Nathan Lynch [off-list ref] writes:

This series began as a single patch[1] to convert the RTAS subsystem's
internal locks to raw spinlocks. The discussion of that patch
identified opportunities to update a few aspects of the RTAS API, so
the series begins with those and ends with a rebased version of the
original patch.

Changes since v1:
- Unexport the singleton 'rtas' struct.
- Remove lock and args fields from 'struct rtas_t', making them
  private to the RTAS subsystem.
- Convert all symbol exports in rtas.c to EXPORT_SYMBOL_GPL.

[1] https://lore.kernel.org/linuxppc-dev/20230110044255.122616-1-nathanl@linux.ibm.com/ (local)

Nathan Lynch (4):
  powerpc/rtas: unexport 'rtas' symbol
  powerpc/rtas: make all exports GPL
  powerpc/rtas: remove lock and args fields from global rtas struct
  powerpc/rtas: upgrade internal arch spinlocks

 arch/powerpc/include/asm/rtas-types.h |   2 -
 arch/powerpc/kernel/rtas.c            | 127 +++++++++++---------------
 2 files changed, 55 insertions(+), 74 deletions(-)

Note this series conflicts with my earlier series "[PATCH v2 0/4] RTAS
function table and tracepoints":

https://lore.kernel.org/linuxppc-dev/20221212230154.851325-1-nathanl@linux.ibm.com/ (local)

I'll plan on rebasing the tracepoint series, which is more
disruptive/ambitious, on this one. Let me know if I should do
otherwise.

To be transparent, I have a fair amount of RTAS-oriented but otherwise
loosely related work in progress and I'm struggling to keep it organized
and establish a submission/review cadence. Having conflicting series
pending probably is not great :-(

Should I maintain a single stack of patches over time to avoid conflicts
like this, even though there may not be a unifying theme beyond it all
being generally RTAS-related?