linux.git/tools/lib/bpf/libbpf.c, branch v4.19

tools lib bpf: Provide wrapper for strerror_r to build in !_GNU_SOURCE systems

2018-09-18T13:16:18+00:00

Same problem that got fixed in a similar fashion in tools/perf/ in
c8b5f2c96d1b ("tools: Introduce str_error_r()"), fix it in the same
way, licensing needs to be sorted out to libbpf to use libapi, so,
for this simple case, just get the same wrapper in tools/lib/bpf.

This makes libbpf and its users (bpftool, selftests, perf) to build
again in Alpine Linux 3.[45678] and edge.

Acked-by: Alexei Starovoitov 
Cc: Adrian Hunter 
Cc: Daniel Borkmann 
Cc: David Ahern 
Cc: Hendrik Brueckner 
Cc: Jakub Kicinski 
Cc: Jiri Olsa 
Cc: Martin KaFai Lau 
Cc: Namhyung Kim 
Cc: Quentin Monnet 
Cc: Thomas Richter 
Cc: Wang Nan 
Cc: Yonghong Song 
Fixes: 1ce6a9fc1549 ("bpf: fix build error in libbpf with EXTRA_CFLAGS="-Wp, -D_FORTIFY_SOURCE=2 -O2"")
Link: https://lkml.kernel.org/r/20180917151636.GA21790@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo

bpf: test BPF_MAP_TYPE_REUSEPORT_SOCKARRAY

2018-08-10T23:58:46+00:00

This patch adds tests for the new BPF_MAP_TYPE_REUSEPORT_SOCKARRAY.

Signed-off-by: Martin KaFai Lau 
Acked-by: Alexei Starovoitov 
Signed-off-by: Daniel Borkmann

Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next

2018-08-07T18:02:05+00:00

Daniel Borkmann says:

====================
pull-request: bpf-next 2018-08-07

The following pull-request contains BPF updates for your *net-next* tree.

The main changes are:

1) Add cgroup local storage for BPF programs, which provides a fast
   accessible memory for storing various per-cgroup data like number
   of transmitted packets, etc, from Roman.

2) Support bpf_get_socket_cookie() BPF helper in several more program
   types that have a full socket available, from Andrey.

3) Significantly improve the performance of perf events which are
   reported from BPF offload. Also convert a couple of BPF AF_XDP
   samples overto use libbpf, both from Jakub.

4) seg6local LWT provides the End.DT6 action, which allows to
   decapsulate an outer IPv6 header containing a Segment Routing Header.
   Adds this action now to the seg6local BPF interface, from Mathieu.

5) Do not mark dst register as unbounded in MOV64 instruction when
   both src and dst register are the same, from Arthur.

6) Define u_smp_rmb() and u_smp_wmb() to their respective barrier
   instructions on arm64 for the AF_XDP sample code, from Brian.

7) Convert the tcp_client.py and tcp_server.py BPF selftest scripts
   over from Python 2 to Python 3, from Jeremy.

8) Enable BTF build flags to the BPF sample code Makefile, from Taeung.

9) Remove an unnecessary rcu_read_lock() in run_lwt_bpf(), from Taehee.

10) Several improvements to the README.rst from the BPF documentation
    to make it more consistent with RST format, from Tobin.

11) Replace all occurrences of strerror() by calls to strerror_r()
    in libbpf and fix a FORTIFY_SOURCE build error along with it,
    from Thomas.

12) Fix a bug in bpftool's get_btf() function to correctly propagate
    an error via PTR_ERR(), from Yue.
====================

Signed-off-by: David S. Miller

Merge ra.kernel.org:/pub/scm/linux/kernel/git/davem/net

2018-08-02T17:55:32+00:00

The BTF conflicts were simple overlapping changes.

The virtio_net conflict was an overlap of a fix of statistics counter,
happening alongisde a move over to a bonafide statistics structure
rather than counting value on the stack.

Signed-off-by: David S. Miller

bpf: fix build error in libbpf with EXTRA_CFLAGS="-Wp, -D_FORTIFY_SOURCE=2 -O2"

2018-07-31T00:08:55+00:00

Commit 531b014e7a2f ("tools: bpf: make use of reallocarray") causes
a compiler error when building the perf tool in the linux-next tree.
Compile file tools/lib/bpf/libbpf.c on a FEDORA 28 installation with
gcc compiler version: gcc (GCC) 8.0.1 20180324 (Red Hat 8.0.1-0.20)
shows this error message:

