target/riscv/kvm: implement SBI debug console (DBCN) calls
PATCH 1/1 target/riscv/kvm: implement SBI debug console (DBCN) calls
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get-reg-list is advertising KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it via kvm_set_one_reg() during kvm_arch_init_vcpu().
Finally, change kvm_riscv_handle_sbi() to handle the incoming calls for SBI_EXT_DBCN, reading and writing as required.
A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
--- target/riscv/kvm/kvm-cpu.c | 111 +++++++++++++++++++++++++++++ target/riscv/sbi_ecall_interface.h | 17 +++++ 2 files changed, 128 insertions(+)
target/riscv/kvm/kvm-cpu.c +111 -0
@@ -409,6 +409,12 @@ static KVMCPUConfig kvm_v_vlenb = { KVM_REG_RISCV_VECTOR_CSR_REG(vlenb) }; +static KVMCPUConfig kvm_sbi_dbcn = { + .name ="sbi_dbcn" , + .kvm_reg_id =KVM_REG_RISCV |KVM_REG_SIZE_U64 | +KVM_REG_RISCV_SBI_EXT |KVM_RISCV_SBI_EXT_DBCN +}; +static void kvm_riscv_update_cpu_cfg_isa_ext (RISCVCPU *cpu,CPUState *cs) {CPURISCVState *env = &cpu->env;
@@ -1041,6 +1047,20 @@ static int uint64_cmp(const void *a, const void *b)return 0 ; } +static void kvm_riscv_check_sbi_dbcn_support (RISCVCPU *cpu, +KVMScratchCPU *kvmcpu, +struct kvm_reg_list *reglist) +{ +struct kvm_reg_list *reg_search; + + reg_search =bsearch (&kvm_sbi_dbcn.kvm_reg_id, reglist->reg, reglist->n, +sizeof (uint64_t ), uint64_cmp); + +if (reg_search) { + kvm_sbi_dbcn.supported = true; + } +} +static void kvm_riscv_read_vlenb (RISCVCPU *cpu,KVMScratchCPU *kvmcpu,struct kvm_reg_list *reglist) {
@@ -1146,6 +1166,8 @@ static void kvm_riscv_init_multiext_cfg(RISCVCPU *cpu, KVMScratchCPU *kvmcpu)if (riscv_has_ext (&cpu->env,RVV )) {kvm_riscv_read_vlenb (cpu, kvmcpu, reglist); } + +kvm_riscv_check_sbi_dbcn_support (cpu, kvmcpu, reglist); }static void riscv_init_kvm_registers (Object *cpu_obj)
@@ -1320,6 +1342,17 @@ static int kvm_vcpu_set_machine_ids(RISCVCPU *cpu, CPUState *cs)return ret; } +static int kvm_vcpu_enable_sbi_dbcn (RISCVCPU *cpu,CPUState *cs) +{ +target_ulong reg =1 ; + +if (!kvm_sbi_dbcn.supported) { +return 0 ; + } + +return kvm_set_one_reg (cs, kvm_sbi_dbcn.kvm_reg_id, ®); +} +int kvm_arch_init_vcpu (CPUState *cs) {int ret =0 ;
@@ -1337,6 +1370,8 @@ int kvm_arch_init_vcpu(CPUState *cs)kvm_riscv_update_cpu_misa_ext (cpu, cs);kvm_riscv_update_cpu_cfg_isa_ext (cpu, cs); + ret =kvm_vcpu_enable_sbi_dbcn (cpu, cs); +return ret; }
@@ -1394,6 +1429,79 @@ bool kvm_arch_stop_on_emulation_error(CPUState *cs)return true; } +static void kvm_riscv_handle_sbi_dbcn (CPUState *cs,struct kvm_run *run) +{ + g_autofree uint8_t *buf = NULL; +RISCVCPU *cpu =RISCV_CPU (cs); +target_ulong num_bytes; +uint64_t addr; +unsigned char ch; +int ret; + +switch (run->riscv_sbi.function_id) { +case SBI_EXT_DBCN_CONSOLE_READ : +case SBI_EXT_DBCN_CONSOLE_WRITE : + num_bytes = run->riscv_sbi.args[0 ]; + +if (num_bytes ==0 ) { + run->riscv_sbi.ret[0 ] =SBI_SUCCESS ; + run->riscv_sbi.ret[1 ] =0 ; +break ; + } + + addr = run->riscv_sbi.args[1 ]; + +/* +* Handle the case where a 32 bit CPU is running in a +* 64 bit addressing env. +*/ +if (riscv_cpu_mxl (&cpu->env) ==MXL_RV32 ) { + addr |= (uint64_t )run->riscv_sbi.args[2 ] <<32 ; + } + + buf =g_malloc0 (num_bytes); + +if (run->riscv_sbi.function_id ==SBI_EXT_DBCN_CONSOLE_READ ) { + ret =qemu_chr_fe_read_all (serial_hd (0 )->be, buf, num_bytes); +if (ret <0 ) { +error_report ("SBI_EXT_DBCN_CONSOLE_READ: error when " +"reading chardev" ); +exit (1 ); + } + +cpu_physical_memory_write (addr, buf, ret); + }else { +cpu_physical_memory_read (addr, buf, num_bytes); + + ret =qemu_chr_fe_write_all (serial_hd (0 )->be, buf, num_bytes); +if (ret <0 ) { +error_report ("SBI_EXT_DBCN_CONSOLE_WRITE: error when " +"writing chardev" ); +exit (1 ); + } + } + + run->riscv_sbi.ret[0 ] =SBI_SUCCESS ; + run->riscv_sbi.ret[1 ] = ret; +break ; +case SBI_EXT_DBCN_CONSOLE_WRITE_BYTE : + ch = run->riscv_sbi.args[0 ]; + ret =qemu_chr_fe_write (serial_hd (0 )->be, &ch,sizeof (ch)); + +if (ret <0 ) { +error_report ("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: error when " +"writing chardev" ); +exit (1 ); + } + + run->riscv_sbi.ret[0 ] =SBI_SUCCESS ; + run->riscv_sbi.ret[1 ] =0 ; +break ; +default : + run->riscv_sbi.ret[0 ] =SBI_ERR_NOT_SUPPORTED ; + } +} +static int kvm_riscv_handle_sbi (CPUState *cs,struct kvm_run *run) {int ret =0 ;
@@ -1412,6 +1520,9 @@ static int kvm_riscv_handle_sbi(CPUState *cs, struct kvm_run *run) } ret =0 ;break ; +case SBI_EXT_DBCN : +kvm_riscv_handle_sbi_dbcn (cs, run); +break ;default : qemu_log_mask(LOG_UNIMP, "%s: un-handled SBI EXIT, specific reasons is %lu\n",
target/riscv/sbi_ecall_interface.h +17 -0
@@ -12,6 +12,17 @@/* clang-format off */ +#define SBI_SUCCESS 0 +#define SBI_ERR_FAILED -1 +#define SBI_ERR_NOT_SUPPORTED -2 +#define SBI_ERR_INVALID_PARAM -3 +#define SBI_ERR_DENIED -4 +#define SBI_ERR_INVALID_ADDRESS -5 +#define SBI_ERR_ALREADY_AVAILABLE -6 +#define SBI_ERR_ALREADY_STARTED -7 +#define SBI_ERR_ALREADY_STOPPED -8 +#define SBI_ERR_NO_SHMEM -9 +/* SBI Extension IDs */ #define SBI_EXT_0_1_SET_TIMER 0x0#define SBI_EXT_0_1_CONSOLE_PUTCHAR 0x1
@@ -27,6 +38,7 @@#define SBI_EXT_IPI 0x735049#define SBI_EXT_RFENCE 0x52464E43#define SBI_EXT_HSM 0x48534D +#define SBI_EXT_DBCN 0x4442434E/* SBI function IDs for BASE extension */ #define SBI_EXT_BASE_GET_SPEC_VERSION 0x0
@@ -57,6 +69,11 @@#define SBI_EXT_HSM_HART_STOP 0x1#define SBI_EXT_HSM_HART_GET_STATUS 0x2 +/* SBI function IDs for DBCN extension */ +#define SBI_EXT_DBCN_CONSOLE_WRITE 0x0 +#define SBI_EXT_DBCN_CONSOLE_READ 0x1 +#define SBI_EXT_DBCN_CONSOLE_WRITE_BYTE 0x2 +#define SBI_HSM_HART_STATUS_STARTED 0x0#define SBI_HSM_HART_STATUS_STOPPED 0x1#define SBI_HSM_HART_STATUS_START_PENDING 0x2
Signature · 1 line
2.44.0
Quoted 23 lines · Daniel Henrique Barboza
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get-reg-list is advertising
KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it viakvm_set_one_reg()duringkvm_arch_init_vcpu().Finally, change
kvm_riscv_handle_sbi()to handle the incoming calls forSBI_EXT_DBCN, reading and writing as required.A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
Quoted 6 lines
Signed-off-by Daniel Henrique Barboza <[email protected]>--- target/riscv/kvm/kvm-cpu.c | 111 +++++++++++++++++++++++++++++ target/riscv/sbi_ecall_interface.h | 17 +++++ 2 files changed, 128 insertions(+)
Alistair Francis
AppliedQuoted 18 lines · Daniel Henrique Barboza
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get-reg-list is advertising
KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it viakvm_set_one_reg()duringkvm_arch_init_vcpu().Finally, change
kvm_riscv_handle_sbi()to handle the incoming calls forSBI_EXT_DBCN, reading and writing as required.
