Re: [PATCH v2 1/2] LoongArch: KVM: Add LSX support
From: maobibo
Date: Tue Dec 05 2023 - 08:26:02 EST
On 2023/12/1 下午4:46, Tianrui Zhao wrote:
This patch adds LSX support for LoongArch KVM.
There will be LSX exception in KVM when guest use the LSX
instruction. KVM will enable LSX and restore the vector
registers for guest then return to guest to continue running.
Signed-off-by: Tianrui Zhao <zhaotianrui@xxxxxxxxxxx>
---
arch/loongarch/include/asm/kvm_host.h | 12 ++-
arch/loongarch/include/asm/kvm_vcpu.h | 12 +++
arch/loongarch/include/uapi/asm/kvm.h | 19 +++--
arch/loongarch/kvm/exit.c | 18 +++++
arch/loongarch/kvm/switch.S | 21 +++++
arch/loongarch/kvm/trace.h | 4 +-
arch/loongarch/kvm/vcpu.c | 109 ++++++++++++++++++++++++--
7 files changed, 179 insertions(+), 16 deletions(-)
diff --git a/arch/loongarch/include/asm/kvm_host.h b/arch/loongarch/include/asm/kvm_host.h
index 11328700d4f..a53b47093f4 100644
--- a/arch/loongarch/include/asm/kvm_host.h
+++ b/arch/loongarch/include/asm/kvm_host.h
@@ -21,8 +21,10 @@
#include <asm/loongarch.h>
/* Loongarch KVM register ids */
-#define KVM_GET_IOC_CSR_IDX(id) ((id & KVM_CSR_IDX_MASK) >> LOONGARCH_REG_SHIFT)
-#define KVM_GET_IOC_CPUCFG_IDX(id) ((id & KVM_CPUCFG_IDX_MASK) >> LOONGARCH_REG_SHIFT)
+#define KVM_GET_IOC_CSR_IDX(id) ((id & KVM_CSR_IDX_MASK) >> LOONGARCH_REG_SHIFT)
+#define KVM_GET_IOC_CPUCFG_IDX(id) ((id & KVM_CPUCFG_IDX_MASK) >> LOONGARCH_REG_SHIFT)
+#define KVM_GET_IOC_CPUCFG_SUPPORTED_IDX(id) ((id & KVM_CPUCFG_SUPPORTED_IDX_MASK) >> \
+ LOONGARCH_REG_SHIFT)
#define KVM_MAX_VCPUS 256
#define KVM_MAX_CPUCFG_REGS 21
@@ -94,6 +96,7 @@ enum emulation_result {
#define KVM_LARCH_FPU (0x1 << 0)
#define KVM_LARCH_SWCSR_LATEST (0x1 << 1)
#define KVM_LARCH_HWCSR_USABLE (0x1 << 2)
+#define KVM_LARCH_LSX (0x1 << 3)
struct kvm_vcpu_arch {
/*
@@ -175,6 +178,11 @@ static inline void writel_sw_gcsr(struct loongarch_csrs *csr, int reg, unsigned
csr->csrs[reg] = val;
}
+static inline bool kvm_guest_has_lsx(struct kvm_vcpu_arch *arch)
+{
+ return arch->cpucfg[2] & CPUCFG2_LSX;
+}
+
/* Debug: dump vcpu state */
int kvm_arch_vcpu_dump_regs(struct kvm_vcpu *vcpu);
diff --git a/arch/loongarch/include/asm/kvm_vcpu.h b/arch/loongarch/include/asm/kvm_vcpu.h
index 553cfa2b2b1..c629771e122 100644
--- a/arch/loongarch/include/asm/kvm_vcpu.h
