This series adds ONE_REG interface for SBI FWFT extension implemented by KVM RISC-V. This was missed out in accepted SBI FWFT patches for KVM RISC-V.
These patches can also be found in the riscv_kvm_fwft_one_reg_v3 branch at: https://github.com/avpatel/linux.git
Changes since v2: - Re-based on latest KVM RISC-V queue - Improved FWFT ONE_REG interface to allow enabling/disabling each FWFT feature from KVM userspace
Changes since v1: - Dropped have_state in PATCH4 as suggested by Drew - Added Drew's Reviewed-by in appropriate patches
Anup Patel (6): RISC-V: KVM: Set initial value of hedeleg in kvm_arch_vcpu_create() RISC-V: KVM: Introduce feature specific reset for SBI FWFT RISC-V: KVM: Introduce optional ONE_REG callbacks for SBI extensions RISC-V: KVM: Move copy_sbi_ext_reg_indices() to SBI implementation RISC-V: KVM: Implement ONE_REG interface for SBI FWFT state KVM: riscv: selftests: Add SBI FWFT to get-reg-list test
arch/riscv/include/asm/kvm_vcpu_sbi.h | 22 +- arch/riscv/include/asm/kvm_vcpu_sbi_fwft.h | 1 + arch/riscv/include/uapi/asm/kvm.h | 15 ++ arch/riscv/kvm/vcpu.c | 3 +- arch/riscv/kvm/vcpu_onereg.c | 60 +---- arch/riscv/kvm/vcpu_sbi.c | 172 +++++++++++-- arch/riscv/kvm/vcpu_sbi_fwft.c | 227 ++++++++++++++++-- arch/riscv/kvm/vcpu_sbi_sta.c | 63 +++-- .../selftests/kvm/riscv/get-reg-list.c | 32 +++ 9 files changed, 467 insertions(+), 128 deletions(-)
The hedeleg may be updated by ONE_REG interface before the VCPU is run at least once hence set the initial value of hedeleg in kvm_arch_vcpu_create() instead of kvm_riscv_vcpu_setup_config().
Signed-off-by: Anup Patel apatel@ventanamicro.com Reviewed-by: Andrew Jones ajones@ventanamicro.com --- arch/riscv/kvm/vcpu.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/arch/riscv/kvm/vcpu.c b/arch/riscv/kvm/vcpu.c index 3ebcfffaa978..47bcf190ccc5 100644 --- a/arch/riscv/kvm/vcpu.c +++ b/arch/riscv/kvm/vcpu.c @@ -133,6 +133,8 @@ int kvm_arch_vcpu_create(struct kvm_vcpu *vcpu)
/* Mark this VCPU never ran */ vcpu->arch.ran_atleast_once = false; + + vcpu->arch.cfg.hedeleg = KVM_HEDELEG_DEFAULT; vcpu->arch.mmu_page_cache.gfp_zero = __GFP_ZERO; bitmap_zero(vcpu->arch.isa, RISCV_ISA_EXT_MAX);
@@ -570,7 +572,6 @@ static void kvm_riscv_vcpu_setup_config(struct kvm_vcpu *vcpu) cfg->hstateen0 |= SMSTATEEN0_SSTATEEN0; }
- cfg->hedeleg = KVM_HEDELEG_DEFAULT; if (vcpu->guest_debug) cfg->hedeleg &= ~BIT(EXC_BREAKPOINT); }
The SBI FWFT feature values must be reset upon VCPU reset so introduce feature specific reset callback for this purpose.
Signed-off-by: Anup Patel apatel@ventanamicro.com Reviewed-by: Andrew Jones ajones@ventanamicro.com --- arch/riscv/kvm/vcpu_sbi_fwft.c | 30 ++++++++++++++++++++++++++++-- 1 file changed, 28 insertions(+), 2 deletions(-)
diff --git a/arch/riscv/kvm/vcpu_sbi_fwft.c b/arch/riscv/kvm/vcpu_sbi_fwft.c index 164a01288b0a..5a3bad0f9330 100644 --- a/arch/riscv/kvm/vcpu_sbi_fwft.c +++ b/arch/riscv/kvm/vcpu_sbi_fwft.c @@ -30,6 +30,13 @@ struct kvm_sbi_fwft_feature { */ bool (*supported)(struct kvm_vcpu *vcpu);
+ /** + * @reset: Reset the feature value irrespective whether feature is supported or not + * + * This callback is mandatory + */ + void (*reset)(struct kvm_vcpu *vcpu); + /** * @set: Set the feature value * @@ -75,6 +82,13 @@ static bool kvm_sbi_fwft_misaligned_delegation_supported(struct kvm_vcpu *vcpu) return misaligned_traps_can_delegate(); }
+static void kvm_sbi_fwft_reset_misaligned_delegation(struct kvm_vcpu *vcpu) +{ + struct kvm_vcpu_config *cfg = &vcpu->arch.cfg; + + cfg->hedeleg &= ~MIS_DELEG; +} + static long kvm_sbi_fwft_set_misaligned_delegation(struct kvm_vcpu *vcpu, struct kvm_sbi_fwft_config *conf, unsigned long value) @@ -124,6 +138,11 @@ static bool kvm_sbi_fwft_pointer_masking_pmlen_supported(struct kvm_vcpu *vcpu) return fwft->have_vs_pmlen_7 || fwft->have_vs_pmlen_16; }
+static void kvm_sbi_fwft_reset_pointer_masking_pmlen(struct kvm_vcpu *vcpu) +{ + vcpu->arch.cfg.henvcfg &= ~ENVCFG_PMM; +} + static long kvm_sbi_fwft_set_pointer_masking_pmlen(struct kvm_vcpu *vcpu, struct kvm_sbi_fwft_config *conf, unsigned long value) @@ -180,6 +199,7 @@ static const struct kvm_sbi_fwft_feature features[] = { { .id = SBI_FWFT_MISALIGNED_EXC_DELEG, .supported = kvm_sbi_fwft_misaligned_delegation_supported, + .reset = kvm_sbi_fwft_reset_misaligned_delegation, .set = kvm_sbi_fwft_set_misaligned_delegation, .get = kvm_sbi_fwft_get_misaligned_delegation, }, @@ -187,6 +207,7 @@ static const struct kvm_sbi_fwft_feature features[] = { { .id = SBI_FWFT_POINTER_MASKING_PMLEN, .supported = kvm_sbi_fwft_pointer_masking_pmlen_supported, + .reset = kvm_sbi_fwft_reset_pointer_masking_pmlen, .set = kvm_sbi_fwft_set_pointer_masking_pmlen, .get = kvm_sbi_fwft_get_pointer_masking_pmlen, }, @@ -321,11 +342,16 @@ static void kvm_sbi_ext_fwft_deinit(struct kvm_vcpu *vcpu)
static void kvm_sbi_ext_fwft_reset(struct kvm_vcpu *vcpu) { - int i; struct kvm_sbi_fwft *fwft = vcpu_to_fwft(vcpu); + const struct kvm_sbi_fwft_feature *feature; + int i;
- for (i = 0; i < ARRAY_SIZE(features); i++) + for (i = 0; i < ARRAY_SIZE(features); i++) { fwft->configs[i].flags = 0; + feature = &features[i]; + if (feature->reset) + feature->reset(vcpu); + } }
const struct kvm_vcpu_sbi_extension vcpu_sbi_ext_fwft = {
SBI extensions can have per-VCPU state which needs to be saved/restored through ONE_REG interface for Guest/VM migration. Introduce optional ONE_REG callbacks for SBI extensions so that ONE_REG implementation for an SBI extenion is part of the extension sources.
