Re: [PATCH V2 07/13] selftests/resctrl: Only support measured read operation

[Reinette Chatre <reinette.chatre@intel.com>]

Hi Ilpo,

On 9/30/24 6:52 AM, Ilpo Järvinen wrote:
> On Thu, 12 Sep 2024, Reinette Chatre wrote:

>> diff --git a/tools/testing/selftests/resctrl/resctrl.h b/tools/testing/selftests/resctrl/resctrl.h
>> index 51f5f4b25e06..ba1ce1b35699 100644
>> --- a/tools/testing/selftests/resctrl/resctrl.h
>> +++ b/tools/testing/selftests/resctrl/resctrl.h
>> @@ -142,7 +142,7 @@ int perf_event_open(struct perf_event_attr *hw_event, pid_t pid, int cpu,
>>  unsigned char *alloc_buffer(size_t buf_size, int memflush);
>>  void mem_flush(unsigned char *buf, size_t buf_size);
>>  void fill_cache_read(unsigned char *buf, size_t buf_size, bool once);
>> -int run_fill_buf(size_t buf_size, int memflush, int op);
>> +int run_fill_buf(size_t buf_size, int memflush);
>>  int initialize_mem_bw_imc(void);
>>  int measure_mem_bw(const struct user_params *uparams,
>>  		   struct resctrl_val_param *param, pid_t bm_pid,
>> diff --git a/tools/testing/selftests/resctrl/resctrl_tests.c b/tools/testing/selftests/resctrl/resctrl_tests.c
>> index bee4123a5a9b..60627dbae20a 100644
>> --- a/tools/testing/selftests/resctrl/resctrl_tests.c
>> +++ b/tools/testing/selftests/resctrl/resctrl_tests.c
>> @@ -265,13 +265,16 @@ int main(int argc, char **argv)
>>  			ksft_exit_fail_msg("Out of memory!\n");
>>  		uparams.benchmark_cmd[1] = span_str;
>>  		uparams.benchmark_cmd[2] = "1";
>> -		uparams.benchmark_cmd[3] = "0";
>>  		/*
>> +		 * Third parameter was previously used for "operation"
>> +		 * (read/write) of which only (now default) "read"/"0"
>> +		 * works.
>>  		 * Fourth parameter was previously used to indicate
>>  		 * how long "fill_buf" should run for, with "false"
>>  		 * ("fill_buf" will keep running until terminated)
>>  		 * the only option that works.
>>  		 */
>> +		uparams.benchmark_cmd[3] = NULL;
>>  		uparams.benchmark_cmd[4] = NULL;
>>  		uparams.benchmark_cmd[5] = NULL;
> 
> The same question as with the previous patch, why is [4] = NULL kept 
> around?
> 

You are correct that functionally this is not required. If this parameter
disappears at this point then there is no record of parameter 4 ever
being used. Even though this is user space I do still have my kernel view
that we should aim to maintain ABI. This means that parameter 4 will always
be "used" to indicate how long fill_buf should run for and if "fill_buf" ever
needs a new parameter, it cannot use parameter 4 since that already has
a meaning.
While the above may seem unnecessary, I think it makes the more robust
parameter processing found in patch #9 that replaces it easier to understand.
In that patch the comments above are coded to ensure parameter values are as
expected and parameter 4 continue to be dedicated to how long "fill_buf"
should run for.

As you mention in similar feedback to patch #6, the [5] assignment is
also unnecessary. Since it is just used as termination I can remove it.

Reinette