bpf_override_return is a very powerful mechanism for error injection,
with the caveat that it requires whitelisting of the functions to be
overriden.

inject.py will take a call chain and optional set of predicates as
input, and inject the appropriate error when both the call chain and all
predicates are satisfied.
Signed-off-by: Howard McLauchlan <hmclauchlan@fb.com>

cmake: search for iperf3 if there is no iperf

Some destro just provide iperf3 like openSUSE Tumbleweed
Signed-off-by: Nirmoy Das <ndas@suse.de>

tools/slabratetop: detect the current memory allocator

fix fd leakage

Currently, slabratetop always included slub_def.h. For the system uses
SLAB, the definition of 'struct kmem_cache' is different and the eBPF
program would fetch the wrong field and print garbages.
Signed-off-by: Gary Lin <glin@suse.com>

bpf_prog_compute_tag() isn't closing socket fds
before successful return
Signed-off-by: Nirmoy Das <ndas@suse.de>

Perf Event interface with raw perf_event_attr argument

docs: fix commit hash and link for bpf_redirect_map()

The link wrongly pointed to the same commit as for
bpf_sk_redirect_map().

Links to recent articles on LWN.net

fix CMakefile for lua

Issue #1621 reported an issue where bcc-lua has an unresolved
symbol bcc_resolve_symname. This is a side effect of previous
cmake file refactorization which permits better control
over what features can be enabled during build time.
The refactorization caused some symbols in bcc_syms.cc
got lost in the final link which produced bcc-lua.

This patch fixed the issue by add a new static library target
bcc-lua-static and adds bcc_syms.cc to it when the library
is created and hence preserved the symbol.

Also removed unnecessary linking for clang_frontend library.
Signed-off-by: Yonghong Song <yhs@fb.com>

C++ API: Allow obtaining BPFPerfBuffer pointer for polling

Tool tcpsubnet

Fix BPFModule allocation / de-allocation mismatch

Rename kprobe_poll to perf_buffer_poll

Port fixes of biosnoop from Python to Lua

Fix bugs introduced in Lua Perf Buffer handling

docs: fix commit hash and links for bpf_set_hash() and bpf_setsockopt()

tools/cachestat: Rewrite it so it makes more sense

cachstat separation between read and writes doesn't make much sense. During tests
its seen that a workload that is supposed to be completely in the missed category
appears to be 50% hit.

This patch rewrites the BCC tool to be more inline with Brendan Gregg's original
cachestat perf tool which makes more sense:
https://github.com/brendangregg/perf-tools/blob/master/fs/cachestat
http://www.brendangregg.com/blog/2014-12-31/linux-page-cache-hit-ratio.html

Fixes Issue: #1586
Signed-off-by: Joel Fernandes <joelaf@google.com>