Re: [PATCH V2] vhost-user-test: fix a memory leak

[Thomas Huth]

On 10/01/2020 15.07, Thomas Huth wrote:
> On 20/12/2019 02.26, address@hidden wrote:
>> From: Pan Nengyuan <address@hidden>
>>
>> Spotted by ASAN.
>>
>> Reported-by: Euler Robot <address@hidden>
>> Signed-off-by: Pan Nengyuan <address@hidden>
>> ---
>> Changes V2 to V1:
>> - use a "goto cleanup", instead of duplicating the "free" functions.
>> - free "dest_cmdline" at the end.
>> ---
>>  tests/vhost-user-test.c | 5 ++++-
>>  1 file changed, 4 insertions(+), 1 deletion(-)
>>
>> diff --git a/tests/vhost-user-test.c b/tests/vhost-user-test.c
>> index 91ea373..dcb8617 100644
>> --- a/tests/vhost-user-test.c
>> +++ b/tests/vhost-user-test.c
>> @@ -717,7 +717,7 @@ static void test_migrate(void *obj, void *arg, 
>> QGuestAllocator *alloc)
>>      guint64 size;
>>  
>>      if (!wait_for_fds(s)) {
>> -        return;
>> +        goto cleanup;
>>      }
>>  
>>      size = get_log_size(s);
>> @@ -776,8 +776,11 @@ static void test_migrate(void *obj, void *arg, 
>> QGuestAllocator *alloc)
>>      g_source_unref(source);
>>  
>>      qtest_quit(to);
>> +
>> + cleanup:
>>      test_server_free(dest);
>>      g_free(uri);
>> +    g_string_free(dest_cmdline, true);
>>  }
>>  
>>  static void wait_for_rings_started(TestServer *s, size_t count)
>>
> 
> Reviewed-by: Thomas Huth <address@hidden>
> 
> ... and picked up to my qtest-next tree.

... and now I had to unqueue the patch again. It is reproducibly causing
one of the gitlab CI pipelines to fail with a timeout, e.g.:

 https://gitlab.com/huth/qemu/-/jobs/400101552

Not sure what is going on here, though, there is no obvious error
message in the output... this needs some more investigation... do you
have a gitlab account and could have a look?

 Thomas