Quoted 7 lines
A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
Signed-off-by Daniel Henrique Barboza <[email protected]>
Thanks!
Applied to riscv-to-apply.next
Alistair
Quoted 215 lines
--- target/riscv/kvm/kvm-cpu.c | 111 +++++++++++++++++++++++++++++ target/riscv/sbi_ecall_interface.h | 17 +++++ 2 files changed, 128 insertions(+)target/riscv/kvm/kvm-cpu.c +111
-0@@ -409,6 +409,12 @@ static KVMCPUConfig kvm_v_vlenb = { KVM_REG_RISCV_VECTOR_CSR_REG(vlenb) }; +static KVMCPUConfig kvm_sbi_dbcn = { + .name ="sbi_dbcn" , + .kvm_reg_id =KVM_REG_RISCV |KVM_REG_SIZE_U64 | +KVM_REG_RISCV_SBI_EXT |KVM_RISCV_SBI_EXT_DBCN +}; +static void kvm_riscv_update_cpu_cfg_isa_ext (RISCVCPU *cpu,CPUState *cs) {CPURISCVState *env = &cpu->env;@@ -1041,6 +1047,20 @@ static int uint64_cmp(const void *a, const void *b)return 0 ; } +static void kvm_riscv_check_sbi_dbcn_support (RISCVCPU *cpu, +KVMScratchCPU *kvmcpu, +struct kvm_reg_list *reglist) +{ +struct kvm_reg_list *reg_search; + + reg_search =bsearch (&kvm_sbi_dbcn.kvm_reg_id, reglist->reg, reglist->n, +sizeof (uint64_t ), uint64_cmp); + +if (reg_search) { + kvm_sbi_dbcn.supported = true; + } +} +static void kvm_riscv_read_vlenb (RISCVCPU *cpu,KVMScratchCPU *kvmcpu,struct kvm_reg_list *reglist) {@@ -1146,6 +1166,8 @@ static void kvm_riscv_init_multiext_cfg(RISCVCPU *cpu, KVMScratchCPU *kvmcpu)if (riscv_has_ext (&cpu->env,RVV )) {kvm_riscv_read_vlenb (cpu, kvmcpu, reglist); } + +kvm_riscv_check_sbi_dbcn_support (cpu, kvmcpu, reglist); }static void riscv_init_kvm_registers (Object *cpu_obj)@@ -1320,6 +1342,17 @@ static int kvm_vcpu_set_machine_ids(RISCVCPU *cpu, CPUState *cs)return ret; } +static int kvm_vcpu_enable_sbi_dbcn (RISCVCPU *cpu,CPUState *cs) +{ +target_ulong reg =1 ; + +if (!kvm_sbi_dbcn.supported) { +return 0 ; + } + +return kvm_set_one_reg (cs, kvm_sbi_dbcn.kvm_reg_id, ®); +} +int kvm_arch_init_vcpu (CPUState *cs) {int ret =0 ;@@ -1337,6 +1370,8 @@ int kvm_arch_init_vcpu(CPUState *cs)kvm_riscv_update_cpu_misa_ext (cpu, cs);kvm_riscv_update_cpu_cfg_isa_ext (cpu, cs); + ret =kvm_vcpu_enable_sbi_dbcn (cpu, cs); +return ret; }@@ -1394,6 +1429,79 @@ bool kvm_arch_stop_on_emulation_error(CPUState *cs)return true; } +static void kvm_riscv_handle_sbi_dbcn (CPUState *cs,struct kvm_run *run) +{ + g_autofree uint8_t *buf = NULL; +RISCVCPU *cpu =RISCV_CPU (cs); +target_ulong num_bytes; +uint64_t addr; +unsigned char ch; +int ret; + +switch (run->riscv_sbi.function_id) { +case SBI_EXT_DBCN_CONSOLE_READ : +case SBI_EXT_DBCN_CONSOLE_WRITE : + num_bytes = run->riscv_sbi.args[0 ]; + +if (num_bytes ==0 ) { + run->riscv_sbi.ret[0 ] =SBI_SUCCESS ; + run->riscv_sbi.ret[1 ] =0 ; +break ; + } + + addr = run->riscv_sbi.args[1 ]; + +/* +* Handle the case where a 32 bit CPU is running in a +* 64 bit addressing env. +*/ +if (riscv_cpu_mxl (&cpu->env) ==MXL_RV32 ) { + addr |= (uint64_t )run->riscv_sbi.args[2 ] <<32 ; + } + + buf =g_malloc0 (num_bytes); + +if (run->riscv_sbi.function_id ==SBI_EXT_DBCN_CONSOLE_READ ) { + ret =qemu_chr_fe_read_all (serial_hd (0 )->be, buf, num_bytes); +if (ret <0 ) { +error_report ("SBI_EXT_DBCN_CONSOLE_READ: error when " +"reading chardev" ); +exit (1 ); + } + +cpu_physical_memory_write (addr, buf, ret); + }else { +cpu_physical_memory_read (addr, buf, num_bytes); + + ret =qemu_chr_fe_write_all (serial_hd (0 )->be, buf, num_bytes); +if (ret <0 ) { +error_report ("SBI_EXT_DBCN_CONSOLE_WRITE: error when " +"writing chardev" ); +exit (1 ); + } + } + + run->riscv_sbi.ret[0 ] =SBI_SUCCESS ; + run->riscv_sbi.ret[1 ] = ret; +break ; +case SBI_EXT_DBCN_CONSOLE_WRITE_BYTE : + ch = run->riscv_sbi.args[0 ]; + ret =qemu_chr_fe_write (serial_hd (0 )->be, &ch,sizeof (ch)); + +if (ret <0 ) { +error_report ("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: error when " +"writing chardev" ); +exit (1 ); + } + + run->riscv_sbi.ret[0 ] =SBI_SUCCESS ; + run->riscv_sbi.ret[1 ] =0 ; +break ; +default : + run->riscv_sbi.ret[0 ] =SBI_ERR_NOT_SUPPORTED ; + } +} +static int kvm_riscv_handle_sbi (CPUState *cs,struct kvm_run *run) {int ret =0 ;@@ -1412,6 +1520,9 @@ static int kvm_riscv_handle_sbi(CPUState *cs, struct kvm_run *run) } ret =0 ;break ; +case SBI_EXT_DBCN : +kvm_riscv_handle_sbi_dbcn (cs, run); +break ;default : qemu_log_mask(LOG_UNIMP, "%s: un-handled SBI EXIT, specific reasons is %lu\n",target/riscv/sbi_ecall_interface.h +17
-1@@ -12,6 +12,17 @@/* clang-format off */ +#define SBI_SUCCESS 0 +#define SBI_ERR_FAILED -1 +#define SBI_ERR_NOT_SUPPORTED -2 +#define SBI_ERR_INVALID_PARAM -3 +#define SBI_ERR_DENIED -4 +#define SBI_ERR_INVALID_ADDRESS -5 +#define SBI_ERR_ALREADY_AVAILABLE -6 +#define SBI_ERR_ALREADY_STARTED -7 +#define SBI_ERR_ALREADY_STOPPED -8 +#define SBI_ERR_NO_SHMEM -9 +/* SBI Extension IDs */ #define SBI_EXT_0_1_SET_TIMER 0x0#define SBI_EXT_0_1_CONSOLE_PUTCHAR 0x1@@ -27,6 +38,7 @@#define SBI_EXT_IPI 0x735049#define SBI_EXT_RFENCE 0x52464E43#define SBI_EXT_HSM 0x48534D +#define SBI_EXT_DBCN 0x4442434E/* SBI function IDs for BASE extension */ #define SBI_EXT_BASE_GET_SPEC_VERSION 0x0@@ -57,6 +69,11 @@#define SBI_EXT_HSM_HART_STOP 0x1#define SBI_EXT_HSM_HART_GET_STATUS 0x2 +/* SBI function IDs for DBCN extension */ +#define SBI_EXT_DBCN_CONSOLE_WRITE 0x0 +#define SBI_EXT_DBCN_CONSOLE_READ 0x1 +#define SBI_EXT_DBCN_CONSOLE_WRITE_BYTE 0x2 +#define SBI_HSM_HART_STATUS_STARTED 0x0#define SBI_HSM_HART_STATUS_STOPPED 0x1#define SBI_HSM_HART_STATUS_START_PENDING 0x2 --2.44.0
