[opt](pipeline) Add passthrough between repeat and streaming aggregation

[Mryange]

What problem does this PR solve?

Problem Summary: Repeat expands grouping sets before local streaming aggregation. Without a passthrough local exchange, independent pipeline tasks create duplicate partial aggregation states, while hash shuffling every expanded row adds unnecessary hashing and skew. Request a passthrough exchange for a direct Repeat-to-StreamingAgg edge in both FE-planned and BE-planned local shuffle paths, and cover both distribution planning paths with tests.

Release note

None

Check List (For Author)

• Test

  • Regression test
  • Unit Test
  • Manual test (add detailed scripts or steps below)
  • No need to test or manual test. Explain why:
    • This is a refactor/code format and no logic has been changed.
    • Previous test can cover this change.
    • No code files have been changed.
    • Other reason

• Behavior changed:

  • No.
  • Yes.

• Does this need documentation?

  • No.
  • Yes.

Check List (For Reviewer who merge this PR)

• Confirm the release note
• Confirm test cases
• Confirm document
• Add branch pick label

[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?

[Mryange]

run buildall

[hello-stephen]

TPC-H: Total hot run time: 29091 ms

machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
scripts: https://github.com/apache/doris/tree/master/tools/tpch-tools
Tpch sf100 test result on commit 9c2a7c8e6671dd8e34f9c3bcf09f0c6a4897eda8, data reload: false

------ Round 1 ----------------------------------
============================================
q1	17620	4033	3959	3959
q2	2101	324	209	209
q3	10244	1492	831	831
q4	4675	470	353	353
q5	7520	841	588	588
q6	180	168	136	136
q7	730	775	595	595
q8	9378	1423	1528	1423
q9	5335	4107	4079	4079
q10	6747	1633	1352	1352
q11	499	354	330	330
q12	724	597	460	460
q13	18109	3233	2719	2719
q14	262	255	237	237
q15	q16	741	732	658	658
q17	997	990	906	906
q18	7078	5619	5566	5566
q19	1333	1301	1081	1081
q20	803	697	593	593
q21	6424	2831	2700	2700
q22	448	366	316	316
Total cold run time: 101948 ms
Total hot run time: 29091 ms

----- Round 2, with runtime_filter_mode=off -----
============================================
q1	4817	4626	4614	4614
q2	293	346	211	211
q3	4884	5256	4715	4715
q4	2173	2269	1399	1399
q5	4504	4651	4397	4397
q6	229	181	135	135
q7	1800	1751	1494	1494
q8	2333	2016	2025	2016
q9	7216	6930	6708	6708
q10	4285	4181	3812	3812
q11	513	367	344	344
q12	693	703	495	495
q13	3001	3272	2750	2750
q14	273	283	249	249
q15	q16	657	682	597	597
q17	1228	1212	1197	1197
q18	7050	6836	6851	6836
q19	1051	1069	1100	1069
q20	2203	2172	1894	1894
q21	5213	4591	4319	4319
q22	520	470	409	409
Total cold run time: 54936 ms
Total hot run time: 49660 ms

[hello-stephen]

TPC-DS: Total hot run time: 167474 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 9c2a7c8e6671dd8e34f9c3bcf09f0c6a4897eda8, data reload: false

query5	4327	605	463	463
query6	484	228	202	202
query7	4892	600	340	340
query8	339	185	170	170
query9	8803	4086	4133	4086
query10	481	361	330	330
query11	5879	2182	2025	2025
query12	164	102	103	102
query13	1284	603	450	450
query14	6404	4703	4369	4369
query14_1	3834	3800	3817	3800
query15	216	211	179	179
query16	1067	480	475	475
query17	1149	683	532	532
query18	2718	460	335	335
query19	218	190	142	142
query20	107	102	100	100
query21	239	156	134	134
query22	13551	13501	13388	13388
query23	17568	16438	16007	16007
query23_1	16058	16065	16091	16065
query24	7479	1686	1224	1224
query24_1	1254	1269	1235	1235
query25	550	430	359	359
query26	1313	373	213	213
query27	2469	591	399	399
query28	4361	1994	2006	1994
query29	1061	582	468	468
query30	363	263	223	223
query31	1114	1076	957	957
query32	116	65	61	61
query33	689	320	262	262
query34	1219	1170	628	628
query35	748	772	641	641
query36	1170	1170	1012	1012
query37	168	107	96	96
query38	1828	1661	1618	1618
query39	842	832	806	806
query39_1	788	796	801	796
query40	257	167	148	148
query41	74	70	73	70
query42	98	94	93	93
query43	323	326	284	284
query44	1472	798	772	772
query45	193	178	175	175
query46	1058	1176	742	742
query47	1516	1528	1489	1489
query48	400	406	305	305
query49	601	415	313	313
query50	1127	435	339	339
query51	10446	10531	10676	10531
query52	89	90	78	78
query53	262	278	203	203
query54	308	253	237	237
query55	79	71	74	71
query56	335	300	300	300
query57	1042	995	908	908
query58	303	264	289	264
query59	1591	1621	1494	1494
query60	328	284	271	271
query61	181	182	176	176
query62	430	323	270	270
query63	244	202	252	202
query64	2796	1022	842	842
query65	3897	3883	3821	3821
query66	1824	453	356	356
query67	24775	24710	24511	24511
query68	3273	1592	997	997
query69	428	308	259	259
query70	1102	983	947	947
query71	393	341	333	333
query72	3044	2666	2394	2394
query73	816	761	405	405
query74	4624	4469	4312	4312
query75	2375	2352	1989	1989
query76	2338	1159	785	785
query77	353	369	278	278
query78	11186	11201	10640	10640
query79	1399	1162	779	779
query80	1184	568	457	457
query81	529	324	274	274
query82	569	151	114	114
query83	402	323	298	298
query84	281	157	132	132
query85	983	604	511	511
query86	409	306	272	272
query87	1853	1790	1711	1711
query88	3806	2832	2801	2801
query89	414	315	288	288
query90	1879	196	199	196
query91	205	194	167	167
query92	66	62	59	59
query93	1609	1525	1047	1047
query94	650	357	306	306
query95	840	605	478	478
query96	1080	741	366	366
query97	2475	2421	2348	2348
query98	222	192	192	192
query99	716	719	614	614
Total cold run time: 254978 ms
Total hot run time: 167474 ms

