| Message ID | 20200325162026.254799-1-sdf@google.com |
|---|---|
| State | Changes Requested |
| Delegated to: | BPF Maintainers |
| Headers | show |
| Series | [bpf-next,v2] libbpf: don't allocate 16M for log buffer by default | expand |
On Wed, Mar 25, 2020 at 9:20 AM Stanislav Fomichev <sdf@google.com> wrote: > > For each prog/btf load we allocate and free 16 megs of verifier buffer. > On production systems it doesn't really make sense because the > programs/btf have gone through extensive testing and (mostly) guaranteed > to successfully load. > > Let's assume successful case by default and skip buffer allocation > on the first try. If there is an error, start with BPF_LOG_BUF_SIZE > and double it on each ENOSPC iteration. > > v2: > * Don't allocate the buffer at all on the first try (Andrii Nakryiko) > > Signed-off-by: Stanislav Fomichev <sdf@google.com> > --- > tools/lib/bpf/btf.c | 20 +++++++++++++++----- > tools/lib/bpf/libbpf.c | 22 ++++++++++++++-------- > 2 files changed, 29 insertions(+), 13 deletions(-) > > diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c > index 3d1c25fc97ae..bfef3d606b54 100644 > --- a/tools/lib/bpf/btf.c > +++ b/tools/lib/bpf/btf.c > @@ -657,22 +657,32 @@ int btf__finalize_data(struct bpf_object *obj, struct btf *btf) > > int btf__load(struct btf *btf) > { > - __u32 log_buf_size = BPF_LOG_BUF_SIZE; > + __u32 log_buf_size = 0; > char *log_buf = NULL; > int err = 0; > > if (btf->fd >= 0) > return -EEXIST; > > - log_buf = malloc(log_buf_size); > - if (!log_buf) > - return -ENOMEM; > +retry_load: > + if (log_buf_size) { > + log_buf = malloc(log_buf_size); > + if (!log_buf) > + return -ENOMEM; > > - *log_buf = 0; > + *log_buf = 0; > + } > > btf->fd = bpf_load_btf(btf->data, btf->data_size, > log_buf, log_buf_size, false); > if (btf->fd < 0) { > + if (!log_buf || errno == ENOSPC) { > + log_buf_size = max((__u32)BPF_LOG_BUF_SIZE, > + log_buf_size << 1); > + free(log_buf); > + goto retry_load; > + } > + > err = -errno; > pr_warn("Error loading BTF: %s(%d)\n", strerror(errno), errno); > if (*log_buf) > diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c > index 085e41f9b68e..b2fd43a03708 100644 > --- a/tools/lib/bpf/libbpf.c > +++ b/tools/lib/bpf/libbpf.c > @@ -4855,8 +4855,8 @@ load_program(struct bpf_program *prog, struct bpf_insn *insns, int insns_cnt, > { > struct bpf_load_program_attr load_attr; > char *cp, errmsg[STRERR_BUFSIZE]; > - int log_buf_size = BPF_LOG_BUF_SIZE; > - char *log_buf; > + size_t log_buf_size = 0; > + char *log_buf = NULL; > int btf_fd, ret; > > if (!insns || !insns_cnt) > @@ -4896,22 +4896,28 @@ load_program(struct bpf_program *prog, struct bpf_insn *insns, int insns_cnt, > load_attr.prog_flags = prog->prog_flags; > > retry_load: > - log_buf = malloc(log_buf_size); > - if (!log_buf) > - pr_warn("Alloc log buffer for bpf loader error, continue without log\n"); > + if (log_buf_size) { > + log_buf = malloc(log_buf_size); > + if (!log_buf) > + pr_warn("Alloc log buffer for bpf loader error, continue without log\n"); > + considering that log_buf is not allocated first time, if it fails to allocate on retry then it should be an error, no? Otherwise looks good, please add my ack after fixing this: Acked-by: Andrii Nakryiko <andriin@fb.com> > + *log_buf = 0; > + } > > ret = bpf_load_program_xattr(&load_attr, log_buf, log_buf_size); > > if (ret >= 0) { > - if (load_attr.log_level) > + if (log_buf && load_attr.log_level) > pr_debug("verifier log:\n%s", log_buf); > *pfd = ret; > ret = 0; > goto out; > } > > - if (errno == ENOSPC) { > - log_buf_size <<= 1; > + if (!log_buf || errno == ENOSPC) { > + log_buf_size = max((size_t)BPF_LOG_BUF_SIZE, > + log_buf_size << 1); > + > free(log_buf); > goto retry_load; > } > -- > 2.25.1.696.g5e7596f4ac-goog >
