| Message ID | 20210728165525.19104-1-hengqi.chen@gmail.com (mailing list archive) |
|---|---|
| State | Superseded |
| Delegated to: | BPF |
| Headers | show |
| Series | [bpf-next,v2] libbpf: add libbpf_load_vmlinux_btf/libbpf_load_module_btf APIs | expand |
| Context | Check | Description |
|---|---|---|
| netdev/cover_letter | success | Link |
| netdev/fixes_present | success | Link |
| netdev/patch_count | success | Link |
| netdev/tree_selection | success | Clearly marked for bpf-next |
| netdev/subject_prefix | success | Link |
| netdev/cc_maintainers | warning | 3 maintainers not CCed: netdev@vger.kernel.org songliubraving@fb.com kpsingh@kernel.org |
| netdev/source_inline | success | Was 0 now: 0 |
| netdev/verify_signedoff | success | Link |
| netdev/module_param | success | Was 0 now: 0 |
| netdev/build_32bit | success | Errors and warnings before: 0 this patch: 0 |
| netdev/kdoc | success | Errors and warnings before: 0 this patch: 0 |
| netdev/verify_fixes | success | Link |
| netdev/checkpatch | warning | WARNING: line length of 84 exceeds 80 columns WARNING: line length of 96 exceeds 80 columns |
| netdev/build_allmodconfig_warn | success | Errors and warnings before: 0 this patch: 0 |
| netdev/header_inline | success | Link |
On Wed, Jul 28, 2021 at 9:55 AM Hengqi Chen <hengqi.chen@gmail.com> wrote: > > Add two new APIs: libbpf_load_vmlinux_btf/libbpf_load_module_btf. > libbpf_load_vmlinux_btf is just an alias to the existing API named > libbpf_find_kernel_btf, rename it to be more precisely. > libbpf_load_module_btf can be used to load module BTF, add it for > completeness. These two APIs are useful for implementing tracing > tools and introspection tools. > This is part of the efforts towards libbpf 1.0. [1] > > [1] https://github.com/libbpf/libbpf/issues/280 > > Signed-off-by: Hengqi Chen <hengqi.chen@gmail.com> > --- > tools/lib/bpf/btf.c | 15 ++++++++++++++- > tools/lib/bpf/btf.h | 2 ++ > tools/lib/bpf/libbpf.c | 4 ++-- > tools/lib/bpf/libbpf.map | 2 ++ > 4 files changed, 20 insertions(+), 3 deletions(-) > > diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c > index b46760b93bb4..5f801739a1a2 100644 > --- a/tools/lib/bpf/btf.c > +++ b/tools/lib/bpf/btf.c > @@ -4021,7 +4021,7 @@ static void btf_dedup_merge_hypot_map(struct btf_dedup *d) > */ > if (d->hypot_adjust_canon) > continue; > - > + > if (t_kind == BTF_KIND_FWD && c_kind != BTF_KIND_FWD) > d->map[t_id] = c_id; > > @@ -4395,6 +4395,11 @@ static int btf_dedup_remap_types(struct btf_dedup *d) > * data out of it to use for target BTF. > */ > struct btf *libbpf_find_kernel_btf(void) > +{ > + return libbpf_load_vmlinux_btf(); > +} > + > +struct btf *libbpf_load_vmlinux_btf(void) > { > struct { > const char *path_fmt; > @@ -4440,6 +4445,14 @@ struct btf *libbpf_find_kernel_btf(void) > return libbpf_err_ptr(-ESRCH); > } > > +struct btf *libbpf_load_module_btf(const char *module_name, struct btf *vmlinux_btf) In the light of Quentin's btf__load_from_kernel_by_id(), I'm now wondering if it's better to keep the naming consistent as btf__load_vmlinux_btf() and btf__load_module_btf()? And we should put them after btf__parse() family of constructors, as just another set of (special, but still) constructors. WDYT? Sorry for a bit of back and forth... Otherwise everything looks good, thanks. > +{ > + char path[80]; > + > + snprintf(path, sizeof(path), "/sys/kernel/btf/%s", module_name); > + return