+++ b/arch/loongarch/include/asm/kvm_vcpu.h
@@ -55,6 +55,18 @@ void kvm_save_fpu(struct loongarch_fpu *fpu);
void kvm_restore_fpu(struct loongarch_fpu *fpu);
void kvm_restore_fcsr(struct loongarch_fpu *fpu);
+#ifdef CONFIG_CPU_HAS_LSX
+void kvm_own_lsx(struct kvm_vcpu *vcpu);
+void kvm_save_lsx(struct loongarch_fpu *fpu);
+void kvm_restore_lsx(struct loongarch_fpu *fpu);
+void kvm_restore_lsx_upper(struct loongarch_fpu *fpu);
+#else
+static inline void kvm_own_lsx(struct kvm_vcpu *vcpu) { }
+static inline void kvm_save_lsx(struct loongarch_fpu *fpu) { }
+static inline void kvm_restore_lsx(struct loongarch_fpu *fpu) { }
+static inline void kvm_restore_lsx_upper(struct loongarch_fpu *fpu) { }
+#endif
+
void kvm_acquire_timer(struct kvm_vcpu *vcpu);
void kvm_init_timer(struct kvm_vcpu *vcpu, unsigned long hz);
void kvm_reset_timer(struct kvm_vcpu *vcpu);
diff --git a/arch/loongarch/include/uapi/asm/kvm.h b/arch/loongarch/include/uapi/asm/kvm.h
index c6ad2ee6106..cfea150be8e 100644
--- a/arch/loongarch/include/uapi/asm/kvm.h
+++ b/arch/loongarch/include/uapi/asm/kvm.h
@@ -59,14 +59,16 @@ struct kvm_fpu {
* have its own identifier in bits[31..16].
*/
-#define KVM_REG_LOONGARCH_GPR (KVM_REG_LOONGARCH | 0x00000ULL)
-#define KVM_REG_LOONGARCH_CSR (KVM_REG_LOONGARCH | 0x10000ULL)
-#define KVM_REG_LOONGARCH_KVM (KVM_REG_LOONGARCH | 0x20000ULL)
-#define KVM_REG_LOONGARCH_FPSIMD (KVM_REG_LOONGARCH | 0x30000ULL)
-#define KVM_REG_LOONGARCH_CPUCFG (KVM_REG_LOONGARCH | 0x40000ULL)
-#define KVM_REG_LOONGARCH_MASK (KVM_REG_LOONGARCH | 0x70000ULL)
-#define KVM_CSR_IDX_MASK 0x7fff
-#define KVM_CPUCFG_IDX_MASK 0x7fff
+#define KVM_REG_LOONGARCH_GPR (KVM_REG_LOONGARCH | 0x00000ULL)
+#define KVM_REG_LOONGARCH_CSR (KVM_REG_LOONGARCH | 0x10000ULL)
+#define KVM_REG_LOONGARCH_KVM (KVM_REG_LOONGARCH | 0x20000ULL)
+#define KVM_REG_LOONGARCH_FPSIMD (KVM_REG_LOONGARCH | 0x30000ULL)
+#define KVM_REG_LOONGARCH_CPUCFG (KVM_REG_LOONGARCH | 0x40000ULL)
+#define KVM_REG_LOONGARCH_CPUCFG_SUPPORTED (KVM_REG_LOONGARCH | 0x50000ULL)
+#define KVM_REG_LOONGARCH_MASK (KVM_REG_LOONGARCH | 0x70000ULL)
+#define KVM_CSR_IDX_MASK 0x7fff
+#define KVM_CPUCFG_IDX_MASK 0x7fff
+#define KVM_CPUCFG_SUPPORTED_IDX_MASK 0x7fff
/*
* KVM_REG_LOONGARCH_KVM - KVM specific control registers.
