[PATCH 2/4] ws2_32/tests: Test ICMP.

[Zebediah Figura]

On 7/9/22 02:47, Saulius Krasuckas wrote:
> On 2022-07-09 01:25, Zebediah Figura wrote:
>> On 7/5/22 14:53, Paul Gofman wrote:
>     ...
>>> +static void test_icmp(const char *dest_addr)
>>
>> Any reason to pass this as a parameter? I'm guessing it's convenient
>> to test ping over a real adapter, but presumably it can also just be
>> edited inside the function...
> 
> Now it's doing:
> 
>>> +    test_icmp("127.0.0.1");
> 
> What if someone adds a real adapter identifying logic and invokes test
> for it too?  Eg. to see the differences between it and the loopback.
> Will it require to duplicate all of the function's body?  This would
> seem like a wasted efforts to me.  It does at least without having
> actual differences at hand (proving to be too many inconsistencies
> between both runs and it being worth to rewrite the body for a real
> adapter).

No, why would it? It'd be simple to turn that back into a parameter.