perf!: stop setting optimization options per header and per function by gabrielfrasantos · Pull Request #348 · embedded-pro/numerical-toolbox-cpp

gabrielfrasantos · 2026-10-04T16:55:38Z

Summary

Two commits:

fix(test) — numerical.math_test locks up QEMU at TestGradientCheck.expect_gradient_near_reports_wrong_component, on main as well.
perf! — the toolbox stops setting optimization options per header and per function; OPTIMIZE_FOR_SPEED only forces inlining.

QEMU test lockup

The QEMU test executables link newlib-nano, whose printf has no float conversions unless _printf_float is linked. std::ostream << float formats through vsnprintf("%.*g"), which then returned a bogus length — the address of its output buffer, 0x203ffce0. libstdc++'s num_put::_M_insert_float allocas that many bytes, the stack pointer wraps below zero, the next push bus-faults, the fault entry cannot stack, and QEMU stops with Lockup: can't escalate 3 to HardFault.

That test is the first whose failure message prints a float; any failing float expectation on QEMU would have crashed the same way instead of reporting. numerical_link_qemu_runtime now links _printf_float into every QEMU test target.

Optimization options

GCC does not inline a callee whose optimization options differ from its caller's. Every header bracketed its body with #pragma GCC optimize("O3", "fast-math"), and OPTIMIZE_FOR_SPEED added optimize("-O3") and optimize("-ffast-math"), so each toolbox function had options of its own: it could not be inlined into the consumer's code, and it could not inline the consumer's or the standard library's small helpers (element access, accessors, unit wrappers) either.

OPTIMIZE_FOR_SPEED expands to __attribute__((always_inline, hot)) inline when NumericalToolbox_ENABLE_OPTIMIZATIONS is defined, for GCC and Clang alike, and to nothing otherwise.
The push_options/optimize/pop_options bracket is gone from all 104 headers.
AGENTS.md, CLAUDE.md, the Copilot instructions, the agents, the roadmap and doc/performance-optimization/README.md forbid optimize pragmas and attributes, and leave the optimization level and floating-point model to the consumer, for whole translation units — for embedded targets -O2 or -O3 with -ffast-math -fno-finite-math-only. math::IsFinite stays the way to test finiteness, since a consumer may still build with -ffinite-math-only.

Breaking change: toolbox code no longer builds at O3 with fast-math on its own. Consumers that want those optimizations pass them for whole translation units.

Effect in e-foc (embedded-pro/e-foc#333)

20 kHz FOC inner loop measured in SIL: a QEMU plugin counts the instructions between the DWT_CYCCNT reads that bracket the FOC callback and weights them with Cortex-M4 timings; 19 controller-matrix scenarios, ~118k iterations per build. Weighted cycles per iteration:

e-foc build	Mean	p99	Calls to ≤ 12-byte helpers on the gated paths
per-function pragmas, this repo's `main`	1282	1665	256
one set of options, this repo's `main`	1110	1486	233
one set of options, this PR	949	1301	11

Validation (local)

host-single-Debug: build clean, ctest 14/14.
qemu-cortex-m4 and qemu-cortex-m7: ctest 15/15 each. numerical.math_test runs 552/552 on M4; before the fix it aborted at the gradient-check test, on main too.
e-foc built against this branch: host ctest 25/25, EK-TM4C1294XL, EK-TM4C123GXL, STM32F407G-DISC1, NUCLEO-H563ZI and the QEMU firmware build, SIL 128/128, cycle budgets within their limits.

a13f42d stands on its own if you would rather merge the test fix separately.

🤖 Generated with Claude Code

https://claude.ai/code/session_01HKhRVwu8yvxdKJgNo17oNc

Generated by Claude Code

The QEMU test executables link newlib-nano, whose printf has no float conversions unless _printf_float is linked. std::ostream << float formats through vsnprintf("%.*g"), which then returned a bogus length (the address of its output buffer); libstdc++'s num_put allocas that many bytes, the stack pointer wraps below zero, the next push bus-faults, the fault entry cannot stack, and QEMU stops with "Lockup: can't escalate 3 to HardFault". TestGradientCheck.expect_gradient_near_reports_wrong_component is the first test whose failure message prints a float, so numerical.math_test aborted there; any failing float expectation would have done the same instead of reporting. numerical_link_qemu_runtime now links _printf_float into every QEMU test target. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01HKhRVwu8yvxdKJgNo17oNc

