Tarantool development patches archive
 help / color / mirror / Atom feed
From: Sergey Bronnikov via Tarantool-patches <tarantool-patches@dev.tarantool.org>
To: Sergey Kaplun <skaplun@tarantool.org>
Cc: tarantool-patches@dev.tarantool.org
Subject: Re: [Tarantool-patches] [PATCH v1 luajit 05/41] perf: adjust binary-trees in LuaJIT-benches
Date: Thu, 13 Nov 2025 14:06:36 +0300	[thread overview]
Message-ID: <5d871254-9296-4eea-9b85-b636c1f80a7e@tarantool.org> (raw)
In-Reply-To: <3e3085c7a9bcb7d1a185e683c8d774a57cb7816a.1761301736.git.skaplun@tarantool.org>

[-- Attachment #1: Type: text/plain, Size: 4716 bytes --]

Hi, Sergey!

thanks for the patch!

On 10/24/25 13:50, Sergey Kaplun wrote:
> This patch adjusts the aforementioned test to use the benchmark
> framework introduced before. The default arguments are adjusted
> according to the <PARAM_x86.txt> file. The arguments to the script still
> can be provided in the command line run.
>
> The test cases are split by the different types of trees:
> 1) stretched tree,
> 2) long-lived tree,
> 3) several trees with a depth of the power of 2,
> 4) iteration over all trees in the third test case.
>
> The number of items is the number of `ItemCheck()` first-level calls
> performed in the payload.
> ---
>
> I'm not sure that we should distinguish different subtests here.
> OTOH, how to calculate the amount of items correctly for the whole test
> instead?
>
>   perf/LuaJIT-benches/binary-trees.lua | 94 ++++++++++++++++++++++------
>   1 file changed, 76 insertions(+), 18 deletions(-)
>
> diff --git a/perf/LuaJIT-benches/binary-trees.lua b/perf/LuaJIT-benches/binary-trees.lua
> index bf040466..9d4dc7b4 100644
> --- a/perf/LuaJIT-benches/binary-trees.lua
> +++ b/perf/LuaJIT-benches/binary-trees.lua
> @@ -1,3 +1,4 @@
> +local bench = require("bench").new(arg)
>   
>   local function BottomUpTree(item, depth)
>     if depth > 0 then
> @@ -18,30 +19,87 @@ local function ItemCheck(tree)
>     end
>   end
>   
> -local N = tonumber(arg and arg[1]) or 0
> +local N = tonumber(arg and arg[1]) or 16
Why 16?
>   local mindepth = 4
>   local maxdepth = mindepth + 2
>   if maxdepth < N then maxdepth = N end
>   
> -do
> -  local stretchdepth = maxdepth + 1
> -  local stretchtree = BottomUpTree(0, stretchdepth)
> -  io.write(string.format("stretch tree of depth %d\t check: %d\n",
> -    stretchdepth, ItemCheck(stretchtree)))
> -end
> +local stretchdepth = maxdepth + 1
> +
> +bench:add({
> +  name = "stretch_depth_" .. tostring(stretchdepth),
> +  payload = function()
> +    local stretchtree = BottomUpTree(0, stretchdepth)
> +    local check = ItemCheck(stretchtree)
> +    return check
> +  end,
> +  items = 1,
> +  checker = function(check)
> +    return check == -1
it deserves a comment
> +  end,
> +})
>   
> -local longlivedtree = BottomUpTree(0, maxdepth)
> +-- This tree created once on the setup for the first test.
> +local longlivedtree

I don't like that we should save a benchmark state in global variables.

What if we allow setting a user-defined object that will have a state

and this state will be passed to checker/payload functions?

