On 8/30/25 01:31, Paolo Bonzini wrote:
There is no reason for some accelerators to use qemu_wait_io_event_common
(which is separated from qemu_wait_io_event() specifically for round
robin). They can also check for events directly on the first pass through
the loop, instead of setting cpu->exit_request to true.
Signed-off-by: Paolo Bonzini<pbonz...@redhat.com>
---
accel/dummy-cpus.c | 2 +-
accel/hvf/hvf-accel-ops.c | 2 +-
accel/kvm/kvm-accel-ops.c | 3 ++-
accel/tcg/tcg-accel-ops-mttcg.c | 7 ++---
accel/tcg/tcg-accel-ops-rr.c | 43 ++++++++++++++-----------------
target/i386/nvmm/nvmm-accel-ops.c | 6 ++---
target/i386/whpx/whpx-accel-ops.c | 6 ++---
7 files changed, 30 insertions(+), 39 deletions(-)
Reviewed-by: Richard Henderson <richard.hender...@linaro.org>
To-do for myself: It appears as if we can reduce the number of checks for cpu == NULL in
the rr loop by moving the cpu=first_cpu assignment to the right place.
r~
