[Piglit] [PATCH] util-cl: Fix non-portable sizeof.

[Ilia Mirkin]

On Wed, Dec 17, 2014 at 2:17 PM, Matt Turner <mattst88 at gmail.com> wrote:
> On Tue, Dec 16, 2014 at 10:47 PM, Vinson Lee <vlee at freedesktop.org> wrote:
>> Fix Coverity "sizeof not portable" defect.
>>
>> Signed-off-by: Vinson Lee <vlee at freedesktop.org>
>> ---
>>  tests/util/piglit-framework-cl-program.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/tests/util/piglit-framework-cl-program.c b/tests/util/piglit-framework-cl-program.c
>> index d335486..f94525b 100644
>> --- a/tests/util/piglit-framework-cl-program.c
>> +++ b/tests/util/piglit-framework-cl-program.c
>> @@ -287,7 +287,7 @@ piglit_cl_program_test_run(const int argc,
>>         } else if(config->program_binary_file != NULL) {
>>                 unsigned int length;
>>                 size_t* lengths = malloc(sizeof(size_t) * env.context->num_devices);
>> -               unsigned char** program_binaries = malloc(sizeof(unsigned char**) * env.context->num_devices);
>> +               unsigned char** program_binaries = malloc(sizeof(unsigned char*) * env.context->num_devices);
>
> Someone explain this one to me?
>
> program_binaries is an unsigned char ** and that's what we're giving
> to sizeof. What's wrong with sizeof(unsigned char **)?

It's an array of (unsigned char *), not (unsigned char **). For
example if it were a 1d array, you might have

unsigned char *foo = malloc(sizeof(unsigned char) * num_elements)

right? Of course it doesn't end up _actually_ mattering, since
sizeof(char*) == sizeof(char**).

  -ilia