ci: use 26.04 image instead of latest

[lvaroqui]

No description provided.

[not-matthias]

LGTM

[greptile-apps]

Confidence Score: 5/5

[High impact] The PR appears safe to merge; no actionable defect was established.

Summary

Pins the Linux release runners to Ubuntu 26.04.

• Linux wheel builds now use named Ubuntu 26.04 runners.

_{Reviews (1) · Last reviewed commit: "ci: use 26.04 image instead of latest" · Reviewed by Greptile}

[codspeed]

Merging this PR will degrade performance by 0.29%

⚠️ 6 benchmarks spent significant time in system calls

System calls cannot be consistently instrumented, so they are not included in the measure, which understates the real cost. Please switch to the Walltime instrument to accurately measure system calls.

Measurement and system calls

• test_n_queens_solve[board0]
• test_generate_all_subsequences[sequence0]
• test_generate_all_permutations[sequence0]
• test_generate_all_subsequences[sequence1]
• test_generate_all_permutations[sequence1]
• test_recursive_cached_fibo_100

ℹ️ 46 benchmarks have no flame graph

The flamegraph could not be generated because we failed to resolve the symbols in the profiling trace. This happens when we fail to harvest information about language frames.

• test_multiprocessing_map[1000]
• test_multiprocessing_map[100000]
• test_multiprocessing_map[10]
• test_multiprocessing_map[10000]
• test_multiprocessing_map[100]
• test_threadpool_map[1000]
• test_threadpool_map[10000]
• test_threadpool_map[100000]
• test_process_creation[ls -l]
• test_fs_write[1000]
• …and 36 more

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 14 improved benchmarks
❌ 19 (👁 19) regressed benchmarks
✅ 244 untouched benchmarks

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	WallTime	test_sum_of_squares[sum_of_squares_for_loop_product]	200.7 µs	164.3 µs	+22.12%
⚡ ℹ️	Memory	test_multiprocessing_map[1000]	3 MB	2.5 MB	+18.51%
⚡	WallTime	test_noop_pass	551.9 ns	486 ns	+13.56%
⚡	WallTime	test_multiprocessing_map[1000]	56.3 ms	51.8 ms	+8.8%
⚡	WallTime	test_make_peak	6.6 µs	6.3 µs	+4.52%
⚡	WallTime	test_make_lowpass	5.5 µs	5.3 µs	+3.78%
⚡	WallTime	test_sum_of_squares[sum_of_squares_sum_comprehension_power]	233.3 µs	225.7 µs	+3.38%
⚡	WallTime	test_generate_all_subsequences[sequence0]	191.4 µs	185.2 µs	+3.34%
⚡	WallTime	test_multiprocessing_map[10]	48.3 ms	47.2 ms	+2.26%
⚡ ℹ️	Memory	test_multiprocessing_map[100000]	5.7 MB	5.6 MB	+2.23%
⚡	WallTime	test_pipe_communication[10]	7.8 ms	7.7 ms	+2.03%
👁	WallTime	test_recursive_fibo_20	5.2 ms	5.3 ms	-2.06%
👁	WallTime	test_noop_lambda_decorated	1.2 µs	1.3 µs	-5.16%
👁	WallTime	test_color[graph0-3]	30.4 µs	32 µs	-5.15%
👁	WallTime	test_open_knight_tour[1]	4.2 µs	4.3 µs	-3.1%
👁	WallTime	test_mmap_operation[409600]	131.5 µs	135.9 µs	-3.26%
👁	WallTime	test_tcp_connection[1.1.1.1-53]	824.5 µs	963.6 µs	-14.43%
👁	WallTime	test_noop_pass_decorated	680 ns	727.8 ns	-6.57%
👁	WallTime	test_generate_sum_of_subsets_soln[nums0-9]	26.5 µs	27.7 µs	-4.22%
👁	WallTime	test_solve_power_sum[13-2]	6.6 µs	7 µs	-5.61%
...	...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

Tip

Curious why performance improved? Comment @codspeedbot explain why performance improved on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing cod-3837-migrate-from-codspeedhq-arm64-ubuntu-204-github-runners-to (e82a6bc) with master (3027063)}