Message ID | 157140968634.9073.6407090804163937103.stgit@john-XPS-13-9370 |
---|---|
State | Accepted |
Delegated to: | BPF Maintainers |
Headers | show |
Series | [bpf-next] bpf: libbpf, add kernel version section parsing back | expand |
On 10/18/19 7:41 AM, John Fastabend wrote: > With commit "libbpf: stop enforcing kern_version,..." we removed the > kernel version section parsing in favor of querying for the kernel > using uname() and populating the version using the result of the > query. After this any version sections were simply ignored. > > Unfortunately, the world of kernels is not so friendly. I've found some > customized kernels where uname() does not match the in kernel version. > To fix this so programs can load in this environment this patch adds > back parsing the section and if it exists uses the user specified > kernel version to override the uname() result. However, keep most the > kernel uname() discovery bits so users are not required to insert the > version except in these odd cases. > > Fixes: 5e61f27070292 ("libbpf: stop enforcing kern_version, populate it for users") > Signed-off-by: John Fastabend <john.fastabend@gmail.com> > --- In the name of not breaking users of weird kernels :) Acked-by: Andrii Nakryiko <andriin@fb.com> > tools/lib/bpf/libbpf.c | 21 ++++++++++++++++++++- > 1 file changed, 20 insertions(+), 1 deletion(-) > [...]
On Fri, Oct 18, 2019 at 07:41:26AM -0700, John Fastabend wrote: > With commit "libbpf: stop enforcing kern_version,..." we removed the > kernel version section parsing in favor of querying for the kernel > using uname() and populating the version using the result of the > query. After this any version sections were simply ignored. > > Unfortunately, the world of kernels is not so friendly. I've found some > customized kernels where uname() does not match the in kernel version. > To fix this so programs can load in this environment this patch adds > back parsing the section and if it exists uses the user specified > kernel version to override the uname() result. However, keep most the > kernel uname() discovery bits so users are not required to insert the > version except in these odd cases. > > Fixes: 5e61f27070292 ("libbpf: stop enforcing kern_version, populate it for users") > Signed-off-by: John Fastabend <john.fastabend@gmail.com> Applied, thanks!
On Fri, Oct 18, 2019 at 9:52 AM Andrii Nakryiko <andriin@fb.com> wrote: > > On 10/18/19 7:41 AM, John Fastabend wrote: > > With commit "libbpf: stop enforcing kern_version,..." we removed the > > kernel version section parsing in favor of querying for the kernel > > using uname() and populating the version using the result of the > > query. After this any version sections were simply ignored. > > > > Unfortunately, the world of kernels is not so friendly. I've found some > > customized kernels where uname() does not match the in kernel version. > > To fix this so programs can load in this environment this patch adds > > back parsing the section and if it exists uses the user specified > > kernel version to override the uname() result. However, keep most the > > kernel uname() discovery bits so users are not required to insert the > > version except in these odd cases. > > > > Fixes: 5e61f27070292 ("libbpf: stop enforcing kern_version, populate it for users") > > Signed-off-by: John Fastabend <john.fastabend@gmail.com> > > --- > > In the name of not breaking users of weird kernels :) > > Acked-by: Andrii Nakryiko <andriin@fb.com> What does it mean that uname is cheated? Can libbpf read it from /proc/sys/kernel/osrelease ? or /proc/version? Is that read only or user space can somehow overwrite it?
Alexei Starovoitov wrote: > On Fri, Oct 18, 2019 at 9:52 AM Andrii Nakryiko <andriin@fb.com> wrote: > > > > On 10/18/19 7:41 AM, John Fastabend wrote: > > > With commit "libbpf: stop enforcing kern_version,..." we removed the > > > kernel version section parsing in favor of querying for the kernel > > > using uname() and populating the version using the result of the > > > query. After this any version sections were simply ignored. > > > > > > Unfortunately, the world of kernels is not so friendly. I've found some > > > customized kernels where uname() does not match the in kernel version. > > > To fix this so programs can load in this environment this patch adds > > > back parsing the section and if it exists uses the user specified > > > kernel version to override the uname() result. However, keep most the > > > kernel uname() discovery bits so users are not required to insert the > > > version except in these odd cases. > > > > > > Fixes: 5e61f27070292 ("libbpf: stop enforcing kern_version, populate it for users") > > > Signed-off-by: John Fastabend <john.fastabend@gmail.com> > > > --- > > > > In the name of not breaking users of weird kernels :) > > > > Acked-by: Andrii Nakryiko <andriin@fb.com> > > What does it mean that uname is cheated? > Can libbpf read it from /proc/sys/kernel/osrelease ? > or /proc/version? > Is that read only or user space can somehow overwrite it? In this case LINUX_VERSION_CODE as shown in version.h from linux-headers does not much what is being reported by /proc/version or osrelease. So its a bit surprising to me as well but I haven't got to the bottom of it. Maybe something to do with how proc is mounted in this kubernetes setup?
diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index fcea6988f962..675383131179 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -656,6 +656,21 @@ bpf_object__init_license(struct bpf_object *obj, void *data, size_t size) return 0; } +static int +bpf_object__init_kversion(struct bpf_object *obj, void *data, size_t size) +{ + __u32 kver; + + if (size != sizeof(kver)) { + pr_warning("invalid kver section in %s\n", obj->path); + return -LIBBPF_ERRNO__FORMAT; + } + memcpy(&kver, data, sizeof(kver)); + obj->kern_version = kver; + pr_debug("kernel version of %s is %x\n", obj->path, obj->kern_version); + return 0; +} + static int compare_bpf_map(const void *_a, const void *_b) { const struct bpf_map *a = _a; @@ -1573,7 +1588,11 @@ static int bpf_object__elf_collect(struct bpf_object *obj, int flags) if (err) return err; } else if (strcmp(name, "version") == 0) { - /* skip, we don't need it anymore */ + err = bpf_object__init_kversion(obj, + data->d_buf, + data->d_size); + if (err) + return err; } else if (strcmp(name, "maps") == 0) { obj->efile.maps_shndx = idx; } else if (strcmp(name, MAPS_ELF_SEC) == 0) {
With commit "libbpf: stop enforcing kern_version,..." we removed the kernel version section parsing in favor of querying for the kernel using uname() and populating the version using the result of the query. After this any version sections were simply ignored. Unfortunately, the world of kernels is not so friendly. I've found some customized kernels where uname() does not match the in kernel version. To fix this so programs can load in this environment this patch adds back parsing the section and if it exists uses the user specified kernel version to override the uname() result. However, keep most the kernel uname() discovery bits so users are not required to insert the version except in these odd cases. Fixes: 5e61f27070292 ("libbpf: stop enforcing kern_version, populate it for users") Signed-off-by: John Fastabend <john.fastabend@gmail.com> --- tools/lib/bpf/libbpf.c | 21 ++++++++++++++++++++- 1 file changed, 20 insertions(+), 1 deletion(-)