Expand Windows CI coverage

[timsehn]

Summary

• add a Windows path smoke test for temp paths, paths with spaces, and MSYS Windows-style paths
• run the plaintext HTTP remotesrv suite on Windows, including the amalgamation HTTP remote test
• run the native C regression harness on Windows by adding the needed Windows link libraries
• raise the build/test job timeout so the expanded Windows surface can complete

Validation

• bash -n test/windows_path_smoke.sh test/run_doltlite_regression_case.sh test/amalgamation_http_remote_test.sh test/http_remote_content_length_test.sh test/remotesrv_http_test.sh test/remotesrv_http_concurrency_test.sh test/remotesrv_http_partial_failure_test.sh test/remotesrv_http_force_push_test.sh
• bash test/windows_path_smoke.sh build/doltlite
• bash test/run_doltlite_regression_case.sh chunk_store_full_pathname
• bash test/run_doltlite_regression_case.sh all
• cd build && bash ../test/http_remote_content_length_test.sh ./doltlite ./doltlite-remotesrv && bash ../test/remotesrv_http_test.sh ./doltlite ./doltlite-remotesrv && bash ../test/remotesrv_http_concurrency_test.sh ./doltlite ./doltlite-remotesrv && bash ../test/remotesrv_http_partial_failure_test.sh ./doltlite ./doltlite-remotesrv && bash ../test/remotesrv_http_force_push_test.sh ./doltlite ./doltlite-remotesrv
• cd build && make DOLTLITE_PROLLY=1 DOLTLITE_PROLLY_CHECK=1 sqlite3.c sqlite3.h && REMOTESRV=./doltlite-remotesrv bash ../test/amalgamation_http_remote_test.sh .
• git diff --check

[github-actions]

Version-Control Performance Ceilings

Runs: median of 3 executions per benchmark, excluding fixture setup. The
working-set fixtures use 800 committed tables with 125 rows each
(100000 rows total); data-dirty cases update 1 row in 40
tables, and schema-dirty cases update 1 row plus add a column in 20 tables.
Branch tests use 300 branches, checkout uses a clean 400-table
branch switch over 100000 rows, and merge
tests use 100000-row tables with 2000 changed or conflicting
rows per side.

Benchmark	Median ms	Ceiling ms	Result
status_clean_many_tables	84	200	PASS
status_dirty_many_tables	88	200	PASS
diff_regular_working_one_table	80	150	PASS
diff_regular_working_many_tables	93	200	PASS
diff_stat_working_many_tables	92	200	PASS
diff_schema_working_many_tables	93	200	PASS
branch_list_many_branches	23	100	PASS
branch_create_delete	26	100	PASS
checkout_branch_clean	58	200	PASS
merge_data_no_conflicts	31	150	PASS
merge_schema_no_conflicts	24	100	PASS
merge_data_conflicts	128	250	PASS
merge_data_conflicts_with_resolve	128	250	PASS

[github-actions]

Sysbench-Style Benchmark (BLOB PK): Doltlite vs SQLite

Companion to the classic Sysbench-Style Benchmark. Every workload here
runs against tables with a 16-byte big-endian BLOB PRIMARY KEY.
Individual ratios gated at 2.5×; section averages gated at 2×. Autocommit writes use 6× / 5×.

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	31,434	37,008	1.18
oltp_range_select	13,566	14,271	1.05
oltp_sum_range	12,390	14,078	1.14
oltp_order_range	2,932	3,155	1.08
oltp_distinct_range	3,924	4,233	1.08
oltp_index_scan	4,448	6,071	1.36
select_random_points	18,502	20,346	1.10
select_random_ranges	4,004	5,158	1.29
covering_index_scan	4,317	4,489	1.04
groupby_scan	31,429	35,094	1.12
index_join	6,688	8,881	1.33
index_join_scan	4,154	5,243	1.26
types_table_scan	1,040,201	1,222,841	1.18
table_scan	1,191,377	1,316,887	1.11
oltp_read_only	118,123	136,931	1.16
Average			1.16

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	242,102	357,255	1.48
oltp_insert	19,796	39,062	1.97
oltp_update_index	68,620	129,385	1.89
oltp_update_non_index	49,096	84,926	1.73
oltp_delete_insert	49,845	102,149	2.05
oltp_write_only	28,179	61,338	2.18
types_delete_insert	33,616	54,240	1.61
oltp_read_write	88,860	138,827	1.56
Average			1.81