On 03/25, Andrii Nakryiko wrote: > On Wed, Mar 25, 2020 at 9:20 AM Stanislav Fomichev <sdf@google.com> wrote: > > > > For each prog/btf load we allocate and free 16 megs of verifier buffer. > > On production systems it doesn't really make sense because the > > programs/btf have gone through extensive testing and (mostly) guaranteed > > to successfully load. > > > > Let's assume successful case by default and skip buffer allocation > > on the first try. If there is an error, start with BPF_LOG_BUF_SIZE > > and double it on each ENOSPC iteration. > > > > v2: > > * Don't allocate the buffer at all on the first try (Andrii Nakryiko) > > > > Signed-off-by: Stanislav Fomichev <sdf@google.com> > > --- > > tools/lib/bpf/btf.c | 20 +++++++++++++++----- > > tools/lib/bpf/libbpf.c | 22 ++++++++++++++-------- > > 2 files changed, 29 insertions(+), 13 deletions(-) > > > > diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c > > index 3d1c25fc97ae..bfef3d606b54 100644 > > --- a/tools/lib/bpf/btf.c > > +++ b/tools/lib/bpf/btf.c > > @@ -657,22 +657,32 @@ int btf__finalize_data(struct bpf_object *obj, struct btf *btf) > > > > int btf__load(struct btf *btf) > > { > > - __u32 log_buf_size = BPF_LOG_BUF_SIZE; > > + __u32 log_buf_size = 0; > > char *log_buf = NULL; > > int err = 0; > > > > if (btf->fd >= 0) > > return -EEXIST; > > > > - log_buf = malloc(log_buf_size); > > - if (!log_buf) > > - return -ENOMEM; > > +retry_load: > > + if (log_buf_size) { > > + log_buf = malloc(log_buf_size); > > + if (!log_buf) > > + return -ENOMEM; > > > > - *log_buf = 0; > > + *log_buf = 0; > > + } > > > > btf->fd = bpf_load_btf(btf->data, btf->data_size, > > log_buf, log_buf_size, false); > > if (btf->fd < 0) { > > + if (!log_buf || errno == ENOSPC) { > > + log_buf_size = max((__u32)BPF_LOG_BUF_SIZE, > > + log_buf_size << 1); > > + free(log_buf); > > + goto retry_load; > > + } > > + > > err = -errno; > > pr_warn("Error loading BTF: %s(%d)\n", strerror(errno), errno); > > if (*log_buf) > > diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c > > index 085e41f9b68e..b2fd43a03708 100644 > > --- a/tools/lib/bpf/libbpf.c > > +++ b/tools/lib/bpf/libbpf.c > > @@ -4855,8 +4855,8 @@ load_program(struct bpf_program *prog, struct bpf_insn *insns, int insns_cnt, > > { > > struct bpf_load_program_attr load_attr; > > char *cp, errmsg[STRERR_BUFSIZE]; > > - int log_buf_size = BPF_LOG_BUF_SIZE; > > - char *log_buf; > > + size_t log_buf_size = 0; > > + char *log_buf = NULL; > > int btf_fd, ret; > > > > if (!insns || !insns_cnt) > > @@ -4896,22 +4896,28 @@ load_program(struct bpf_program *prog, struct bpf_insn *insns, int insns_cnt, > > load_attr.prog_flags = prog->prog_flags; > > > > retry_load: > > - log_buf = malloc(log_buf_size); > > - if (!log_buf) > > - pr_warn("Alloc log buffer for bpf loader error, continue without log\n"); > > + if (log_buf_size) { > > + log_buf = malloc(log_buf_size); > > + if (!log_buf) > > + pr_warn("Alloc log buffer for bpf loader error, continue without log\n"); > > + > > considering that log_buf is not allocated first time, if it fails to > allocate on retry then it should be an error, no? > > Otherwise looks good, please add my ack after fixing this: > > Acked-by: Andrii Nakryiko <andriin@fb.com> Thx, will respin a v3. Returning -ENOMEM also makes it more