[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <2c96c61a-2685-4cee-9cef-963ed833bf92@rivosinc.com>
Date: Wed, 21 Feb 2024 00:37:21 -0800
From: Atish Patra <atishp@...osinc.com>
To: Haibo Xu <xiaobo55x@...il.com>, Atish Patra <atishp@...shpatra.org>
Cc: Haibo Xu <haibo1.xu@...el.com>, ajones@...tanamicro.com,
Paul Walmsley <paul.walmsley@...ive.com>, Palmer Dabbelt
<palmer@...belt.com>, Albert Ou <aou@...s.berkeley.edu>,
Paolo Bonzini <pbonzini@...hat.com>, Shuah Khan <shuah@...nel.org>,
Marc Zyngier <maz@...nel.org>, Oliver Upton <oliver.upton@...ux.dev>,
James Morse <james.morse@....com>, Suzuki K Poulose
<suzuki.poulose@....com>, Zenghui Yu <yuzenghui@...wei.com>,
Anup Patel <anup@...infault.org>, Guo Ren <guoren@...nel.org>,
Mayuresh Chitale <mchitale@...tanamicro.com>,
Daniel Henrique Barboza <dbarboza@...tanamicro.com>,
Conor Dooley <conor.dooley@...rochip.com>,
Samuel Holland <samuel@...lland.org>,
Minda Chen <minda.chen@...rfivetech.com>, Jisheng Zhang
<jszhang@...nel.org>, Sean Christopherson <seanjc@...gle.com>,
Peter Xu <peterx@...hat.com>, Like Xu <likexu@...cent.com>,
Vipin Sharma <vipinsh@...gle.com>, Thomas Huth <thuth@...hat.com>,
Aaron Lewis <aaronlewis@...gle.com>,
Maciej Wieczor-Retman <maciej.wieczor-retman@...el.com>,
linux-kernel@...r.kernel.org, linux-riscv@...ts.infradead.org,
kvm@...r.kernel.org, linux-kselftest@...r.kernel.org,
linux-arm-kernel@...ts.infradead.org, kvmarm@...ts.linux.dev,
kvm-riscv@...ts.infradead.org
Subject: Re: [PATCH v5 11/12] KVM: riscv: selftests: Change vcpu_has_ext to a
common function
On 2/20/24 18:13, Haibo Xu wrote:
> On Wed, Feb 21, 2024 at 7:03 AM Atish Patra <atishp@...shpatra.org> wrote:
>>
>> On Mon, Jan 22, 2024 at 1:48 AM Haibo Xu <haibo1.xu@...el.com> wrote:
>>>
>>> Move vcpu_has_ext to the processor.c and rename it to __vcpu_has_ext
>>> so that other test cases can use it for vCPU extension check.
>>>
>>> Signed-off-by: Haibo Xu <haibo1.xu@...el.com>
>>> Reviewed-by: Andrew Jones <ajones@...tanamicro.com>
>>> ---
>>> tools/testing/selftests/kvm/include/riscv/processor.h | 2 ++
>>> tools/testing/selftests/kvm/lib/riscv/processor.c | 10 ++++++++++
>>> tools/testing/selftests/kvm/riscv/get-reg-list.c | 11 +----------
>>> 3 files changed, 13 insertions(+), 10 deletions(-)
>>>
>>> diff --git a/tools/testing/selftests/kvm/include/riscv/processor.h b/tools/testing/selftests/kvm/include/riscv/processor.h
>>> index b68b1b731a34..bd27e1c67579 100644
>>> --- a/tools/testing/selftests/kvm/include/riscv/processor.h
>>> +++ b/tools/testing/selftests/kvm/include/riscv/processor.h
>>> @@ -42,6 +42,8 @@ static inline uint64_t __kvm_reg_id(uint64_t type, uint64_t idx,
>>> #define RISCV_ISA_EXT_REG(idx) __kvm_reg_id(KVM_REG_RISCV_ISA_EXT, \
>>> idx, KVM_REG_SIZE_ULONG)
>>>
>>> +bool __vcpu_has_ext(struct kvm_vcpu *vcpu, int ext);
>>> +
>>> struct ex_regs {
>>> unsigned long ra;
>>> unsigned long sp;
>>> diff --git a/tools/testing/selftests/kvm/lib/riscv/processor.c b/tools/testing/selftests/kvm/lib/riscv/processor.c
>>> index 39a1e9902dec..dad73ce18164 100644
>>> --- a/tools/testing/selftests/kvm/lib/riscv/processor.c
>>> +++ b/tools/testing/selftests/kvm/lib/riscv/processor.c
>>> @@ -15,6 +15,16 @@
>>>
>>> static vm_vaddr_t exception_handlers;
>>>
>>> +bool __vcpu_has_ext(struct kvm_vcpu *vcpu, int ext)
>>> +{
>>> + unsigned long value = 0;
>>> + int ret;
>>> +
>>> + ret = __vcpu_get_reg(vcpu, RISCV_ISA_EXT_REG(ext), &value);
>>> +
>>> + return !ret && !!value;
>>> +}
>>> +
>>
>> Not sure what was the base patch on which this was rebased. The actual
>> commit in the queue branch looks different.
>>
>
> This patch set was based on 6.7-rc8.
>
>> https://github.com/kvm-riscv/linux/commit/5563517cc2012e3326411b360c9924d3f2706c8d
>>
>> Both seem to have the same bug though the tests fail now and require
>> the following fix.
>> The ext id should be uint64_t and we need to pass ext directly so that
>> SBI extension tests can also pass.
>>
>
> It's weird that 6.7-rc8 has already included Andrew's change on the ISA ext reg,
> but this patch was not generated against his change.
>
> commit bdf6aa328f137e184b0fce607fd585354c3742f1
> Author: Andrew Jones <ajones@...tanamicro.com>
> Date: Wed Dec 13 18:09:58 2023 +0100
>
> RISC-V: KVM: selftests: Treat SBI ext regs like ISA ext regs
>
> Anyway, your changes were right. Please go ahead to include them when merging.
>
I am not sure what happened. Probably, a merge conflict issue.
I just realized I forgot to copy paste another fix in arch timer
+++ b/tools/testing/selftests/kvm/riscv/arch_timer.c
@@ -85,7 +85,7 @@ struct kvm_vm *test_vm_create(void)
int nr_vcpus = test_args.nr_vcpus;
vm = vm_create_with_vcpus(nr_vcpus, guest_code, vcpus);
- __TEST_REQUIRE(__vcpu_has_ext(vcpus[0], KVM_RISCV_ISA_EXT_SSTC),
+ __TEST_REQUIRE(__vcpu_has_ext(vcpus[0],
RISCV_ISA_EXT_REG(KVM_RISCV_ISA_EXT_SSTC)),
> Thanks,
> Haibo
>
>>
>> --- a/tools/testing/selftests/kvm/include/riscv/processor.h
>> +++ b/tools/testing/selftests/kvm/include/riscv/processor.h
>> @@ -48,7 +48,7 @@ static inline uint64_t __kvm_reg_id(uint64_t type,
>> uint64_t subtype,
>> KVM_REG_RISCV_SBI_SINGLE, \
>> idx, KVM_REG_SIZE_ULONG)
>>
>> -bool __vcpu_has_ext(struct kvm_vcpu *vcpu, int ext);
>> +bool __vcpu_has_ext(struct kvm_vcpu *vcpu, uint64_t ext);
>>
>> struct ex_regs {
>> unsigned long ra;
>> diff --git a/tools/testing/selftests/kvm/lib/riscv/processor.c
>> b/tools/testing/selftests/kvm/lib/riscv/processor.c
>> index 282587cd4bbc..ec66d331a127 100644
>> --- a/tools/testing/selftests/kvm/lib/riscv/processor.c
>> +++ b/tools/testing/selftests/kvm/lib/riscv/processor.c
>> @@ -15,12 +15,12 @@
>>
>> static vm_vaddr_t exception_handlers;
>>
>> -bool __vcpu_has_ext(struct kvm_vcpu *vcpu, int ext)
>> +bool __vcpu_has_ext(struct kvm_vcpu *vcpu, uint64_t ext)
>> {
>> unsigned long value = 0;
>> int ret;
>>
>> - ret = __vcpu_get_reg(vcpu, RISCV_ISA_EXT_REG(ext), &value);
>> + ret = __vcpu_get_reg(vcpu, ext, &value);
>>
>> return !ret && !!value;
>> }
>>
>> With the above the fix, Both SBI/ISA extension tests pass.