Hi Daniel,
(now merged as commit a6b53378f537)
Quoted 84 lines · Daniel Henrique Barboza
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get-reg-list is advertising
KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it viakvm_set_one_reg()duringkvm_arch_init_vcpu().Finally, change
kvm_riscv_handle_sbi()to handle the incoming calls forSBI_EXT_DBCN, reading and writing as required.A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
Signed-off-by Daniel Henrique Barboza <[email protected]>--- target/riscv/kvm/kvm-cpu.c | 111 +++++++++++++++++++++++++++++ target/riscv/sbi_ecall_interface.h | 17 +++++ 2 files changed, 128 insertions(+)target/riscv/kvm/kvm-cpu.c +54
-0+static void kvm_riscv_handle_sbi_dbcn (CPUState *cs,struct kvm_run *run) +{ + g_autofree uint8_t *buf = NULL; +RISCVCPU *cpu =RISCV_CPU (cs); +target_ulong num_bytes; +uint64_t addr; +unsigned char ch; +int ret; + +switch (run->riscv_sbi.function_id) { +case SBI_EXT_DBCN_CONSOLE_READ : +case SBI_EXT_DBCN_CONSOLE_WRITE : + num_bytes = run->riscv_sbi.args[0 ]; + +if (num_bytes ==0 ) { + run->riscv_sbi.ret[0 ] =SBI_SUCCESS ; + run->riscv_sbi.ret[1 ] =0 ; +break ; + } + + addr = run->riscv_sbi.args[1 ]; + +/* +* Handle the case where a 32 bit CPU is running in a +* 64 bit addressing env. +*/ +if (riscv_cpu_mxl (&cpu->env) ==MXL_RV32 ) { + addr |= (uint64_t )run->riscv_sbi.args[2 ] <<32 ; + } + + buf =g_malloc0 (num_bytes); + +if (run->riscv_sbi.function_id ==SBI_EXT_DBCN_CONSOLE_READ ) { + ret =qemu_chr_fe_read_all (serial_hd (0 )->be, buf, num_bytes); +if (ret <0 ) { +error_report ("SBI_EXT_DBCN_CONSOLE_READ: error when " +"reading chardev" ); +exit (1 ); + } + +cpu_physical_memory_write (addr, buf, ret); + }else { +cpu_physical_memory_read (addr, buf, num_bytes); + + ret =qemu_chr_fe_write_all (serial_hd (0 )->be, buf, num_bytes); +if (ret <0 ) { +error_report ("SBI_EXT_DBCN_CONSOLE_WRITE: error when " +"writing chardev" ); +exit (1 ); + } + } + + run->riscv_sbi.ret[0 ] =SBI_SUCCESS ; + run->riscv_sbi.ret[1 ] = ret;
target/riscv/kvm/kvm-cpu.c bool kvm_arch_stop_on_emulation_error(CPUState *cs) · line 1495
Quoted 10 lines
target/riscv/kvm/kvm-cpu.c +10
-0+break ; +case SBI_EXT_DBCN_CONSOLE_WRITE_BYTE : + ch = run->riscv_sbi.args[0 ]; + ret =qemu_chr_fe_write (serial_hd (0 )->be, &ch,sizeof (ch)); + +if (ret <0 ) { +error_report ("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: error when " +"writing chardev" ); +exit (1 ); + }
We are ignoring partial writes (non-blocking call returning 0 byte written), is that expected? If so, is it OK to add a comment we can safely discard not-yet-written DBCN_CONSOLE_WRITE_BYTE?
Quoted 8 lines
target/riscv/kvm/kvm-cpu.c +8
-0+ + run->riscv_sbi.ret[0 ] =SBI_SUCCESS ; + run->riscv_sbi.ret[1 ] =0 ; +break ; +default : + run->riscv_sbi.ret[0 ] =SBI_ERR_NOT_SUPPORTED ; + } +}
Quoted 95 lines · Philippe Mathieu-Daudé
Hi Daniel,
(now merged as commit
a6b53378f537)Quoted 90 lines · Daniel Henrique Barboza
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get-reg-list is advertising
KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it viakvm_set_one_reg()duringkvm_arch_init_vcpu().Finally, change
kvm_riscv_handle_sbi()to handle the incoming calls forSBI_EXT_DBCN, reading and writing as required.A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
Signed-off-by Daniel Henrique Barboza <[email protected]>---
target/riscv/kvm/kvm-cpu.c | 111 +++++++++++++++++++++++++++++target/riscv/sbi_ecall_interface.h| 17 +++++2 files changed, 128 insertions(+)target/riscv/kvm/kvm-cpu.c +60
-0+static void kvm_riscv_handle_sbi_dbcn (CPUState *cs,struct kvm_run *run) +{ + g_autofree uint8_t *buf = NULL; + RISCVCPU *cpu = RISCV_CPU(cs); + target_ulong num_bytes; + uint64_t addr; + unsigned char ch; + int ret; + + switch (run->riscv_sbi.function_id) { + case SBI_EXT_DBCN_CONSOLE_READ: + case SBI_EXT_DBCN_CONSOLE_WRITE: + num_bytes = run->riscv_sbi.args[0]; + + if (num_bytes == 0) { + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = 0; + break; + } + + addr = run->riscv_sbi.args[1]; + + /* + * Handle the case where a 32 bit CPU is running in a + * 64 bit addressing env. + */ + if (riscv_cpu_mxl(&cpu->env) == MXL_RV32) { + addr |= (uint64_t)run->riscv_sbi.args[2] << 32; + } + + buf = g_malloc0(num_bytes); + + if (run->riscv_sbi.function_id == SBI_EXT_DBCN_CONSOLE_READ) { + ret = qemu_chr_fe_read_all(serial_hd(0)->be, buf, num_bytes); + if (ret < 0) { + error_report("SBI_EXT_DBCN_CONSOLE_READ: error when " + "reading chardev"); + exit(1); + } + + cpu_physical_memory_write(addr, buf, ret); + } else { + cpu_physical_memory_read(addr, buf, num_bytes); + + ret = qemu_chr_fe_write_all(serial_hd(0)->be, buf, num_bytes); + if (ret < 0) { + error_report("SBI_EXT_DBCN_CONSOLE_WRITE: error when " + "writing chardev"); + exit(1); + } + } + + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = ret; + break; + case SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: + ch = run->riscv_sbi.args[0]; + ret = qemu_chr_fe_write(serial_hd(0)->be, &ch, sizeof(ch)); + + if (ret < 0) {
Quoted 8 lines
Quoted 4 lines
+ error_report("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: error when " + "writing chardev");
+ exit(1);
+ }We are ignoring partial writes (non-blocking call returning 0 byte written), is that expected? If so, is it OK to add a comment we can safely discard not-yet-written
DBCN_CONSOLE_WRITE_BYTE?