btf__parse_split(path, vmlinux_btf); > +} > + > int btf_type_visit_type_ids(struct btf_type *t, type_id_visit_fn visit, void *ctx) > { > int i, n, err; > diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h > index 374e9f15de2e..1abf94e3bd9e 100644 > --- a/tools/lib/bpf/btf.h > +++ b/tools/lib/bpf/btf.h > @@ -90,6 +90,8 @@ LIBBPF_API __u32 btf_ext__func_info_rec_size(const struct btf_ext *btf_ext); > LIBBPF_API __u32 btf_ext__line_info_rec_size(const struct btf_ext *btf_ext); > > LIBBPF_API struct btf *libbpf_find_kernel_btf(void); > +LIBBPF_API struct btf *libbpf_load_vmlinux_btf(void); > +LIBBPF_API struct btf *libbpf_load_module_btf(const char *module_name, struct btf *vmlinux_btf); as mentioned above, let's move these right after btf__parse() family, so that all BTF constructor APIs are listed first. > > LIBBPF_API int btf__find_str(struct btf *btf, const char *s); > LIBBPF_API int btf__add_str(struct btf *btf, const char *s); > diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c > index a1ca6fb0c6d8..321d8f4889af 100644 > --- a/tools/lib/bpf/libbpf.c > +++ b/tools/lib/bpf/libbpf.c > @@ -2680,7 +2680,7 @@ static int bpf_object__load_vmlinux_btf(struct bpf_object *obj, bool force) > if (!force && !obj_needs_vmlinux_btf(obj)) > return 0; > > - obj->btf_vmlinux = libbpf_find_kernel_btf(); > + obj->btf_vmlinux = libbpf_load_vmlinux_btf(); > err = libbpf_get_error(obj->btf_vmlinux); > if (err) { > pr_warn("Error loading vmlinux BTF: %d\n", err); > @@ -8297,7 +8297,7 @@ int libbpf_find_vmlinux_btf_id(const char *name, > struct btf *btf; > int err; > > - btf = libbpf_find_kernel_btf(); > + btf = libbpf_load_vmlinux_btf(); > err = libbpf_get_error(btf); > if (err) { > pr_warn("vmlinux BTF is not found\n"); > diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map > index c240d488eb5e..2088bdbc0f50 100644 > --- a/tools/lib/bpf/libbpf.map > +++ b/tools/lib/bpf/libbpf.map > @@ -377,4 +377,6 @@ LIBBPF_0.5.0 { > bpf_object__gen_loader; > btf_dump__dump_type_data; > libbpf_set_strict_mode; > + libbpf_load_vmlinux_btf; > + libbpf_load_module_btf; > } LIBBPF_0.4.0; > -- > 2.25.1 >
On 2021/7/30 5:35 AM, Andrii Nakryiko wrote: > On Wed, Jul 28, 2021 at 9:55 AM Hengqi Chen <hengqi.chen@gmail.com> wrote: >> >> Add two new APIs: libbpf_load_vmlinux_btf/libbpf_load_module_btf. >> libbpf_load_vmlinux_btf is just an alias to the existing API named >> libbpf_find_kernel_btf, rename it to be more precisely. >> libbpf_load_module_btf can be used to load module BTF, add it for >> completeness. These two APIs are useful for implementing tracing >> tools and introspection tools. >> This is part of the efforts towards libbpf 1.0. [1] >> >> [1] https://github.com/libbpf/libbpf/issues/280 >> >> Signed-off-by: Hengqi Chen <hengqi.chen@gmail.com> >> --- >> tools/lib/bpf/btf.c | 15 ++++++++++++++- >> tools/lib/bpf/btf.h | 2 ++ >> tools/lib/bpf/libbpf.c | 4 ++-- >> tools/lib/bpf/libbpf.map | 2 ++ >> 4 files changed, 20 insertions(+), 3 deletions(-) >> >> diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c >> index b46760b93bb4..5f801739a1a2 100644 >> --- a/tools/lib/bpf/btf.c >> +++ b/tools/lib/bpf/btf.c >> @@ -4021,7 +4021,7 @@ static void btf_dedup_merge_hypot_map(struct btf_dedup *d) >> */ >> if (d->hypot_adjust_canon) >> continue; >> - >> + >> if (t_kind == BTF_KIND_FWD && c_kind != BTF_KIND_FWD) >> d->map[t_id] = c_id; >> >> @@ -4395,6 +4395,11 @@ static int btf_dedup_remap_types(struct btf_dedup *d) >> * data out of it to use for target BTF. >> */ >> struct btf *libbpf_find_kernel_btf(void) >> +{ >> + return libbpf_load_vmlinux_btf(); >> +} >> + >> +struct btf *libbpf_load_vmlinux_btf(void) >> { >> struct { >> const char *path_fmt; >> @@ -4440,6 +4445,14 @@ struct btf *libbpf_find_kernel_btf(void) >> return libbpf_err_ptr(-ESRCH); >> } >> >> +struct btf *libbpf_load_module_btf(const char *module_name, struct btf *vmlinux_btf) > > In the light of Quentin's btf__load_from_kernel_by_id(), I'm now > wondering if it's better to keep the naming consistent as > btf__load_vmlinux_btf() and btf__load_module_btf()? And we should put > them after btf__parse() family of constructors, as just another set of > (special, but still) constructors. WDYT? > > Sorry for a bit of back and forth... > > Otherwise everything looks good, thanks. > Thanks for the review. Will update as you suggested. Sometimes naming is the hardest part of programming, I am not good at that. :) (BTW, Quentin's tweet led me to here) >> +{ >> + char path[80]; >> + >> + snprintf(path, sizeof(path), "/sys/kernel/btf/%s", module_name); >> + return btf__parse_split(path, vmlinux_btf); >> +} >> + >> int btf_type_visit_type_ids(struct btf_type *t, type_id_visit_fn visit, void *ctx) >> { >> int i, n, err; >> diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h >> index 374e9f15de2e..1abf94e3bd9e 100644 >> --- a/tools/lib/bpf/btf.h >> +++ b/tools/lib/bpf/btf.h >> @@ -90,6 +90,8 @@ LIBBPF_API __u32 btf_ext__func_info_rec_size(const struct btf_ext *btf_ext); >> LIBBPF_API __u32 btf_ext__line_info_rec_size(const struct btf_ext *btf_ext); >> >> LIBBPF_API struct btf *libbpf_find_kernel_btf(void); >> +LIBBPF_API struct btf *libbpf_load_vmlinux_btf(void); >> +LIBBPF_API struct btf *libbpf_load_module_btf(const char *module_name, struct btf *vmlinux_btf); > > as mentioned above, let's move these right after btf__parse() family, > so that all BTF constructor APIs are listed first. > Sure, will do. >> >> LIBBPF_API int btf__find_str(struct btf *btf, const char *s); >> LIBBPF_API int btf__add_str(struct btf *btf, const char *s); >> diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c >> index a1ca6fb0c6d8..321d8f4889af 100644 >> --- a/tools/lib/bpf/libbpf.c >> +++ b/tools/lib/bpf/libbpf.c >> @@ -2680,7 +2680,7 @@ static int bpf_object__load_vmlinux_btf(struct bpf_object *obj, bool force) >> if (!force && !obj_needs_vmlinux_btf(obj)) >> return 0; >> >> - obj->btf_vmlinux = libbpf_find_kernel_btf(); >> + obj->btf_vmlinux = libbpf_load_vmlinux_btf(); >> err = libbpf_get_error(obj->btf_vmlinux); >> if (err) { >> pr_warn("Error loading vmlinux BTF: %d\n", err); >> @@ -8297,7 +8297,7 @@ int libbpf_find_vmlinux_btf_id(const char *name, >> struct btf *btf; >> int err; >> >> - btf = libbpf_find_kernel_btf(); >> + btf = libbpf_load_vmlinux_btf(); >> err = libbpf_get_error(btf); >> if (err) { >> pr_warn("vmlinux BTF is not found\n"); >> diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map >> index c240d488eb5e..2088bdbc0f50 100644 >> --- a/tools/lib/bpf/libbpf.map >> +++ b/tools/lib/bpf/libbpf.map >> @@ -377,4 +377,6 @@ LIBBPF_0.5.0 { >> bpf_object__gen_loader; >> btf_dump__dump_type_data; >> libbpf_set_strict_mode; >> + libbpf_load_vmlinux_btf; >> + libbpf_load_module_btf; >> } LIBBPF_0.4.0; >> -- >> 2.25.1 >>
diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index b46760b93bb4..5f801739a1a2 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -4021,7 +4021,7 @@ static void btf_dedup_merge_hypot_map(struct btf_dedup *d) */ if (d->hypot_adjust_canon) continue; - + if (t_kind == BTF_KIND_FWD && c_kind != BTF_KIND_FWD) d->map[t_id] = c_id; @@ -4395,6 +4395,11 @@ static int btf_dedup_remap_types(struct btf_dedup *d) * data out of it to use for target BTF. */ struct btf *libbpf_find_kernel_btf(void) +{ + return libbpf_load_vmlinux_btf(); +} + +struct btf *libbpf_load_vmlinux_btf(void) { struct { const char *path_fmt; @@ -4440,6 +4445,14 @@ struct btf *libbpf_find_kernel_btf(void) return libbpf_err_ptr(-ESRCH); } +struct btf *libbpf_load_module_btf(const char *module_name, struct btf *vmlinux_btf) +{ + char path[80]; + + snprintf(path, sizeof(path), "/sys/kernel/btf/%s", module_name); + return btf__parse_split(path, vmlinux_btf); +} + int btf_type_visit_type_ids(struct btf_type *t, type_id_visit_fn visit, void *ctx) { int i, n, err; diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h index 374e9f15de2e..1abf94e3bd9e 100644 --- a/tools/lib/bpf/btf.h +++ b/tools/lib/bpf/btf.h @@ -90,6 +90,8 @@ LIBBPF_API __u32 btf_ext__func_info_rec_size(const struct btf_ext *btf_ext); LIBBPF_API __u32 btf_ext__line_info_rec_size(const struct btf_ext *btf_ext); LIBBPF_API struct btf *libbpf_find_kernel_btf(void); +LIBBPF_API struct btf *libbpf_load_vmlinux_btf(void); +LIBBPF_API struct btf *libbpf_load_module_btf(const char *module_name, struct btf *vmlinux_btf); LIBBPF_API int btf__find_str(struct btf *btf, const char *s); LIBBPF_API int btf__add_str(struct btf *btf, const char *s); diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index a1ca6fb0c6d8..321d8f4889af 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -2680,7 +2680,7 @@ static int bpf_object__load_vmlinux_btf(struct bpf_object *obj, bool force) if (!force && !obj_needs_vmlinux_btf(obj)) return 0; - obj->btf_vmlinux = libbpf_find_kernel_btf(); + obj->btf_vmlinux = libbpf_load_vmlinux_btf(); err = libbpf_get_error(obj->btf_vmlinux); if (err) { pr_warn("Error loading vmlinux BTF: %d\n", err); @@ -8297,7 +8297,7 @@ int libbpf_find_vmlinux_btf_id(const char *name, struct btf *btf; int err; - btf = libbpf_find_kernel_btf(); + btf = libbpf_load_vmlinux_btf(); err = libbpf_get_error(btf); if (err) { pr_warn("vmlinux BTF is not found\n"); diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map index c240d488eb5e..2088bdbc0f50 100644 --- a/tools/lib/bpf/libbpf.map +++ b/tools/lib/bpf/libbpf.map @@ -377,4 +377,6 @@ LIBBPF_0.5.0 { bpf_object__gen_loader; btf_dump__dump_type_data; libbpf_set_strict_mode; + libbpf_load_vmlinux_btf; + libbpf_load_module_btf; } LIBBPF_0.4.0;
Add two new APIs: libbpf_load_vmlinux_btf/libbpf_load_module_btf. libbpf_load_vmlinux_btf is just an alias to the existing API named libbpf_find_kernel_btf, rename it to be more precisely. libbpf_load_module_btf can be used to load module BTF, add it for completeness. These two APIs are useful for implementing tracing tools and introspection tools. This is part of the efforts towards libbpf 1.0. [1] [1] https://github.com/libbpf/libbpf/issues/280 Signed-off-by: Hengqi Chen <hengqi.chen@gmail.com> --- tools/lib/bpf/btf.c | 15 ++++++++++++++- tools/lib/bpf/btf.h | 2 ++ tools/lib/bpf/libbpf.c | 4 ++-- tools/lib/bpf/libbpf.map | 2 ++ 4 files changed, 20 insertions(+), 3 deletions(-)