>   
> -for depth=mindepth,maxdepth,2 do
> +for depth = mindepth, maxdepth, 2 do
>     local iterations = 2 ^ (maxdepth - depth + mindepth)
> -  local check = 0
> -  for i=1,iterations do
> -    check = check + ItemCheck(BottomUpTree(1, depth)) +
> -            ItemCheck(BottomUpTree(-1, depth))
> -  end
> -  io.write(string.format("%d\t trees of depth %d\t check: %d\n",
> -    iterations*2, depth, check))
> +  local tree_bench
> +  tree_bench = {
> +    name = "tree_depth_" .. tostring(depth),
> +    setup = function()
> +      if not longlivedtree then
> +        longlivedtree = BottomUpTree(0, maxdepth)
> +      end
> +      tree_bench.items = iterations * 2
> +    end,
> +    checker = function(check)
> +      return check == -iterations * 2
> +    end,
> +    payload = function()
> +      local check = 0
> +      for i = 1, iterations do
> +        check = check + ItemCheck(BottomUpTree(1, depth)) +
> +                ItemCheck(BottomUpTree(-1, depth))
> +      end
> +      return check
> +    end,
> +  }
> +
> +bench:add(tree_bench)
>   end
>   
> -io.write(string.format("long lived tree of depth %d\t check: %d\n",
> -  maxdepth, ItemCheck(longlivedtree)))
> +bench:add({
> +  name = "longlived_depth_" .. tostring(maxdepth),
> +  payload = function()
> +    local check = ItemCheck(longlivedtree)
> +    return check
> +  end,
> +  items = 1,
> +  checker = function(check)
> +    return check == -1
> +  end,
> +})
> +
> +bench:add({
> +  name = "all_in_once",
s/all_in_once/all_in_one/?
> +  payload = function()
> +    for depth = mindepth, maxdepth, 2 do
> +      local iterations = 2 ^ (maxdepth - depth + mindepth)
> +      local tree_bench
> +      local check = 0
> +      for i = 1, iterations do
> +        check = check + ItemCheck(BottomUpTree(1, depth)) +
> +                ItemCheck(BottomUpTree(-1, depth))
> +      end
> +      assert(check == -iterations * 2)
> +    end
> +  end,
> +  -- Geometric progression, starting at maxdepth trees with the
> +  -- corresponding step.
> +  items = (2 * maxdepth) * (4 ^ ((maxdepth - mindepth) / 2 + 1) - 1) / 3,
> +  -- Correctness is checked in the payload function.
> +  skip_check = true,
> +})
> +
> +bench:run_and_report()

[-- Attachment #2: Type: text/html, Size: 5781 bytes --]

  reply	other threads:[~2025-11-13 11:06 UTC|newest]