GCC does not inline a callee whose optimization options differ from its caller's. Every header bracketed its body with #pragma GCC optimize("O3", "fast-math"), and OPTIMIZE_FOR_SPEED added optimize("-O3") and optimize("-ffast-math"), so each toolbox function had options of its own: it could not be inlined into the consumer's code, and it could not inline the consumer's or the standard library's small helpers either. In e-foc's FOC firmware, together with e-foc's own scoped pragmas, that left 256 out-of-line calls to helpers of 12 bytes or less on the three cycle-gated paths. With one set of options for the whole build there are 11, and the 20 kHz inner loop costs 26 % fewer cycles. - OPTIMIZE_FOR_SPEED expands to always_inline, hot and inline when NumericalToolbox_ENABLE_OPTIMIZATIONS is defined, for GCC and Clang alike, and to nothing otherwise. - Drop the push_options/optimize/pop_options bracket from all 104 headers. - AGENTS.md, CLAUDE.md, the Copilot instructions, the agents, the roadmap and the performance guide forbid optimize pragmas and attributes, and leave the optimization level and floating-point model to the consumer, for whole translation units: -O2 or -O3 with -ffast-math -fno-finite-math-only for embedded targets. BREAKING CHANGE: toolbox code no longer builds at O3 with fast-math on its own. Consumers that want those optimizations pass them for whole translation units, e.g. -ffast-math -fno-finite-math-only. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01HKhRVwu8yvxdKJgNo17oNc

github-actions · 2026-10-04T16:57:30Z

✅⚠️MegaLinter analysis: Success with warnings

Descriptor	Linter	Files	Fixed	Errors	Elapsed time
✅ ACTION	actionlint	7		0	0.41s
✅ CPP	clang-format	430	11	0	2.56s
✅ CPP	cppcheck	430		0	6.02s
✅ DOCKERFILE	hadolint	1		0	0.33s
✅ JSON	jsonlint	8		0	0.07s
✅ JSON	prettier	8	8	0	0.37s
⚠️ MARKDOWN	markdownlint	132	3	178	2.1s
✅ MARKDOWN	markdown-table-formatter	132	24	0	0.24s
⚠️ SPELL	lychee	241		26	25.94s
✅ YAML	prettier	12	3	0	0.42s
✅ YAML	yamllint	12		0	0.44s

Detailed Issues

⚠️ SPELL / lychee - 26 errors

📝 Summary
---------------------
🔍 Total..........567
🔗 Unique.........436
✅ Successful.....529
⏳ Timeouts.........0
🔀 Redirected.....148
👻 Excluded........11
❓ Unknown..........0
🚫 Errors..........26
⛔ Unsupported.....26

Errors in .github/workflows/static-analysis.yml
[403] https://binaries.sonarsource.com/Distribution/sonar-scanner-cli/sonar-scanner-cli-$ (at 37:21) | Rejected status code: 403 Forbidden

Errors in CHANGELOG.md
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/1 (at 206:31) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/12 (at 221:8) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/13 (at 220:7) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/15 (at 215:18) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/18 (at 205:62) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/2 (at 213:34) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/24 (at 210:19) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/26 (at 211:30) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/27 (at 208:23) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/3 (at 204:35) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/4 (at 214:34) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/6 (at 212:53) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/7 (at 209:32) | Rejected status code: 404 Not Found
[404] https://github.com/embedded-pro/embedded-numerical-toolbox/issues/9 (at 207:15) | Rejected status code: 404 Not Found

Errors in doc/filters/active/AhrsMadgwickMahony.md
[ERROR] file://doc/filters/ComplementaryFilter.md (at 131:3) | File not found. Check if file exists and path is correct

Errors in doc/filters/active/ExtendedKalmanFilter.md
[ERROR] file://doc/dynamics/EulerLagrange.md (at 130:3) | File not found. Check if file exists and path is correct

Errors in doc/filters/active/KalmanFilter.md
[ERROR] file://doc/filters/controllers/Lqr.md (at 114:77) | File not found. Check if file exists and path is correct
[ERROR] file://doc/filters/controllers/Lqr.md (at 133:3) | File not found. Check if file exists and path is correct
[ERROR] file://doc/filters/controllers/Pid.md (at 135:3) | File not found. Check if file exists and path is correct
[ERROR] file://doc/filters/solvers/DiscreteAlgebraicRiccatiEquation.md (at 134:3) | File not found. Check if file exists and path is correct

Errors in doc/performance-optimization/README.md
[403] https://developer.arm.com/documentation/100166/latest/ (at 472:3) | Rejected status code: 403 Forbidden
[403] https://developer.arm.com/documentation/dui0473/latest (at 474:3) | Rejected status code: 403 Forbidden

