ci: 计时表只留每平台最重的 10 个,加测试不再需要动它

这张表之前列出每个成员,于是每加一个测试它就过期一次 —— 而这次撑爆分片的
正是"表里没有的成员按中位数计价":mysql-connector-cpp 实测 881s,被当成 60s。

但打包决策其实只由重的那几个做出。linux 上 top10 占总量的 69%,其余 57 个平
均 86s、中位 24s。把它们从表里拿掉,最慢分片一分钟都不动:

    完整表 67 行,4 片    74 / 64 / 63 / 63    最慢 74 分
    top10 表,4 片        74 / 73 / 60 / 58    最慢 74 分

## 兜底价必须是固定常数,不能再取中位数

直接缩表会当场炸:表里只剩 10 个重的,中位数就变成"第 5 重的那个",在 linux
上是 875s,于是 57 个小成员每个都按 875s 计价 ——

    top10 表 + 中位数兜底   90 / 48 / 81 / 46    最慢 90 分  ← 正好撞 job 上限

改成固定 90s(未计时成员的实测均值 86s 取整)。这个值不吃调参:兜底价从 30s
到 300s,最慢分片始终在 72–84 分之间,全都在 90 以下。这种不敏感正是"表可以
长期不动"的依据。

## shards_for 也得跟着改,否则片数会掉回去

它原本直接把表里的行加起来当工作量。表一缩,linux 总量从 15891s 读成 10965s
→ 3 片,而 3 片在这张表下最慢 107 分,直接超时。

给 plan_shards 加一个 `<platform> 0 0` 模式返回总量,shards_for 改调它。这样
"有哪些成员"和"未计时的算多少钱"只有一份定义,而不是 yaml 和 lua 各一份等着
漂移。估算精度:linux 16095s vs 实测 15891s,+1.3%。macOS / windows 的成员比
默认价便宜,总量偏高 —— 两者本来就顶着 cap 2,而且偏多分片是安全方向。

## timings job 同步产出 top10

否则下次刷新又胖回 197 行。

端到端复核:总量 16095s → 4 片,四片覆盖 67/67,最慢 74 分。

Author: Sunrisepeak

.github/workflows/validate.yml

@@ -478,20 +478,25 @@ jobs:
           # `libmysqlclient` (329s) landed on shard 0 next to grpc-codegen
           # exactly that way and pushed it past the cap, and the stale table
           # then hid the overflow by under-counting the total.
+          # `plan_shards.lua <platform> 0 0` reports the platform's total cost.
+          #
+          # This used to sum the timing table's rows inline, which was the same
+          # answer only while the table listed every member. It now lists just
+          # the heavy ones, so summing rows measures a fraction of the work:
+          # 10965s instead of 16095s on linux, i.e. 3 shards where 4 are
+          # needed, i.e. a slowest shard of 107 minutes against a 90-minute
+          # cap. Asking plan_shards keeps ONE definition of both "which members
+          # are there" and "what does an untimed one cost".
+          #
+          # The estimate is tuned on linux (16095s vs 15891s measured, +1.3%).
+          # macOS and windows members are cheaper on average than the default,
+          # so their totals come out high — harmless here because both are
+          # already at their cap, and erring toward more shards is the safe
+          # direction. Revisit if either cap moves.
           shards_for() {   # platform cap -> shard count
-            local secs
-            secs=$(XPLAT="$1" lua5.4 -e '
-              local plat = os.getenv("XPLAT")
-              local sel, all = {}, (os.getenv("MEMBERS") == "__ALL__")
-              if not all then
-                for m in (os.getenv("MEMBERS") or ""):gmatch("%S+") do sel[m] = true end
-              end
-              local total = 0
-              for line in io.lines("tests/member-timings.tsv") do
-                local p, m, s = line:match("^(%S+)\t(%S+)\t(%d+)$")
-                if p == plat and m and (all or sel[m]) then total = total + tonumber(s) end
-              end
-              print(total)')
+            local secs sel=""
+            [ "$MEMBERS" = "__ALL__" ] || sel="$MEMBERS"
+            secs=$(lua5.4 tests/plan_shards.lua "$1" 0 0 $sel)
             local n=$(( secs / 4200 + 1 ))
             [ "$n" -gt "$2" ] && n="$2"
             echo "$1 work: ${secs}s (measured) -> $n shard(s)" >&2
@@ -983,25 +988,37 @@ jobs:
           # silently absorb a one-off slow runner. Refresh it deliberately —
           # download this artifact and replace tests/member-timings.tsv when
           # the numbers have actually moved.
+          # Top 10 per platform, not every member.
+          #
+          # plan_shards prices anything absent at a fixed default, and the
+          # packing decisions are made by the heavy members anyway — on linux
+          # the top 10 of 67 are 69% of the total, and dropping the other 57
+          # moves the slowest shard by 0 minutes. A table that lists everyone
+          # goes stale the moment someone adds a test; one that lists ten only
+          # goes stale when the heavy set actually changes. That is the whole
+          # point: adding a member must not require touching this file.
           {
-            echo "# <platform>\t<member>\t<seconds> — from run ${{ github.run_id }}"
-            echo "# refresh: download the member-timings artifact and replace this file"
+            echo "# <platform>\t<member>\t<seconds> — the HEAVY members only, top 10"
+            echo "# per platform, from run ${{ github.run_id }}. Everything else is"
+            echo "# priced at plan_shards.lua's fixed default; adding a test does NOT"
+            echo "# require touching this file. Refresh when a member becomes heavy"
+            echo "# enough to enter the top 10, or one of these numbers moves."
+            # ONLY the default-toolchain leg feeds this table. The llvm leg
+            # runs the SAME members again, so globbing every leg would put two
+            # rows per (platform, member) into the file — `sort -u` keeps both,
+            # since the seconds differ. Both legs are planned from these
+            # numbers, so the default leg is the right single baseline.
+            #
+            # Then: heaviest first, keep 10. `sort -u` before that so a member
+            # reported by two shards cannot occupy two of the ten slots.
             for plat in linux macos windows; do
-              # ONLY the default-toolchain leg feeds this table. The llvm leg
-              # runs the SAME members again, so globbing every leg would put
-              # two rows per (platform, member) into the file — `sort -u`
-              # keeps both, since the seconds differ — and shards_for sums
-              # every matching row. Linux work would read as roughly double
-              # and its shard count would be permanently pinned at the cap.
-              # Both legs are planned from these numbers, so the default leg
-              # is the right single baseline.
               for f in timings/timings-$plat-default-*/timings.tsv; do
                 [ -f "$f" ] || continue
                 awk -F'\t' -v p="$plat" '{ printf "%s\t%s\t%s\n", p, $2, $1 }' "$f"
-              done
+              done | sort -u | sort -t"$(printf '\t')" -k3,3nr | head -10
             done
-          } | sort -u > member-timings.tsv
-          echo "wrote member-timings.tsv ($(grep -vc '^#' member-timings.tsv) rows)"
+          } > member-timings.tsv
+          echo "wrote member-timings.tsv ($(grep -vc '^#' member-timings.tsv) rows, top 10/platform)"
 
