftool: Add {i,d}tlb_misses support for bpftool profile Commit 47c09d6a9f67("bpftool: Introduce "prog profile" command") introduced "bpftool prog profile" command which can be used to profile bpf program with metrics like # of instructions, This patch added support for itlb_misses and dtlb_misses. During an internal bpf program performance evaluation, I found these two metrics are also very useful. The following is an example output: $ bpftool prog profile id 324 duration 3 cycles itlb_misses 1885029 run_cnt 5134686073 cycles 306893 itlb_misses $ bpftool prog profile id 324 duration 3 cycles dtlb_misses 1827382 run_cnt 4943593648 cycles 5975636 dtlb_misses $ bpftool prog profile id 324 duration 3 cycles llc_misses 1836527 run_cnt 5019612972 cycles 4161041 llc_misses From the above, we can see quite some dtlb misses, 3 dtlb misses perf prog run. This might be something worth further investigation. Signed-off-by: Yonghong Song Signed-off-by: Daniel Borkmann Acked-by: Song Liu Link: https://lore.kernel.org/bpf/20201119073039.4060095-1-yhs@fb.com --- tools/bpf/bpftool/prog.c | 30 +++++++++++++++++++++++++++++- 1 file changed, 29 insertions(+), 1 deletion(-) (limited to 'tools/bpf/bpftool') diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c index acdb2c245f0a..1fe3ba255bad 100644 --- a/tools/bpf/bpftool/prog.c +++ b/tools/bpf/bpftool/prog.c @@ -1717,6 +1717,34 @@ struct profile_metric { .ratio_desc = "LLC misses per million insns", .ratio_mul = 1e6, }, + { + .name = "itlb_misses", + .attr = { + .type = PERF_TYPE_HW_CACHE, + .config = + PERF_COUNT_HW_CACHE_ITLB | + (PERF_COUNT_HW_CACHE_OP_READ << 8) | + (PERF_COUNT_HW_CACHE_RESULT_MISS << 16), + .exclude_user = 1 + }, + .ratio_metric = 2, + .ratio_desc = "itlb misses per million insns", + .ratio_mul = 1e6, + }, + { + .name = "dtlb_misses", + .attr = { + .type = PERF_TYPE_HW_CACHE, + .config = + PERF_COUNT_HW_CACHE_DTLB | + (PERF_COUNT_HW_CACHE_OP_READ << 8) | + (PERF_COUNT_HW_CACHE_RESULT_MISS << 16), + .exclude_user = 1 + }, + .ratio_metric = 2, + .ratio_desc = "dtlb misses per million insns", + .ratio_mul = 1e6, + }, }; static __u64 profile_total_count; @@ -2109,7 +2137,7 @@ static int do_help(int argc, char **argv) " struct_ops | fentry | fexit | freplace | sk_lookup }\n" " ATTACH_TYPE := { msg_verdict | stream_verdict | stream_parser |\n" " flow_dissector }\n" - " METRIC := { cycles | instructions | l1d_loads | llc_misses }\n" + " METRIC := { cycles | instructions | l1d_loads | llc_misses | itlb_misses | dtlb_misses }\n" " " HELP_SPEC_OPTIONS "\n" "", bin_name, argv[-2]); -- cgit v1.2.3 From 71ccb50074f31a50a1da4c1d8306d54da0907b00 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Tue, 1 Dec 2020 22:52:41 -0800 Subject: tools/bpftool: Emit name for anonymous BTFs For consistency of output, emit "name " for BTFs without the name. This keeps output more consistent and obvious. Suggested-by: Song Liu Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20201202065244.530571-2-andrii@kernel.org --- tools/bpf/bpftool/btf.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'tools/bpf/bpftool') diff --git a/tools/bpf/bpftool/btf.c b/tools/bpf/bpftool/btf.c index ed5e97157241..bd46af6a61cc 100644 --- a/tools/bpf/bpftool/btf.c +++ b/tools/bpf/bpftool/btf.c @@ -750,6 +750,8 @@ show_btf_plain(struct bpf_btf_info *info, int fd, printf("name [%s] ", name); else if (name && name[0]) printf("name %s ", name); + else + printf("name "); printf("size %uB", info->btf_size); n = 0; -- cgit v1.2.3 From fa4528379a51ff8d5271e1bcfa0d5ef71657869f Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Tue, 1 Dec 2020 22:52:43 -0800 Subject: tools/bpftool: Auto-detect split BTFs in common cases In case of working with module's split BTF from /sys/kernel/btf/*, auto-substitute /sys/kernel/btf/vmlinux as the base BTF. This makes using bpftool with module BTFs faster and simpler. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20201202065244.530571-4-andrii@kernel.org --- tools/bpf/bpftool/btf.c | 25 +++++++++++++++++++++---- 1 file changed, 21 insertions(+), 4 deletions(-) (limited to 'tools/bpf/bpftool') diff --git a/tools/bpf/bpftool/btf.c b/tools/bpf/bpftool/btf.c index bd46af6a61cc..94cd3bad5430 100644 --- a/tools/bpf/bpftool/btf.c +++ b/tools/bpf/bpftool/btf.c @@ -357,11 +357,13 @@ static int dump_btf_raw(const struct btf *btf, dump_btf_type(btf, root_type_ids[i], t); } } else { + const struct btf *base; int cnt = btf__get_nr_types(btf); int start_id = 1; - if (base_btf) - start_id = btf__get_nr_types(base_btf) + 1; + base = btf__base_btf(btf); + if (base) + start_id = btf__get_nr_types(base) + 1; for (i = start_id; i <= cnt; i++) { t = btf__type_by_id(btf, i); @@ -428,7 +430,7 @@ done: static int do_dump(int argc, char **argv) { - struct btf *btf = NULL; + struct btf *btf = NULL, *base = NULL; __u32 root_type_ids[2]; int root_type_cnt = 0; bool dump_c = false; @@ -502,7 +504,21 @@ static int do_dump(int argc, char **argv) } NEXT_ARG(); } else if (is_prefix(src, "file")) { - btf = btf__parse_split(*argv, base_btf); + const char sysfs_prefix[] = "/sys/kernel/btf/"; + const char sysfs_vmlinux[] = "/sys/kernel/btf/vmlinux"; + + if (!base_btf && + strncmp(*argv, sysfs_prefix, sizeof(sysfs_prefix) - 1) == 0 && + strcmp(*argv, sysfs_vmlinux) != 0) { + base = btf__parse(sysfs_vmlinux, NULL); + if (libbpf_get_error(base)) { + p_err("failed to parse vmlinux BTF at '%s': %ld\n", + sysfs_vmlinux, libbpf_get_error(base)); + base = NULL; + } + } + + btf = btf__parse_split(*argv, base ?: base_btf); if (IS_ERR(btf)) { err = -PTR_ERR(btf); btf = NULL; @@ -567,6 +583,7 @@ static int do_dump(int argc, char **argv) done: close(fd); btf__free(btf); + btf__free(base); return err; } -- cgit v1.2.3