Not sure what you meant. IIUC qemu_chr_fe_write() returns the number of bytes consumed, 0 if no chardev is found, and -1 on error. Are you saying that we should do a loop when there's no chardev found (ret = 0) and wait a certain time until there's one available?
In fact, seeing how SBI_EXT_DBCN_CONSOLE_WRITE is written, I wonder if we could use qemu_chr_fe_write_all() in this case too. Thanks,
Daniel
Quoted 9 lines
Quoted 9 lines
target/riscv/kvm/kvm-cpu.c +8
-0+ + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = 0; + break; + default: + run->riscv_sbi.ret[0] = SBI_ERR_NOT_SUPPORTED; + } +}
Quoted 110 lines · Daniel Henrique Barboza
Quoted 98 lines · Philippe Mathieu-Daudé
Hi Daniel,
(now merged as commit
a6b53378f537)Quoted 93 lines · Daniel Henrique Barboza
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get-reg-list is advertising
KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it viakvm_set_one_reg()duringkvm_arch_init_vcpu().Finally, change
kvm_riscv_handle_sbi()to handle the incoming calls forSBI_EXT_DBCN, reading and writing as required.A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
Signed-off-by Daniel Henrique Barboza <[email protected]>---
target/riscv/kvm/kvm-cpu.c | 111 +++++++++++++++++++++++++++++target/riscv/sbi_ecall_interface.h| 17 +++++2 files changed, 128 insertions(+)+static void
kvm_riscv_handle_sbi_dbcn(CPUState *cs, struct kvm_run *run)target/riscv/kvm/kvm-cpu.c +33
-0+{ + g_autofree uint8_t *buf = NULL; + RISCVCPU *cpu = RISCV_CPU(cs); + target_ulong num_bytes; + uint64_t addr; + unsigned char ch; + int ret; + + switch (run->riscv_sbi.function_id) { + case SBI_EXT_DBCN_CONSOLE_READ: + case SBI_EXT_DBCN_CONSOLE_WRITE: + num_bytes = run->riscv_sbi.args[0]; + + if (num_bytes == 0) { + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = 0; + break; + } + + addr = run->riscv_sbi.args[1]; + + /* + * Handle the case where a 32 bit CPU is running in a + * 64 bit addressing env. + */ + if (riscv_cpu_mxl(&cpu->env) == MXL_RV32) { + addr |= (uint64_t)run->riscv_sbi.args[2] << 32; + } + + buf = g_malloc0(num_bytes); + + if (run->riscv_sbi.function_id == SBI_EXT_DBCN_CONSOLE_READ) { + ret = qemu_chr_fe_read_all(serial_hd(0)->be, buf,
num_bytes);
+ if (ret < 0) {
+ error_report("SBI_EXT_DBCN_CONSOLE_READ: error when " + "reading chardev");
+ exit(1);
+ }target/riscv/kvm/kvm-cpu.c +6
-0+ + cpu_physical_memory_write(addr, buf, ret); + } else { + cpu_physical_memory_read(addr, buf, num_bytes); + + ret = qemu_chr_fe_write_all(serial_hd(0)->be, buf,
num_bytes);
+ if (ret < 0) {
+ error_report("SBI_EXT_DBCN_CONSOLE_WRITE: error when " + "writing chardev");
+ exit(1);
+ }
+ }target/riscv/kvm/kvm-cpu.c +9
-0+ + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = ret; + break; + case SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: + ch = run->riscv_sbi.args[0]; + ret = qemu_chr_fe_write(serial_hd(0)->be, &ch, sizeof(ch)); + + if (ret < 0) {Quoted 8 lines
Quoted 4 lines
+ error_report("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: error when " + "writing chardev");
+ exit(1);
+ }We are ignoring partial writes (non-blocking call returning 0 byte written), is that expected? If so, is it OK to add a comment we can safely discard not-yet-written
DBCN_CONSOLE_WRITE_BYTE?
Quoted 2 lines
Not sure what you meant. IIUC
qemu_chr_fe_write()returns the number of bytes consumed, 0 if no chardev is found, and -1 on error.
I'm trying to address an issue Peter reported with qemu_chr_fe_write():
https://lore.kernel.org/qemu-devel/CAFEAcA_kEndvNtw4EHySXWwQPoGs029yAzZGGBcV=zGHaj7KUQ@mail.gmail.com/
Basically upon introduction in commit cd18720a294 in 2013 ("char: introduce a blocking version of qemu_chr_fe_write") the API contract was "Returns: the number of bytes consumed" which could be 0, so some frontends return 0 for "wrote no bytes".
Later in 2016 in commit fa394ed6257 ("char: make some qemu_chr_fe skip if no driver") the API documentation was changed:
- * Returns: the number of bytes consumed
+ * Returns: the number of bytes consumed (0 if no assicated CharDriver)
After this commit, some frontends started to handle '<=0' as error, while 0 is not an error.
Quoted 7 lines
Are you
saying that we should do a loop when there's no chardev found (ret = 0) and wait a certain time until there's one available?In fact, seeing how
SBI_EXT_DBCN_CONSOLE_WRITEis written, I wonder if we could useqemu_chr_fe_write_all()in this case too.
This is certainly simpler.
Regards,
Phil.