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	100,637	61,664	0.61
oltp_range_select	21,457	16,742	0.78
oltp_sum_range	20,453	16,601	0.81
oltp_order_range	3,706	3,452	0.93
oltp_distinct_range	4,777	4,729	0.99
oltp_index_scan	11,845	8,815	0.74
select_random_points	26,910	23,973	0.89
select_random_ranges	11,253	7,643	0.68
covering_index_scan	12,293	7,226	0.59
groupby_scan	32,785	35,554	1.08
index_join	11,276	11,154	0.99
index_join_scan	5,206	5,726	1.10
types_table_scan	1,038,715	1,220,078	1.17
table_scan	1,199,663	1,317,437	1.10
oltp_read_only	223,693	174,031	0.78
Average			0.88

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	255,083	381,380	1.50
oltp_insert	32,100	51,562	1.61
oltp_update_index	100,701	160,323	1.59
oltp_update_non_index	78,882	106,348	1.35
oltp_delete_insert	79,541	129,653	1.63
oltp_write_only	55,486	82,597	1.49
types_delete_insert	52,513	71,808	1.37
oltp_read_write	115,429	158,682	1.37
Average			1.49

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	54,602	61,184	1.12
oltp_range_select	16,277	16,939	1.04
oltp_sum_range	15,231	16,912	1.11
oltp_order_range	3,212	3,505	1.09
oltp_distinct_range	4,332	4,692	1.08
oltp_index_scan	6,980	8,827	1.26
select_random_points	22,208	24,107	1.09
select_random_ranges	6,666	7,712	1.16
covering_index_scan	7,087	7,194	1.02
groupby_scan	32,031	35,406	1.11
index_join	8,485	11,119	1.31
index_join_scan	4,807	5,740	1.19
types_table_scan	1,035,322	1,216,658	1.18
table_scan	1,210,181	1,311,364	1.08
oltp_read_only	155,425	172,504	1.11
Average			1.13

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	21,682	77,484	3.57
oltp_insert_ac	23,969	99,035	4.13
oltp_update_index_ac	25,813	108,931	4.22
oltp_update_non_index_ac	22,502	89,131	3.96
oltp_delete_insert_ac	23,682	102,164	4.31
oltp_write_only_ac	23,868	100,191	4.20
types_delete_insert_ac	20,698	89,404	4.32
oltp_read_write_ac	30,932	110,834	3.58
Average			4.04

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.

[github-actions]

Sysbench-Style Benchmark (composite PK): Doltlite vs SQLite

Companion to the classic Sysbench-Style Benchmark. Every workload here
runs against tables with a 2-column INTEGER PRIMARY KEY(a, b) WITHOUT ROWID.
Individual ratios gated at 2.5×; section averages gated at 2×. Autocommit writes use 6× / 5×.

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	35,649	41,774	1.17
oltp_range_select	19,282	22,242	1.15
oltp_sum_range	18,420	21,821	1.18
oltp_order_range	3,610	3,942	1.09
oltp_distinct_range	4,662	5,109	1.10
oltp_index_scan	4,825	6,612	1.37
select_random_points	28,983	33,092	1.14
select_random_ranges	7,831	9,265	1.18
covering_index_scan	4,346	4,692	1.08
groupby_scan	36,656	40,982	1.12
index_join	8,263	11,372	1.38
index_join_scan	4,249	5,777	1.36
types_table_scan	1,149,293	1,268,314	1.10
table_scan	1,385,790	1,366,640	0.99
oltp_read_only	155,647	175,597	1.13
Average			1.17

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	246,288	364,526	1.48
oltp_insert	19,174	37,273	1.94
oltp_update_index	70,649	121,020	1.71
oltp_update_non_index	51,783	85,388	1.65
oltp_delete_insert	51,764	98,426	1.90
oltp_write_only	28,301	60,139	2.12
types_delete_insert	32,879	55,734	1.70
oltp_read_write	109,802	161,006	1.47
Average			1.75

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	105,395	66,259	0.63
oltp_range_select	27,195	25,056	0.92
oltp_sum_range	25,935	24,422	0.94
oltp_order_range	4,400	4,296	0.98
oltp_distinct_range	5,464	5,512	1.01
oltp_index_scan	11,757	9,085	0.77
select_random_points	37,307	36,444	0.98
select_random_ranges	15,417	12,038	0.78
covering_index_scan	11,482	7,230	0.63
groupby_scan	36,771	41,229	1.12
index_join	12,305	13,149	1.07
index_join_scan	5,219	6,109	1.17
types_table_scan	1,063,584	1,244,936	1.17
table_scan	1,331,605	1,346,079	1.01
oltp_read_only	255,621	213,360	0.83
Average			0.93