Signed-off-by: Anup Patel apatel@ventanamicro.com Reviewed-by: Andrew Jones ajones@ventanamicro.com --- arch/riscv/include/asm/kvm_vcpu_sbi.h | 20 ++-- arch/riscv/kvm/vcpu_onereg.c | 31 +----- arch/riscv/kvm/vcpu_sbi.c | 145 ++++++++++++++++++++++---- arch/riscv/kvm/vcpu_sbi_sta.c | 63 +++++++---- 4 files changed, 176 insertions(+), 83 deletions(-)
diff --git a/arch/riscv/include/asm/kvm_vcpu_sbi.h b/arch/riscv/include/asm/kvm_vcpu_sbi.h index 766031e80960..8970cc7530c4 100644 --- a/arch/riscv/include/asm/kvm_vcpu_sbi.h +++ b/arch/riscv/include/asm/kvm_vcpu_sbi.h @@ -59,6 +59,14 @@ struct kvm_vcpu_sbi_extension { void (*deinit)(struct kvm_vcpu *vcpu);
void (*reset)(struct kvm_vcpu *vcpu); + + unsigned long state_reg_subtype; + unsigned long (*get_state_reg_count)(struct kvm_vcpu *vcpu); + int (*get_state_reg_id)(struct kvm_vcpu *vcpu, int index, u64 *reg_id); + int (*get_state_reg)(struct kvm_vcpu *vcpu, unsigned long reg_num, + unsigned long reg_size, void *reg_val); + int (*set_state_reg)(struct kvm_vcpu *vcpu, unsigned long reg_num, + unsigned long reg_size, const void *reg_val); };
void kvm_riscv_vcpu_sbi_forward(struct kvm_vcpu *vcpu, struct kvm_run *run); @@ -73,10 +81,9 @@ int kvm_riscv_vcpu_set_reg_sbi_ext(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg); int kvm_riscv_vcpu_get_reg_sbi_ext(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg); -int kvm_riscv_vcpu_set_reg_sbi(struct kvm_vcpu *vcpu, - const struct kvm_one_reg *reg); -int kvm_riscv_vcpu_get_reg_sbi(struct kvm_vcpu *vcpu, - const struct kvm_one_reg *reg); +int kvm_riscv_vcpu_reg_indices_sbi(struct kvm_vcpu *vcpu, u64 __user *uindices); +int kvm_riscv_vcpu_set_reg_sbi(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg); +int kvm_riscv_vcpu_get_reg_sbi(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg); const struct kvm_vcpu_sbi_extension *kvm_vcpu_sbi_find_ext( struct kvm_vcpu *vcpu, unsigned long extid); bool riscv_vcpu_supports_sbi_ext(struct kvm_vcpu *vcpu, int idx); @@ -85,11 +92,6 @@ void kvm_riscv_vcpu_sbi_init(struct kvm_vcpu *vcpu); void kvm_riscv_vcpu_sbi_deinit(struct kvm_vcpu *vcpu); void kvm_riscv_vcpu_sbi_reset(struct kvm_vcpu *vcpu);
-int kvm_riscv_vcpu_get_reg_sbi_sta(struct kvm_vcpu *vcpu, unsigned long reg_num, - unsigned long *reg_val); -int kvm_riscv_vcpu_set_reg_sbi_sta(struct kvm_vcpu *vcpu, unsigned long reg_num, - unsigned long reg_val); - #ifdef CONFIG_RISCV_SBI_V01 extern const struct kvm_vcpu_sbi_extension vcpu_sbi_ext_v01; #endif diff --git a/arch/riscv/kvm/vcpu_onereg.c b/arch/riscv/kvm/vcpu_onereg.c index 9f8b196e7697..0f4e444e5e10 100644 --- a/arch/riscv/kvm/vcpu_onereg.c +++ b/arch/riscv/kvm/vcpu_onereg.c @@ -1112,36 +1112,9 @@ static unsigned long num_sbi_ext_regs(struct kvm_vcpu *vcpu) return copy_sbi_ext_reg_indices(vcpu, NULL); }
-static int copy_sbi_reg_indices(struct kvm_vcpu *vcpu, u64 __user *uindices) -{ - struct kvm_vcpu_sbi_context *scontext = &vcpu->arch.sbi_context; - int total = 0; - - if (scontext->ext_status[KVM_RISCV_SBI_EXT_STA] == KVM_RISCV_SBI_EXT_STATUS_ENABLED) { - u64 size = IS_ENABLED(CONFIG_32BIT) ? KVM_REG_SIZE_U32 : KVM_REG_SIZE_U64; - int n = sizeof(struct kvm_riscv_sbi_sta) / sizeof(unsigned long); - - for (int i = 0; i < n; i++) { - u64 reg = KVM_REG_RISCV | size | - KVM_REG_RISCV_SBI_STATE | - KVM_REG_RISCV_SBI_STA | i; - - if (uindices) { - if (put_user(reg, uindices)) - return -EFAULT; - uindices++; - } - } - - total += n; - } - - return total; -} - static inline unsigned long num_sbi_regs(struct kvm_vcpu *vcpu) { - return copy_sbi_reg_indices(vcpu, NULL); + return kvm_riscv_vcpu_reg_indices_sbi(vcpu, NULL); }
static inline unsigned long num_vector_regs(const struct kvm_vcpu *vcpu) @@ -1269,7 +1242,7 @@ int kvm_riscv_vcpu_copy_reg_indices(struct kvm_vcpu *vcpu, return ret; uindices += ret;
- ret = copy_sbi_reg_indices(vcpu, uindices); + ret = kvm_riscv_vcpu_reg_indices_sbi(vcpu, uindices); if (ret < 0) return ret; uindices += ret; diff --git a/arch/riscv/kvm/vcpu_sbi.c b/arch/riscv/kvm/vcpu_sbi.c index 01a93f4fdb16..04903e5012d6 100644 --- a/arch/riscv/kvm/vcpu_sbi.c +++ b/arch/riscv/kvm/vcpu_sbi.c @@ -364,64 +364,163 @@ int kvm_riscv_vcpu_get_reg_sbi_ext(struct kvm_vcpu *vcpu, return 0; }
-int kvm_riscv_vcpu_set_reg_sbi(struct kvm_vcpu *vcpu, - const struct kvm_one_reg *reg) +int kvm_riscv_vcpu_reg_indices_sbi(struct kvm_vcpu *vcpu, u64 __user *uindices) +{ + struct kvm_vcpu_sbi_context *scontext = &vcpu->arch.sbi_context; + const struct kvm_riscv_sbi_extension_entry *entry; + const struct kvm_vcpu_sbi_extension *ext; + unsigned long state_reg_count; + int i, j, rc, count = 0; + u64 reg; + + for (i = 0; i < ARRAY_SIZE(sbi_ext); i++) { + entry = &sbi_ext[i]; + ext = entry->ext_ptr; + + if (!ext->get_state_reg_count || + scontext->ext_status[entry->ext_idx] != KVM_RISCV_SBI_EXT_STATUS_ENABLED) + continue; + + state_reg_count = ext->get_state_reg_count(vcpu); + if (!uindices) + goto skip_put_user; + + for (j = 0; j < state_reg_count; j++) { + if (ext->get_state_reg_id) { + rc = ext->get_state_reg_id(vcpu, j, ®); + if (rc) + return rc; + } else { + reg = KVM_REG_RISCV | + (IS_ENABLED(CONFIG_32BIT) ? + KVM_REG_SIZE_U32 : KVM_REG_SIZE_U64) | + KVM_REG_RISCV_SBI_STATE | + ext->state_reg_subtype | j; + } + + if (put_user(reg, uindices)) + return -EFAULT; + uindices++; + } + +skip_put_user: + count += state_reg_count; + } + + return