Quoted 119 lines · Philippe Mathieu-Daudé
Quoted 107 lines · Daniel Henrique Barboza
Quoted 95 lines · Philippe Mathieu-Daudé
Hi Daniel,
(now merged as commit
a6b53378f537)Quoted 90 lines · Daniel Henrique Barboza
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get-reg-list is advertising
KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it viakvm_set_one_reg()duringkvm_arch_init_vcpu().Finally, change
kvm_riscv_handle_sbi()to handle the incoming calls forSBI_EXT_DBCN, reading and writing as required.A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
Signed-off-by Daniel Henrique Barboza <[email protected]>---
target/riscv/kvm/kvm-cpu.c | 111 +++++++++++++++++++++++++++++target/riscv/sbi_ecall_interface.h| 17 +++++2 files changed, 128 insertions(+)target/riscv/kvm/kvm-cpu.c +60
-0+static void kvm_riscv_handle_sbi_dbcn (CPUState *cs,struct kvm_run *run) +{ + g_autofree uint8_t *buf = NULL; + RISCVCPU *cpu = RISCV_CPU(cs); + target_ulong num_bytes; + uint64_t addr; + unsigned char ch; + int ret; + + switch (run->riscv_sbi.function_id) { + case SBI_EXT_DBCN_CONSOLE_READ: + case SBI_EXT_DBCN_CONSOLE_WRITE: + num_bytes = run->riscv_sbi.args[0]; + + if (num_bytes == 0) { + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = 0; + break; + } + + addr = run->riscv_sbi.args[1]; + + /* + * Handle the case where a 32 bit CPU is running in a + * 64 bit addressing env. + */ + if (riscv_cpu_mxl(&cpu->env) == MXL_RV32) { + addr |= (uint64_t)run->riscv_sbi.args[2] << 32; + } + + buf = g_malloc0(num_bytes); + + if (run->riscv_sbi.function_id == SBI_EXT_DBCN_CONSOLE_READ) { + ret = qemu_chr_fe_read_all(serial_hd(0)->be, buf, num_bytes); + if (ret < 0) { + error_report("SBI_EXT_DBCN_CONSOLE_READ: error when " + "reading chardev"); + exit(1); + } + + cpu_physical_memory_write(addr, buf, ret); + } else { + cpu_physical_memory_read(addr, buf, num_bytes); + + ret = qemu_chr_fe_write_all(serial_hd(0)->be, buf, num_bytes); + if (ret < 0) { + error_report("SBI_EXT_DBCN_CONSOLE_WRITE: error when " + "writing chardev"); + exit(1); + } + } + + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = ret; + break; + case SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: + ch = run->riscv_sbi.args[0]; + ret = qemu_chr_fe_write(serial_hd(0)->be, &ch, sizeof(ch)); + + if (ret < 0) {Quoted 8 lines
Quoted 4 lines
+ error_report("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: error when " + "writing chardev");
+ exit(1);
+ }We are ignoring partial writes (non-blocking call returning 0 byte written), is that expected? If so, is it OK to add a comment we can safely discard not-yet-written
DBCN_CONSOLE_WRITE_BYTE?Quoted 2 lines
Not sure what you meant. IIUC
qemu_chr_fe_write()returns the number of bytes consumed, 0 if no chardev is found, and -1 on error.I'm trying to address an issue Peter reported with
qemu_chr_fe_write():
https://lore.kernel.org/qemu-devel/CAFEAcA_kEndvNtw4EHySXWwQPoGs029yAzZGGBcV=zGHaj7KUQ@mail.gmail.com/Basically upon introduction in commit
cd18720a294in 2013 ("char: introduce a blocking version ofqemu_chr_fe_write") the API contract was "Returns: the number of bytes consumed" which could be 0, so some frontends return 0 for "wrote no bytes".
Quoted 8 lines
Later in 2016 in commit
fa394ed6257("char: make someqemu_chr_feskip if no driver") the API documentation was changed:- * Returns: the number of bytes consumed
+ * Returns: the number of bytes consumed (0 if no assicated CharDriver)After this commit, some frontends started to handle '<=0' as error, while 0 is not an error.
I think I got the gist of it, thanks.
For this particular console call the spec says:
"This is a blocking SBI call and it will only return after writing the specified byte to the debug console. It will also return, with SBI_ERR_FAILED, if there are I/O errors."
So I think it pairs well with the blocking version qemu_chr_fe_write_all() instead. I can do this change and get out of your way in changing the callers of qemu_chr_fe_write().
But I still have questions, hehe. This blocking version has the following doc:
"(...) Unlike @qemu_chr_fe_write, this function will block if the back end cannot consume all of the data attempted to be written. This function is thread-safe.
Returns: the number of bytes consumed (0 if no associated Chardev) or -1 on error."
Do we have plans to change this API like we're doing with the non-blocking version? Because being a blocking call that promises "block until all bytes are written", and I have len > 0, I don't expect a ret = 0 to be interpret as "no bytes were written". I am ok with ret = 0 being 'no associated chardev' and not handling it as an error (for now at least) but I would like to confirm that qemu_chr_fe_write_all() will not interpret ret = 0 as a zero byte write. In other words, if for some reason other than "no chardev present" we ended up with zero bytes written I would like a ret < 0 return.
Thanks,
Daniel
Quoted 14 lines
Quoted 8 lines
Are you
saying that we should do a loop when there's no chardev found (ret = 0) and wait a certain time until there's one available?In fact, seeing how
SBI_EXT_DBCN_CONSOLE_WRITEis written, I wonder if we could useqemu_chr_fe_write_all()in this case too.This is certainly simpler.
Regards,
Phil.
(+Marc-André and Paolo who I forgot to Cc first)
Quoted 142 lines · Daniel Henrique Barboza
Quoted 126 lines · Philippe Mathieu-Daudé
Quoted 114 lines · Daniel Henrique Barboza
Quoted 102 lines · Philippe Mathieu-Daudé
Hi Daniel,
(now merged as commit
a6b53378f537)Quoted 97 lines · Daniel Henrique Barboza
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get-reg- list is advertising
KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it viakvm_set_one_reg()duringkvm_arch_init_vcpu().Finally, change
kvm_riscv_handle_sbi()to handle the incoming calls forSBI_EXT_DBCN, reading and writing as required.A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
Signed-off-by Daniel Henrique Barboza <[email protected]>---
target/riscv/kvm/kvm-cpu.c | 111 ++++++++++++++++++++++++
+++++
target/riscv/sbi_ecall_interface.h| 17 +++++2 files changed, 128 insertions(+)+static void
kvm_riscv_handle_sbi_dbcn(CPUState *cs, struct kvm_run *run)target/riscv/kvm/kvm-cpu.c +32
-0+{ + g_autofree uint8_t *buf = NULL; + RISCVCPU *cpu = RISCV_CPU(cs); + target_ulong num_bytes; + uint64_t addr; + unsigned char ch; + int ret; + + switch (run->riscv_sbi.function_id) { + case SBI_EXT_DBCN_CONSOLE_READ: + case SBI_EXT_DBCN_CONSOLE_WRITE: + num_bytes = run->riscv_sbi.args[0]; + + if (num_bytes == 0) { + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = 0; + break; + } + + addr = run->riscv_sbi.args[1]; + + /* + * Handle the case where a 32 bit CPU is running in a + * 64 bit addressing env. + */ + if (riscv_cpu_mxl(&cpu->env) == MXL_RV32) { + addr |= (uint64_t)run->riscv_sbi.args[2] << 32; + } + + buf = g_malloc0(num_bytes); + + if (run->riscv_sbi.function_id ==
SBI_EXT_DBCN_CONSOLE_READ) {
+ ret =qemu_chr_fe_read_all(serial_hd(0)->be, buf, num_bytes);
+ if (ret < 0) {
+ error_report("SBI_EXT_DBCN_CONSOLE_READ: error when " + "reading chardev");
+ exit(1);
+ }target/riscv/kvm/kvm-cpu.c +6
-0+ + cpu_physical_memory_write(addr, buf, ret); + } else { + cpu_physical_memory_read(addr, buf, num_bytes); + + ret = qemu_chr_fe_write_all(serial_hd(0)->be, buf,
num_bytes);
+ if (ret < 0) {
+ error_report("SBI_EXT_DBCN_CONSOLE_WRITE: error when "
+ "writing chardev");
+ exit(1);
+ }
+ }target/riscv/kvm/kvm-cpu.c +10
-0+ + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = ret; + break; + case SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: + ch = run->riscv_sbi.args[0]; + ret = qemu_chr_fe_write(serial_hd(0)->be, &ch, sizeof(ch)); + + if (ret < 0) { + error_report("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: errorQuoted 8 lines
Quoted 4 lines
when "
+ "writing chardev");
+ exit(1);
+ }We are ignoring partial writes (non-blocking call returning 0 byte written), is that expected? If so, is it OK to add a comment we can safely discard not-yet-written
DBCN_CONSOLE_WRITE_BYTE?Quoted 2 lines
Not sure what you meant. IIUC
qemu_chr_fe_write()returns the number of bytes consumed, 0 if no chardev is found, and -1 on error.I'm trying to address an issue Peter reported with
qemu_chr_fe_write():
https://lore.kernel.org/qemu-devel/ CAFEAcA_kEndvNtw4EHySXWwQPoGs029yAzZGGBcV=zGHaj7KUQ@mail.gmail.com/Basically upon introduction in commit
cd18720a294in 2013 ("char: introduce a blocking version ofqemu_chr_fe_write") the API contract was "Returns: the number of bytes consumed" which could be 0, so some frontends return 0 for "wrote no bytes".Quoted 8 lines
Later in 2016 in commit
fa394ed6257("char: make someqemu_chr_feskip if no driver") the API documentation was changed:- * Returns: the number of bytes consumed
+ * Returns: the number of bytes consumed (0 if no assicated CharDriver)After this commit, some frontends started to handle '<=0' as error, while 0 is not an error.