@@ -79,6 +81,7 @@ struct kvm_fpu {
#define LOONGARCH_REG_64(TYPE, REG) (TYPE | KVM_REG_SIZE_U64 | (REG << LOONGARCH_REG_SHIFT))
#define KVM_IOC_CSRID(REG) LOONGARCH_REG_64(KVM_REG_LOONGARCH_CSR, REG)
#define KVM_IOC_CPUCFG(REG) LOONGARCH_REG_64(KVM_REG_LOONGARCH_CPUCFG, REG)
+#define KVM_IOC_CPUCFG_SUPPORTED(REG) LOONGARCH_REG_64(KVM_REG_LOONGARCH_CPUCFG_SUPPORTED, REG)
struct kvm_debug_exit_arch {
};
diff --git a/arch/loongarch/kvm/exit.c b/arch/loongarch/kvm/exit.c
index ce8de3fa472..1b1c58ccc83 100644
--- a/arch/loongarch/kvm/exit.c
+++ b/arch/loongarch/kvm/exit.c
@@ -659,6 +659,23 @@ static int kvm_handle_fpu_disabled(struct kvm_vcpu *vcpu)
return RESUME_GUEST;
}
+/*
+ * kvm_handle_lsx_disabled() - Guest used LSX while disabled in root.
+ * @vcpu: Virtual CPU context.
+ *
+ * Handle when the guest attempts to use LSX when it is disabled in the root
+ * context.
+ */
+static int kvm_handle_lsx_disabled(struct kvm_vcpu *vcpu)
+{
+ if (!kvm_guest_has_lsx(&vcpu->arch))
+ kvm_queue_exception(vcpu, EXCCODE_INE, 0);
+ else
+ kvm_own_lsx(vcpu);
+
+ return RESUME_GUEST;
+}
+
/*
* LoongArch KVM callback handling for unimplemented guest exiting
*/
@@ -687,6 +704,7 @@ static exit_handle_fn kvm_fault_tables[EXCCODE_INT_START] = {
[EXCCODE_TLBS] = kvm_handle_write_fault,
[EXCCODE_TLBM] = kvm_handle_write_fault,
[EXCCODE_FPDIS] = kvm_handle_fpu_disabled,
+ [EXCCODE_LSXDIS] = kvm_handle_lsx_disabled,
[EXCCODE_GSPR] = kvm_handle_gspr,
};
diff --git a/arch/loongarch/kvm/switch.S b/arch/loongarch/kvm/switch.S
index 0ed9040307b..6c48f7d1ca5 100644
--- a/arch/loongarch/kvm/switch.S
+++ b/arch/loongarch/kvm/switch.S
@@ -245,6 +245,27 @@ SYM_FUNC_START(kvm_restore_fpu)
jr ra
SYM_FUNC_END(kvm_restore_fpu)
+#ifdef CONFIG_CPU_HAS_LSX
+SYM_FUNC_START(kvm_save_lsx)
+ fpu_save_csr a0 t1
+ fpu_save_cc a0 t1 t2
+ lsx_save_data a0 t1
+ jr ra
+SYM_FUNC_END(kvm_save_lsx)
+
+SYM_FUNC_START(kvm_restore_lsx)
+ lsx_restore_data a0 t1
+ fpu_restore_cc a0 t1 t2
+ fpu_restore_csr a0 t1 t2
+ jr ra
+SYM_FUNC_END(kvm_restore_lsx)
+
+SYM_FUNC_START(kvm_restore_lsx_upper)
+ lsx_restore_all_upper a0 t0 t1
+ jr ra
+SYM_FUNC_END(kvm_restore_lsx_upper)
+#endif
+
.section ".rodata"
SYM_DATA(kvm_exception_size, .quad kvm_exc_entry_end - kvm_exc_entry)
SYM_DATA(kvm_enter_guest_size, .quad kvm_enter_guest_end - kvm_enter_guest)
diff --git a/arch/loongarch/kvm/trace.h b/arch/loongarch/kvm/trace.h
index a1e35d65541..7da4e230e89 100644
--- a/arch/loongarch/kvm/trace.h
+++ b/arch/loongarch/kvm/trace.h
@@ -102,6 +102,7 @@ TRACE_EVENT(kvm_exit_gspr,
