Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [PATCH 2/2] selftests: kvm: Add mem_slot_test test
To:     Andrew Jones <drjones@redhat.com>
Cc:     kvm@vger.kernel.org, pbonzini@redhat.com,
        linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org,
        david@redhat.com
References: <20200330204310.21736-1-wainersm@redhat.com>
 <20200330204310.21736-3-wainersm@redhat.com>
 <20200331081632.ithcwuzjyjhiwphy@kamzik.brq.redhat.com>
From:   Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-ID: <b261aa4f-87d5-2ac8-9f66-9f10e1a0803a@redhat.com>
Date:   Tue, 31 Mar 2020 18:42:21 -0300
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101
 Thunderbird/60.4.0
MIME-Version: 1.0
In-Reply-To: <20200331081632.ithcwuzjyjhiwphy@kamzik.brq.redhat.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Content-Language: en-US
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi Andrew,

Nice review. Few comments below.

On 3/31/20 5:16 AM, Andrew Jones wrote:
> On Mon, Mar 30, 2020 at 05:43:10PM -0300, Wainer dos Santos Moschetta wrote:
>> This patch introduces the mem_slot_test test which checks
>> an VM can have added memory slots up to the limit defined in
>> KVM_CAP_NR_MEMSLOTS. Then attempt to add one more slot to
>> verify it fails as expected.
>>
>> Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
>> ---
>>   tools/testing/selftests/kvm/.gitignore      |  1 +
>>   tools/testing/selftests/kvm/Makefile        |  3 +
>>   tools/testing/selftests/kvm/mem_slot_test.c | 92 +++++++++++++++++++++
>>   3 files changed, 96 insertions(+)
>>   create mode 100644 tools/testing/selftests/kvm/mem_slot_test.c
>>
>> diff --git a/tools/testing/selftests/kvm/.gitignore b/tools/testing/selftests/kvm/.gitignore
>> index 30072c3f52fb..b1b94d50f6a2 100644
>> --- a/tools/testing/selftests/kvm/.gitignore
>> +++ b/tools/testing/selftests/kvm/.gitignore
>> @@ -17,3 +17,4 @@
>>   /clear_dirty_log_test
>>   /dirty_log_test
>>   /kvm_create_max_vcpus
>> +/mem_slot_test
>> diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
>> index d91c53b726e6..070133349403 100644
>> --- a/tools/testing/selftests/kvm/Makefile
>> +++ b/tools/testing/selftests/kvm/Makefile
>> @@ -30,16 +30,19 @@ TEST_GEN_PROGS_x86_64 += x86_64/svm_vmcall_test
>>   TEST_GEN_PROGS_x86_64 += clear_dirty_log_test
>>   TEST_GEN_PROGS_x86_64 += dirty_log_test
>>   TEST_GEN_PROGS_x86_64 += kvm_create_max_vcpus
>> +TEST_GEN_PROGS_x86_64 += mem_slot_test
>>   
>>   TEST_GEN_PROGS_aarch64 += clear_dirty_log_test
>>   TEST_GEN_PROGS_aarch64 += dirty_log_test
>>   TEST_GEN_PROGS_aarch64 += kvm_create_max_vcpus
>> +TEST_GEN_PROGS_aarch64 += mem_slot_test
>>   
>>   TEST_GEN_PROGS_s390x = s390x/memop
>>   TEST_GEN_PROGS_s390x += s390x/sync_regs_test
>>   TEST_GEN_PROGS_s390x += s390x/resets
>>   TEST_GEN_PROGS_s390x += dirty_log_test
>>   TEST_GEN_PROGS_s390x += kvm_create_max_vcpus
>> +TEST_GEN_PROGS_s390x += mem_slot_test
>>   
>>   TEST_GEN_PROGS += $(TEST_GEN_PROGS_$(UNAME_M))
>>   LIBKVM += $(LIBKVM_$(UNAME_M))
>> diff --git a/tools/testing/selftests/kvm/mem_slot_test.c b/tools/testing/selftests/kvm/mem_slot_test.c
>> new file mode 100644
>> index 000000000000..75d2bbd71642
>> --- /dev/null
>> +++ b/tools/testing/selftests/kvm/mem_slot_test.c
>> @@ -0,0 +1,92 @@
>> +// SPDX-License-Identifier: GPL-2.0-only
>> +/*
>> + * mem_slot_test
>> + *
>> + * Copyright (C) 2020, Red Hat, Inc.
>> + *
>> + * Test it can be added memory slots up to KVM_CAP_NR_MEMSLOTS, then any
>> + * tentative to add further slots should fail.
>> + */
>> +#define _GNU_SOURCE /* for program_invocation_short_name */
>> +#include <linux/kvm.h>
>> +#include <sys/mman.h>
>> +#include <unistd.h>
>> +
>> +#include "test_util.h"
>> +#include "kvm_util.h"
>> +
>> +/* Memory region flags */
>> +#define MEM_REG_FLAGS KVM_MEM_LOG_DIRTY_PAGES
>> +
>> +/* Guest VM mode */
>> +#define GUEST_VM_MODE VM_MODE_DEFAULT
> I'm not sure what the value of the two defines above are. I'd prefer we
> avoid unnecessary renaming. Also, do we need KVM_MEM_LOG_DIRTY_PAGES for
> this test?