I think I got the gist of it, thanks.
For this particular console call the spec says:
Quoted 8 lines
"This is a blocking SBI call and it will only return after writing the specified byte to the debug console. It will also return, with
SBI_ERR_FAILED, if there are I/O errors."So I think it pairs well with the blocking version
qemu_chr_fe_write_all()instead. I can do this change and get out of your way in changing the callers ofqemu_chr_fe_write().
I appreciate if you post the patch (this is a 1 line change, but what matters here is the justification you just provided), but if you are busy I can do it, I have enough information to write the commit desc.
Quoted 11 lines
But I still have questions, hehe. This blocking version has the following doc:
"(...) Unlike @qemu_chr_fe_write, this function will block if the back end cannot consume all of the data attempted to be written. This function is thread-safe.
Returns: the number of bytes consumed (0 if no associated Chardev) or -1 on error."
Quoted 8 lines
Do we have plans to change this API like we're doing with the non-blocking version? Because being a blocking call that promises "block until all bytes are written", and I have len > 0, I don't expect a ret = 0 to be interpret as "no bytes were written". I am ok with ret = 0 being 'no associated chardev' and not handling it as an error (for now at least) but I would like to confirm that
qemu_chr_fe_write_all()will not interpret ret = 0 as a zero byte write. In other words, if for some reason other than "no chardev present" we ended up with zero bytes written I would like a ret < 0 return.
Correct. Clarifying this method is in my TODO. What about this example:
- frontend wants to block to write 8 bytes
- backend writes 5 bytes, hits an unrecoverable link error
Should the backend support be responsible to re-establish link and retry (when possible) to complete?
If we return -1 for error, could the frontend try to reconnect and write the 8 bytes, ending with the first 5 bytes being transmitted twice?
Meanwhile I'm thinking to document as:
"Returns @len on success, or -1 on error (some data might has been written)".
Simpler API could be: "Returns 0 on success, otherwise -errno" but we'd need to rework all the callers.
Quoted 23 lines
Thanks,
Daniel
Quoted 14 lines
Quoted 8 lines
Are you
saying that we should do a loop when there's no chardev found (ret = 0) and wait a certain time until there's one available?In fact, seeing how
SBI_EXT_DBCN_CONSOLE_WRITEis written, I wonder if we could useqemu_chr_fe_write_all()in this case too.This is certainly simpler.
Regards,
Phil.
Quoted 143 lines · Philippe Mathieu-Daudé
(+Marc-André and Paolo who I forgot to Cc first)
Quoted 140 lines · Daniel Henrique Barboza
Quoted 119 lines · Philippe Mathieu-Daudé
Quoted 107 lines · Daniel Henrique Barboza
Quoted 95 lines · Philippe Mathieu-Daudé
Hi Daniel,
(now merged as commit
a6b53378f537)Quoted 90 lines · Daniel Henrique Barboza
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get-reg- list is advertising
KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it viakvm_set_one_reg()duringkvm_arch_init_vcpu().Finally, change
kvm_riscv_handle_sbi()to handle the incoming calls forSBI_EXT_DBCN, reading and writing as required.A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
Signed-off-by Daniel Henrique Barboza <[email protected]>---
target/riscv/kvm/kvm-cpu.c | 111 ++++++++++++++++++++++++ +++++target/riscv/sbi_ecall_interface.h| 17 +++++2 files changed, 128 insertions(+)target/riscv/kvm/kvm-cpu.c +60
-0+static void kvm_riscv_handle_sbi_dbcn (CPUState *cs,struct kvm_run *run) +{ + g_autofree uint8_t *buf = NULL; + RISCVCPU *cpu = RISCV_CPU(cs); + target_ulong num_bytes; + uint64_t addr; + unsigned char ch; + int ret; + + switch (run->riscv_sbi.function_id) { + case SBI_EXT_DBCN_CONSOLE_READ: + case SBI_EXT_DBCN_CONSOLE_WRITE: + num_bytes = run->riscv_sbi.args[0]; + + if (num_bytes == 0) { + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = 0; + break; + } + + addr = run->riscv_sbi.args[1]; + + /* + * Handle the case where a 32 bit CPU is running in a + * 64 bit addressing env. + */ + if (riscv_cpu_mxl(&cpu->env) == MXL_RV32) { + addr |= (uint64_t)run->riscv_sbi.args[2] << 32; + } + + buf = g_malloc0(num_bytes); + + if (run->riscv_sbi.function_id == SBI_EXT_DBCN_CONSOLE_READ) { + ret = qemu_chr_fe_read_all(serial_hd(0)->be, buf, num_bytes); + if (ret < 0) { + error_report("SBI_EXT_DBCN_CONSOLE_READ: error when " + "reading chardev"); + exit(1); + } + + cpu_physical_memory_write(addr, buf, ret); + } else { + cpu_physical_memory_read(addr, buf, num_bytes); + + ret = qemu_chr_fe_write_all(serial_hd(0)->be, buf, num_bytes); + if (ret < 0) { + error_report("SBI_EXT_DBCN_CONSOLE_WRITE: error when " + "writing chardev"); + exit(1); + } + } + + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = ret; + break; + case SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: + ch = run->riscv_sbi.args[0]; + ret = qemu_chr_fe_write(serial_hd(0)->be, &ch, sizeof(ch)); + + if (ret < 0) {Quoted 8 lines
Quoted 4 lines
+ error_report("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: error when " + "writing chardev");
+ exit(1);
+ }We are ignoring partial writes (non-blocking call returning 0 byte written), is that expected? If so, is it OK to add a comment we can safely discard not-yet-written
DBCN_CONSOLE_WRITE_BYTE?Quoted 2 lines
Not sure what you meant. IIUC
qemu_chr_fe_write()returns the number of bytes consumed, 0 if no chardev is found, and -1 on error.I'm trying to address an issue Peter reported with
qemu_chr_fe_write():
https://lore.kernel.org/qemu-devel/ CAFEAcA_kEndvNtw4EHySXWwQPoGs029yAzZGGBcV=zGHaj7KUQ@mail.gmail.com/Basically upon introduction in commit
cd18720a294in 2013 ("char: introduce a blocking version ofqemu_chr_fe_write") the API contract was "Returns: the number of bytes consumed" which could be 0, so some frontends return 0 for "wrote no bytes".Quoted 8 lines
Later in 2016 in commit
fa394ed6257("char: make someqemu_chr_feskip if no driver") the API documentation was changed:- * Returns: the number of bytes consumed
+ * Returns: the number of bytes consumed (0 if no assicated CharDriver)After this commit, some frontends started to handle '<=0' as error, while 0 is not an error.
I think I got the gist of it, thanks.
For this particular console call the spec says:
"This is a blocking SBI call and it will only return after writing the specified byte to the debug console. It will also return, with
SBI_ERR_FAILED, if there are I/O errors."
Quoted 7 lines
Quoted 3 lines
So I think it pairs well with the blocking version
qemu_chr_fe_write_all()instead. I can do this change and get out of your way in changing the callers ofqemu_chr_fe_write().I appreciate if you post the patch (this is a 1 line change, but what matters here is the justification you just provided), but if you are busy I can do it, I have enough information to write the commit desc.
I can handle this change, don't worry about it. I'll send it out today.