consistent with btf retry logic. > > + *log_buf = 0; > > + } > > > > ret = bpf_load_program_xattr(&load_attr, log_buf, log_buf_size); > > > > if (ret >= 0) { > > - if (load_attr.log_level) > > + if (log_buf && load_attr.log_level) > > pr_debug("verifier log:\n%s", log_buf); > > *pfd = ret; > > ret = 0; > > goto out; > > } > > > > - if (errno == ENOSPC) { > > - log_buf_size <<= 1; > > + if (!log_buf || errno == ENOSPC) { > > + log_buf_size = max((size_t)BPF_LOG_BUF_SIZE, > > + log_buf_size << 1); > > + > > free(log_buf); > > goto retry_load; > > } > > -- > > 2.25.1.696.g5e7596f4ac-goog > >
diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index 3d1c25fc97ae..bfef3d606b54 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -657,22 +657,32 @@ int btf__finalize_data(struct bpf_object *obj, struct btf *btf) int btf__load(struct btf *btf) { - __u32 log_buf_size = BPF_LOG_BUF_SIZE; + __u32 log_buf_size = 0; char *log_buf = NULL; int err = 0; if (btf->fd >= 0) return -EEXIST; - log_buf = malloc(log_buf_size); - if (!log_buf) - return -ENOMEM; +retry_load: + if (log_buf_size) { + log_buf = malloc(log_buf_size); + if (!log_buf) + return -ENOMEM; - *log_buf = 0; + *log_buf = 0; + } btf->fd = bpf_load_btf(btf->data, btf->data_size, log_buf, log_buf_size, false); if (btf->fd < 0) { + if (!log_buf || errno == ENOSPC) { + log_buf_size = max((__u32)BPF_LOG_BUF_SIZE, + log_buf_size << 1); + free(log_buf); + goto retry_load; + } + err = -errno; pr_warn("Error loading BTF: %s(%d)\n", strerror(errno), errno); if (*log_buf) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 085e41f9b68e..b2fd43a03708 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -4855,8 +4855,8 @@ load_program(struct bpf_program *prog, struct bpf_insn *insns, int insns_cnt, { struct bpf_load_program_attr load_attr; char *cp, errmsg[STRERR_BUFSIZE]; - int log_buf_size = BPF_LOG_BUF_SIZE; - char *log_buf; + size_t log_buf_size = 0; + char *log_buf = NULL; int btf_fd, ret; if (!insns || !insns_cnt) @@ -4896,22 +4896,28 @@ load_program(struct bpf_program *prog, struct bpf_insn *insns, int insns_cnt, load_attr.prog_flags = prog->prog_flags; retry_load: - log_buf = malloc(log_buf_size); - if (!log_buf) - pr_warn("Alloc log buffer for bpf loader error, continue without log\n"); + if (log_buf_size) { + log_buf = malloc(log_buf_size); + if (!log_buf) + pr_warn("Alloc log buffer for bpf loader error, continue without log\n"); + + *log_buf = 0; + } ret = bpf_load_program_xattr(&load_attr, log_buf, log_buf_size); if (ret >= 0) { - if (load_attr.log_level) + if (log_buf && load_attr.log_level) pr_debug("verifier log:\n%s", log_buf); *pfd = ret; ret = 0; goto out; } - if (errno == ENOSPC) { - log_buf_size <<= 1; + if (!log_buf || errno == ENOSPC) { + log_buf_size = max((size_t)BPF_LOG_BUF_SIZE, + log_buf_size << 1); + free(log_buf); goto retry_load; }
For each prog/btf load we allocate and free 16 megs of verifier buffer. On production systems it doesn't really make sense because the programs/btf have gone through extensive testing and (mostly) guaranteed to successfully load. Let's assume successful case by default and skip buffer allocation on the first try. If there is an error, start with BPF_LOG_BUF_SIZE and double it on each ENOSPC iteration. v2: * Don't allocate the buffer at all on the first try (Andrii Nakryiko) Signed-off-by: Stanislav Fomichev <sdf@google.com> --- tools/lib/bpf/btf.c | 20 +++++++++++++++----- tools/lib/bpf/libbpf.c | 22 ++++++++++++++-------- 2 files changed, 29 insertions(+), 13 deletions(-)