[fix](be) Stabilize conjunct cost ordering

[BiteTheDDDDt]

What problem does this PR solve?

Issue Number: None

Related PR: None

Problem Summary: When enable_adjust_conjunct_order_by_cost is enabled, conjuncts are sorted by execute_cost with std::ranges::sort. Equal-cost conjuncts may have their original relative order changed, which can introduce unnecessary evaluation-order side effects. This patch switches the cost ordering paths to std::ranges::stable_sort so lower-cost conjuncts are still prioritized while equal-cost conjuncts keep their original order.

Release note

None

Check List (For Author)

• Test: Manual test
  • Ran clang-format v16 on changed C++ files
  • Ran git diff --check
  • BE build was started but interrupted per request; not used as validation
• Behavior changed: Yes. Equal-cost conjuncts now preserve their original order when enable_adjust_conjunct_order_by_cost is enabled.
• Does this need documentation: No

[hello-stephen]

Thank you for your contribution to Apache Doris.
Don't know what should be done next? See How to process your PR.

Please clearly describe your PR:

1. What problem was fixed (it's best to include specific error reporting information). How it was fixed.
2. Which behaviors were modified. What was the previous behavior, what is it now, why was it modified, and what possible impacts might there be.
3. What features were added. Why was this function added?
4. Which code was refactored and why was this part of the code refactored?
5. Which functions were optimized and what is the difference before and after the optimization?

[BiteTheDDDDt]

run buildall

[BiteTheDDDDt]

/review

[github-actions]

No blocking issues found in this review.

Code-review checkpoint conclusions:

• Goal: The PR switches all existing enable_adjust_conjunct_order_by_cost() cost-ordering sites from unstable sort to stable sort, so equal-cost conjuncts preserve the current input order while lower-cost conjuncts are still prioritized.
• Test coverage: No automated test specifically proves equal-cost tie preservation; the PR reports manual validation only. I consider this a residual test gap, not a blocker for this small mechanical fix.
• Scope: The modification is small and focused: three sort calls plus the required <algorithm> includes.
• Concurrency: The only changed concurrent path is late-arrival runtime filters in ScanLocalStateBase::update_late_arrival_runtime_filter; it remains serialized by _conjuncts_lock, and scanners still clone under the same lock. No new lock-order issue found.
• Lifecycle/static initialization: No new ownership, static initialization, or lifecycle coupling.
• Configuration/compatibility: No new config, protocol, persistence, storage format, or FE-BE compatibility changes.
• Parallel paths: I searched the changed option and execute_cost() ordering paths; the three existing config-gated sort paths are all covered.
• Error handling/data correctness: Status propagation and expression prepare/open/evaluation flow are unchanged; no silent failure path introduced.
• Observability: No new observability appears necessary for this local ordering behavior.
• Performance: stable_sort may do slightly more work than sort, but these conjunct vectors are expected to be small and the existing hot-path structure is unchanged.
• User focus: No additional user-provided focus points were present.

CI note: the visible macOS BE UT failure fails before build/test execution due to the runner using JDK 25 while the script requires JDK 17, so I did not treat that as caused by this patch.

[hello-stephen]

BE Regression && UT Coverage Report

Increment line coverage 100.00% (3/3) 🎉

Increment coverage report
Complete coverage report

Category	Coverage
Function Coverage	74.18% (28429/38325)
Line Coverage	58.12% (310132/533652)
Region Coverage	55.01% (260121/472864)
Branch Coverage	56.29% (112826/200446)

[hello-stephen]

TPC-H: Total hot run time: 29690 ms

machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
scripts: https://github.com/apache/doris/tree/master/tools/tpch-tools
Tpch sf100 test result on commit 872a784e4e0116da738c314602c91db2e0c87537, data reload: false
------ Round 1 ----------------------------------
============================================
q1	17781	4115	4036	4036
q2	1991	333	192	192
q3	10300	1440	843	843
q4	4683	468	335	335
q5	7510	856	588	588
q6	183	174	138	138
q7	794	841	613	613
q8	9374	1704	1659	1659
q9	5807	4596	4534	4534
q10	6824	1791	1535	1535
q11	434	273	249	249
q12	654	436	312	312
q13	18186	3473	2724	2724
q14	271	266	259	259
q15	q16	792	782	706	706
q17	1028	975	1070	975
q18	7090	5896	5645	5645
q19	1305	1337	1147	1147
q20	489	403	266	266
q21	5913	2684	2633	2633
q22	436	370	301	301
Total cold run time: 101845 ms
Total hot run time: 29690 ms
----- Round 2, with runtime_filter_mode=off -----
============================================
q1	4408	4322	4341	4322
q2	330	361	240	240
q3	4608	5002	4470	4470
q4	2087	2165	1385	1385
q5	4493	4339	4350	4339
q6	226	176	129	129
q7	1747	2286	1693	1693
q8	2655	2400	2189	2189
q9	8189	8237	8164	8164
q10	4826	4734	4269	4269
q11	581	436	375	375
q12	773	765	530	530
q13	3446	3668	2967	2967
q14	300	326	264	264
q15	q16	716	731	658	658
q17	1345	1333	1486	1333
q18	7845	7416	7301	7301
q19	1190	1085	1095	1085
q20	2217	2214	1940	1940
q21	5327	4666	4455	4455
q22	527	447	407	407
Total cold run time: 57836 ms
Total hot run time: 52515 ms

[hello-stephen]

