Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH] KVM: selftests: Speed up set_memory_region_test
To:     Zenghui Yu <yuzenghui@huawei.com>,
        Vitaly Kuznetsov <vkuznets@redhat.com>
Cc:     kvm@vger.kernel.org, Sean Christopherson <seanjc@google.com>,
        Wanpeng Li <wanpengli@tencent.com>,
        Jim Mattson <jmattson@google.com>,
        kernel test robot <oliver.sang@intel.com>,
        linux-kernel@vger.kernel.org, wanghaibin.wang@huawei.com
References: <20210426130121.758229-1-vkuznets@redhat.com>
 <91a2d145-fd3c-6e8d-6478-60f62dff07fe@huawei.com>
From:   Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <3dc9bd69-f38a-daed-4ac3-84b280ef5901@redhat.com>
Date:   Wed, 23 Jun 2021 23:45:30 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.10.1
MIME-Version: 1.0
In-Reply-To: <91a2d145-fd3c-6e8d-6478-60f62dff07fe@huawei.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Precedence: bulk

On 23/06/21 11:03, Zenghui Yu wrote:
> On 2021/4/26 21:01, Vitaly Kuznetsov wrote:
>> After commit 4fc096a99e01 ("KVM: Raise the maximum number of user 
>> memslots")
>> set_memory_region_test may take too long, reports are that the default
>> timeout value we have (120s) may not be enough even on a physical host.
>>
>> Speed things up a bit by throwing away vm_userspace_mem_region_add() 
>> usage
>> from test_add_max_memory_regions(), we don't really need to do the 
>> majority
>> of the stuff it does for the sake of this test.
>>
>> On my AMD EPYC 7401P, # time ./set_memory_region_test
>> pre-patch:
>>  Testing KVM_RUN with zero added memory regions
>>  Allowed number of memory slots: 32764
>>  Adding slots 0..32763, each memory region with 2048K size
>>  Testing MOVE of in-use region, 10 loops
>>  Testing DELETE of in-use region, 10 loops
>>
>>  real    0m44.917s
>>  user    0m7.416s
>>  sys    0m34.601s
>>
>> post-patch:
>>  Testing KVM_RUN with zero added memory regions
>>  Allowed number of memory slots: 32764
>>  Adding slots 0..32763, each memory region with 2048K size
>>  Testing MOVE of in-use region, 10 loops
>>  Testing DELETE of in-use region, 10 loops
>>
>>  real    0m20.714s
>>  user    0m0.109s
>>  sys    0m18.359s
>>
>> Reported-by: kernel test robot <oliver.sang@intel.com>
>> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
> 
> I've seen the failure on my arm64 server, # ./set_memory_region_test
> 
> Allowed number of memory slots: 32767
> Adding slots 0..32766, each memory region with 2048K size
> ==== Test Assertion Failure ====
>    set_memory_region_test.c:391: ret == 0
>    pid=42696 tid=42696 errno=22 - Invalid argument
>       1    0x00000000004015a7: test_add_max_memory_regions at 
> set_memory_region_test.c:389
>       2     (inlined by) main at set_memory_region_test.c:426
>       3    0x0000ffffb7c63bdf: ?? ??:0
>       4    0x00000000004016db: _start at :?
>    KVM_SET_USER_MEMORY_REGION IOCTL failed,
>    rc: -1 errno: 22 slot: 2624
> 
>> +    mem = mmap(NULL, MEM_REGION_SIZE * max_mem_slots + alignment,
> 
> The problem is that max_mem_slots is declared as uint32_t, the result
> of (MEM_REGION_SIZE * max_mem_slots) is unexpectedly truncated to be
> 0xffe00000.
> 
>> +           PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
>> +    TEST_ASSERT(mem != MAP_FAILED, "Failed to mmap() host");
>> +    mem_aligned = (void *)(((size_t) mem + alignment - 1) & 
>> ~(alignment - 1));
>> +
>>      for (slot = 0; slot < max_mem_slots; slot++) {
>> -        vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS,
>> -                        guest_addr, slot, mem_reg_npages,
>> -                        0);
>> -        guest_addr += MEM_REGION_SIZE;
>> +        ret = test_memory_region_add(vm, mem_aligned +
>> +                         ((uint64_t)slot * MEM_REGION_SIZE),
> 
> These unmapped VAs got caught by access_ok() checker in
> __kvm_set_memory_region() as they happen to go beyond the task's
> address space on arm64. Casting max_mem_slots to size_t in both
> mmap() and munmap() fixes the issue for me.

Can you provide a patch for both?

Paolo