       - name: Upload the timing table for the next run's sharding
         if: always() && hashFiles('member-timings.tsv') != ''

tests/member-timings.tsv

@@ -1,203 +1,43 @@
-# <platform>\t<member>\t<seconds> — from run 31260545520
-# refresh: download the member-timings artifact and replace this file
-# libmysqlclient / mysql-connector-cpp: measured from run 31266814148's
-# linux default 0/3 step durations — that shard was cancelled at the 90m
-# cap so its timings artifact never uploaded, and these two are exactly
-# the members whose absence caused the overflow. Chicken-and-egg broken by hand, once.
-linux	abseil	96
-linux	archive	28
-linux	asio-module	15
-linux	asio-ssl	175
-linux	boost-ext.ut	4
-linux	build-mcpp	1
-linux	c-ares	5
-linux	catch2	59
-linux	catch2-main	57
-linux	catch2-v2	14
-linux	catch2-v2-main	15
-linux	cjson	19
-linux	cli11	8
-linux	cmdline	6
-linux	core	78
-linux	curl	264
-linux	eigen	24
-linux	eui-neo	43
-linux	eui-neo-app-main	48
-linux	eui-neo-markdown	144
-linux	eui-neo-sdl2	591
-linux	eui-neo-vulkan	143
-linux	eui-neo-window	143
-linux	ffmpeg	296
-linux	ffmpeg-module	277
-linux	fmtlib.fmt	4
-linux	freetype	11
-linux	glad	3
-linux	godot-cpp	564
-linux	godot-cpp-module	627
-linux	godot-cpp-module-v10	568
-linux	godot-cpp-v10	504
+# <platform>\t<member>\t<seconds> — the HEAVY members only, top 10 per
+# platform. Everything else is priced at plan_shards.lua's fixed default;
+# on linux the 57 members not listed here average 86s and are 31% of the
+# total, and listing them changes the slowest shard by 0 minutes.
+#
+# So: adding a test does NOT require touching this file. Refresh it when
+# a member becomes heavy enough to enter the top 10, or when one of these
+# numbers moves materially — download the member-timings artifact from a
+# full run and keep the top 10 per platform.
+#
+# Sources: run 31260545520; libmysqlclient / mysql-connector-cpp measured
+# from run 31266814148's linux default 0/3 step durations, whose shard was
+# cancelled at the 90m cap before its timings artifact could upload.
 linux	grpc-codegen	3463
 linux	grpc-module	1771
-linux	gui-stack	128
-linux	imgui	7
-linux	imgui-module	106
-linux	imgui-window	7
-linux	libmysqlclient	329
-linux	libpng	7
-linux	llamacpp	132
-linux	llamacpp-metal	21
-linux	llmapi	19
-linux	magic_enum	4
-linux	marzer.tomlplusplus	7
-linux	md4c	23
+linux	protobuf-protoc	885
 linux	mysql-connector-cpp	881
-linux	nlohmann.json	10
-linux	openblas	0
-linux	opencv-module	649
 linux	opencv-module-dnn	875
 linux	opencv-module-unifont	655
-linux	openssl	166
-linux	protobuf	175
-linux	protobuf-gzip	267
-linux	protobuf-protoc	885
-linux	protobuf-upb	179
-linux	re2	16
-linux	sdl2	108
-linux	spdlog	9
-linux	spdlog-compiled	13
-linux	tinyhttps	11
-linux	tray	2
-linux	vulkan	76
-linux	websocket	20
-linux	websocket-features	26
-linux	yyjson	10
-macos	abseil	48
-macos	archive	21
-macos	asio-module	10
-macos	asio-ssl	65
-macos	boost-ext.ut	4
-macos	build-mcpp	3
-macos	c-ares	5
-macos	catch2	24
-macos	catch2-main	27