count; +} + +static const struct kvm_vcpu_sbi_extension *kvm_vcpu_sbi_find_ext_withstate(struct kvm_vcpu *vcpu, + unsigned long subtype) +{ + struct kvm_vcpu_sbi_context *scontext = &vcpu->arch.sbi_context; + const struct kvm_riscv_sbi_extension_entry *entry; + const struct kvm_vcpu_sbi_extension *ext; + int i; + + for (i = 0; i < ARRAY_SIZE(sbi_ext); i++) { + entry = &sbi_ext[i]; + ext = entry->ext_ptr; + + if (ext->get_state_reg_count && + ext->state_reg_subtype == subtype && + scontext->ext_status[entry->ext_idx] == KVM_RISCV_SBI_EXT_STATUS_ENABLED) + return ext; + } + + return NULL; +} + +int kvm_riscv_vcpu_set_reg_sbi(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg) { unsigned long __user *uaddr = (unsigned long __user *)(unsigned long)reg->addr; unsigned long reg_num = reg->id & ~(KVM_REG_ARCH_MASK | KVM_REG_SIZE_MASK | KVM_REG_RISCV_SBI_STATE); - unsigned long reg_subtype, reg_val; - - if (KVM_REG_SIZE(reg->id) != sizeof(unsigned long)) + const struct kvm_vcpu_sbi_extension *ext; + unsigned long reg_subtype; + void *reg_val; + u64 data64; + u32 data32; + u16 data16; + u8 data8; + + switch (KVM_REG_SIZE(reg->id)) { + case 1: + reg_val = &data8; + break; + case 2: + reg_val = &data16; + break; + case 4: + reg_val = &data32; + break; + case 8: + reg_val = &data64; + break; + default: return -EINVAL; + }
- if (copy_from_user(®_val, uaddr, KVM_REG_SIZE(reg->id))) + if (copy_from_user(reg_val, uaddr, KVM_REG_SIZE(reg->id))) return -EFAULT;
reg_subtype = reg_num & KVM_REG_RISCV_SUBTYPE_MASK; reg_num &= ~KVM_REG_RISCV_SUBTYPE_MASK;
- switch (reg_subtype) { - case KVM_REG_RISCV_SBI_STA: - return kvm_riscv_vcpu_set_reg_sbi_sta(vcpu, reg_num, reg_val); - default: + ext = kvm_vcpu_sbi_find_ext_withstate(vcpu, reg_subtype); + if (!ext || !ext->set_state_reg) return -EINVAL; - }
- return 0; + return ext->set_state_reg(vcpu, reg_num, KVM_REG_SIZE(reg->id), reg_val); }
-int kvm_riscv_vcpu_get_reg_sbi(struct kvm_vcpu *vcpu, - const struct kvm_one_reg *reg) +int kvm_riscv_vcpu_get_reg_sbi(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg) { unsigned long __user *uaddr = (unsigned long __user *)(unsigned long)reg->addr; unsigned long reg_num = reg->id & ~(KVM_REG_ARCH_MASK | KVM_REG_SIZE_MASK | KVM_REG_RISCV_SBI_STATE); - unsigned long reg_subtype, reg_val; + const struct kvm_vcpu_sbi_extension *ext; + unsigned long reg_subtype; + void *reg_val; + u64 data64; + u32 data32; + u16 data16; + u8 data8; int ret;
- if (KVM_REG_SIZE(reg->id) != sizeof(unsigned long)) + switch (KVM_REG_SIZE(reg->id)) { + case 1: + reg_val = &data8; + break; + case 2: + reg_val = &data16; + break; + case 4: + reg_val = &data32; + break; + case 8: + reg_val = &data64; + break; + default: return -EINVAL; + }
reg_subtype = reg_num & KVM_REG_RISCV_SUBTYPE_MASK; reg_num &= ~KVM_REG_RISCV_SUBTYPE_MASK;
- switch (reg_subtype) { - case KVM_REG_RISCV_SBI_STA: - ret = kvm_riscv_vcpu_get_reg_sbi_sta(vcpu, reg_num, ®_val); - break; - default: + ext = kvm_vcpu_sbi_find_ext_withstate(vcpu, reg_subtype); + if (!ext || !ext->get_state_reg) return -EINVAL; - }
+ ret = ext->get_state_reg(vcpu, reg_num, KVM_REG_SIZE(reg->id), reg_val); if (ret) return ret;
- if (copy_to_user(uaddr, ®_val, KVM_REG_SIZE(reg->id))) + if (copy_to_user(uaddr, reg_val, KVM_REG_SIZE(reg->id))) return -EFAULT;
return 0; diff --git a/arch/riscv/kvm/vcpu_sbi_sta.c b/arch/riscv/kvm/vcpu_sbi_sta.c index cc6cb7c8f0e4..68486f90a61e 100644 --- a/arch/riscv/kvm/vcpu_sbi_sta.c +++ b/arch/riscv/kvm/vcpu_sbi_sta.c @@ -151,63 +151,82 @@ static unsigned long kvm_sbi_ext_sta_probe(struct kvm_vcpu *vcpu) return !!sched_info_on(); }
-const struct kvm_vcpu_sbi_extension vcpu_sbi_ext_sta = { - .extid_start = SBI_EXT_STA, - .extid_end = SBI_EXT_STA, - .handler = kvm_sbi_ext_sta_handler, - .probe = kvm_sbi_ext_sta_probe, - .reset = kvm_riscv_vcpu_sbi_sta_reset, -}; +static unsigned long kvm_sbi_ext_sta_get_state_reg_count(struct kvm_vcpu *vcpu) +{ + return sizeof(struct kvm_riscv_sbi_sta) / sizeof(unsigned long); +}
-int kvm_riscv_vcpu_get_reg_sbi_sta(struct kvm_vcpu *vcpu, - unsigned long reg_num, - unsigned long *reg_val) +static int kvm_sbi_ext_sta_get_reg(struct kvm_vcpu *vcpu, unsigned long reg_num, + unsigned long reg_size, void *reg_val) { + unsigned long *value; + + if (reg_size != sizeof(unsigned long)) + return -EINVAL; + value = reg_val; + switch (reg_num) { case KVM_REG_RISCV_SBI_STA_REG(shmem_lo): - *reg_val = (unsigned long)vcpu->arch.sta.shmem; + *value = (unsigned long)vcpu->arch.sta.shmem; break; case KVM_REG_RISCV_SBI_STA_REG(shmem_hi): if (IS_ENABLED(CONFIG_32BIT)) - *reg_val = upper_32_bits(vcpu->arch.sta.shmem); + *value = upper_32_bits(vcpu->arch.sta.shmem); else - *reg_val = 0; + *value = 0; break; default: - return -EINVAL; + return -ENOENT; }
return 0; }
-int kvm_riscv_vcpu_set_reg_sbi_sta(struct kvm_vcpu *vcpu, - unsigned long reg_num, - unsigned long reg_val) +static int kvm_sbi_ext_sta_set_reg(struct kvm_vcpu *vcpu, unsigned long reg_num, + unsigned long reg_size, const void *reg_val) { + unsigned long value; + + if (reg_size != sizeof(unsigned long)) + return -EINVAL; + value = *(const unsigned long *)reg_val; + switch (reg_num) { case KVM_REG_RISCV_SBI_STA_REG(shmem_lo): if (IS_ENABLED(CONFIG_32BIT)) { gpa_t hi = upper_32_bits(vcpu->arch.sta.shmem);
