| Message ID | 20240229010130.1380926-11-atishp@rivosinc.com |
|---|---|
| State | Superseded |
| Headers | show |
| Series | RISC-V SBI v2.0 PMU improvements and Perf sampling in KVM guest | expand |
On Wed, Feb 28, 2024 at 05:01:25PM -0800, Atish Patra wrote: > The SBI v2.0 introduced a fw_read_hi function to read 64 bit firmware > counters for RV32 based systems. > > Add infrastructure to support that. > > Reviewed-by: Anup Patel <anup@brainfault.org> > Signed-off-by: Atish Patra <atishp@rivosinc.com> > --- > arch/riscv/include/asm/kvm_vcpu_pmu.h | 4 ++- > arch/riscv/kvm/vcpu_pmu.c | 37 ++++++++++++++++++++++++++- > arch/riscv/kvm/vcpu_sbi_pmu.c | 6 +++++ > 3 files changed, 45 insertions(+), 2 deletions(-) > > diff --git a/arch/riscv/include/asm/kvm_vcpu_pmu.h b/arch/riscv/include/asm/kvm_vcpu_pmu.h > index 8cb21a4f862c..e0ad27dea46c 100644 > --- a/arch/riscv/include/asm/kvm_vcpu_pmu.h > +++ b/arch/riscv/include/asm/kvm_vcpu_pmu.h > @@ -20,7 +20,7 @@ static_assert(RISCV_KVM_MAX_COUNTERS <= 64); > > struct kvm_fw_event { > /* Current value of the event */ > - unsigned long value; > + u64 value; > > /* Event monitoring status */ > bool started; > @@ -91,6 +91,8 @@ int kvm_riscv_vcpu_pmu_ctr_cfg_match(struct kvm_vcpu *vcpu, unsigned long ctr_ba > struct kvm_vcpu_sbi_return *retdata); > int kvm_riscv_vcpu_pmu_ctr_read(struct kvm_vcpu *vcpu, unsigned long cidx, > struct kvm_vcpu_sbi_return *retdata); > +int kvm_riscv_vcpu_pmu_fw_ctr_read_hi(struct kvm_vcpu *vcpu, unsigned long cidx, > + struct kvm_vcpu_sbi_return *retdata); > void kvm_riscv_vcpu_pmu_init(struct kvm_vcpu *vcpu); > int kvm_riscv_vcpu_pmu_setup_snapshot(struct kvm_vcpu *vcpu, unsigned long saddr_low, > unsigned long saddr_high, unsigned long flags, > diff --git a/arch/riscv/kvm/vcpu_pmu.c b/arch/riscv/kvm/vcpu_pmu.c > index a02f7b981005..469bb430cf97 100644 > --- a/arch/riscv/kvm/vcpu_pmu.c > +++ b/arch/riscv/kvm/vcpu_pmu.c > @@ -196,6 +196,29 @@ static int pmu_get_pmc_index(struct kvm_pmu *pmu, unsigned long eidx, > return kvm_pmu_get_programmable_pmc_index(pmu, eidx, cbase, cmask); > } > > +static int pmu_fw_ctr_read_hi(struct kvm_vcpu *vcpu, unsigned long cidx, > + unsigned long *out_val) > +{ > + struct kvm_pmu *kvpmu = vcpu_to_pmu(vcpu); > + struct kvm_pmc *pmc; > + int fevent_code; > + > + if (!IS_ENABLED(CONFIG_32BIT)) Let's remove the CONFIG_32BIT check in kvm_sbi_ext_pmu_handler() and then set *out_val to zero here and return success. Either that, or we should WARN or something here since it's a KVM bug to get here with !CONFIG_32BIT. > + return -EINVAL; > + > + pmc = &kvpmu->pmc[cidx]; Uh oh! We're missing range validation of cidx! And I see we're missing it in pmu_ctr_read() too. We need the same check we have in kvm_riscv_vcpu_pmu_ctr_info(). I think the other SBI functions are OK, but it's worth a triple check. > + > + if (pmc->cinfo.type != SBI_PMU_CTR_TYPE_FW) > + return -EINVAL; > + > + fevent_code = get_event_code(pmc->event_idx); > + pmc->counter_val = kvpmu->fw_event[fevent_code].value; > + > + *out_val = pmc->counter_val >> 32; > + > + return 0; > +} > + > static int pmu_ctr_read(struct kvm_vcpu *vcpu, unsigned long cidx, > unsigned long *out_val) > { > @@ -702,6 +725,18 @@ int kvm_riscv_vcpu_pmu_ctr_cfg_match(struct kvm_vcpu *vcpu, unsigned long ctr_ba > return 0; > } > > +int kvm_riscv_vcpu_pmu_fw_ctr_read_hi(struct kvm_vcpu *vcpu, unsigned long cidx, > + struct kvm_vcpu_sbi_return *retdata) > +{ > + int ret; > + > + ret = pmu_fw_ctr_read_hi(vcpu, cidx, &retdata->out_val); > + if (ret == -EINVAL) > + retdata->err_val = SBI_ERR_INVALID_PARAM; > + > + return 0; I see this follows the pattern we have with kvm_riscv_vcpu_pmu_ctr_read and pmu_ctr_read, but I wonder if we really need the kvm_riscv_vcpu_pmu_ctr_read() and kvm_riscv_vcpu_pmu_fw_ctr_read_hi() wrapper functions? > +} > + > int kvm_riscv_vcpu_pmu_ctr_read(struct kvm_vcpu *vcpu, unsigned long cidx, > struct kvm_vcpu_sbi_return *retdata) > { > @@ -775,7 +810,7 @@ void kvm_riscv_vcpu_pmu_init(struct kvm_vcpu *vcpu) > pmc->cinfo.csr = CSR_CYCLE + i; > } else { > pmc->cinfo.type = SBI_PMU_CTR_TYPE_FW; > - pmc->cinfo.width = BITS_PER_LONG - 1; > + pmc->cinfo.width = 63; > } > } > > diff --git a/arch/riscv/kvm/vcpu_sbi_pmu.c b/arch/riscv/kvm/vcpu_sbi_pmu.c > index 9f61136e4bb1..58a0e5587e2a 100644 > --- a/arch/riscv/kvm/vcpu_sbi_pmu.c > +++ b/arch/riscv/kvm/vcpu_sbi_pmu.c > @@ -64,6 +64,12 @@ static int kvm_sbi_ext_pmu_handler(struct kvm_vcpu *vcpu, struct kvm_run *run, > case SBI_EXT_PMU_COUNTER_FW_READ: > ret = kvm_riscv_vcpu_pmu_ctr_read(vcpu, cp->a0, retdata); > break; > + case SBI_EXT_PMU_COUNTER_FW_READ_HI: > + if (IS_ENABLED(CONFIG_32BIT)) > + ret = kvm_riscv_vcpu_pmu_fw_ctr_read_hi(vcpu, cp->a0, retdata); > + else > + retdata->out_val = 0; > + break; > case SBI_EXT_PMU_SNAPSHOT_SET_SHMEM: > ret = kvm_riscv_vcpu_pmu_setup_snapshot(vcpu, cp->a0, cp->a1, cp->a2, retdata); > break; > -- > 2.34.1 > Thanks, drew
On Sat, Mar 2, 2024 at 2:52 AM Andrew Jones <ajones@ventanamicro.com> wrote: > > On Wed, Feb 28, 2024 at 05:01:25PM -0800, Atish Patra wrote: > > The SBI v2.0 introduced a fw_read_hi function to read 64 bit firmware > > counters for RV32 based systems. > > > > Add infrastructure to support that. > > > > Reviewed-by: Anup Patel <anup@brainfault.org> > > Signed-off-by: Atish Patra <atishp@rivosinc.com> > > --- > > arch/riscv/include/asm/kvm_vcpu_pmu.h | 4 ++- > > arch/riscv/kvm/vcpu_pmu.c | 37 ++++++++++++++++++++++++++- > > arch/riscv/kvm/vcpu_sbi_pmu.c | 6 +++++ > > 3 files changed, 45 insertions(+), 2 deletions(-) > > > > diff --git a/arch/riscv/include/asm/kvm_vcpu_pmu.h b/arch/riscv/include/asm/kvm_vcpu_pmu.h > > index 8cb21a4f862c..e0ad27dea46c 100644 > > --- a/arch/riscv/include/asm/kvm_vcpu_pmu.h > > +++ b/arch/riscv/include/asm/kvm_vcpu_pmu.h > > @@ -20,7 +20,7 @@ static_assert(RISCV_KVM_MAX_COUNTERS <= 64); > > > > struct kvm_fw_event { > > /* Current value of the event */ > > - unsigned long value; > > + u64 value; > > > > /* Event monitoring status */ > > bool started; > > @@ -91,6 +91,8 @@ int kvm_riscv_vcpu_pmu_ctr_cfg_match(struct kvm_vcpu *vcpu, unsigned long ctr_ba > > struct kvm_vcpu_sbi_return *retdata); > > int kvm_riscv_vcpu_pmu_ctr_read(struct kvm_vcpu *vcpu, unsigned long cidx, > > struct kvm_vcpu_sbi_return *retdata); > > +int kvm_riscv_vcpu_pmu_fw_ctr_read_hi(struct kvm_vcpu *vcpu, unsigned long cidx, > > + struct kvm_vcpu_sbi_return *retdata); > > void kvm_riscv_vcpu_pmu_init(struct kvm_vcpu *vcpu); > > int kvm_riscv_vcpu_pmu_setup_snapshot(struct kvm_vcpu *vcpu, unsigned long saddr_low, > > unsigned long saddr_high, unsigned long flags, > > diff --git a/arch/riscv/kvm/vcpu_pmu.c b/arch/riscv/kvm/vcpu_pmu.c > > index a02f7b981005..469bb430cf97 100644 > > --- a/arch/riscv/kvm/vcpu_pmu.c > > +++ b/arch/riscv/kvm/vcpu_pmu.c > > @@ -196,6 +196,29 @@ static int pmu_get_pmc_index(struct kvm_pmu *pmu, unsigned long eidx, > > return kvm_pmu_get_programmable_pmc_index(pmu, eidx, cbase, cmask); > > } > > > > +static int pmu_fw_ctr_read_hi(struct kvm_vcpu *vcpu, unsigned long cidx, > > + unsigned long *out_val) > > +{ > > + struct kvm_pmu *kvpmu = vcpu_to_pmu(vcpu); > > + struct kvm_pmc *pmc; > > + int fevent_code; > > + > > + if (!IS_ENABLED(CONFIG_32BIT)) > > Let's remove the CONFIG_32BIT check in kvm_sbi_ext_pmu_handler() and then > set *out_val to zero here and return success. Either that, or we should > WARN or something here since it's a KVM bug to get here with > !CONFIG_32BIT. > I added a warning here to prevent any sort of kvm bug. Returning silently with out_val to zero from here may hide that. The CONFIG_32BIT check in kvm_sbi_ext_pmu_handler also avoids unnecessary code execution (even though they are few) in case the lower privilege mode software invokes the read_hi by mistake for non RV32. > > + return -EINVAL; > > + > > + pmc = &kvpmu->pmc[cidx]; > > Uh oh! We're missing range validation of cidx! And I see we're missing it > in pmu_ctr_read() too. We need the same check we have in > kvm_riscv_vcpu_pmu_ctr_info(). I think the other SBI functions are OK, > but it's worth a triple check. > Good catch. Thanks. Fixed it. > > + > > + if (pmc->cinfo.type != SBI_PMU_CTR_TYPE_FW) > > + return -EINVAL; > > + > > + fevent_code = get_event_code(pmc->event_idx); > > + pmc->counter_val = kvpmu->fw_event[fevent_code].value; > > + > > + *out_val = pmc->counter_val >> 32; > > + > > + return 0; > > +} > > + > > static int pmu_ctr_read(struct kvm_vcpu *vcpu, unsigned long cidx, > > unsigned long *out_val) > > { > > @@ -702,6 +725,18 @@ int kvm_riscv_vcpu_pmu_ctr_cfg_match(struct