>> # ./get-reg-list
>> sbi-base: PASS
>> sbi-sta: PASS
>> sbi-pmu: PASS
>> sbi-dbcn: PASS
>> aia: PASS
>> fp_f: PASS
>> fp_d: PASS
>> 1..0 # SKIP - h not available, skipping tests
>> smstateen: PASS
>> sscofpmf: PASS
>> sstc: PASS
>> 1..0 # SKIP - svinval not available, skipping tests
>> 1..0 # SKIP - svnapot not available, skipping tests
>> 1..0 # SKIP - svpbmt not available, skipping tests
>> zba: PASS
>> zbb: PASS
>> zbc: PASS
>> 1..0 # SKIP - zbkb not available, skipping tests
>> 1..0 # SKIP - zbkc not available, skipping tests
>> 1..0 # SKIP - zbkx not available, skipping tests
>> zbs: PASS
>> zfa: PASS
>> 1..0 # SKIP - zfh not available, skipping tests
>> 1..0 # SKIP - zfhmin not available, skipping tests
>> zicbom: PASS
>> zicboz: PASS
>> zicntr: PASS
>> 1..0 # SKIP - zicond not available, skipping tests
>> zicsr: PASS
>> zifencei: PASS
>> zihintntl: PASS
>> zihintpause: PASS
>> zihpm: PASS
>>
>>
>>> static uint64_t page_align(struct kvm_vm *vm, uint64_t v)
>>> {
>>> return (v + vm->page_size) & ~(vm->page_size - 1);
>>> diff --git a/tools/testing/selftests/kvm/riscv/get-reg-list.c b/tools/testing/selftests/kvm/riscv/get-reg-list.c
>>> index 25de4b8bc347..ed29ba45588c 100644
>>> --- a/tools/testing/selftests/kvm/riscv/get-reg-list.c
>>> +++ b/tools/testing/selftests/kvm/riscv/get-reg-list.c
>>> @@ -75,15 +75,6 @@ bool check_reject_set(int err)
>>> return err == EINVAL;
>>> }
>>>
>>> -static inline bool vcpu_has_ext(struct kvm_vcpu *vcpu, int ext)
>>> -{
>>> - int ret;
>>> - unsigned long value;
>>> -
>>> - ret = __vcpu_get_reg(vcpu, RISCV_ISA_EXT_REG(ext), &value);
>>> - return (ret) ? false : !!value;
>>> -}
>>> -
>>> void finalize_vcpu(struct kvm_vcpu *vcpu, struct vcpu_reg_list *c)
>>> {
>>> unsigned long isa_ext_state[KVM_RISCV_ISA_EXT_MAX] = { 0 };
>>> @@ -111,7 +102,7 @@ void finalize_vcpu(struct kvm_vcpu *vcpu, struct vcpu_reg_list *c)
>>> __vcpu_set_reg(vcpu, RISCV_ISA_EXT_REG(s->feature), 1);
>>>
>>> /* Double check whether the desired extension was enabled */
>>> - __TEST_REQUIRE(vcpu_has_ext(vcpu, s->feature),
>>> + __TEST_REQUIRE(__vcpu_has_ext(vcpu, s->feature),
>>> "%s not available, skipping tests\n", s->name);
>>> }
>>> }
>>> --
>>> 2.34.1
>>>
>>
>>
>> --
>> Regards,
>> Atish
Powered by blists - more mailing lists