Quoted 29 lines
Quoted 12 lines
But I still have questions, hehe. This blocking version has the following doc:
"(...) Unlike @qemu_chr_fe_write, this function will block if the back end cannot consume all of the data attempted to be written. This function is thread-safe.
Returns: the number of bytes consumed (0 if no associated Chardev) or -1 on error."
Quoted 8 lines
Do we have plans to change this API like we're doing with the non-blocking version? Because being a blocking call that promises "block until all bytes are written", and I have len > 0, I don't expect a ret = 0 to be interpret as "no bytes were written". I am ok with ret = 0 being 'no associated chardev' and not handling it as an error (for now at least) but I would like to confirm that
qemu_chr_fe_write_all()will not interpret ret = 0 as a zero byte write. In other words, if for some reason other than "no chardev present" we ended up with zero bytes written I would like a ret < 0 return.Correct. Clarifying this method is in my TODO. What about this example:
- frontend wants to block to write 8 bytes
- backend writes 5 bytes, hits an unrecoverable link errorShould the backend support be responsible to re-establish link and retry (when possible) to complete?
Quoted 7 lines
If we return -1 for error, could the frontend try to reconnect and write the 8 bytes, ending with the first 5 bytes being transmitted twice?
Meanwhile I'm thinking to document as:
"Returns @len on success, or -1 on error (some data might has been written)".
This make sense to me. It also goes in line with what Daniel mentioned in his reply. I'll make the change with this semantic in mind.
Thanks,
Daniel
Quoted 27 lines
Simpler API could be: "Returns 0 on success, otherwise -errno" but we'd need to rework all the callers.
Quoted 23 lines
Thanks,
Daniel
Quoted 14 lines
Quoted 8 lines
Are you
saying that we should do a loop when there's no chardev found (ret = 0) and wait a certain time until there's one available?In fact, seeing how
SBI_EXT_DBCN_CONSOLE_WRITEis written, I wonder if we could useqemu_chr_fe_write_all()in this case too.This is certainly simpler.
Regards,
Phil.
Hi Daniel,
Quoted 78 lines · Daniel Henrique Barboza
Quoted 75 lines · Philippe Mathieu-Daudé
(+Marc-André and Paolo who I forgot to Cc first)
Quoted 72 lines · Daniel Henrique Barboza
Quoted 69 lines · Philippe Mathieu-Daudé
Quoted 68 lines · Daniel Henrique Barboza
Quoted 65 lines · Philippe Mathieu-Daudé
Hi Daniel,
(now merged as commit
a6b53378f537)Quoted 60 lines · Daniel Henrique Barboza
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get- reg- list is advertising
KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it viakvm_set_one_reg()duringkvm_arch_init_vcpu().Finally, change
kvm_riscv_handle_sbi()to handle the incoming calls forSBI_EXT_DBCN, reading and writing as required.A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
Signed-off-by Daniel Henrique Barboza <[email protected]>---
target/riscv/kvm/kvm-cpu.c | 111 ++++++++++++++++++++++ ++ +++++
target/riscv/sbi_ecall_interface.h| 17 +++++2 files changed, 128 insertions(+)+static void
kvm_riscv_handle_sbi_dbcn(CPUState *cs, struct kvm_run *run)(quoted diff) +27
-0+{ + g_autofree uint8_t *buf = NULL; + RISCVCPU *cpu = RISCV_CPU(cs); + target_ulong num_bytes; + uint64_t addr; + unsigned char ch; + int ret; + + switch (run->riscv_sbi.function_id) { + case SBI_EXT_DBCN_CONSOLE_READ: + case SBI_EXT_DBCN_CONSOLE_WRITE: + num_bytes = run->riscv_sbi.args[0]; + + if (num_bytes == 0) { + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = 0; + break; + } + + addr = run->riscv_sbi.args[1]; + + /* + * Handle the case where a 32 bit CPU is running in a + * 64 bit addressing env. + */ + if (riscv_cpu_mxl(&cpu->env) == MXL_RV32) { + addr |= (uint64_t)run->riscv_sbi.args[2] << 32;
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
+ }
(quoted diff) +4
-0+ + buf = g_malloc0(num_bytes); + + if (run->riscv_sbi.function_id ==
SBI_EXT_DBCN_CONSOLE_READ) {
+ ret =qemu_chr_fe_read_all(serial_hd(0)->be, buf, num_bytes);
Still following up on that patch :) Do we have one SBI debug console per system / soc / cluster / hart? Instead of calling serial_hd(0) each time we should set it once when realizing the vcpu IMO. If it is per hart we would add this property:
DEFINE_PROP_CHR("sbi-chardev", RISCVCPU, sbi_console),
WDYT?
Quoted 34 lines
Quoted 34 lines
Quoted 34 lines
Quoted 34 lines
Quoted 34 lines
Quoted 34 lines
Quoted 34 lines
+ if (ret < 0) {
+ error_report("SBI_EXT_DBCN_CONSOLE_READ: error when "
+ "reading chardev");
+ exit(1);
+ }(quoted diff) +6
-0+ + cpu_physical_memory_write(addr, buf, ret); + } else { + cpu_physical_memory_read(addr, buf, num_bytes); + + ret = qemu_chr_fe_write_all(serial_hd(0)->be, buf,
num_bytes);
+ if (ret < 0) {
+ error_report("SBI_EXT_DBCN_CONSOLE_WRITE: error when "
+ "writing chardev");
+ exit(1);
+ }
+ }(quoted diff) +10
-0+ + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = ret; + break; + case SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: + ch = run->riscv_sbi.args[0]; + ret = qemu_chr_fe_write(serial_hd(0)->be, &ch, sizeof(ch)); + + if (ret < 0) { + error_report("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: errorwhen "
+ "writing chardev");
+ exit(1);
+ }
Quoted 96 lines · Philippe Mathieu-Daudé
Hi Daniel,
Quoted 78 lines · Daniel Henrique Barboza
Quoted 75 lines · Philippe Mathieu-Daudé
(+Marc-André and Paolo who I forgot to Cc first)
Quoted 72 lines · Daniel Henrique Barboza
Quoted 69 lines · Philippe Mathieu-Daudé
Quoted 68 lines · Daniel Henrique Barboza
Quoted 65 lines · Philippe Mathieu-Daudé
Hi Daniel,
(now merged as commit
a6b53378f537)Quoted 60 lines · Daniel Henrique Barboza
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get- reg- list is advertising
KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it viakvm_set_one_reg()duringkvm_arch_init_vcpu().Finally, change
kvm_riscv_handle_sbi()to handle the incoming calls forSBI_EXT_DBCN, reading and writing as required.A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
Signed-off-by Daniel Henrique Barboza <[email protected]>---
target/riscv/kvm/kvm-cpu.c | 111 +++++++++++++++++++++
+ ++ +++++
target/riscv/sbi_ecall_interface.h| 17 +++++2 files changed, 128 insertions(+)+static void
kvm_riscv_handle_sbi_dbcn(CPUState *cs, struct kvm_run *run)(quoted diff) +27
-0+{ + g_autofree uint8_t *buf = NULL; + RISCVCPU *cpu = RISCV_CPU(cs); + target_ulong num_bytes; + uint64_t addr; + unsigned char ch; + int ret; + + switch (run->riscv_sbi.function_id) { + case SBI_EXT_DBCN_CONSOLE_READ: + case SBI_EXT_DBCN_CONSOLE_WRITE: + num_bytes = run->riscv_sbi.args[0]; + + if (num_bytes == 0) { + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = 0; + break; + } + + addr = run->riscv_sbi.args[1]; + + /* + * Handle the case where a 32 bit CPU is running in a + * 64 bit addressing env. + */ + if (riscv_cpu_mxl(&cpu->env) == MXL_RV32) { + addr |= (uint64_t)run->riscv_sbi.args[2] << 32;Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
+ }
(quoted diff) +4
-0+ + buf = g_malloc0(num_bytes); + + if (run->riscv_sbi.function_id ==
SBI_EXT_DBCN_CONSOLE_READ) {
+ ret =qemu_chr_fe_read_all(serial_hd(0)->be, buf, num_bytes);Still following up on that patch :) Do we have one SBI debug console per system / soc / cluster / hart? Instead of calling
serial_hd(0)each time we should set it once when realizing the vcpu IMO. If it is per hart we would add this property:
Quoted 3 lines
DEFINE_PROP_CHR("sbi-chardev", RISCVCPU, sbi_console),WDYT?