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	259,787	381,888	1.47
oltp_insert	26,342	47,270	1.79
oltp_update_index	99,253	144,920	1.46
oltp_update_non_index	80,481	106,479	1.32
oltp_delete_insert	77,836	120,918	1.55
oltp_write_only	52,194	78,389	1.50
types_delete_insert	49,818	71,262	1.43
oltp_read_write	129,346	177,578	1.37
Average			1.49

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	56,426	65,066	1.15
oltp_range_select	21,670	24,836	1.15
oltp_sum_range	20,945	24,386	1.16
oltp_order_range	3,973	4,333	1.09
oltp_distinct_range	4,959	5,426	1.09
oltp_index_scan	7,341	9,227	1.26
select_random_points	31,741	36,783	1.16
select_random_ranges	10,461	11,925	1.14
covering_index_scan	6,953	7,175	1.03
groupby_scan	35,769	41,147	1.15
index_join	9,904	13,027	1.32
index_join_scan	4,606	5,850	1.27
types_table_scan	1,062,709	1,237,273	1.16
table_scan	1,252,422	1,334,073	1.07
oltp_read_only	188,773	212,505	1.13
Average			1.16

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	22,081	85,791	3.89
oltp_insert_ac	27,743	105,101	3.79
oltp_update_index_ac	26,848	116,864	4.35
oltp_update_non_index_ac	23,584	93,996	3.99
oltp_delete_insert_ac	24,697	103,860	4.21
oltp_write_only_ac	24,560	105,848	4.31
types_delete_insert_ac	22,737	97,292	4.28
oltp_read_write_ac	32,586	115,446	3.54
Average			4.04

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.

[github-actions]

Sysbench-Style Benchmark: Doltlite vs SQLite

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	25,903	27,868	1.08
oltp_range_select	11,525	12,250	1.06
oltp_sum_range	10,022	11,857	1.18
oltp_order_range	2,689	3,037	1.13
oltp_distinct_range	3,701	4,122	1.11
oltp_index_scan	3,972	5,263	1.33
select_random_points	11,203	11,696	1.04
select_random_ranges	3,169	4,058	1.28
covering_index_scan	4,322	4,325	1.00
groupby_scan	31,904	36,620	1.15
index_join	5,968	8,232	1.38
index_join_scan	3,550	4,885	1.38
types_table_scan	1,132,641	1,261,144	1.11
table_scan	1,322,119	1,396,299	1.06
oltp_read_only	109,087	121,480	1.11
Average			1.16

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	181,496	252,557	1.39
oltp_insert	15,887	28,710	1.81
oltp_update_index	52,392	92,299	1.76
oltp_update_non_index	36,691	60,127	1.64
oltp_delete_insert	45,811	71,461	1.56
oltp_write_only	22,681	47,952	2.11
types_delete_insert	25,752	41,259	1.60
oltp_read_write	69,711	108,823	1.56
Average			1.68

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	105,879	56,478	0.53
oltp_range_select	19,955	15,092	0.76
oltp_sum_range	18,414	14,866	0.81
oltp_order_range	3,602	3,368	0.94
oltp_distinct_range	4,587	4,551	0.99
oltp_index_scan	12,695	8,411	0.66
select_random_points	20,221	14,798	0.73
select_random_ranges	11,490	6,985	0.61
covering_index_scan	13,004	7,442	0.57
groupby_scan	33,147	37,091	1.12
index_join	10,707	10,286	0.96
index_join_scan	4,678	5,407	1.16
types_table_scan	1,184,554	1,278,523	1.08
table_scan	1,418,337	1,397,702	0.99
oltp_read_only	225,963	161,165	0.71
Average			0.84

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	197,255	271,416	1.38
oltp_insert	21,727	36,434	1.68
oltp_update_index	79,203	119,060	1.50
oltp_update_non_index	59,431	84,366	1.42
oltp_delete_insert	68,243	95,517	1.40
oltp_write_only	43,427	66,332	1.53
types_delete_insert	40,752	54,185	1.33
oltp_read_write	88,445	124,946	1.41
Average			1.46

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	52,171	56,676	1.09
oltp_range_select	14,305	14,978	1.05
oltp_sum_range	12,988	14,726	1.13
oltp_order_range	3,133	3,337	1.07
oltp_distinct_range	4,091	4,456	1.09
oltp_index_scan	6,909	8,190	1.19
select_random_points	14,327	14,243	0.99
select_random_ranges	5,938	6,934	1.17
covering_index_scan	7,317	7,389	1.01
groupby_scan	32,247	36,813	1.14
index_join	7,554	10,069	1.33
index_join_scan	3,935	5,295	1.35
types_table_scan	1,134,443	1,276,258	1.13
table_scan	1,421,889	1,405,111	0.99
oltp_read_only	150,440	163,493	1.09
Average			1.12

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	16,518	63,975	3.87
oltp_insert_ac	18,658	91,243	4.89
oltp_update_index_ac	20,859	104,539	5.01
oltp_update_non_index_ac	17,588	73,276	4.17
oltp_delete_insert_ac	18,620	89,075	4.78
oltp_write_only_ac	18,620	88,067	4.73
types_delete_insert_ac	17,368	80,792	4.65
oltp_read_write_ac	24,398	97,210	3.98
Average			4.51

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.

