On Tue, Jan 18, 2022 at 3:21 PM Kui-Feng Lee <kuifeng@xxxxxx> wrote: > > Add a hashmap to map the string offsets from a source btf to the > string offsets from a target btf to reduce overheads. > > btf__add_btf() calls btf__add_str() to add strings from a source to a > target btf. It causes many string comparisons, and it is a major > hotspot when adding a big btf. btf__add_str() uses strcmp() to check > if a hash entry is the right one. The extra hashmap here compares > offsets of strings, that are much cheaper. It remembers the results > of btf__add_str() for later uses to reduce the cost. > > We are parallelizing BTF encoding for pahole by creating separated btf > instances for worker threads. These per-thread btf instances will be > added to the btf instance of the main thread by calling btf__add_str() > to deduplicate and write out. With this patch and -j4, the running > time of pahole drops to about 6.0s from 6.6s. > > The following lines are the summary of 'perf stat' w/o the change. > > 6.668126396 seconds time elapsed > > 13.451054000 seconds user > 0.715520000 seconds sys > > The following lines are the summary w/ the change. > > 5.986973919 seconds time elapsed > > 12.939903000 seconds user > 0.724152000 seconds sys > > V3 removes an unnecssary check against str_off_map, and merges the > declarations of two variables into one line. > > [v2] https://lore.kernel.org/bpf/20220114193713.461349-1-kuifeng@xxxxxx/ > > Signed-off-by: Kui-Feng Lee <kuifeng@xxxxxx> > --- > tools/lib/bpf/btf.c | 31 ++++++++++++++++++++++++++++++- > 1 file changed, 30 insertions(+), 1 deletion(-) > [...] > @@ -1680,6 +1697,9 @@ static int btf_rewrite_type_ids(__u32 *type_id, void *ctx) > return 0; > } > > +static size_t btf_dedup_identity_hash_fn(const void *key, void *ctx); > +static bool btf_dedup_equal_fn(const void *k1, const void *k2, void *ctx); > + > int btf__add_btf(struct btf *btf, const struct btf *src_btf) > { > struct btf_pipe p = { .src = src_btf, .dst = btf }; > @@ -1713,6 +1733,11 @@ int btf__add_btf(struct btf *btf, const struct btf *src_btf) > if (!off) > return libbpf_err(-ENOMEM); > > + /* Map the string offsets from src_btf to the offsets from btf to improve performance */ > + p.str_off_map = hashmap__new(btf_dedup_identity_hash_fn, btf_dedup_equal_fn, NULL); > + if (p.str_off_map == NULL) Sorry, I didn't catch this the first time. hashmap__new() returns ERR_PTR() on error (it's an internal API so we use ERR_PTR() for pointer-returning APIs), so you need to check for IS_ERR(p.str_off_map) instead. > + return libbpf_err(-ENOMEM); > + > /* bulk copy types data for all types from src_btf */ > memcpy(t, src_btf->types_data, data_sz); > > @@ -1754,6 +1779,8 @@ int btf__add_btf(struct btf *btf, const struct btf *src_btf) > btf->hdr->str_off += data_sz; > btf->nr_types += cnt; > > + hashmap__free(p.str_off_map); > + > /* return type ID of the first added BTF type */ > return btf->start_id + btf->nr_types - cnt; > err_out: > @@ -1767,6 +1794,8 @@ int btf__add_btf(struct btf *btf, const struct btf *src_btf) > * wasn't modified, so doesn't need restoring, see big comment above */ > btf->hdr->str_len = old_strs_len; > > + hashmap__free(p.str_off_map); > + > return libbpf_err(err); > } > > -- > 2.30.2 >