- vcpu->arch.sta.shmem = reg_val; + vcpu->arch.sta.shmem = value; vcpu->arch.sta.shmem |= hi << 32; } else { - vcpu->arch.sta.shmem = reg_val; + vcpu->arch.sta.shmem = value; } break; case KVM_REG_RISCV_SBI_STA_REG(shmem_hi): if (IS_ENABLED(CONFIG_32BIT)) { gpa_t lo = lower_32_bits(vcpu->arch.sta.shmem);
- vcpu->arch.sta.shmem = ((gpa_t)reg_val << 32); + vcpu->arch.sta.shmem = ((gpa_t)value << 32); vcpu->arch.sta.shmem |= lo; - } else if (reg_val != 0) { + } else if (value != 0) { return -EINVAL; } break; default: - return -EINVAL; + return -ENOENT; }
return 0; } + +const struct kvm_vcpu_sbi_extension vcpu_sbi_ext_sta = { + .extid_start = SBI_EXT_STA, + .extid_end = SBI_EXT_STA, + .handler = kvm_sbi_ext_sta_handler, + .probe = kvm_sbi_ext_sta_probe, + .reset = kvm_riscv_vcpu_sbi_sta_reset, + .state_reg_subtype = KVM_REG_RISCV_SBI_STA, + .get_state_reg_count = kvm_sbi_ext_sta_get_state_reg_count, + .get_state_reg = kvm_sbi_ext_sta_get_reg, + .set_state_reg = kvm_sbi_ext_sta_set_reg, +};
The ONE_REG handling of SBI extension enable/disable registers and SBI extension state registers is already under SBI implementation. On similar lines, let's move copy_sbi_ext_reg_indices() under SBI implementation.
Signed-off-by: Anup Patel apatel@ventanamicro.com Reviewed-by: Andrew Jones ajones@ventanamicro.com --- arch/riscv/include/asm/kvm_vcpu_sbi.h | 2 +- arch/riscv/kvm/vcpu_onereg.c | 29 ++------------------------- arch/riscv/kvm/vcpu_sbi.c | 27 ++++++++++++++++++++++++- 3 files changed, 29 insertions(+), 29 deletions(-)
diff --git a/arch/riscv/include/asm/kvm_vcpu_sbi.h b/arch/riscv/include/asm/kvm_vcpu_sbi.h index 8970cc7530c4..d75ca45c0152 100644 --- a/arch/riscv/include/asm/kvm_vcpu_sbi.h +++ b/arch/riscv/include/asm/kvm_vcpu_sbi.h @@ -77,6 +77,7 @@ void kvm_riscv_vcpu_sbi_request_reset(struct kvm_vcpu *vcpu, unsigned long pc, unsigned long a1); void kvm_riscv_vcpu_sbi_load_reset_state(struct kvm_vcpu *vcpu); int kvm_riscv_vcpu_sbi_return(struct kvm_vcpu *vcpu, struct kvm_run *run); +int kvm_riscv_vcpu_reg_indices_sbi_ext(struct kvm_vcpu *vcpu, u64 __user *uindices); int kvm_riscv_vcpu_set_reg_sbi_ext(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg); int kvm_riscv_vcpu_get_reg_sbi_ext(struct kvm_vcpu *vcpu, @@ -86,7 +87,6 @@ int kvm_riscv_vcpu_set_reg_sbi(struct kvm_vcpu *vcpu, const struct kvm_one_reg * int kvm_riscv_vcpu_get_reg_sbi(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg); const struct kvm_vcpu_sbi_extension *kvm_vcpu_sbi_find_ext( struct kvm_vcpu *vcpu, unsigned long extid); -bool riscv_vcpu_supports_sbi_ext(struct kvm_vcpu *vcpu, int idx); int kvm_riscv_vcpu_sbi_ecall(struct kvm_vcpu *vcpu, struct kvm_run *run); void kvm_riscv_vcpu_sbi_init(struct kvm_vcpu *vcpu); void kvm_riscv_vcpu_sbi_deinit(struct kvm_vcpu *vcpu); diff --git a/arch/riscv/kvm/vcpu_onereg.c b/arch/riscv/kvm/vcpu_onereg.c index 0f4e444e5e10..865dae903aa0 100644 --- a/arch/riscv/kvm/vcpu_onereg.c +++ b/arch/riscv/kvm/vcpu_onereg.c @@ -1082,34 +1082,9 @@ static inline unsigned long num_isa_ext_regs(const struct kvm_vcpu *vcpu) return copy_isa_ext_reg_indices(vcpu, NULL); }
-static int copy_sbi_ext_reg_indices(struct kvm_vcpu *vcpu, u64 __user *uindices) -{ - unsigned int n = 0; - - for (int i = 0; i < KVM_RISCV_SBI_EXT_MAX; i++) { - u64 size = IS_ENABLED(CONFIG_32BIT) ? - KVM_REG_SIZE_U32 : KVM_REG_SIZE_U64; - u64 reg = KVM_REG_RISCV | size | KVM_REG_RISCV_SBI_EXT | - KVM_REG_RISCV_SBI_SINGLE | i; - - if (!riscv_vcpu_supports_sbi_ext(vcpu, i)) - continue; - - if (uindices) { - if (put_user(reg, uindices)) - return -EFAULT; - uindices++; - } - - n++; - } - - return n; -} - static unsigned long num_sbi_ext_regs(struct kvm_vcpu *vcpu) { - return copy_sbi_ext_reg_indices(vcpu, NULL); + return kvm_riscv_vcpu_reg_indices_sbi_ext(vcpu, NULL); }
static inline unsigned long num_sbi_regs(struct kvm_vcpu *vcpu) @@ -1237,7 +1212,7 @@ int kvm_riscv_vcpu_copy_reg_indices(struct kvm_vcpu *vcpu, return ret; uindices += ret;
- ret = copy_sbi_ext_reg_indices(vcpu, uindices); + ret = kvm_riscv_vcpu_reg_indices_sbi_ext(vcpu, uindices); if (ret < 0) return ret; uindices += ret; diff --git a/arch/riscv/kvm/vcpu_sbi.c b/arch/riscv/kvm/vcpu_sbi.c index 04903e5012d6..1b13623380e1 100644 --- a/arch/riscv/kvm/vcpu_sbi.c +++ b/arch/riscv/kvm/vcpu_sbi.c @@ -110,7 +110,7 @@ riscv_vcpu_get_sbi_ext(struct kvm_vcpu *vcpu, unsigned long idx) return sext; }
-bool riscv_vcpu_supports_sbi_ext(struct kvm_vcpu *vcpu, int idx) +static bool riscv_vcpu_supports_sbi_ext(struct kvm_vcpu *vcpu, int idx) { struct kvm_vcpu_sbi_context *scontext = &vcpu->arch.sbi_context; const struct kvm_riscv_sbi_extension_entry *sext; @@ -288,6 +288,31 @@ static int riscv_vcpu_get_sbi_ext_multi(struct kvm_vcpu *vcpu, return 0; }
+int kvm_riscv_vcpu_reg_indices_sbi_ext(struct kvm_vcpu *vcpu, u64 __user *uindices) +{ + unsigned int n = 0; + + for (int i = 0; i < KVM_RISCV_SBI_EXT_MAX; i++) { + u64 size = IS_ENABLED(CONFIG_32BIT) ? + KVM_REG_SIZE_U32 : KVM_REG_SIZE_U64; + u64 reg = KVM_REG_RISCV | size | KVM_REG_RISCV_SBI_EXT | + KVM_REG_RISCV_SBI_SINGLE | i; + + if (!riscv_vcpu_supports_sbi_ext(vcpu, i)) + continue; + + if (uindices) { + if (put_user(reg, uindices)) + return -EFAULT; + uindices++; + } + + n++; + } + + return n; +} + int kvm_riscv_vcpu_set_reg_sbi_ext(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg) {
The KVM user-space needs a way to save/restore the state of SBI FWFT features so implement SBI extension ONE_REG callbacks.