Thread overview: 58+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-10-24 10:50 [Tarantool-patches] [PATCH v1 luajit 00/41] LuaJIT performance testing Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 01/41] perf: add LuaJIT-test-cleanup perf suite Sergey Kaplun via Tarantool-patches
2025-11-11 14:28   ` Sergey Bronnikov via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 02/41] perf: introduce clock module Sergey Kaplun via Tarantool-patches
2025-11-11 14:28   ` Sergey Bronnikov via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 03/41] perf: introduce bench module Sergey Kaplun via Tarantool-patches
2025-11-11 15:41   ` Sergey Bronnikov via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 04/41] perf: adjust array3d in LuaJIT-benches Sergey Kaplun via Tarantool-patches
2025-11-13 11:06   ` Sergey Bronnikov via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 05/41] perf: adjust binary-trees " Sergey Kaplun via Tarantool-patches
2025-11-13 11:06   ` Sergey Bronnikov via Tarantool-patches [this message]
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 06/41] perf: adjust chameneos " Sergey Kaplun via Tarantool-patches
2025-11-13 11:11   ` Sergey Bronnikov via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 07/41] perf: adjust coroutine-ring " Sergey Kaplun via Tarantool-patches
2025-11-13 11:17   ` Sergey Bronnikov via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 08/41] perf: adjust euler14-bit " Sergey Kaplun via Tarantool-patches
2025-11-13 11:44   ` Sergey Bronnikov via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 09/41] perf: adjust fannkuch " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 10/41] perf: adjust fasta " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 11/41] perf: adjust k-nucleotide " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 12/41] perf: adjust life " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 13/41] perf: adjust mandelbrot-bit " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 14/41] perf: adjust mandelbrot " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 15/41] perf: adjust md5 " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 16/41] perf: adjust meteor " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 17/41] perf: adjust nbody " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 18/41] perf: adjust nsieve-bit-fp " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 19/41] perf: adjust nsieve-bit " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 20/41] perf: adjust nsieve " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 21/41] perf: adjust partialsums " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 22/41] perf: adjust pidigits-nogmp " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 23/41] perf: adjust ray " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 24/41] perf: adjust recursive-ack " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 25/41] perf: adjust recursive-fib " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 26/41] perf: adjust revcomp " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 27/41] perf: adjust scimark-2010-12-20 " Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 28/41] perf: move <scimark_lib.lua> to <libs/> directory Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 29/41] perf: adjust scimark-fft in LuaJIT-benches Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 30/41] perf: adjust scimark-lu " Sergey Kaplun via Tarantool-patches
2025-10-24 11:00   ` Sergey Kaplun via Tarantool-patches
2025-10-24 11:01   ` Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 31/41] perf: add scimark-mc " Sergey Kaplun via Tarantool-patches
2025-10-24 11:00   ` Sergey Kaplun via Tarantool-patches
2025-10-24 11:02   ` Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 32/41] perf: adjust scimark-sor " Sergey Kaplun via Tarantool-patches
2025-10-24 11:00   ` Sergey Kaplun via Tarantool-patches
2025-10-24 11:02   ` Sergey Kaplun via Tarantool-patches
2025-10-24 10:50 ` [Tarantool-patches] [PATCH v1 luajit 33/41] perf: adjust scimark-sparse " Sergey Kaplun via Tarantool-patches
2025-10-24 11:00   ` Sergey Kaplun via Tarantool-patches
2025-10-24 11:03   ` Sergey Kaplun via Tarantool-patches
2025-10-24 11:00 ` [Tarantool-patches] [PATCH v1 luajit 34/41] perf: adjust series " Sergey Kaplun via Tarantool-patches
2025-10-24 11:00 ` [Tarantool-patches] [PATCH v1 luajit 35/41] perf: adjust spectral-norm " Sergey Kaplun via Tarantool-patches
2025-10-24 11:00 ` [Tarantool-patches] [PATCH v1 luajit 36/41] perf: adjust sum-file " Sergey Kaplun via Tarantool-patches
2025-10-24 11:00 ` [Tarantool-patches] [PATCH v1 luajit 37/41] perf: add CMake infrastructure Sergey Kaplun via Tarantool-patches
2025-10-24 11:00 ` [Tarantool-patches] [PATCH v1 luajit 38/41] perf: add aggregator helper for bench statistics Sergey Kaplun via Tarantool-patches
2025-10-24 11:00 ` [Tarantool-patches] [PATCH v1 luajit 39/41] perf: add a script for the environment setup Sergey Kaplun via Tarantool-patches
2025-10-24 11:00 ` [Tarantool-patches] [PATCH v1 luajit 40/41] perf: provide CMake option to setup the benchmark Sergey Kaplun via Tarantool-patches
2025-10-24 11:00 ` [Tarantool-patches] [PATCH v1 luajit 41/41] ci: introduce the performance workflow Sergey Kaplun via Tarantool-patches

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5d871254-9296-4eea-9b85-b636c1f80a7e@tarantool.org \
    --to=tarantool-patches@dev.tarantool.org \
    --cc=sergeyb@tarantool.org \
    --cc=skaplun@tarantool.org \
    --subject='Re: [Tarantool-patches] [PATCH v1 luajit 05/41] perf: adjust binary-trees in LuaJIT-benches' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox