Juan Quintela <[email protected]> writes:
> Fabiano Rosas <[email protected]> wrote:
>> Juan Quintela <[email protected]> writes:
>>
>>> Fabiano Rosas <[email protected]> wrote:
>>> D> Juan Quintela <[email protected]> writes:
>>>>
>>>>> From: Fabiano Rosas <[email protected]>
>>>>>
>>>>> Add basic tests for file-based migration.
>>>>>
>>>>> Note that we cannot use test_precopy_common because that routine
>>>>> expects it to be possible to run the migration live. With the file
>>>>> transport there is no live migration because we must wait for the
>>>>> source to finish writing the migration data to the file before the
>>>>> destination can start reading. Add a new migration function
>>>>> specifically to handle the file migration.
>>>>>
>>>>> Reviewed-by: Peter Xu <[email protected]>
>>>>> Reviewed-by: Juan Quintela <[email protected]>
>>>>> Signed-off-by: Fabiano Rosas <[email protected]>
>>>>> Signed-off-by: Juan Quintela <[email protected]>
>>>>> Message-ID: <[email protected]>
>>>
>>>>> +static void file_offset_finish_hook(QTestState *from, QTestState *to,
>>>>> + void *opaque)
>>>>> +{
>>>>> +#if defined(__linux__)
>>>>> + g_autofree char *path = g_strdup_printf("%s/%s", tmpfs,
>>>>> FILE_TEST_FILENAME);
>>>>> + size_t size = FILE_TEST_OFFSET + sizeof(QEMU_VM_FILE_MAGIC);
>>>>> + uintptr_t *addr, *p;
>>>>> + int fd;
>>>>> +
>>>>> + fd = open(path, O_RDONLY);
>>>>> + g_assert(fd != -1);
>>>>> + addr = mmap(NULL, size, PROT_READ, MAP_SHARED, fd, 0);
>>>>> + g_assert(addr != MAP_FAILED);
>>>>> +
>>>>> + /*
>>>>> + * Ensure the skipped offset contains zeros and the migration
>>>>> + * stream starts at the right place.
>>>>> + */
>>>>> + p = addr;
>>>>> + while (p < addr + FILE_TEST_OFFSET / sizeof(uintptr_t)) {
>>>>> + g_assert(*p == 0);
>>>>> + p++;
>>>>> + }
>>>>> + g_assert_cmpint(cpu_to_be32(*p), ==, QEMU_VM_FILE_MAGIC);
>>>>
>>>> This truncates to 32-bits, so it breaks on a BE host. We need this:
>>>>
>>>> -->8--
>>>> From ea0c2d1c988add48d9754891a9fc7f6854a9718a Mon Sep 17 00:00:00 2001
>>>> From: Fabiano Rosas <[email protected]>
>>>> Date: Mon, 16 Oct 2023 15:21:49 -0300
>>>> Subject: [PATCH] fixup! tests/qtest: migration-test: Add tests for
>>>> file-based
>>>> migration
>>>>
>>>> ---
>>>> tests/qtest/migration-test.c | 2 +-
>>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>>
>>>> diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
>>>> index da02b6d692..e1c110537b 100644
>>>> --- a/tests/qtest/migration-test.c
>>>> +++ b/tests/qtest/migration-test.c
>>>> @@ -1966,7 +1966,7 @@ static void file_offset_finish_hook(QTestState
>>>> *from, QTestState *to,
>>>> g_assert(*p == 0);
>>>> p++;
>>>> }
>>>> - g_assert_cmpint(cpu_to_be32(*p), ==, QEMU_VM_FILE_MAGIC);
>>>> + g_assert_cmpint(cpu_to_be64(*p) >> 32, ==, QEMU_VM_FILE_MAGIC);
>>>>
>>>> munmap(addr, size);
>>>> close(fd);
>>>
>>> I am resubmitting with this change.
>>>
>>> But I think we need to change this:
>>>
>>>>> + g_autofree char *path = g_strdup_printf("%s/%s", tmpfs,
>>>>> FILE_TEST_FILENAME);
>>>>> + size_t size = FILE_TEST_OFFSET + sizeof(QEMU_VM_FILE_MAGIC);
>>>>> + uintptr_t *addr, *p;
>>>
>>> I think we should change the test so the file is 64 bits on every
>>> architecture.
>>> Then we can cast to void * or uintptr_t as needed.
>>
>> Hm, I don't get what you mean here. What needs to be 64 bits?
>
> size_t is 32 bits on 32bits host, and 64 bits on 64 bits host.
> uintprt_t is the same.
Right, I have thought of that when writing this fix yesterday, but I
dismissed it because I thought we were never have a 32bit host running
these tests.
> So using explicit sizes:
>
> static void file_offset_finish_hook(QTestState *from, QTestState *to,
> void *opaque)
> {
> #if defined(__linux__)
> g_autofree char *path = g_strdup_printf("%s/%s", tmpfs,
> FILE_TEST_FILENAME);
> size_t size = FILE_TEST_OFFSET + sizeof(QEMU_VM_FILE_MAGIC);
> uint64_t *addr, *p;
> int fd;
>
> fd = open(path, O_RDONLY);
> g_assert(fd != -1);
> addr = mmap(NULL, size, PROT_READ, MAP_SHARED, fd, 0);
> g_assert(addr != MAP_FAILED);
>
> /*
> * Ensure the skipped offset contains zeros and the migration
> * stream starts at the right place.
> */
> p = addr;
> while (p < (uintprt_t)addr + FILE_TEST_OFFSET) {
> g_assert(*p == 0);
> p++;
> }
> g_assert_cmpint(cpu_to_be64(*p) >> 32, ==, QEMU_VM_FILE_MAGIC);
>
> munmap(addr, size);
> close(fd);
> #endif
> }
>
> This is completely untested, but it should make sure that we are reading
> 64bits integers in both 32 and 64 bits hosts, no?
Looks like it. I can give it a try and send an update as a separate
patch.