Errors in README.md
[ERROR] file:///releases/latest (at 38:14) | File not found. Check if file exists and path is correct
[404] https://embedded-pro.github.io/embedded-dsp-control/ (at 37:1) | Rejected status code: 404 Not Found

Errors in roadmap/README.md
[ERROR] file://roadmap/filters/passive/ExponentialMovingAverage/implementation.md (at 24:1) | File not found. Check if file exists and path is correct

Hint: Followed 148 redirects. You might want to consider replacing redirecting URLs with the resolved URLs. Use verbose mode (`-v`/`-vv`) to see redirection details.
Hint: You can configure accepted/rejected response codes with `-a` or `--accept`

⚠️ MARKDOWN / markdownlint - 178 errors

rithm.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 492]
doc/analysis/GoertzelAlgorithm.md:107:401 error MD013/line-length Line length [Expected: 400; Actual: 451]
doc/analysis/HilbertTransform.md:31:34 error MD056/table-column-count Table column count [Expected: 2; Actual: 4; Too many cells, extra data will be missing]
doc/analysis/PowerDensitySpectrum.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 509]
doc/analysis/PowerDensitySpectrum.md:104:20 error MD056/table-column-count Table column count [Expected: 2; Actual: 6; Too many cells, extra data will be missing]
doc/analysis/PowerDensitySpectrum.md:104:17 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/analysis/PowerDensitySpectrum.md:104:20 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/analysis/PowerDensitySpectrum.md:104:32 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/analysis/PowerDensitySpectrum.md:104:38 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/analysis/PowerDensitySpectrum.md:104:45 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/analysis/PowerDensitySpectrum.md:104:135 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/analysis/SignalDetectors.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 481]
doc/control_analysis/FrequencyResponse.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 469]
doc/control_analysis/README.md:11:68 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/control_analysis/README.md:11:147 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/control_analysis/README.md:12:53 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/control_analysis/README.md:12:172 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/control_analysis/README.md:13:68 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/control_analysis/README.md:13:151 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/control_analysis/RootLocus.md:69 error MD040/fenced-code-language Fenced code blocks should have a language specified [Context: "```"]
doc/control_analysis/TransferFunctionStateSpace.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 589]
doc/controllers/Feedforward2Dof.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 627]
doc/controllers/GainScheduledController.md:7:401 error MD013/line-length Line length [Expected: 400; Actual: 454]
doc/controllers/LeadLagCompensator.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 639]
doc/controllers/Lqg.md:7:401 error MD013/line-length Line length [Expected: 400; Actual: 428]
doc/controllers/SaturationRateLimiter.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 451]
doc/controllers/SaturationRateLimiter.md:7:401 error MD013/line-length Line length [Expected: 400; Actual: 477]
doc/estimators/ExpectationMaximization.md:104:103 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/estimators/LmsAdaptiveFilter.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 547]
doc/estimators/README.md:11:122 error MD056/table-column-count Table column count [Expected: 2; Actual: 3; Too many cells, extra data will be missing]
doc/estimators/README.md:11:122 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/AhrsMadgwickMahony.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 538]
doc/filters/active/AhrsMadgwickMahony.md:7:401 error MD013/line-length Line length [Expected: 400; Actual: 554]
doc/filters/active/AlphaBetaFilter.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 419]
doc/filters/active/AlphaBetaFilter.md:7:401 error MD013/line-length Line length [Expected: 400; Actual: 451]
doc/filters/active/KalmanSmoother.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 421]
doc/filters/active/KalmanSmoother.md:111:28 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:111:61 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:111:110 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:122:19 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:122:27 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:122:67 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:123:19 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:123:27 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:123:67 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:124:19 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:124:27 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:124:67 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:125:19 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:125:27 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/KalmanSmoother.md:125:67 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/UnscentedKalmanFilter.md:11:401 error MD013/line-length Line length [Expected: 400; Actual: 468]
doc/filters/active/UnscentedKalmanFilter.md:149:65 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/active/UnscentedKalmanFilter.md:149:153 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/passive/MedianFilter.md:69:25 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/passive/MedianFilter.md:69:43 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/passive/MedianFilter.md:70:51 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/passive/MedianFilter.md:70:96 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/README.md:13:74 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/README.md:13:179 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/README.md:14:74 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/README.md:14:179 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/README.md:15:74 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/README.md:15:179 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/README.md:16:74 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/README.md:16:179 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/README.md:17:74 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/filters/README.md:17:179 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/GivensRotation.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 429]
doc/math/HouseholderTransform.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 455]
doc/math/MatrixExponential.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 555]
doc/math/MatrixNorms.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 452]