We have a single SBI debug console per platform. So, using this patch lingo, all CPUs will use serial_hd(0). In theory this would be a board property, but that makes things messier to deal with - we don't want to keep poking at board properties when initializing CPUs.
So I think the CPU property you suggested works nicely. Thanks,
Daniel
Quoted 36 lines
Quoted 36 lines
Quoted 36 lines
Quoted 36 lines
Quoted 36 lines
Quoted 36 lines
Quoted 36 lines
Quoted 36 lines
+ if (ret < 0) {
+ error_report("SBI_EXT_DBCN_CONSOLE_READ: error when "
+ "reading chardev");
+ exit(1);
+ }(quoted diff) +6
-0+ + cpu_physical_memory_write(addr, buf, ret); + } else { + cpu_physical_memory_read(addr, buf, num_bytes); + + ret = qemu_chr_fe_write_all(serial_hd(0)->be, buf,
num_bytes);
+ if (ret < 0) {
+ error_report("SBI_EXT_DBCN_CONSOLE_WRITE: error when "
+ "writing chardev");
+ exit(1);
+ }
+ }
+
+ run->riscv_sbi.ret[0] =SBI_SUCCESS;
+ run->riscv_sbi.ret[1] = ret;
+ break;
+ caseSBI_EXT_DBCN_CONSOLE_WRITE_BYTE:
+ ch = run->riscv_sbi.args[0];
+ ret =qemu_chr_fe_write(serial_hd(0)->be, &ch, sizeof(ch));
+
+ if (ret < 0) {
+ error_report("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE:
error when "
+ "writing chardev");
+ exit(1);
+ }
Quoted 122 lines · Philippe Mathieu-Daudé
Quoted 110 lines · Daniel Henrique Barboza
Quoted 98 lines · Philippe Mathieu-Daudé
Hi Daniel,
(now merged as commit
a6b53378f537)Quoted 93 lines · Daniel Henrique Barboza
SBI defines a Debug Console extension "DBCN" that will, in time, replace the legacy console putchar and getchar SBI extensions.
The appeal of the DBCN extension is that it allows multiple bytes to be read/written in the SBI console in a single SBI call.
As far as KVM goes, the DBCN calls are forwarded by an in-kernel KVM module to userspace. But this will only happens if the KVM module actually supports this SBI extension and we activate it.
We'll check for DBCN support during init time, checking if get-reg-list is advertising
KVM_RISCV_SBI_EXT_DBCN. In that case, we'll enable it viakvm_set_one_reg()duringkvm_arch_init_vcpu().Finally, change
kvm_riscv_handle_sbi()to handle the incoming calls forSBI_EXT_DBCN, reading and writing as required.A simple KVM guest with 'earlycon=sbi', running in an emulated RISC-V host, takes around 20 seconds to boot without using DBCN. With this patch we're taking around 14 seconds to boot due to the speed-up in the terminal output. There's no change in boot time if the guest isn't using earlycon.
Signed-off-by Daniel Henrique Barboza <[email protected]>---
target/riscv/kvm/kvm-cpu.c | 111 +++++++++++++++++++++++++++++target/riscv/sbi_ecall_interface.h| 17 +++++2 files changed, 128 insertions(+)+static void
kvm_riscv_handle_sbi_dbcn(CPUState *cs, struct kvm_run *run)target/riscv/kvm/kvm-cpu.c +33
-0+{ + g_autofree uint8_t *buf = NULL; + RISCVCPU *cpu = RISCV_CPU(cs); + target_ulong num_bytes; + uint64_t addr; + unsigned char ch; + int ret; + + switch (run->riscv_sbi.function_id) { + case SBI_EXT_DBCN_CONSOLE_READ: + case SBI_EXT_DBCN_CONSOLE_WRITE: + num_bytes = run->riscv_sbi.args[0]; + + if (num_bytes == 0) { + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = 0; + break; + } + + addr = run->riscv_sbi.args[1]; + + /* + * Handle the case where a 32 bit CPU is running in a + * 64 bit addressing env. + */ + if (riscv_cpu_mxl(&cpu->env) == MXL_RV32) { + addr |= (uint64_t)run->riscv_sbi.args[2] << 32; + } + + buf = g_malloc0(num_bytes); + + if (run->riscv_sbi.function_id == SBI_EXT_DBCN_CONSOLE_READ) { + ret = qemu_chr_fe_read_all(serial_hd(0)->be, buf,
num_bytes);
+ if (ret < 0) {
+ error_report("SBI_EXT_DBCN_CONSOLE_READ: error when " + "reading chardev");
+ exit(1);
+ }target/riscv/kvm/kvm-cpu.c +6
-0+ + cpu_physical_memory_write(addr, buf, ret); + } else { + cpu_physical_memory_read(addr, buf, num_bytes); + + ret = qemu_chr_fe_write_all(serial_hd(0)->be, buf,
num_bytes);
+ if (ret < 0) {
+ error_report("SBI_EXT_DBCN_CONSOLE_WRITE: error when " + "writing chardev");
+ exit(1);
+ }
+ }target/riscv/kvm/kvm-cpu.c +9
-0+ + run->riscv_sbi.ret[0] = SBI_SUCCESS; + run->riscv_sbi.ret[1] = ret; + break; + case SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: + ch = run->riscv_sbi.args[0]; + ret = qemu_chr_fe_write(serial_hd(0)->be, &ch, sizeof(ch)); + + if (ret < 0) {Quoted 8 lines
Quoted 4 lines
+ error_report("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: error when " + "writing chardev");
+ exit(1);
+ }We are ignoring partial writes (non-blocking call returning 0 byte written), is that expected? If so, is it OK to add a comment we can safely discard not-yet-written
DBCN_CONSOLE_WRITE_BYTE?Quoted 2 lines
Not sure what you meant. IIUC
qemu_chr_fe_write()returns the number of bytes consumed, 0 if no chardev is found, and -1 on error.I'm trying to address an issue Peter reported with
qemu_chr_fe_write():
https://lore.kernel.org/qemu-devel/CAFEAcA_kEndvNtw4EHySXWwQPoGs029yAzZGGBcV=zGHaj7KUQ@mail.gmail.com/Basically upon introduction in commit
cd18720a294in 2013 ("char: introduce a blocking version ofqemu_chr_fe_write") the API contract was "Returns: the number of bytes consumed" which could be 0, so some frontends return 0 for "wrote no bytes".
Quoted 5 lines
Later in 2016 in commit
fa394ed6257("char: make someqemu_chr_feskip if no driver") the API documentation was changed:- * Returns: the number of bytes consumed
+ * Returns: the number of bytes consumed (0 if no assicated CharDriver)
IMHO those semantics are broken for the write methods.
0 for a write() attempt indicates that the caller must be prepared to retry again later.
0 for a write_all() attempt should not be permitted - write_all() must always write everythnig, or return an error.
When no CharDriver is present, these semantics falls down. We need to be returning "len" when no driver is present, ie no associated CharDriver should be like writing to /dev/null, nor /dev/full - everything should be reported as consumed, but discarded internally.
Quoted 2 lines
After this commit, some frontends started to handle '<=0' as error, while 0 is not an error.
Yep, that's wrong.
With regards,
Daniel
Signature · 3 lines
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|