kvm_vcpu *vcpu, unsigned long ctr_ba > > return 0; > > } > > > > +int kvm_riscv_vcpu_pmu_fw_ctr_read_hi(struct kvm_vcpu *vcpu, unsigned long cidx, > > + struct kvm_vcpu_sbi_return *retdata) > > +{ > > + int ret; > > + > > + ret = pmu_fw_ctr_read_hi(vcpu, cidx, &retdata->out_val); > > + if (ret == -EINVAL) > > + retdata->err_val = SBI_ERR_INVALID_PARAM; > > + > > + return 0; > > I see this follows the pattern we have with kvm_riscv_vcpu_pmu_ctr_read > and pmu_ctr_read, but I wonder if we really need the > kvm_riscv_vcpu_pmu_ctr_read() and kvm_riscv_vcpu_pmu_fw_ctr_read_hi() > wrapper functions? > pmu_ctr_read is invoked from kvm_riscv_vcpu_pmu_read_hpm as well. That's why I have a wrapper to read the counters in the SBI path. kvm_riscv_vcpu_pmu_ctr_read kvm_riscv_vcpu_pmu_fw_ctr_read_hi just followed the pattern. If we refactor the firmware counter read and hpmcounter read to be separate functions, we won't need the wrapper though. But I am not sure if it will actually improve the code readability. If you think it's better that way, I will modify it. Looking at this code, we should definitely change the kvm_riscv_vcpu_pmu_ctr_read to kvm_riscv_vcpu_pmu_fw_ctr_read to reflect the real purpose. > > +} > > + > > int kvm_riscv_vcpu_pmu_ctr_read(struct kvm_vcpu *vcpu, unsigned long cidx, > > struct kvm_vcpu_sbi_return *retdata) > > { > > @@ -775,7 +810,7 @@ void kvm_riscv_vcpu_pmu_init(struct kvm_vcpu *vcpu) > > pmc->cinfo.csr = CSR_CYCLE + i; > > } else { > > pmc->cinfo.type = SBI_PMU_CTR_TYPE_FW; > > - pmc->cinfo.width = BITS_PER_LONG - 1; > > + pmc->cinfo.width = 63; > > } > > } > > > > diff --git a/arch/riscv/kvm/vcpu_sbi_pmu.c b/arch/riscv/kvm/vcpu_sbi_pmu.c > > index 9f61136e4bb1..58a0e5587e2a 100644 > > --- a/arch/riscv/kvm/vcpu_sbi_pmu.c > > +++ b/arch/riscv/kvm/vcpu_sbi_pmu.c > > @@ -64,6 +64,12 @@ static int kvm_sbi_ext_pmu_handler(struct kvm_vcpu *vcpu, struct kvm_run *run, > > case SBI_EXT_PMU_COUNTER_FW_READ: > > ret = kvm_riscv_vcpu_pmu_ctr_read(vcpu, cp->a0, retdata); > > break; > > + case SBI_EXT_PMU_COUNTER_FW_READ_HI: > > + if (IS_ENABLED(CONFIG_32BIT)) > > + ret = kvm_riscv_vcpu_pmu_fw_ctr_read_hi(vcpu, cp->a0, retdata); > > + else > > + retdata->out_val = 0; > > + break; > > case SBI_EXT_PMU_SNAPSHOT_SET_SHMEM: > > ret = kvm_riscv_vcpu_pmu_setup_snapshot(vcpu, cp->a0, cp->a1, cp->a2, retdata); > > break; > > -- > > 2.34.1 > > > > Thanks, > drew
diff --git a/arch/riscv/include/asm/kvm_vcpu_pmu.h b/arch/riscv/include/asm/kvm_vcpu_pmu.h index 8cb21a4f862c..e0ad27dea46c 100644 --- a/arch/riscv/include/asm/kvm_vcpu_pmu.h +++ b/arch/riscv/include/asm/kvm_vcpu_pmu.h @@ -20,7 +20,7 @@ static_assert(RISCV_KVM_MAX_COUNTERS <= 64); struct kvm_fw_event { /* Current value of the event */ - unsigned long value; + u64 value; /* Event monitoring status */ bool started; @@ -91,6 +91,8 @@ int