linux/tools/testing
Linus Torvalds fa3889d970 user-event updates for v6.10:
- Minor update to the user_events interface
   The ABI of creating a user event states that the fields
   are separated by semicolons, and spaces should be ignored.
   But the parsing expected at least one space to be there (which was incorrect).
   Fix the reading of the string to handle fields separated by
   semicolons but no space between them.
 
   This does extend the API sightly as now "field;field" will now be
   parsed and not cause an error. But it should not cause any regressions
   as no logic should expect it to fail.
 
   Note, that the logic that parses the event fields to create the
   trace_event works with no spaces after the semi-colon. It is
   the logic that tests against existing events that is inconsistent.
   This causes registering an event without using spaces to succeed
   if it doesn't exist, but makes the same call that tries to register
   to the same event, but doesn't use spaces, fail.
 -----BEGIN PGP SIGNATURE-----
 
 iIoEABYIADIWIQRRSw7ePDh/lE+zeZMp5XQQmuv6qgUCZkZN1hQccm9zdGVkdEBn
 b29kbWlzLm9yZwAKCRAp5XQQmuv6qvCXAQDO8b2GeCuAMa2SW7PMFdpB2Tc2F5v4
 WPBEKaLb0TU+7AEAwR0rCm22p9rpke754lcpZDz7xJNcyiyMkyXeJWCauQA=
 =PYwP
 -----END PGP SIGNATURE-----

Merge tag 'trace-user-events-v6.10' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace

Pull tracing user-event updates from Steven Rostedt:

 - Minor update to the user_events interface

  The ABI of creating a user event states that the fields are separated
  by semicolons, and spaces should be ignored.

  But the parsing expected at least one space to be there (which was
  incorrect). Fix the reading of the string to handle fields separated
  by semicolons but no space between them.

  This does extend the API sightly as now "field;field" will now be
  parsed and not cause an error. But it should not cause any regressions
  as no logic should expect it to fail.

  Note, that the logic that parses the event fields to create the
  trace_event works with no spaces after the semi-colon. It is
  the logic that tests against existing events that is inconsistent.
  This causes registering an event without using spaces to succeed
  if it doesn't exist, but makes the same call that tries to register
  to the same event, but doesn't use spaces, fail.

* tag 'trace-user-events-v6.10' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
  selftests/user_events: Add non-spacing separator check
  tracing/user_events: Fix non-spaced field matching
2024-05-17 18:46:30 -07:00
..
crypto/chacha20-s390
cxl CXL changes for v6.10 merge window 2024-05-15 14:32:27 -07:00
fault-injection
ktest ktest: force $buildonly = 1 for 'make_warnings_file' test type 2024-03-15 12:36:19 -04:00
kunit RISC-V SoC Kconfig Updates for v6.10 2024-05-07 11:02:56 +02:00
memblock
nvdimm ndtest: Convert to platform remove callback returning void 2024-04-25 12:37:12 -07:00
radix-tree tools: fix userspace compilation with new test_xarray changes 2024-05-05 17:28:05 -07:00
scatterlist
selftests user-event updates for v6.10: 2024-05-17 18:46:30 -07:00
vsock vsock/test: print type for SOCK_SEQPACKET 2024-01-25 16:39:21 -08:00