Signed-off-by: Anup Patel apatel@ventanamicro.com --- arch/riscv/include/asm/kvm_vcpu_sbi_fwft.h | 1 + arch/riscv/include/uapi/asm/kvm.h | 15 ++ arch/riscv/kvm/vcpu_sbi_fwft.c | 197 +++++++++++++++++++-- 3 files changed, 200 insertions(+), 13 deletions(-)
diff --git a/arch/riscv/include/asm/kvm_vcpu_sbi_fwft.h b/arch/riscv/include/asm/kvm_vcpu_sbi_fwft.h index afb6305e46bb..5604cec79902 100644 --- a/arch/riscv/include/asm/kvm_vcpu_sbi_fwft.h +++ b/arch/riscv/include/asm/kvm_vcpu_sbi_fwft.h @@ -16,6 +16,7 @@ struct kvm_sbi_fwft_feature; struct kvm_sbi_fwft_config { const struct kvm_sbi_fwft_feature *feature; bool supported; + bool enabled; unsigned long flags; };
diff --git a/arch/riscv/include/uapi/asm/kvm.h b/arch/riscv/include/uapi/asm/kvm.h index d2ebc4d0be4b..11fe4a89600c 100644 --- a/arch/riscv/include/uapi/asm/kvm.h +++ b/arch/riscv/include/uapi/asm/kvm.h @@ -220,6 +220,18 @@ struct kvm_riscv_sbi_sta { unsigned long shmem_hi; };
+struct kvm_riscv_sbi_fwft_feature { + unsigned long enable; + unsigned long flags; + unsigned long value; +}; + +/* SBI FWFT extension registers for KVM_GET_ONE_REG and KVM_SET_ONE_REG */ +struct kvm_riscv_sbi_fwft { + struct kvm_riscv_sbi_fwft_feature misaligned_deleg; + struct kvm_riscv_sbi_fwft_feature pointer_masking; +}; + /* Possible states for kvm_riscv_timer */ #define KVM_RISCV_TIMER_STATE_OFF 0 #define KVM_RISCV_TIMER_STATE_ON 1 @@ -303,6 +315,9 @@ struct kvm_riscv_sbi_sta { #define KVM_REG_RISCV_SBI_STA (0x0 << KVM_REG_RISCV_SUBTYPE_SHIFT) #define KVM_REG_RISCV_SBI_STA_REG(name) \ (offsetof(struct kvm_riscv_sbi_sta, name) / sizeof(unsigned long)) +#define KVM_REG_RISCV_SBI_FWFT (0x1 << KVM_REG_RISCV_SUBTYPE_SHIFT) +#define KVM_REG_RISCV_SBI_FWFT_REG(name) \ + (offsetof(struct kvm_riscv_sbi_fwft, name) / sizeof(unsigned long))
/* Device Control API: RISC-V AIA */ #define KVM_DEV_RISCV_APLIC_ALIGN 0x1000 diff --git a/arch/riscv/kvm/vcpu_sbi_fwft.c b/arch/riscv/kvm/vcpu_sbi_fwft.c index 5a3bad0f9330..cacb3d4410a5 100644 --- a/arch/riscv/kvm/vcpu_sbi_fwft.c +++ b/arch/riscv/kvm/vcpu_sbi_fwft.c @@ -22,6 +22,11 @@ struct kvm_sbi_fwft_feature { */ enum sbi_fwft_feature_t id;
+ /** + * @first_reg_num: ONE_REG index of the first ONE_REG register + */ + unsigned long first_reg_num; + /** * @supported: Check if the feature is supported on the vcpu * @@ -44,7 +49,8 @@ struct kvm_sbi_fwft_feature { * * This callback is mandatory */ - long (*set)(struct kvm_vcpu *vcpu, struct kvm_sbi_fwft_config *conf, unsigned long value); + long (*set)(struct kvm_vcpu *vcpu, struct kvm_sbi_fwft_config *conf, + bool one_reg_access, unsigned long value);
/** * @get: Get the feature current value @@ -53,7 +59,8 @@ struct kvm_sbi_fwft_feature { * * This callback is mandatory */ - long (*get)(struct kvm_vcpu *vcpu, struct kvm_sbi_fwft_config *conf, unsigned long *value); + long (*get)(struct kvm_vcpu *vcpu, struct kvm_sbi_fwft_config *conf, + bool one_reg_access, unsigned long *value); };
static const enum sbi_fwft_feature_t kvm_fwft_defined_features[] = { @@ -91,16 +98,18 @@ static void kvm_sbi_fwft_reset_misaligned_delegation(struct kvm_vcpu *vcpu)
static long kvm_sbi_fwft_set_misaligned_delegation(struct kvm_vcpu *vcpu, struct kvm_sbi_fwft_config *conf, - unsigned long value) + bool one_reg_access, unsigned long value) { struct kvm_vcpu_config *cfg = &vcpu->arch.cfg;
if (value == 1) { cfg->hedeleg |= MIS_DELEG; - csr_set(CSR_HEDELEG, MIS_DELEG); + if (!one_reg_access) + csr_set(CSR_HEDELEG, MIS_DELEG); } else if (value == 0) { cfg->hedeleg &= ~MIS_DELEG; - csr_clear(CSR_HEDELEG, MIS_DELEG); + if (!one_reg_access) + csr_clear(CSR_HEDELEG, MIS_DELEG); } else { return SBI_ERR_INVALID_PARAM; } @@ -110,10 +119,11 @@ static long kvm_sbi_fwft_set_misaligned_delegation(struct kvm_vcpu *vcpu,
static long kvm_sbi_fwft_get_misaligned_delegation(struct kvm_vcpu *vcpu, struct kvm_sbi_fwft_config *conf, - unsigned long *value) + bool one_reg_access, unsigned long *value) { - *value = (csr_read(CSR_HEDELEG) & MIS_DELEG) == MIS_DELEG; + struct kvm_vcpu_config *cfg = &vcpu->arch.cfg;
+ *value = (cfg->hedeleg & MIS_DELEG) == MIS_DELEG; return SBI_SUCCESS; }
@@ -145,7 +155,7 @@ static void kvm_sbi_fwft_reset_pointer_masking_pmlen(struct kvm_vcpu *vcpu)
static long kvm_sbi_fwft_set_pointer_masking_pmlen(struct kvm_vcpu *vcpu, struct kvm_sbi_fwft_config *conf, - unsigned long value) + bool one_reg_access, unsigned long value) { struct kvm_sbi_fwft *fwft = vcpu_to_fwft(vcpu); unsigned long pmm; @@ -167,14 +177,15 @@ static long kvm_sbi_fwft_set_pointer_masking_pmlen(struct kvm_vcpu *vcpu, * update here so that VCPU see's pointer masking mode change * immediately. */ - csr_write(CSR_HENVCFG, vcpu->arch.cfg.henvcfg); + if (!one_reg_access) + csr_write(CSR_HENVCFG, vcpu->arch.cfg.henvcfg);