#define KVM_TRACE_AUX_DISCARD 4
#define KVM_TRACE_AUX_FPU 1
+#define KVM_TRACE_AUX_LSX 2
#define kvm_trace_symbol_aux_op \
{ KVM_TRACE_AUX_SAVE, "save" }, \
@@ -111,7 +112,8 @@ TRACE_EVENT(kvm_exit_gspr,
{ KVM_TRACE_AUX_DISCARD, "discard" }
#define kvm_trace_symbol_aux_state \
- { KVM_TRACE_AUX_FPU, "FPU" }
+ { KVM_TRACE_AUX_FPU, "FPU" }, \
+ { KVM_TRACE_AUX_LSX, "LSX" }
TRACE_EVENT(kvm_aux,
TP_PROTO(struct kvm_vcpu *vcpu, unsigned int op,
diff --git a/arch/loongarch/kvm/vcpu.c b/arch/loongarch/kvm/vcpu.c
index 73d0c2b9c1a..4820c95091f 100644
--- a/arch/loongarch/kvm/vcpu.c
+++ b/arch/loongarch/kvm/vcpu.c
@@ -309,6 +309,34 @@ static int _kvm_setcsr(struct kvm_vcpu *vcpu, unsigned int id, u64 val)
return ret;
}
+static int kvm_get_cpucfg_supported(int id, u64 *v)
+{
+ int ret = 0;
+
+ if (id < 0 && id >= KVM_MAX_CPUCFG_REGS)
+ return -EINVAL;
+
+ switch (id) {
+ /* Only CPUCFG2 checking are supported, we will add all CPUCFGs checking later. */
+ case 2:
+ /* return CPUCFG2 features which have been supported by KVM */
+ *v = CPUCFG2_FP | CPUCFG2_FPSP | CPUCFG2_FPDP |
+ CPUCFG2_FPVERS | CPUCFG2_LLFTP | CPUCFG2_LLFTPREV |
+ CPUCFG2_LAM;
+ /*
+ * if LSX is supported by CPU, it is also supported by KVM,
+ * as we implement it.
+ */
+ if (cpu_has_lsx)
+ *v |= CPUCFG2_LSX;
+ break;
+ default:
+ ret = -EINVAL;
+ break;
+ }
+ return ret;
+}
+
static int kvm_get_one_reg(struct kvm_vcpu *vcpu,
const struct kvm_one_reg *reg, u64 *v)
{
@@ -327,6 +355,10 @@ static int kvm_get_one_reg(struct kvm_vcpu *vcpu,
else
ret = -EINVAL;
break;
+ case KVM_REG_LOONGARCH_CPUCFG_SUPPORTED:
+ id = KVM_GET_IOC_CPUCFG_SUPPORTED_IDX(reg->id);
+ ret = kvm_get_cpucfg_supported(id, v);
+ break;
To be frankly, KVM_REG_LOONGARCH_CPUCFG_SUPPORTED is a little strange, I
prefer to DEVICE_ATTR for vcpu such as KVM_GET_DEVICE_ATTR, since it
exists already and also pv stealtime can be used by this ioctl command.
case KVM_REG_LOONGARCH_KVM:
switch (reg->id) {
case KVM_REG_LOONGARCH_COUNTER:
@@ -365,6 +397,36 @@ static int kvm_get_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg)
return ret;
}
+static int kvm_check_cpucfg(int id, u64 val)
+{
+ u64 mask;
+ int ret = 0;
+
+ if (id < 0 && id >= KVM_MAX_CPUCFG_REGS)
+ return -EINVAL;
+
+ switch (id) {
+ case 2:
+ ret = kvm_get_cpucfg_supported(id, &mask);
+ /*
+ * CPUCFG2 features checking, if features are not supported, return false.
+ * Features version checking, if feature enabled but without version bits
+ * or feature disabled but along with version bits, return false.
