> -----Original Message-----
> From: Anup Patel [mailto:a...@brainfault.org]
> Sent: Tuesday, March 31, 2020 1:17 PM
> To: Jiangyifei <jiangyi...@huawei.com>
> Cc: QEMU Developers <qemu-devel@nongnu.org>; open list:RISC-V <qemu-
> ri...@nongnu.org>; Anup Patel <anup.pa...@wdc.com>; Zhanghailiang
> <zhang.zhanghaili...@huawei.com>; Sagar Karandikar
> <sag...@eecs.berkeley.edu>; Bastian Koppelmann <kbast...@mail.uni-
> paderborn.de>; Zhangxiaofeng (F) <victor.zhangxiaof...@huawei.com>;
> Alistair Francis <alistair.fran...@wdc.com>; yinyipeng
> <yinyipe...@huawei.com>; Palmer Dabbelt <pal...@dabbelt.com>;
> dengkai (A) <dengk...@huawei.com>
> Subject: Re: [PATCH RFC 8/9] target/riscv: Handler KVM_EXIT_RISCV_SBI exit
>
> On Fri, Mar 13, 2020 at 9:23 AM Yifei Jiang <jiangyi...@huawei.com> wrote:
> >
> > Use char-fe handler console sbi call, which implement early console io
> > while apply 'earlycon=sbi' into kernel parameters.
> >
> > Signed-off-by: Yifei Jiang <jiangyi...@huawei.com>
> > Signed-off-by: Yipeng Yin <yinyipe...@huawei.com>
> > ---
> > target/riscv/kvm.c | 54
> > +++++++++++++++++++++++++++++++++++++++++++++-
> > 1 file changed, 53 insertions(+), 1 deletion(-)
> >
> > diff --git a/target/riscv/kvm.c b/target/riscv/kvm.c index
> > 0f429fd802..1df70fbb29 100644
> > --- a/target/riscv/kvm.c
> > +++ b/target/riscv/kvm.c
> > @@ -38,6 +38,7 @@
> > #include "qemu/log.h"
> > #include "hw/loader.h"
> > #include "kvm_riscv.h"
> > +#include "chardev/char-fe.h"
> >
> > static __u64 kvm_riscv_reg_id(__u64 type, __u64 idx) { @@ -61,6
> > +62,19 @@ static __u64 kvm_riscv_reg_id(__u64 type, __u64 idx)
> >
> > #define RISCV_FP_D_REG(idx) kvm_riscv_reg_id(KVM_REG_RISCV_FP_D,
> > idx)
> >
> > +enum sbi_ext_id {
> > + SBI_EXT_0_1_SET_TIMER = 0x0,
> > + SBI_EXT_0_1_CONSOLE_PUTCHAR = 0x1,
> > + SBI_EXT_0_1_CONSOLE_GETCHAR = 0x2,
> > + SBI_EXT_0_1_CLEAR_IPI = 0x3,
> > + SBI_EXT_0_1_SEND_IPI = 0x4,
> > + SBI_EXT_0_1_REMOTE_FENCE_I = 0x5,
> > + SBI_EXT_0_1_REMOTE_SFENCE_VMA = 0x6,
> > + SBI_EXT_0_1_REMOTE_SFENCE_VMA_ASID = 0x7,
> > + SBI_EXT_0_1_SHUTDOWN = 0x8,
> > + SBI_EXT_BASE = 0x10,
> > +};
> > +
>
> Please add separate SBI ecall interface header (similar to OpenSBI).
> This header will only have SBI spec related defines.
>
> Refer,
> https://github.com/riscv/opensbi/blob/master/include/sbi/sbi_ecall_interfa
> ce.h
Yes, thanks for your review. I'll update it at next series after you review
this series all.
>
> > static int kvm_riscv_get_regs_core(CPUState *cs) {
> > int ret = 0;
> > @@ -423,9 +437,47 @@ bool
> kvm_arch_stop_on_emulation_error(CPUState *cs)
> > return true;
> > }
> >
> > +static int kvm_riscv_handle_sbi(struct kvm_run *run) {
> > + int ret = 0;
> > + unsigned char ch;
> > + switch (run->riscv_sbi.extension_id) {
> > + case SBI_EXT_0_1_CONSOLE_PUTCHAR:
> > + ch = run->riscv_sbi.args[0];
> > + qemu_chr_fe_write(serial_hd(0)->be, &ch, sizeof(ch));
> > + break;
> > + case SBI_EXT_0_1_CONSOLE_GETCHAR:
> > + ret = qemu_chr_fe_read_all(serial_hd(0)->be, &ch, sizeof(ch));
> > + if (ret == sizeof(ch)) {
> > + run->riscv_sbi.args[0] = ch;
> > + } else {
> > + run->riscv_sbi.args[0] = -1;
> > + }
> > + break;
>
> Please emulate SBI v0.1 Shutdown call as well.
It seems that there is no need for emulating SHUTDOWN call at QEMU.
SBI_EXT_0_1_SHUTDOWN is redirect to KVM_SYSTEM_EVENT_SHUTDOWN in KVM,
which is handled by common kvm-exit handler in QEMU.
And, It is normal to perform `poweroff` in the virtual machine.
>
> > + default:
> > + qemu_log_mask(LOG_UNIMP,
> > + "%s: un-handled SBI EXIT, specific reasons is %lu\n",
> > + __func__, run->riscv_sbi.extension_id);
> > + ret = -1;
> > + break;
> > + }
> > + return ret;
> > +}
> > +
> > int kvm_arch_handle_exit(CPUState *cs, struct kvm_run *run) {
> > - return 0;
> > + int ret = 0;
> > + switch (run->exit_reason) {
> > + case KVM_EXIT_RISCV_SBI:
> > + ret = kvm_riscv_handle_sbi(run);
> > + break;
> > + default:
> > + qemu_log_mask(LOG_UNIMP, "%s: un-handled exit reason %d\n",
> > + __func__, run->exit_reason);
> > + ret = -1;
> > + break;
> > + }
> > + return ret;
> > }
> >
> > void kvm_riscv_reset_vcpu(RISCVCPU *cpu)
> > --
> > 2.19.1
> >
> >
> >
>
> Regards,
> Anup
Best regards,
Yifei Jiang