[github-actions]

Sysbench-Style Benchmark (TEXT PK): Doltlite vs SQLite

Companion to the classic Sysbench-Style Benchmark. Every workload here
runs against tables with a 32-char hex TEXT PRIMARY KEY (UUID-shaped).
Individual ratios gated at 2.5×; section averages gated at 2×. Autocommit writes use 6× / 5×.

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	25,189	29,182	1.16
oltp_range_select	11,533	12,381	1.07
oltp_sum_range	10,933	11,835	1.08
oltp_order_range	2,721	2,883	1.06
oltp_distinct_range	3,604	3,833	1.06
oltp_index_scan	3,863	5,039	1.30
select_random_points	15,886	18,980	1.19
select_random_ranges	3,323	4,532	1.36
covering_index_scan	3,694	3,565	0.97
groupby_scan	30,737	31,955	1.04
index_join	6,546	8,181	1.25
index_join_scan	4,132	5,214	1.26
types_table_scan	1,013,427	1,124,411	1.11
table_scan	1,200,644	1,229,660	1.02
oltp_read_only	100,634	115,210	1.14
Average			1.14

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	176,511	269,027	1.52
oltp_insert	17,552	30,795	1.75
oltp_update_index	60,035	110,142	1.83
oltp_update_non_index	39,975	70,583	1.77
oltp_delete_insert	42,670	84,242	1.97
oltp_write_only	24,128	49,266	2.04
types_delete_insert	26,467	43,670	1.65
oltp_read_write	68,552	108,877	1.59
Average			1.77

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	53,229	39,333	0.74
oltp_range_select	15,456	13,638	0.88
oltp_sum_range	14,567	13,173	0.90
oltp_order_range	3,222	3,112	0.97
oltp_distinct_range	4,139	4,177	1.01
oltp_index_scan	7,181	6,355	0.88
select_random_points	19,576	20,660	1.06
select_random_ranges	6,431	5,571	0.87
covering_index_scan	7,779	5,162	0.66
groupby_scan	30,980	32,133	1.04
index_join	8,382	9,670	1.15
index_join_scan	4,737	5,525	1.17
types_table_scan	1,040,981	1,150,369	1.11
table_scan	1,204,168	1,248,437	1.04
oltp_read_only	149,812	131,630	0.88
Average			0.96

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	250,975	411,923	1.64
oltp_insert	99,746	75,705	0.76
oltp_update_index	219,428	213,986	0.98
oltp_update_non_index	160,983	135,604	0.84
oltp_delete_insert	231,701	164,151	0.71
oltp_write_only	193,284	134,410	0.70
types_delete_insert	221,117	100,104	0.45
oltp_read_write	168,409	269,486	1.60
Average			0.96

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	34,583	40,102	1.16
oltp_range_select	13,506	14,239	1.05
oltp_sum_range	13,226	13,417	1.01
oltp_order_range	3,084	3,214	1.04
oltp_distinct_range	4,008	4,163	1.04
oltp_index_scan	5,729	6,760	1.18
select_random_points	18,484	20,893	1.13
select_random_ranges	4,643	5,796	1.25
covering_index_scan	5,730	5,295	0.92
groupby_scan	31,833	32,271	1.01
index_join	7,984	9,935	1.24
index_join_scan	4,623	5,575	1.21
types_table_scan	1,048,751	1,146,585	1.09
table_scan	1,235,289	1,261,611	1.02
oltp_read_only	121,740	132,557	1.09
Average			1.10

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	670,957	439,594	0.66
oltp_insert_ac	112,369	241,553	2.15
oltp_update_index_ac	359,755	321,627	0.89
oltp_update_non_index_ac	242,391	327,446	1.35
oltp_delete_insert_ac	262,053	438,349	1.67
oltp_write_only_ac	50,098	260,063	5.19
types_delete_insert_ac	457,129	271,109	0.59
oltp_read_write_ac	235,941	463,547	1.96
Average			1.81

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.