Message ID | 20220204165506.2846058-1-peter.maydell@linaro.org |
---|---|
Headers | show |
Series | target/arm: -cpu host/max KVM and HVF fixes | expand |
On 4/2/22 17:55, Peter Maydell wrote: > Peter Maydell (6): > target/arm: Move '-cpu host' code to cpu64.c > target/arm: Use aarch64_cpu_register() for 'host' CPU type > target/arm: Make KVM -cpu max exactly like -cpu host > target/arm: Unindent unnecessary else-clause > target/arm: Fix '-cpu max' for HVF > target/arm: Support PAuth extension for hvf Series: Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
On Fri, Feb 04, 2022 at 04:55:00PM +0000, Peter Maydell wrote: > This patchset fixes various minor bugs in KVM and HVF > -cpu host and -cpu max: > > (1) KVM -cpu max was incorrectly adding a 'sve-max-vq' property > that wouldn't work and which doesn't exist in KVM -cpu host > > (2) HVF -cpu max was using all the TCG ID fields and thus > promising the guest more than the host CPU can actually do > > (3) HVF -cpu host wasn't setting up the PAuth properties, so > it defaulted to not telling the guest about PAuth support > > This series fixes (1) by moving the '-cpu host' code to > cpu64.c since it's aarch64-specific anyway, which lets us > reuse it from the '-cpu max' init function. It fixes > (2) and (3) mostly by making HVF use the same bits of code > that KVM does for -cpu max and PAuth. > > thanks > -- PMM > > Peter Maydell (6): > target/arm: Move '-cpu host' code to cpu64.c > target/arm: Use aarch64_cpu_register() for 'host' CPU type > target/arm: Make KVM -cpu max exactly like -cpu host > target/arm: Unindent unnecessary else-clause > target/arm: Fix '-cpu max' for HVF > target/arm: Support PAuth extension for hvf > > target/arm/cpu.c | 30 ----- > target/arm/cpu64.c | 330 +++++++++++++++++++++++++-------------------- > 2 files changed, 181 insertions(+), 179 deletions(-) > > -- > 2.25.1 > For the series Reviewed-by: Andrew Jones <drjones@redhat.com> Thanks, drew
On 04.02.22 17:55, Peter Maydell wrote: > This patchset fixes various minor bugs in KVM and HVF > -cpu host and -cpu max: > > (1) KVM -cpu max was incorrectly adding a 'sve-max-vq' property > that wouldn't work and which doesn't exist in KVM -cpu host > > (2) HVF -cpu max was using all the TCG ID fields and thus > promising the guest more than the host CPU can actually do > > (3) HVF -cpu host wasn't setting up the PAuth properties, so > it defaulted to not telling the guest about PAuth support > > This series fixes (1) by moving the '-cpu host' code to > cpu64.c since it's aarch64-specific anyway, which lets us > reuse it from the '-cpu max' init function. It fixes > (2) and (3) mostly by making HVF use the same bits of code > that KVM does for -cpu max and PAuth. Reviewed-by: Alexander Graf <agraf@csgraf.de> Alex