doc/math/MatrixNorms.md:42:67 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/MatrixNorms.md:43:67 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/MatrixNorms.md:44:67 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/MatrixNorms.md:45:67 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/MatrixNorms.md:46:67 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:9:55 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:9:151 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:10:55 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:10:151 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:11:55 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:11:151 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:12:55 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:12:151 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:13:55 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:13:151 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:14:55 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:14:159 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:15:55 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:15:153 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:16:55 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:16:152 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:17:55 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:17:172 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:18:55 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:18:171 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:19:55 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/README.md:19:174 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/math/StepResponseMetrics.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 514]
doc/math/StepResponseMetrics.md:61 error MD040/fenced-code-language Fenced code blocks should have a language specified [Context: "```"]
doc/nonlinear_control/BacksteppingControl.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 594]
doc/nonlinear_control/FeedbackLinearization.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 533]
doc/nonlinear_control/FeedbackLinearization.md:15:401 error MD013/line-length Line length [Expected: 400; Actual: 491]
doc/nonlinear_control/ModelReferenceAdaptiveControl.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 587]
doc/nonlinear_control/README.md:9:80 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/nonlinear_control/README.md:9:258 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/nonlinear_control/README.md:10:80 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/nonlinear_control/README.md:10:258 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/nonlinear_control/README.md:11:80 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/nonlinear_control/README.md:11:270 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/regularization/Regularization.md:100:32 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/regularization/Regularization.md:100:135 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/robust_control/ActiveDisturbanceRejection.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 699]
doc/robust_control/ActiveDisturbanceRejection.md:85:401 error MD013/line-length Line length [Expected: 400; Actual: 575]
doc/robust_control/ActiveDisturbanceRejection.md:93:401 error MD013/line-length Line length [Expected: 400; Actual: 622]
doc/robust_control/DisturbanceObserver.md:7:401 error MD013/line-length Line length [Expected: 400; Actual: 476]
doc/robust_control/README.md:12:259 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/robust_control/SlidingModeControl.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 644]
doc/solvers/ConditionNumber.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 461]
doc/solvers/ConditionNumber.md:51:401 error MD013/line-length Line length [Expected: 400; Actual: 425]
doc/solvers/JacobiEigenSolver.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 443]
doc/solvers/JacobiEigenSolver.md:7:401 error MD013/line-length Line length [Expected: 400; Actual: 609]
doc/solvers/JacobiEigenSolver.md:45:401 error MD013/line-length Line length [Expected: 400; Actual: 427]
doc/solvers/README.md:16:163 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/solvers/README.md:17:163 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/solvers/README.md:18:167 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/solvers/README.md:19:184 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/solvers/README.md:20:184 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
doc/solvers/RungeKuttaIntegrators.md:9:401 error MD013/line-length Line length [Expected: 400; Actual: 509]
doc/solvers/SpectralRadius.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 665]
doc/windowing/window.md:5:401 error MD013/line-length Line length [Expected: 400; Actual: 415]
README.md:1 error MD041/first-line-heading/first-line-h1 First line in a file should be a top-level heading [Context: "[![Quality Gate Status](https:..."]
README.md:9:401 error MD013/line-length Line length [Expected: 400; Actual: 442]
README.md:19:327 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
README.md:20:211 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
README.md:21:303 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
README.md:22:250 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
README.md:23:446 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
README.md:25:140 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
README.md:26:174 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
README.md:27:363 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
README.md:28:158 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
README.md:29:205 error MD060/table-column-style Table column style [Table pipe does not align with header for style "aligned"]
roadmap/math/SE3Transform/implementation.md:7 error MD040/fenced-code-language Fenced code blocks should have a language specified [Context: "```"]
roadmap/math/SE3Transform/implementation.md:21 error MD040/fenced-code-language Fenced code blocks should have a language specified [Context: "```"]
roadmap/math/SE3Transform/implementation.md:35 error MD040/fenced-code-language Fenced code blocks should have a language specified [Context: "```"]
roadmap/math/SE3Transform/tests.md:7 error MD040/fenced-code-language Fenced code blocks should have a language specified [Context: "```"]
roadmap/math/SE3Transform/tests.md:16 error MD040/fenced-code-language Fenced code blocks should have a language specified [Context: "```"]
roadmap/README.md:11 error MD040/fenced-code-language Fenced code blocks should have a language specified [Context: "```"]

(Truncated to last 20000 characters out of 24106)

Notices

⚠️ Your configuration references items that have been removed from MegaLinter and are ignored: REPOSITORY_GITLEAKS, REPOSITORY_KICS. See Removed linters to find their replacements.