It would be nice to exercise the code by adding slots with different 
page flags. But for this test that simple checks the limit, the use of 
KVM_MEM_READONLY is enough. I will change it on v2.

>
>> +
>> +int main(int argc, char *argv[])
>> +{
>> +	struct kvm_vm *vm;
>> +	/* Maximum allowed number of memory slots */
>> +	uint32_t max_mem_slots;
>> +	/* Slot number */
>> +	uint32_t slot;
>> +	/* Number of pages in a memory region */
>> +	uint64_t mem_reg_npages;
>> +	/* Memory region size */
>> +	uint64_t mem_reg_size;
>> +	/* Guest physical memory guest_address */
>> +	uint64_t guest_addr;
>> +	/* VM page size */
>> +	uint64_t vm_page_size;
> nit: IMO, the variable names above are descriptive enough to drop the
> comments.
>
>> +	int ret;
>> +
>> +	max_mem_slots = kvm_check_cap(KVM_CAP_NR_MEMSLOTS);
>> +	TEST_ASSERT(max_mem_slots > 0,
>> +		    "KVM_CAP_NR_MEMSLOTS should be greater than 0");
>> +	DEBUG("Allowed number of memory slots: %i\n", max_mem_slots);
> DEBUG() no longer exists in kvm/queue. This should now be pr_debug().
Great. I will rebase my code to kvm/queue...
>
>> +
>> +	vm = vm_create(GUEST_VM_MODE, 0, O_RDWR);
>> +
>> +	/* Determine the minimal number of pages as possible per region. */
>> +	vm_page_size = vm_get_page_size(vm);
>> +#ifdef __s390x__
>> +	mem_reg_size = 0x100000;
>> +#else
>> +	uint64_t host_page_size = sysconf(_SC_PAGESIZE);
>> +
>> +	mem_reg_size = (host_page_size > vm_page_size) ? host_page_size :
>> +							 vm_page_size;
>> +#endif
>> +	mem_reg_npages = mem_reg_size / vm_page_size;
> On kvm/queue the above 11 lines can now all be done with
>
>    mem_reg_size = SOME_ARBITRARY_MEM_REG_SIZE;
>    mem_reg_npages = vm_calc_num_guest_pages(VM_MODE_DEFAULT, mem_reg_size);
>
>> +	guest_addr = 0x0;
>> +
>> +	/* Check it can be added memory slots up to the maximum allowed */
>> +	DEBUG("Adding slots 0..%i, each memory region with %ldK size\n",
>> +	      (max_mem_slots - 1), mem_reg_size >> 10);
>> +	for (slot = 0; slot < max_mem_slots; slot++) {
>> +		vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS,
>> +					    guest_addr, slot, mem_reg_npages,
>> +					    MEM_REG_FLAGS);
>> +		guest_addr += mem_reg_size;
>> +	}
>> +
>> +	/* Check it cannot be added memory slots beyond the limit */
>> +	guest_addr += mem_reg_size;
> nit: shouldn't be necessary. We already incremented guest_addr on the
> last loop.
Good catch.
>
>> +	void *mem = mmap(NULL, mem_reg_size, PROT_READ | PROT_WRITE,
>> +			 MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
>> +	TEST_ASSERT(mem != NULL, "Failed to mmap() host");
>> +
>> +	struct kvm_userspace_memory_region kvm_region = {
>> +		.slot = slot,
>> +		.flags = MEM_REG_FLAGS,
>> +		.guest_phys_addr = guest_addr,
>> +		.memory_size = mem_reg_size,
>> +		.userspace_addr = (uint64_t) mem,
>> +	};
>> +
>> +	ret = ioctl(vm_get_fd(vm), KVM_SET_USER_MEMORY_REGION, &kvm_region);
>> +	TEST_ASSERT(ret == -1, "Adding one more memory slot should fail");
> Shouldn't we also check that we get the correct errno?

Good idea.

In the kvm API document there isn't any mention to the errno returned 
but looking at code it seems to be EINVAL for all errors (except for one 
EEXIST).

Thanks,

Wainer

>
>> +
>> +	munmap(mem, mem_reg_size);
>> +	kvm_vm_free(vm);
>> +
>> +	return 0;
>> +}
>> -- 
>> 2.17.2
>>
> Thanks,
> drew