TPC-DS: Total hot run time: 176160 ms

machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
scripts: https://github.com/apache/doris/tree/master/tools/tpcds-tools
TPC-DS sf100 test result on commit 872a784e4e0116da738c314602c91db2e0c87537, data reload: false
query5	4324	646	499	499
query6	442	191	174	174
query7	4815	580	313	313
query8	361	222	210	210
query9	8744	4083	4055	4055
query10	445	327	257	257
query11	5924	2352	2148	2148
query12	152	104	101	101
query13	1257	633	438	438
query14	6380	5464	5096	5096
query14_1	4411	4425	4379	4379
query15	206	206	175	175
query16	1016	459	457	457
query17	1146	704	578	578
query18	2519	475	355	355
query19	202	186	155	155
query20	112	107	106	106
query21	222	146	121	121
query22	13595	13498	13321	13321
query23	17522	16622	16235	16235
query23_1	16398	16344	16246	16246
query24	7545	1786	1301	1301
query24_1	1324	1334	1316	1316
query25	520	420	358	358
query26	1316	306	162	162
query27	2663	556	369	369
query28	4388	2024	2029	2024
query29	1059	601	477	477
query30	334	236	198	198
query31	1112	1080	955	955
query32	98	61	57	57
query33	512	318	245	245
query34	1171	1147	632	632
query35	739	782	682	682
query36	1377	1423	1260	1260
query37	151	106	85	85
query38	3206	3136	3042	3042
query39	941	915	894	894
query39_1	865	890	860	860
query40	214	121	97	97
query41	63	60	60	60
query42	93	90	95	90
query43	325	330	290	290
query44	1463	781	789	781
query45	193	188	182	182
query46	1111	1200	759	759
query47	2386	2383	2225	2225
query48	425	438	280	280
query49	617	475	339	339
query50	998	360	249	249
query51	4350	4289	4218	4218
query52	88	89	77	77
query53	259	268	187	187
query54	263	220	203	203
query55	78	74	67	67
query56	262	220	219	219
query57	1449	1433	1304	1304
query58	237	211	205	205
query59	1601	1670	1447	1447
query60	305	276	222	222
query61	152	155	155	155
query62	698	641	590	590
query63	235	192	190	190
query64	2499	766	591	591
query65	4846	4778	4758	4758
query66	1775	446	332	332
query67	29767	29642	29556	29556
query68	3328	1575	982	982
query69	421	300	258	258
query70	1060	928	939	928
query71	296	237	203	203
query72	2905	2635	2351	2351
query73	869	807	421	421
query74	5109	5012	4765	4765
query75	2621	2616	2276	2276
query76	2329	1199	800	800
query77	357	384	282	282
query78	12328	12353	11831	11831
query79	1208	1161	796	796
query80	511	472	391	391
query81	448	282	234	234
query82	236	156	120	120
query83	270	276	251	251
query84	277	145	116	116
query85	825	539	403	403
query86	332	301	285	285
query87	3350	3314	3164	3164
query88	3677	2774	2738	2738
query89	411	384	327	327
query90	2187	186	176	176
query91	171	166	133	133
query92	62	63	58	58
query93	1554	1410	896	896
query94	541	348	309	309
query95	672	371	357	357
query96	1061	789	323	323
query97	2703	2691	2559	2559
query98	215	232	201	201
query99	1161	1153	1018	1018
Total cold run time: 260203 ms
Total hot run time: 176160 ms

[hello-stephen]

ClickBench: Total hot run time: 25.43 s

machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
scripts: https://github.com/apache/doris/tree/master/tools/clickbench-tools
ClickBench test result on commit 872a784e4e0116da738c314602c91db2e0c87537, data reload: false
query1	0.01	0.01	0.01
query2	0.10	0.05	0.04
query3	0.25	0.14	0.14
query4	1.61	0.14	0.14
query5	0.25	0.23	0.26
query6	1.24	1.07	1.06
query7	0.04	0.01	0.01
query8	0.06	0.04	0.04
query9	0.39	0.31	0.31
query10	0.56	0.60	0.56
query11	0.20	0.14	0.15
query12	0.18	0.15	0.15
query13	0.48	0.47	0.48
query14	1.03	1.02	1.01
query15	0.62	0.62	0.60
query16	0.32	0.34	0.33
query17	1.12	1.08	1.12
query18	0.24	0.22	0.22
query19	2.03	1.97	1.98
query20	0.02	0.01	0.01
query21	15.44	0.23	0.14
query22	4.72	0.05	0.05
query23	16.11	0.31	0.12
query24	2.90	0.43	0.32
query25	0.12	0.04	0.04
query26	0.74	0.22	0.16
query27	0.04	0.04	0.04
query28	3.51	0.94	0.54
query29	12.47	4.33	3.46
query30	0.27	0.15	0.14
query31	2.77	0.63	0.32
query32	3.24	0.60	0.50
query33	3.15	3.27	3.22
query34	15.60	4.28	3.52
query35	3.51	3.52	3.53
query36	0.55	0.44	0.43
query37	0.09	0.06	0.07
query38	0.06	0.04	0.04
query39	0.04	0.03	0.03
query40	0.18	0.16	0.16
query41	0.08	0.03	0.03
query42	0.04	0.03	0.03
query43	0.04	0.04	0.04
Total cold run time: 96.42 s
Total hot run time: 25.43 s

[HappenLee]

LGTM

[github-actions]

PR approved by at least one committer and no changes requested.

[github-actions]

PR approved by anyone and no changes requested.