return SBI_SUCCESS; }
static long kvm_sbi_fwft_get_pointer_masking_pmlen(struct kvm_vcpu *vcpu, struct kvm_sbi_fwft_config *conf, - unsigned long *value) + bool one_reg_access, unsigned long *value) { switch (vcpu->arch.cfg.henvcfg & ENVCFG_PMM) { case ENVCFG_PMM_PMLEN_0: @@ -198,6 +209,8 @@ static long kvm_sbi_fwft_get_pointer_masking_pmlen(struct kvm_vcpu *vcpu, static const struct kvm_sbi_fwft_feature features[] = { { .id = SBI_FWFT_MISALIGNED_EXC_DELEG, + .first_reg_num = offsetof(struct kvm_riscv_sbi_fwft, misaligned_deleg.enable) / + sizeof(unsigned long), .supported = kvm_sbi_fwft_misaligned_delegation_supported, .reset = kvm_sbi_fwft_reset_misaligned_delegation, .set = kvm_sbi_fwft_set_misaligned_delegation, @@ -206,6 +219,8 @@ static const struct kvm_sbi_fwft_feature features[] = { #ifndef CONFIG_32BIT { .id = SBI_FWFT_POINTER_MASKING_PMLEN, + .first_reg_num = offsetof(struct kvm_riscv_sbi_fwft, pointer_masking.enable) / + sizeof(unsigned long), .supported = kvm_sbi_fwft_pointer_masking_pmlen_supported, .reset = kvm_sbi_fwft_reset_pointer_masking_pmlen, .set = kvm_sbi_fwft_set_pointer_masking_pmlen, @@ -214,6 +229,20 @@ static const struct kvm_sbi_fwft_feature features[] = { #endif };
+static const struct kvm_sbi_fwft_feature *kvm_sbi_fwft_regnum_to_feature(unsigned long reg_num) +{ + const struct kvm_sbi_fwft_feature *feature; + int i; + + for (i = 0; i < ARRAY_SIZE(features); i++) { + feature = &features[i]; + if (feature->first_reg_num <= reg_num && reg_num < (feature->first_reg_num + 3)) + return feature; + } + + return NULL; +} + static struct kvm_sbi_fwft_config * kvm_sbi_fwft_get_config(struct kvm_vcpu *vcpu, enum sbi_fwft_feature_t feature) { @@ -241,7 +270,7 @@ static int kvm_fwft_get_feature(struct kvm_vcpu *vcpu, u32 feature, return SBI_ERR_DENIED; }
- if (!tconf->supported) + if (!tconf->supported || !tconf->enabled) return SBI_ERR_NOT_SUPPORTED;
*conf = tconf; @@ -267,7 +296,7 @@ static int kvm_sbi_fwft_set(struct kvm_vcpu *vcpu, u32 feature,
conf->flags = flags;
- return conf->feature->set(vcpu, conf, value); + return conf->feature->set(vcpu, conf, false, value); }
static int kvm_sbi_fwft_get(struct kvm_vcpu *vcpu, unsigned long feature, @@ -280,7 +309,7 @@ static int kvm_sbi_fwft_get(struct kvm_vcpu *vcpu, unsigned long feature, if (ret) return ret;
- return conf->feature->get(vcpu, conf, value); + return conf->feature->get(vcpu, conf, false, value); }
static int kvm_sbi_ext_fwft_handler(struct kvm_vcpu *vcpu, struct kvm_run *run, @@ -327,6 +356,7 @@ static int kvm_sbi_ext_fwft_init(struct kvm_vcpu *vcpu) else conf->supported = true;
+ conf->enabled = conf->supported; conf->feature = feature; }
@@ -354,6 +384,142 @@ static void kvm_sbi_ext_fwft_reset(struct kvm_vcpu *vcpu) } }
+static unsigned long kvm_sbi_ext_fwft_get_reg_count(struct kvm_vcpu *vcpu) +{ + unsigned long max_reg_count = sizeof(struct kvm_riscv_sbi_fwft) / sizeof(unsigned long); + const struct kvm_sbi_fwft_feature *feature; + struct kvm_sbi_fwft_config *conf; + unsigned long reg, ret = 0; + + for (reg = 0; reg < max_reg_count; reg++) { + feature = kvm_sbi_fwft_regnum_to_feature(reg); + if (!feature) + continue; + + conf = kvm_sbi_fwft_get_config(vcpu, feature->id); + if (!conf || !conf->supported) + continue; + + ret++; + } + + return ret; +} + +static int kvm_sbi_ext_fwft_get_reg_id(struct kvm_vcpu *vcpu, int index, u64 *reg_id) +{ + int reg, max_reg_count = sizeof(struct kvm_riscv_sbi_fwft) / sizeof(unsigned long); + const struct kvm_sbi_fwft_feature *feature; + struct kvm_sbi_fwft_config *conf; + int idx = 0; + + for (reg = 0; reg < max_reg_count; reg++) { + feature = kvm_sbi_fwft_regnum_to_feature(reg); + if (!feature) + continue; + + conf = kvm_sbi_fwft_get_config(vcpu, feature->id); + if (!conf || !conf->supported) + continue; + + if (index == idx) { + *reg_id = KVM_REG_RISCV | + (IS_ENABLED(CONFIG_32BIT) ? + KVM_REG_SIZE_U32 : KVM_REG_SIZE_U64) | + KVM_REG_RISCV_SBI_STATE | + KVM_REG_RISCV_SBI_FWFT | reg; + return 0; + } + + idx++; + } + + return -ENOENT; +} + +static int kvm_sbi_ext_fwft_get_reg(struct kvm_vcpu *vcpu, unsigned long reg_num, + unsigned long reg_size, void *reg_val) +{ + const struct kvm_sbi_fwft_feature *feature; + struct kvm_sbi_fwft_config *conf; + unsigned long *value; + int ret = 0; + + if (reg_size != sizeof(unsigned long)) + return -EINVAL; + value = reg_val; + + feature = kvm_sbi_fwft_regnum_to_feature(reg_num); + if (!feature) + return -ENOENT; + + conf = kvm_sbi_fwft_get_config(vcpu, feature->id); + if (!conf || !conf->supported) + return -ENOENT; + + switch (reg_num - feature->first_reg_num) { + case 