-macos	catch2-v2	7
-macos	catch2-v2-main	9
-macos	cjson	3
-macos	cli11	5
-macos	cmdline	4
-macos	core	44
-macos	curl	66
-macos	eigen	8
-macos	eui-neo	47
-macos	eui-neo-app-main	49
-macos	eui-neo-markdown	64
-macos	eui-neo-sdl2	138
-macos	eui-neo-vulkan	59
-macos	eui-neo-window	48
-macos	ffmpeg	43
-macos	ffmpeg-module	107
-macos	fmtlib.fmt	3
-macos	freetype	10
-macos	glad	5
-macos	godot-cpp	5
-macos	godot-cpp-module	276
-macos	godot-cpp-module-v10	195
-macos	godot-cpp-v10	206
+linux	opencv-module	649
+linux	godot-cpp-module	627
+linux	eui-neo-sdl2	591
+linux	godot-cpp-module-v10	568
 macos	grpc-codegen	1156
 macos	grpc-module	719
-macos	gui-stack	25
-macos	imgui	4
-macos	imgui-module	21
-macos	imgui-window	0
-macos	libpng	6
-macos	llamacpp	63
-macos	llamacpp-metal	71
-macos	llmapi	16
-macos	magic_enum	4
-macos	marzer.tomlplusplus	6
-macos	md4c	3
-macos	nlohmann.json	8
-macos	openblas	0
-macos	opencv-module	223
 macos	opencv-module-dnn	325
-macos	opencv-module-unifont	164
-macos	openssl	83
-macos	protobuf	71
-macos	protobuf-gzip	97
 macos	protobuf-protoc	314
-macos	protobuf-upb	66
-macos	re2	8
-macos	sdl2	23
-macos	spdlog	4
-macos	spdlog-compiled	6
-macos	tinyhttps	14
-macos	tray	3
-macos	vulkan	8
-macos	websocket	10
-macos	websocket-features	11
-macos	yyjson	5
-windows	abseil	103
-windows	archive	62
-windows	asio-module	15
-windows	asio-ssl	0
-windows	boost-ext.ut	4
-windows	build-mcpp	4
-windows	c-ares	16
-windows	catch2	74
-windows	catch2-main	69
-windows	catch2-v2	10
-windows	catch2-v2-main	11
-windows	cjson	3
-windows	cli11	8
-windows	cmdline	5
-windows	core	100
-windows	curl	27
-windows	eigen	12
-windows	eui-neo	75
-windows	eui-neo-app-main	88
-windows	eui-neo-markdown	86
-windows	eui-neo-sdl2	123
-windows	eui-neo-vulkan	86
-windows	eui-neo-window	83
-windows	ffmpeg	693
-windows	ffmpeg-module	526
-windows	fmtlib.fmt	6
-windows	freetype	13
-windows	glad	4
-windows	godot-cpp	7
+macos	godot-cpp-module	276
+macos	opencv-module	223
+macos	godot-cpp-v10	206
+macos	godot-cpp-module-v10	195
+macos	opencv-module-unifont	164
+macos	eui-neo-sdl2	138
+windows	opencv-module	868
+windows	opencv-module-dnn	828
 windows	godot-cpp-module	707
+windows	ffmpeg	693
 windows	godot-cpp-module-v10	657
 windows	godot-cpp-v10	655
-windows	grpc-codegen	10
-windows	grpc-module	0
-windows	gui-stack	19
-windows	imgui	6
-windows	imgui-module	24
-windows	imgui-window	0
-windows	libpng	10
-windows	llamacpp	126
-windows	llamacpp-metal	1
-windows	llmapi	16
-windows	magic_enum	4
-windows	marzer.tomlplusplus	7
-windows	md4c	3
-windows	nlohmann.json	10
-windows	openblas	10
-windows	opencv-module	868
-windows	opencv-module-dnn	828
-windows	opencv-module-unifont	7
-windows	openssl	1
-windows	protobuf	139
+windows	ffmpeg-module	526
 windows	protobuf-gzip	238
 windows	protobuf-protoc	152
 windows	protobuf-upb	151
-windows	re2	17
-windows	sdl2	47
-windows	spdlog	22
-windows	spdlog-compiled	12
-windows	tinyhttps	16
-windows	tray	2
-windows	vulkan	5
-windows	websocket	25
-windows	websocket-features	29
-windows	yyjson	4