[igt-dev] [RFC 2/4] tests/i915/gem_mmap: Add invalid parameters tests

[Antonio Argenziano]

On 12/03/19 15:02, Chris Wilson wrote:
> Quoting Antonio Argenziano (2019-03-12 21:57:36)
>> Add a couple of tests that supply invalid parameters to the mmap IOCTL.
>>
>> Signed-off-by: Antonio Argenziano <antonio.argenziano at intel.com>
>> ---
>>   tests/i915/gem_mmap.c | 18 ++++++++++++++++++
>>   1 file changed, 18 insertions(+)
>>
>> diff --git a/tests/i915/gem_mmap.c b/tests/i915/gem_mmap.c
>> index 0ed15878..2a676b9a 100644
>> --- a/tests/i915/gem_mmap.c
>> +++ b/tests/i915/gem_mmap.c
>> @@ -136,6 +136,24 @@ igt_main
>>                  igt_assert(ret == -1 && errno == ENOENT);
>>          }
>>   
>> +       igt_subtest("bad-offset") {
>> +               memset(&arg, 0, sizeof(arg));
>> +               arg.handle = gem_create(fd, 4096);
>> +               arg.offset = 4096 + 1;
>> +               arg.size = 4096;
>> +               ret = ioctl(fd, DRM_IOCTL_I915_GEM_MMAP, &arg);
>> +               igt_assert(ret == -1 && errno == EINVAL);
> 
> Also arg.offset = -size is a good one.
> 
> Then size=8192 and arg.offset=-4096;
> 
>> +       }
>> +
>> +       igt_subtest("bad-size") {
>> +               memset(&arg, 0, sizeof(arg));
>> +               arg.handle = gem_create(fd, 4096);
>> +               arg.offset = 4096;
>> +               arg.size = 0;
>> +               ret = ioctl(fd, DRM_IOCTL_I915_GEM_MMAP, &arg);
>> +               igt_assert(ret == -1 && errno == EINVAL);
> 
> Also >4096, -4096

Hmmm, >4096 seems not to be returning an error. Where should it be 
checked against the size of the object?

Antonio

> -Chris
>