0: + *value = conf->enabled; + break; + case 1: + *value = conf->flags; + break; + case 2: + ret = conf->feature->get(vcpu, conf, true, value); + break; + default: + return -ENOENT; + } + + return sbi_err_map_linux_errno(ret); +} + +static int kvm_sbi_ext_fwft_set_reg(struct kvm_vcpu *vcpu, unsigned long reg_num, + unsigned long reg_size, const void *reg_val) +{ + const struct kvm_sbi_fwft_feature *feature; + struct kvm_sbi_fwft_config *conf; + unsigned long value; + int ret = 0; + + if (reg_size != sizeof(unsigned long)) + return -EINVAL; + value = *(const unsigned long *)reg_val; + + feature = kvm_sbi_fwft_regnum_to_feature(reg_num); + if (!feature) + return -ENOENT; + + conf = kvm_sbi_fwft_get_config(vcpu, feature->id); + if (!conf || !conf->supported) + return -ENOENT; + + switch (reg_num - feature->first_reg_num) { + case 0: + switch (value) { + case 0: + conf->enabled = false; + break; + case 1: + conf->enabled = true; + break; + default: + return -EINVAL; + } + break; + case 1: + conf->flags = value & SBI_FWFT_SET_FLAG_LOCK; + break; + case 2: + ret = conf->feature->set(vcpu, conf, true, value); + break; + default: + return -ENOENT; + } + + return sbi_err_map_linux_errno(ret); +} + const struct kvm_vcpu_sbi_extension vcpu_sbi_ext_fwft = { .extid_start = SBI_EXT_FWFT, .extid_end = SBI_EXT_FWFT, @@ -361,4 +527,9 @@ const struct kvm_vcpu_sbi_extension vcpu_sbi_ext_fwft = { .init = kvm_sbi_ext_fwft_init, .deinit = kvm_sbi_ext_fwft_deinit, .reset = kvm_sbi_ext_fwft_reset, + .state_reg_subtype = KVM_REG_RISCV_SBI_FWFT, + .get_state_reg_count = kvm_sbi_ext_fwft_get_reg_count, + .get_state_reg_id = kvm_sbi_ext_fwft_get_reg_id, + .get_state_reg = kvm_sbi_ext_fwft_get_reg, + .set_state_reg = kvm_sbi_ext_fwft_set_reg, };
On Sat, Aug 23, 2025 at 09:29:46PM +0530, Anup Patel wrote:
The KVM user-space needs a way to save/restore the state of SBI FWFT features so implement SBI extension ONE_REG callbacks.
Signed-off-by: Anup Patel apatel@ventanamicro.com
arch/riscv/include/asm/kvm_vcpu_sbi_fwft.h | 1 + arch/riscv/include/uapi/asm/kvm.h | 15 ++ arch/riscv/kvm/vcpu_sbi_fwft.c | 197 +++++++++++++++++++-- 3 files changed, 200 insertions(+), 13 deletions(-)
Reviewed-by: Andrew Jones ajones@ventanamicro.com
KVM RISC-V now supports SBI FWFT, so add it to the get-reg-list test.
Signed-off-by: Anup Patel apatel@ventanamicro.com --- .../selftests/kvm/riscv/get-reg-list.c | 32 +++++++++++++++++++ 1 file changed, 32 insertions(+)
diff --git a/tools/testing/selftests/kvm/riscv/get-reg-list.c b/tools/testing/selftests/kvm/riscv/get-reg-list.c index 34456e8cba02..705ab3d7778b 100644 --- a/tools/testing/selftests/kvm/riscv/get-reg-list.c +++ b/tools/testing/selftests/kvm/riscv/get-reg-list.c @@ -132,6 +132,7 @@ bool filter_reg(__u64 reg) case KVM_REG_RISCV_SBI_EXT | KVM_REG_RISCV_SBI_SINGLE | KVM_RISCV_SBI_EXT_DBCN: case KVM_REG_RISCV_SBI_EXT | KVM_REG_RISCV_SBI_SINGLE | KVM_RISCV_SBI_EXT_SUSP: case KVM_REG_RISCV_SBI_EXT | KVM_REG_RISCV_SBI_SINGLE | KVM_RISCV_SBI_EXT_STA: + case KVM_REG_RISCV_SBI_EXT | KVM_REG_RISCV_SBI_SINGLE | KVM_RISCV_SBI_EXT_FWFT: case KVM_REG_RISCV_SBI_EXT | KVM_REG_RISCV_SBI_SINGLE | KVM_RISCV_SBI_EXT_EXPERIMENTAL: case KVM_REG_RISCV_SBI_EXT | KVM_REG_RISCV_SBI_SINGLE | KVM_RISCV_SBI_EXT_VENDOR: return true; @@ -637,6 +638,7 @@ static const char *sbi_ext_single_id_to_str(__u64 reg_off) KVM_SBI_EXT_ARR(KVM_RISCV_SBI_EXT_DBCN), KVM_SBI_EXT_ARR(KVM_RISCV_SBI_EXT_SUSP), KVM_SBI_EXT_ARR(KVM_RISCV_SBI_EXT_STA), + KVM_SBI_EXT_ARR(KVM_RISCV_SBI_EXT_FWFT), KVM_SBI_EXT_ARR(KVM_RISCV_SBI_EXT_EXPERIMENTAL), KVM_SBI_EXT_ARR(KVM_RISCV_SBI_EXT_VENDOR), }; @@ -693,6 +695,19 @@ static const char *sbi_sta_id_to_str(__u64 reg_off) return strdup_printf("KVM_REG_RISCV_SBI_STA | %lld /* UNKNOWN */", reg_off); }
+static const char *sbi_fwft_id_to_str(__u64 reg_off) +{ + switch (reg_off) { + case 0: return "KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(misaligned_deleg.enable)"; + case 1: return "KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(misaligned_deleg.flags)"; + case 2: return "KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(misaligned_deleg.value)"; + case 3: return "KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(pointer_masking.enable)"; + case 4: return "KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(pointer_masking.flags)"; + case 5: return "KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(pointer_masking.value)"; + } + return strdup_printf("KVM_REG_RISCV_SBI_FWFT | %lld /* UNKNOWN */", reg_off); +} + static const char *sbi_id_to_str(const char *prefix, __u64 id) { __u64 reg_off = id & ~(REG_MASK | KVM_REG_RISCV_SBI_STATE); @@ -705,6 +720,8 @@ static const char *sbi_id_to_str(const char *prefix, __u64 id) switch (reg_subtype) { case KVM_REG_RISCV_SBI_STA: return sbi_sta_id_to_str(reg_off); + case KVM_REG_RISCV_SBI_FWFT: + return sbi_fwft_id_to_str(reg_off); }