  [root@p23lp27] # make V=1 EXTRA_CFLAGS="-Wp,-D_FORTIFY_SOURCE=2 -O2"
  [...]
  make -f /home6/tmricht/linux-next/tools/build/Makefile.build
  	dir=./util/scripting-engines obj=libperf
  libbpf.c: In function ‘bpf_object__elf_collect’:
  libbpf.c:811:15: error: ignoring return value of ‘strerror_r’,
  		declared with attribute warn_unused_result [-Werror=unused-result]
       strerror_r(-err, errmsg, sizeof(errmsg));
                 ^
  cc1: all warnings being treated as errors
  mv: cannot stat './.libbpf.o.tmp': No such file or directory
  /home6/tmricht/linux-next/tools/build/Makefile.build:96: recipe for target 'libbpf.o' failed

Replace all occurrences of strerror() by calls to strerror_r(). To
keep the compiler quiet also use the return value from strerror_r()
otherwise a 'variable set but not use' warning which is treated as
error terminates the compile.

Fixes: 531b014e7a2f ("tools: bpf: make use of reallocarray")
Suggested-by: Jakub Kicinski 
Suggested-by: Daniel Borkmann 
Signed-off-by: Thomas Richter 
Reviewed-by: Hendrik Brueckner 
Reviewed-by: Jakub Kicinski 
Signed-off-by: Daniel Borkmann

perf build: Build error in libbpf missing initialization

2018-07-28T19:22:40+00:00

In linux-next tree compiling the perf tool with additional make flags
EXTRA_CFLAGS="-Wp,-D_FORTIFY_SOURCE=2 -O2" causes a compiler error.
It is the warning 'variable may be used uninitialized' which is treated
as error: I compile it using a FEDORA 28 installation, my gcc compiler
version: gcc (GCC) 8.0.1 20180324 (Red Hat 8.0.1-0.20). The file that
causes the error is tools/lib/bpf/libbpf.c.

  [root@p23lp27] # make V=1 EXTRA_CFLAGS="-Wp,-D_FORTIFY_SOURCE=2 -O2"
  [...]
  Makefile.config:849: No openjdk development package found, please
     install JDK package, e.g. openjdk-8-jdk, java-1.8.0-openjdk-devel
  Warning: Kernel ABI header at 'tools/include/uapi/linux/if_link.h'
          differs from latest version at 'include/uapi/linux/if_link.h'
    CC       libbpf.o
  libbpf.c: In function ‘bpf_perf_event_read_simple’:
  libbpf.c:2342:6: error: ‘ret’ may be used uninitialized in this
  			function [-Werror=maybe-uninitialized]
    int ret;
        ^
  cc1: all warnings being treated as errors
  mv: cannot stat './.libbpf.o.tmp': No such file or directory
  /home6/tmricht/linux-next/tools/build/Makefile.build:96: recipe for target 'libbpf.o' failed

Suggested-by: Jakub Kicinski 
Signed-off-by: Thomas Richter 
Signed-off-by: Daniel Borkmann

tools: libbpf: add bpf_object__find_program_by_title()

2018-07-27T05:18:44+00:00

Allow users to find programs by section names.

Signed-off-by: Jakub Kicinski 
Reviewed-by: Quentin Monnet 
Signed-off-by: Daniel Borkmann

tools: libbpf: handle NULL program gracefully in bpf_program__nth_fd()

2018-07-27T05:18:44+00:00

bpf_map__fd() handles NULL map gracefully and returns -EINVAL.
bpf_program__fd() and bpf_program__nth_fd() crash in this case.
Make the behaviour more consistent by validating prog pointer
as well.

Signed-off-by: Jakub Kicinski 
Reviewed-by: Quentin Monnet 
Signed-off-by: Daniel Borkmann

bpf: Introduce BPF_ANNOTATE_KV_PAIR

2018-07-25T04:57:55+00:00

This patch introduces BPF_ANNOTATE_KV_PAIR to signal the
bpf loader about the btf key_type and value_type of a bpf map.
Please refer to the changes in test_btf_haskv.c for its usage.
Both iproute2 and libbpf loader will then have the same
convention to find out the map's btf_key_type_id and
btf_value_type_id from a map's name.

Fixes: 8a138aed4a80 ("bpf: btf: Add BTF support to libbpf")
Suggested-by: Daniel Borkmann 
Signed-off-by: Martin KaFai Lau 
Acked-by: Yonghong Song 
Signed-off-by: Daniel Borkmann

bpf: Replace [u]int32_t and [u]int64_t in libbpf

2018-07-25T04:57:55+00:00

This patch replaces [u]int32_t and [u]int64_t usage with
__[su]32 and __[su]64.  The same change goes for [u]int16_t
and [u]int8_t.

Fixes: 8a138aed4a80 ("bpf: btf: Add BTF support to libbpf")
Signed-off-by: Martin KaFai Lau 
Acked-by: Yonghong Song 
Signed-off-by: Daniel Borkmann