Re: [Tarantool-patches] [PATCH v2 luajit 31/41] perf: add scimark-mc in LuaJIT-benches

[Sergey Bronnikov via Tarantool-patches <tarantool-patches@dev.tarantool.org>]

[-- Attachment #1: Type: text/plain, Size: 1855 bytes --]

Hi, Sergey,

LGTM then

Sergey

On 1/3/26 09:05, Sergey Kaplun wrote:
> Hi, Sergey!
> Thanks for the review!
> Please consider my answer below.
>
> On 02.01.26, Sergey Bronnikov wrote:
>> Hi, Sergey!
>>
>> thanks for the patch! LGTM with a minor comment.
>>
>> Sergey
>>
>> On 12/26/25 12:18, Sergey Kaplun wrote:
>>> This patch adds the aforementioned test with the use of the benchmark
>>> framework introduced before. The default arguments are adjusted
>>> according to the amount of cycles in the <scimark-2010-12-20.lua> file.
>>> The arguments to the script can be provided in the command line run.
>>>
>>> Checks are omitted since they were not present in the original suite,
>>> plus the precise result value depends on the input parameter.
>>> ---
>>>    perf/LuaJIT-benches/scimark-mc.lua | 19 +++++++++++++++++++
>>>    1 file changed, 19 insertions(+)
>>>    create mode 100644 perf/LuaJIT-benches/scimark-mc.lua
>>>
>>> diff --git a/perf/LuaJIT-benches/scimark-mc.lua b/perf/LuaJIT-benches/scimark-mc.lua
>>> new file mode 100644
>>> index 00000000..d26b6e48
>>> --- /dev/null
>>> +++ b/perf/LuaJIT-benches/scimark-mc.lua
>>> @@ -0,0 +1,19 @@
>>> +local bench = require("bench").new(arg)
>>> +
>>> +local cycles = tonumber(arg and arg[1]) or 15e7
>> Do we want to add this to the usage?
> I suppose there is no need for it.
> This still may be done if you are exploring the benchmark behaviour. But
> you will read its sources in this case anyway.
>
>>> +
>>> +local benchmark
>>> +benchmark = {
>>> +  name = "scimark_mc",
>>> +  -- XXX: The description of tests for the function is too
>>> +  -- inconvenient.
>>> +  skip_check = true,
>>> +  payload = function()
>>> +    local flops = require("scimark_lib").MC()(cycles)
>>> +    benchmark.items = flops
>>> +  end,
>>> +}
>>> +
>>> +bench:add(benchmark)
>>> +
>>> +bench:run_and_report()

[-- Attachment #2: Type: text/html, Size: 2623 bytes --]