Script to check that the lists of helpers are in sync

Checks src/cc/libbpf.c, docs/kernel-versions.md, src/cc/compat/linux/bpf.h,
and src/cc/export/helpers.h.

Cosmetic: Make kernel-versions commit hashes monospace

Discourage use of str() type strings in python API

Fix two issues in examples

Implement get_text for Python USDT class

Destruct USDTContext correctly in Python

<linux/blkdev.h> is not needed.
Signed-off-by: Liu Bo <liub.liubo@gmail.com>

A function definition doesn't need a tail semicolon.
Signed-off-by: Liu Bo <liub.liubo@gmail.com>

Conform to bytes encoding for some portion of the tools/tests, such that
smoke tests pass on python3. More conversions are surely required.
Signed-off-by: Brenden Blanco <bblanco@gmail.com>

Move bcc internals to sanitize all arguments as bytes instead of str
type. For now, disable warnings or actual assertion, to be turned on
incrementally.

Introduce some helpers for managing bytes/unicode objects in a way that
bridges the gap from python2 to 3.

1. Add printb() helper for writing bytes output directly to stdout. This
avoids complaints from print() in python3, which expects a unicode
str(). Since python 3.5, `b"" % bytes()` style format strings should
work and we can write tools with common code, once we convert format
strings to bytes.
http://legacy.python.org/dev/peps/pep-0461/

2. Add a class for wrapping command line arguments that are intended for
comparing to debugged memory, for instance running process COMM or
kernel pathname data. The approach takes some of the discussion from
http://legacy.python.org/dev/peps/pep-0383/ into account, though
unfortunately the python2-future implementation of "surrogateescape" is
buggy, therefore this iteration is partial.

The object instance should only ever be coerced into a bytes object.
This silently invokes encode(sys.getfilesystemencoding()), which if it
fails implies that the tool was passed junk characters on the command
line. Thereafter the tool should implement only bytes-bytes comparisons
(for instance re.search(b"", b"")) and bytes stdout printing (see
printb).

3. Add an _assert_is_bytes helper to check for improper usage of str
objects in python arguments. The behavior of the assertion can be
tweaked by changing the bcc.utils._strict_bytes bool.

Going forward, one should never invoke decode() on a bpf data stream,
e.g. the result of a table lookup or perf ring output. Leave that data
in the native bytes() representation.
Signed-off-by: Brenden Blanco <bblanco@gmail.com>

Signed-off-by: Brenden Blanco <bblanco@gmail.com>

BCC cross compilation support

BCC at the moment builds eBPF only considering the local architecture
instead of the one that the user's target kernel is running on.
For cross-compiler environments, the ARCH environment variable is
used to specify which ARCH to build the kernel for. In this patch we
add support to read ARCH and if that's not set, then fallback to
detecting based on local architecture.

This patch borrows some code from a patch doing a similar thing for
eBPF samples in the kenrel that I submitted recently [1]

[1] https://patchwork.kernel.org/patch/9961801/Signed-off-by: Joel Fernandes <joelaf@google.com>

Many developers have kernel sources at an specific location. Allow
clang to build from there by allowing to specify a single absolute
path to the kernel sources through a new env var BCC_KERNEL_SOURCE.
Signed-off-by: Joel Fernandes <joelaf@google.com>

Updated the FAQ with the error produced if python[2-3]-bcc isn't inst…

Document bpf_tail_call()

Signed-off-by: Gary Lin <glin@suse.com>

Define a new macro to make it easier to declare a program array.
Also replace BPF_TABLE("prog") in examples and tests with
BPF_PROG_ARRAY.
Signed-off-by: Gary Lin <glin@suse.com>

Fix missing include

tcptracer: Fix argparse is not defined error

Fixes compile with libc++ on linux:

```
In file included from ../tests/cc/test_array_table.cc:17:
In file included from ../src/cc/api/BPF.h:24:
../src/cc/api/BPFTable.h:118:71: error: use of undeclared identifier 'errno'
      return StatusTuple(-1, "Error getting value: %s", std::strerror(errno));
```

Add missing includes to fix build against libc++

argparse is imported as ap, using argparse explicitly results in a name
not defined error.
Signed-off-by: Gal Pressman <galp@mellanox.com>

continue adding --ebpf option to the python tools/ scripts

Fixes this compile error:

```
../src/cc/frontends/clang/kbuild_helper.h:40:64: error: use of undeclared identifier 'errno'
      fprintf(stderr, "chdir(%s): %s\n", dst.c_str(), strerror(errno));
```

This fixes the compile with `-stdlib=libc++`. Without this header the build fails with:

```
../src/cc/ns_guard.cc:75:30: error: use of undeclared identifier 'CLONE_NEWNS'
```

Found by yonghong-song while reviewing PR #1570.
Signed-off-by: Nathan Scott <nathans@redhat.com>

opensnoop: -d option for duration

Small fixes

Signed-off-by: Song Liu <songliubraving@fb.com>

The semicolon is usually added when the macro is used. Update both the
macro definition and all uses.
Signed-off-by: Song Liu <songliubraving@fb.com>

Several python tools allow their eBPF code to be printed to
stdout for debugging.  There are other projects that would
like to share these program definitions however, instead of
duplicating code.  We previously agreed on an --ebpf option
and we now continue adding it to more tools.
Signed-off-by: Nathan Scott <nathans@redhat.com>

Add option to print virtual address to trace.py