kvm_riscv_vcpu_pmu_ctr_cfg_match(struct kvm_vcpu *vcpu, unsigned long ctr_ba struct kvm_vcpu_sbi_return *retdata); int kvm_riscv_vcpu_pmu_ctr_read(struct kvm_vcpu *vcpu, unsigned long cidx, struct kvm_vcpu_sbi_return *retdata); +int kvm_riscv_vcpu_pmu_fw_ctr_read_hi(struct kvm_vcpu *vcpu, unsigned long cidx, + struct kvm_vcpu_sbi_return *retdata); void kvm_riscv_vcpu_pmu_init(struct kvm_vcpu *vcpu); int kvm_riscv_vcpu_pmu_setup_snapshot(struct kvm_vcpu *vcpu, unsigned long saddr_low, unsigned long saddr_high, unsigned long flags, diff --git a/arch/riscv/kvm/vcpu_pmu.c b/arch/riscv/kvm/vcpu_pmu.c index a02f7b981005..469bb430cf97 100644 --- a/arch/riscv/kvm/vcpu_pmu.c +++ b/arch/riscv/kvm/vcpu_pmu.c @@ -196,6 +196,29 @@ static int pmu_get_pmc_index(struct kvm_pmu *pmu, unsigned long eidx, return kvm_pmu_get_programmable_pmc_index(pmu, eidx, cbase, cmask); } +static int pmu_fw_ctr_read_hi(struct kvm_vcpu *vcpu, unsigned long cidx, + unsigned long *out_val) +{ + struct kvm_pmu *kvpmu = vcpu_to_pmu(vcpu); + struct kvm_pmc *pmc; + int fevent_code; + + if (!IS_ENABLED(CONFIG_32BIT)) + return -EINVAL; + + pmc = &kvpmu->pmc[cidx]; + + if (pmc->cinfo.type != SBI_PMU_CTR_TYPE_FW) + return -EINVAL; + + fevent_code = get_event_code(pmc->event_idx); + pmc->counter_val = kvpmu->fw_event[fevent_code].value; + + *out_val = pmc->counter_val >> 32; + + return 0; +} + static int pmu_ctr_read(struct kvm_vcpu *vcpu, unsigned long cidx, unsigned long *out_val) { @@ -702,6 +725,18 @@ int kvm_riscv_vcpu_pmu_ctr_cfg_match(struct kvm_vcpu *vcpu, unsigned long ctr_ba return 0; } +int kvm_riscv_vcpu_pmu_fw_ctr_read_hi(struct kvm_vcpu *vcpu, unsigned long cidx, + struct kvm_vcpu_sbi_return *retdata) +{ + int ret; + + ret = pmu_fw_ctr_read_hi(vcpu, cidx, &retdata->out_val); + if (ret == -EINVAL) + retdata->err_val = SBI_ERR_INVALID_PARAM; + + return 0; +} + int kvm_riscv_vcpu_pmu_ctr_read(struct kvm_vcpu *vcpu, unsigned long cidx, struct kvm_vcpu_sbi_return *retdata) { @@ -775,7 +810,7 @@ void kvm_riscv_vcpu_pmu_init(struct kvm_vcpu *vcpu) pmc->cinfo.csr = CSR_CYCLE + i; } else { pmc->cinfo.type = SBI_PMU_CTR_TYPE_FW; - pmc->cinfo.width = BITS_PER_LONG - 1; + pmc->cinfo.width = 63; } } diff --git a/arch/riscv/kvm/vcpu_sbi_pmu.c b/arch/riscv/kvm/vcpu_sbi_pmu.c index 9f61136e4bb1..58a0e5587e2a 100644 --- a/arch/riscv/kvm/vcpu_sbi_pmu.c +++ b/arch/riscv/kvm/vcpu_sbi_pmu.c @@ -64,6 +64,12 @@ static int kvm_sbi_ext_pmu_handler(struct kvm_vcpu *vcpu, struct kvm_run *run, case SBI_EXT_PMU_COUNTER_FW_READ: ret = kvm_riscv_vcpu_pmu_ctr_read(vcpu, cp->a0, retdata); break; + case SBI_EXT_PMU_COUNTER_FW_READ_HI: + if (IS_ENABLED(CONFIG_32BIT)) + ret = kvm_riscv_vcpu_pmu_fw_ctr_read_hi(vcpu, cp->a0, retdata); + else + retdata->out_val = 0; + break; case SBI_EXT_PMU_SNAPSHOT_SET_SHMEM: ret = kvm_riscv_vcpu_pmu_setup_snapshot(vcpu, cp->a0, cp->a1, cp->a2, retdata); break;