+ */
+ if (!ret) {
+ if ((val & ~mask) ||
+ (!!(val & CPUCFG2_FP) != !!(val & CPUCFG2_FPVERS)) ||
+ (!!(val & CPUCFG2_LLFTP) != !!(val & CPUCFG2_LLFTPREV)))
+ ret = -EINVAL;
What is the logic for the above sentences? Is there special logic about
FPVERS?
How about such lines like this?
/*
* Reserved bit for future usage can not be set
* FP and LLFTP must be enabled
*/
if ((val & ~mask) || !(val & CPUCFG2_LLFTP) || !(val & CPUCFG2_FP))
Regards
Bibo Mao
+ } else
+ ret = -EINVAL;
+ break;
+ default:
+ break;
+ }
+ return ret;
+}
+
static int kvm_set_one_reg(struct kvm_vcpu *vcpu,
const struct kvm_one_reg *reg, u64 v)
{
@@ -378,10 +440,10 @@ static int kvm_set_one_reg(struct kvm_vcpu *vcpu,
break;
case KVM_REG_LOONGARCH_CPUCFG:
id = KVM_GET_IOC_CPUCFG_IDX(reg->id);
- if (id >= 0 && id < KVM_MAX_CPUCFG_REGS)
- vcpu->arch.cpucfg[id] = (u32)v;
- else
- ret = -EINVAL;
+ ret = kvm_check_cpucfg(id, v);
+ if (ret)
+ break;
+ vcpu->arch.cpucfg[id] = (u32)v;
break;
case KVM_REG_LOONGARCH_KVM:
switch (reg->id) {
@@ -561,12 +623,49 @@ void kvm_own_fpu(struct kvm_vcpu *vcpu)
preempt_enable();
}
+#ifdef CONFIG_CPU_HAS_LSX
+/* Enable LSX for guest and restore context */
+void kvm_own_lsx(struct kvm_vcpu *vcpu)
+{
+ preempt_disable();
+
+ /* Enable LSX for guest */
+ set_csr_euen(CSR_EUEN_LSXEN | CSR_EUEN_FPEN);
+ switch (vcpu->arch.aux_inuse & KVM_LARCH_FPU) {
+ case KVM_LARCH_FPU:
+ /*
+ * Guest FPU state already loaded,
+ * only restore upper LSX state
+ */
+ kvm_restore_lsx_upper(&vcpu->arch.fpu);
+ break;
+ default:
+ /* Neither FP or LSX already active,
+ * restore full LSX state
+ */
+ kvm_restore_lsx(&vcpu->arch.fpu);
+ break;
+ }
+
+ trace_kvm_aux(vcpu, KVM_TRACE_AUX_RESTORE, KVM_TRACE_AUX_LSX);
+ vcpu->arch.aux_inuse |= KVM_LARCH_LSX | KVM_LARCH_FPU;
+ preempt_enable();
+}
+#endif
+
/* Save context and disable FPU */
void kvm_lose_fpu(struct kvm_vcpu *vcpu)
{
preempt_disable();
- if (vcpu->arch.aux_inuse & KVM_LARCH_FPU) {
+ if (vcpu->arch.aux_inuse & KVM_LARCH_LSX) {
+ kvm_save_lsx(&vcpu->arch.fpu);
+ vcpu->arch.aux_inuse &= ~(KVM_LARCH_LSX | KVM_LARCH_FPU);
+ trace_kvm_aux(vcpu, KVM_TRACE_AUX_SAVE, KVM_TRACE_AUX_LSX);
+
+ /* Disable LSX & FPU */
+ clear_csr_euen(CSR_EUEN_FPEN | CSR_EUEN_LSXEN);
+ } else if (vcpu->arch.aux_inuse & KVM_LARCH_FPU) {
kvm_save_fpu(&vcpu->arch.fpu);
vcpu->arch.aux_inuse &= ~KVM_LARCH_FPU;
trace_kvm_aux(vcpu, KVM_TRACE_AUX_SAVE, KVM_TRACE_AUX_FPU);