Here are the last 2 patches of this series rebased on the ppc-for-6.1 tree. I've tidied up the comments and control flow around the reserved values of AIL, so different behaviours/reasons are treated individually which hopefully addresses David's comments.
On real hardware, setting LPCR[AIL] to a reserved value (e.g., 1 on POWER9) causes the register to retain that value but it's treated like 0, which matches what the patch does. Thanks, Nick Nicholas Piggin (2): target/ppc: rework AIL logic in interrupt delivery target/ppc: Add POWER10 exception model hw/ppc/spapr_hcall.c | 8 +- target/ppc/cpu-qom.h | 2 + target/ppc/cpu.h | 13 +- target/ppc/excp_helper.c | 217 +++++++++++++++++++++++--------- target/ppc/translate.c | 3 +- target/ppc/translate_init.c.inc | 4 +- 6 files changed, 171 insertions(+), 76 deletions(-) -- 2.23.0