return strdup_printf("%lld | %lld /* UNKNOWN */", reg_subtype, reg_off); @@ -872,6 +889,16 @@ static __u64 sbi_sta_regs[] = { KVM_REG_RISCV | KVM_REG_SIZE_ULONG | KVM_REG_RISCV_SBI_STATE | KVM_REG_RISCV_SBI_STA | KVM_REG_RISCV_SBI_STA_REG(shmem_hi), };
+static __u64 sbi_fwft_regs[] = { + KVM_REG_RISCV | KVM_REG_SIZE_ULONG | KVM_REG_RISCV_SBI_EXT | KVM_REG_RISCV_SBI_SINGLE | KVM_RISCV_SBI_EXT_FWFT, + KVM_REG_RISCV | KVM_REG_SIZE_ULONG | KVM_REG_RISCV_SBI_STATE | KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(misaligned_deleg.enable), + KVM_REG_RISCV | KVM_REG_SIZE_ULONG | KVM_REG_RISCV_SBI_STATE | KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(misaligned_deleg.flags), + KVM_REG_RISCV | KVM_REG_SIZE_ULONG | KVM_REG_RISCV_SBI_STATE | KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(misaligned_deleg.value), + KVM_REG_RISCV | KVM_REG_SIZE_ULONG | KVM_REG_RISCV_SBI_STATE | KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(pointer_masking.enable), + KVM_REG_RISCV | KVM_REG_SIZE_ULONG | KVM_REG_RISCV_SBI_STATE | KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(pointer_masking.flags), + KVM_REG_RISCV | KVM_REG_SIZE_ULONG | KVM_REG_RISCV_SBI_STATE | KVM_REG_RISCV_SBI_FWFT | KVM_REG_RISCV_SBI_FWFT_REG(pointer_masking.value), +}; + static __u64 zicbom_regs[] = { KVM_REG_RISCV | KVM_REG_SIZE_ULONG | KVM_REG_RISCV_CONFIG | KVM_REG_RISCV_CONFIG_REG(zicbom_block_size), KVM_REG_RISCV | KVM_REG_SIZE_ULONG | KVM_REG_RISCV_ISA_EXT | KVM_REG_RISCV_ISA_SINGLE | KVM_RISCV_ISA_EXT_ZICBOM, @@ -1028,6 +1055,9 @@ static __u64 vector_regs[] = { #define SUBLIST_SBI_STA \ {"sbi-sta", .feature_type = VCPU_FEATURE_SBI_EXT, .feature = KVM_RISCV_SBI_EXT_STA, \ .regs = sbi_sta_regs, .regs_n = ARRAY_SIZE(sbi_sta_regs),} +#define SUBLIST_SBI_FWFT \ + {"sbi-fwft", .feature_type = VCPU_FEATURE_SBI_EXT, .feature = KVM_RISCV_SBI_EXT_FWFT, \ + .regs = sbi_fwft_regs, .regs_n = ARRAY_SIZE(sbi_fwft_regs),} #define SUBLIST_ZICBOM \ {"zicbom", .feature = KVM_RISCV_ISA_EXT_ZICBOM, .regs = zicbom_regs, .regs_n = ARRAY_SIZE(zicbom_regs),} #define SUBLIST_ZICBOP \ @@ -1112,6 +1142,7 @@ KVM_SBI_EXT_SUBLIST_CONFIG(sta, STA); KVM_SBI_EXT_SIMPLE_CONFIG(pmu, PMU); KVM_SBI_EXT_SIMPLE_CONFIG(dbcn, DBCN); KVM_SBI_EXT_SIMPLE_CONFIG(susp, SUSP); +KVM_SBI_EXT_SUBLIST_CONFIG(fwft, FWFT);
KVM_ISA_EXT_SUBLIST_CONFIG(aia, AIA); KVM_ISA_EXT_SUBLIST_CONFIG(fp_f, FP_F); @@ -1191,6 +1222,7 @@ struct vcpu_reg_list *vcpu_configs[] = { &config_sbi_pmu, &config_sbi_dbcn, &config_sbi_susp, + &config_sbi_fwft, &config_aia, &config_fp_f, &config_fp_d,
On Sat, Aug 23, 2025 at 09:29:47PM +0530, Anup Patel wrote:
KVM RISC-V now supports SBI FWFT, so add it to the get-reg-list test.
Signed-off-by: Anup Patel apatel@ventanamicro.com
.../selftests/kvm/riscv/get-reg-list.c | 32 +++++++++++++++++++ 1 file changed, 32 insertions(+)
Reviewed-by: Andrew Jones ajones@ventanamicro.com
On Sat, Aug 23, 2025 at 9:30 PM Anup Patel apatel@ventanamicro.com wrote:
This series adds ONE_REG interface for SBI FWFT extension implemented by KVM RISC-V. This was missed out in accepted SBI FWFT patches for KVM RISC-V.
These patches can also be found in the riscv_kvm_fwft_one_reg_v3 branch at: https://github.com/avpatel/linux.git
Changes since v2:
- Re-based on latest KVM RISC-V queue
- Improved FWFT ONE_REG interface to allow enabling/disabling each FWFT feature from KVM userspace
Changes since v1:
- Dropped have_state in PATCH4 as suggested by Drew
- Added Drew's Reviewed-by in appropriate patches
Anup Patel (6): RISC-V: KVM: Set initial value of hedeleg in kvm_arch_vcpu_create() RISC-V: KVM: Introduce feature specific reset for SBI FWFT RISC-V: KVM: Introduce optional ONE_REG callbacks for SBI extensions RISC-V: KVM: Move copy_sbi_ext_reg_indices() to SBI implementation RISC-V: KVM: Implement ONE_REG interface for SBI FWFT state KVM: riscv: selftests: Add SBI FWFT to get-reg-list test
Queued this series for Linux-6.18
Regards, Anup
arch/riscv/include/asm/kvm_vcpu_sbi.h | 22 +- arch/riscv/include/asm/kvm_vcpu_sbi_fwft.h | 1 + arch/riscv/include/uapi/asm/kvm.h | 15 ++ arch/riscv/kvm/vcpu.c | 3 +- arch/riscv/kvm/vcpu_onereg.c | 60 +---- arch/riscv/kvm/vcpu_sbi.c | 172 +++++++++++-- arch/riscv/kvm/vcpu_sbi_fwft.c | 227 ++++++++++++++++-- arch/riscv/kvm/vcpu_sbi_sta.c | 63 +++-- .../selftests/kvm/riscv/get-reg-list.c | 32 +++ 9 files changed, 467 insertions(+), 128 deletions(-)
-- 2.43.0
linux-kselftest-mirror@lists.linaro.org
-
Andrew Jones -
Anup Patel -
Anup Patel