[Qemu-devel] [PATCH v3 00/10] TCG optimizations for 2.10

Emilio G. Cota posted 10 patches 6 years, 11 months ago
Failed in applying to current master (apply log)
There is a newer version of this series
[Qemu-devel] [PATCH v3 00/10] TCG optimizations for 2.10
Posted by Emilio G. Cota 6 years, 11 months ago
This is the v3 of this series. v2 for context:
  https://lists.gnu.org/archive/html/qemu-devel/2017-04/msg04342.html

Changes from v2 -- all due to Richard's comments:

- Inlined tb_from_jmp_cache into the TCG helper; now that this helper
  is common there is no point in having a separate function for this.

- Compile-tested for all targets.

- Dropped the TCG alignment patches for now.

- Added reviewed-by tags.

- Removed stray goto_tb from tcg_dump_ops. (That was a leftover
  that should have been removed long ago.)

- merged helper_lookup_tb_ptr and goto_ptr into gen_lookup_and_goto_ptr.
  This makes the calling code simpler and also prunes out the lookup
  when goto_ptr isn't implemented by the backend.

- tcg/i386: add the "pre-epilogue" instruction that sets eax to 0,
  instead of setting eax to 0 before every goto_ptr.

- return "r" instead of "ri" for goto_ptr's op def. (*ouch!*)

- target/arm: extend to tl before calling the lookup_and_goto_ptr
  helper, thereby fixing aarch64-* target compilation.

Also re-ran all measurements, and dropped the noinline measurements
since now there's only a single (inlined) caller.
You can inspect/fetch the changes at:
  https://github.com/cota/qemu/tree/tcg-opt-v3

Thanks,

		Emilio