Re: [PATCH v7 00/40] accel: Move has_work() from CPUClass to AccelOpsClass

2021-09-25 Thread Philippe Mathieu-Daudé
On 9/25/21 17:28, Richard Henderson wrote: On 9/25/21 10:50 AM, Philippe Mathieu-Daudé wrote: Philippe Mathieu-Daudé (40):    accel: Simplify qemu_init_vcpu()    hw/core: Restrict cpu_has_work() to sysemu    hw/core: Un-inline cpu_has_work()    hw/core: Move cpu_common_has_work() to cpu_has_work

Re: [PATCH v7 00/40] accel: Move has_work() from CPUClass to AccelOpsClass

2021-09-25 Thread Richard Henderson
On 9/25/21 10:50 AM, Philippe Mathieu-Daudé wrote: Philippe Mathieu-Daudé (40): accel: Simplify qemu_init_vcpu() hw/core: Restrict cpu_has_work() to sysemu hw/core: Un-inline cpu_has_work() hw/core: Move cpu_common_has_work() to cpu_has_work() accel: Introduce AccelOpsClass::has_wo

[PATCH v7 00/40] accel: Move has_work() from CPUClass to AccelOpsClass

2021-09-25 Thread Philippe Mathieu-Daudé
Hi, CPU has_work() is a per-accelerator handler. This series is organized in 2 parts: - Patches 1-15: Move has_work() from CPUClass to AccelOpsClass - Patches 16-40: Move remainging has_work() to TCGCPUOps I prefer to send as a single big series to be sure it is merged at once, since the 2nd part