[hello-stephen]

ClickBench: Total hot run time: 23.99 s

machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
scripts: https://github.com/apache/doris/tree/master/tools/clickbench-tools
ClickBench test result on commit 9c2a7c8e6671dd8e34f9c3bcf09f0c6a4897eda8, data reload: false

query1	0.00	0.00	0.01
query2	0.10	0.05	0.05
query3	0.26	0.13	0.13
query4	1.60	0.14	0.14
query5	0.24	0.21	0.22
query6	1.17	0.82	0.82
query7	0.04	0.00	0.01
query8	0.06	0.03	0.04
query9	0.39	0.31	0.30
query10	0.58	0.53	0.53
query11	0.19	0.13	0.13
query12	0.18	0.14	0.14
query13	0.47	0.46	0.47
query14	0.99	0.99	1.03
query15	0.60	0.58	0.57
query16	0.33	0.33	0.30
query17	1.05	1.10	1.10
query18	0.22	0.19	0.20
query19	2.08	1.96	1.89
query20	0.02	0.01	0.01
query21	15.46	0.22	0.13
query22	4.82	0.06	0.05
query23	16.13	0.31	0.12
query24	2.93	0.42	0.32
query25	0.12	0.06	0.05
query26	0.74	0.21	0.15
query27	0.03	0.04	0.03
query28	3.54	0.76	0.36
query29	12.49	4.19	3.31
query30	0.28	0.16	0.15
query31	2.77	0.57	0.31
query32	3.22	0.58	0.48
query33	3.16	3.24	3.27
query34	15.59	3.91	3.29
query35	3.22	3.24	3.23
query36	0.56	0.43	0.42
query37	0.10	0.06	0.07
query38	0.06	0.04	0.04
query39	0.04	0.03	0.03
query40	0.16	0.15	0.14
query41	0.09	0.04	0.03
query42	0.04	0.03	0.03
query43	0.04	0.03	0.04
Total cold run time: 96.16 s
Total hot run time: 23.99 s

[hello-stephen]

FE UT Coverage Report

Increment line coverage 0.00% (0/3) 🎉
Increment coverage report
Complete coverage report

[hello-stephen]

BE Regression && UT Coverage Report

Increment line coverage 100.00% (5/5) 🎉

Increment coverage report
Complete coverage report

Category	Coverage
Function Coverage	75.90% (32070/42253)
Line Coverage	60.54% (357663/590803)
Region Coverage	57.17% (300436/525502)
Branch Coverage	58.58% (135338/231046)

[hello-stephen]

FE Regression Coverage Report

Increment line coverage 66.67% (2/3) 🎉
Increment coverage report
Complete coverage report

[924060929]

The Repeat-to-StreamingAgg PASSTHROUGH direction is reasonable for avoiding per-row hash work and grouping-set skew, and the FE !needsFinalize guard correctly limits it to partial aggregation. However, the current implementation makes the optimization unconditional and bypasses the dedicated StreamingAgg local-exchange switch in both planning paths. Please preserve the existing opt-out semantics and cover the on/off matrix before merging.

[924060929]

This special case unconditionally inserts a local exchange and bypasses the opt-out semantics of enable_local_exchange_before_streaming_agg.

That dedicated switch was added with a default of false specifically so StreamingAgg preserves its inherited distribution unless local exchange is explicitly enabled. In the BE path, the new Repeat check likewise runs before enable_local_exchange_before_streaming_agg() is checked. In the FE path, this branch also precedes the existing enableLeBeforeAgg fallback and does not consult the streaming-specific switch at all.

A two-phase grouping-sets query therefore gets StreamingAgg <- LE(PASSTHROUGH) <- Repeat even when the streaming-agg switch is false. This should not change query results because this is partial aggregation followed by a final merge, but it makes the extra pipeline boundary, queues, and memory overhead impossible to disable, and silently changes the default behavior established by #66222.

Please gate the Repeat PASSTHROUGH optimization with enableLocalExchangeBeforeStreamingAgg here and with enable_local_exchange_before_streaming_agg() in BE, and add FE-/BE-planned tests for both switch values. If Repeat is intentionally meant to ignore the existing switch, that needs an explicit separately named switch and documented default behavior rather than bypassing the current opt-out.