[PATCH bpf-next v6 0/8] libbpf, selftests/bpf: Support cross-endian usage

[Tony Ambardar <tony.ambardar@gmail.com>]

Hello all,

This patch series targets a long-standing BPF usability issue - the lack
of general cross-compilation support - by enabling cross-endian usage of
libbpf and bpftool, as well as supporting cross-endian build targets for
selftests/bpf.

Benefits include improved BPF development and testing for embedded systems
based on e.g. big-endian MIPS, more build options e.g for s390x systems,
and better accessibility to the very latest test tools e.g. 'test_progs'.

The series touches many functional areas: BTF.ext handling; object access,
introspection, and linking; generation of normal and "light" skeletons.

Initial development and testing used mips64, since this arch makes
switching the build byte-order trivial and is thus very handy for A/B
testing. However, it lacks some key features (bpf2bpf call, kfuncs, etc)
making for poor selftests/bpf coverage.

Final testing takes the kernel and selftests/bpf cross-built from x86_64
to s390x, and runs the result under QEMU/s390x. That same configuration
could also be used on kernel-patches/bpf CI for regression testing endian
support or perhaps load-sharing s390x builds across x86_64 systems.

This thread includes some background regarding testing on QEMU/s390x and
the generally favourable results:
    https://lore.kernel.org/bpf/ZsEcsaa3juxxQBUf@kodidev-ubuntu/

Earlier versions and related discussion of the series are here:

v1: https://lore.kernel.org/bpf/cover.1724216108.git.tony.ambardar@gmail.com/
v2: https://lore.kernel.org/bpf/cover.1724313164.git.tony.ambardar@gmail.com/
v3: https://lore.kernel.org/bpf/cover.1724843049.git.tony.ambardar@gmail.com/
v4: https://lore.kernel.org/bpf/cover.1724976539.git.tony.ambardar@gmail.com/
v5: https://lore.kernel.org/bpf/cover.1725347944.git.tony.ambardar@gmail.com/

Feedback and suggestions are welcome!

Best regards,
Tony


Changelog:
---------
v5 -> v6: (comments from Andrii, Alexei, Eduard)
 - clarify info_blob_bswap() by making it explicitly conditional on
   non-native target endianness, and merge a pair of related debug
   statements
 - reformat debug statement in bpf_object_bswap_progs() on single line
 - update existing info setup functions to validate and parse info
   section metadata prior to any byte-swapping, and drop earlier added
   validation checks
 - rework cross-endian BTF.ext handling by using callback functions to
   byte-swap different types of info records, but after initial parsing
 - fix a bug always outputting BTF.ext raw data in native endianness
 - include v5 "Acked-by:" from Alexei, Yonghong

v4 -> v5: (feedback from Andrii and Eduard)
 - add separate functions to byte-swap info metadata and records, and
   ensure ordering so record bswaps occur when metadata is native endian
 - use new and existing macros to iterate through info sections/records,
   and check embedded record sizes match that of info structs used
 - drop use of <cough> evil callbacks
 - move setting swapped_endian flag to after byte-swapping functions are
   called during initialization, allowing funcs to infer endianness and
   drop a 'bool native' call parameter
 - simplify byte-swapping macro used to generate light skeleton, and use
   internal lib funcs to swap info records instead of assuming all __u32
 - change info bswap library funcs to void return
 - rework/consolidate new debug statements to reduce their number
 - remove some unneeded handling of impossible errors, and drop a safety
   check already handled elsewhere
 - add and clarify some comments

v3 -> v4:
 - fix a use-after-free ELF data-handling error causing rare CI failures
 - move bswap functions for func/line/core-relo records to internal header
 - use bswap functions also for info blobs in light skeleton

v2 -> v3: (feedback from Andrii)
 - improve some log and commit message formatting
 - restructure BTF.ext endianness safety checks and byte-swapping
 - use BTF.ext info record definitions for swapping, require BTF v1
 - follow BTF API implementation more closely for BTF.ext
 - explicitly reject loading non-native endianness program into kernel
 - simplify linker output byte-order setting
 - drop redundant safety checks during linking
 - simplify endianness macro and improve blob setup code for light skel
 - no unexpected test failures after cross-compiling x86_64 -> s390x

v1 -> v2:
 - fixed a light skeleton bug causing test_progs 'map_ptr' failure
 - simplified some BTF.ext related endianness logic
 - remove an 'inline' usage related to CI checkpatch failure
 - improve some formatting noted by checkpatch warnings
 - unexpected 'test_progs' failures drop 3 -> 2 (x86_64 to s390x cross)



Tony Ambardar (8):
  libbpf: Improve log message formatting
  libbpf: Fix header comment typos for BTF.ext
  libbpf: Fix output .symtab byte-order during linking
  libbpf: Support BTF.ext loading and output in either endianness
  libbpf: Support opening bpf objects of either endianness
  libbpf: Support linking bpf objects of either endianness
  libbpf: Support creating light skeleton of either endianness
  selftests/bpf: Support cross-endian building

 tools/lib/bpf/bpf_gen_internal.h     |   1 +
 tools/lib/bpf/btf.c                  | 284 +++++++++++++++++++++------
 tools/lib/bpf/btf.h                  |   3 +
 tools/lib/bpf/btf_dump.c             |   2 +-
 tools/lib/bpf/btf_relocate.c         |   2 +-
 tools/lib/bpf/gen_loader.c           | 187 +++++++++++++-----
 tools/lib/bpf/libbpf.c               |  56 ++++--
 tools/lib/bpf/libbpf.map             |   2 +
 tools/lib/bpf/libbpf_internal.h      |  45 ++++-
 tools/lib/bpf/linker.c               |  80 ++++++--
 tools/lib/bpf/relo_core.c            |   2 +-
 tools/lib/bpf/skel_internal.h        |   3 +-
 tools/testing/selftests/bpf/Makefile |   7 +-
 13 files changed, 527 insertions(+), 147 deletions(-)

-- 
2.34.1