See detailed reports in MegaLinter artifacts

Your project could benefit from a custom flavor, which would allow you to run only the linters you need, and thus improve runtime performances. (Skip this info by defining FLAVOR_SUGGESTIONS: false)

Documentation: Custom Flavors
Command: npx mega-linter-runner@10.1.0 --custom-flavor-setup --custom-flavor-linters ACTION_ACTIONLINT,CPP_CPPCHECK,CPP_CLANG_FORMAT,DOCKERFILE_HADOLINT,JSON_JSONLINT,JSON_PRETTIER,MARKDOWN_MARKDOWNLINT,MARKDOWN_MARKDOWN_TABLE_FORMATTER,SPELL_LYCHEE,YAML_PRETTIER,YAML_YAMLLINT

Show us your support by starring ⭐ the repository

sonarqubecloud · 2026-10-04T17:02:37Z

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

Every Matrix operator that returns a new matrix value-initialised its result (data = {}) and then overwrote every element. Without the O3 pragmas #348 removed, a consumer that builds at -O2 keeps those loops rolled, so the zeroing survives. For a 3x3 result it is a memset call, and the Arm GNU toolchain's newlib-nano memset stores one byte at a time. e-foc runs an RLS update in the control interrupt during mechanical identification. Its covariance update, (P - g g^T / d) / lambda, builds four 3x3 temporaries. In e-foc's SIL scenario that times the control interrupt through a full calibration, the slowest execution rose from 237 to 377 cycles of the emulated 25 MHz clock (arm-none-eabi-gcc 15.2.1). The four memset calls were 596 of its 1892 instructions. - Matrix gains a private constructor that leaves the elements uninitialised. operator+, operator-, the matrix and scalar products, Transpose, GetBlock and GetColumn build their results with it, since each writes every element. - The default and initializer-list constructors still zero-fill, so a default-constructed or partly listed matrix keeps its zeros. - A test evaluates those operators in constant expressions, so a compiler that checks (Clang does, GCC does not) rejects an operator that leaves an element unwritten. The RLS update keeps its matrix expression and now makes no memset calls. The same scenario peaks at 250 cycles. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01HKhRVwu8yvxdKJgNo17oNc

* perf(math): build matrix results without zero-filling them first Every Matrix operator that returns a new matrix value-initialised its result (data = {}) and then overwrote every element. Without the O3 pragmas #348 removed, a consumer that builds at -O2 keeps those loops rolled, so the zeroing survives. For a 3x3 result it is a memset call, and the Arm GNU toolchain's newlib-nano memset stores one byte at a time. e-foc runs an RLS update in the control interrupt during mechanical identification. Its covariance update, (P - g g^T / d) / lambda, builds four 3x3 temporaries. In e-foc's SIL scenario that times the control interrupt through a full calibration, the slowest execution rose from 237 to 377 cycles of the emulated 25 MHz clock (arm-none-eabi-gcc 15.2.1). The four memset calls were 596 of its 1892 instructions. - Matrix gains a private constructor that leaves the elements uninitialised. operator+, operator-, the matrix and scalar products, Transpose, GetBlock and GetColumn build their results with it, since each writes every element. - The default and initializer-list constructors still zero-fill, so a default-constructed or partly listed matrix keeps its zeros. - A test evaluates those operators in constant expressions, so a compiler that checks (Clang does, GCC does not) rejects an operator that leaves an element unwritten. The RLS update keeps its matrix expression and now makes no memset calls. The same scenario peaks at 250 cycles. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01HKhRVwu8yvxdKJgNo17oNc * fix(math): build the matrix product's result in a member function The matrix product is a friend function of its left operand's Matrix specialisation. It built its result, a Matrix of another shape, with that shape's private constructor. GCC and Clang allowed it, but MSVC rejects it (C2248): friendship goes to the Matrix specialisations, not to their friend functions. The operator now calls a private member function, Multiply, which builds the result in place, as Transpose, GetBlock and GetColumn already do. Members of every Matrix specialisation are friends of the others, so every compiler accepts it. A static helper that returned the uninitialised result was not used: GCC folds such an argument-less constexpr call into a zero-filled constant, which brings back the memset calls. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01HKhRVwu8yvxdKJgNo17oNc --------- Co-authored-by: Claude <noreply@anthropic.com>

claude added 2 commits October 4, 2026 16:34

gabrielfrasantos merged commit 31c7ac9 into main Oct 4, 2026
14 checks passed

gabrielfrasantos deleted the ccr-90fcd432-u9pqe7 branch October 4, 2026 18:05

embedded-releaser-app Bot mentioned this pull request Oct 1, 2026

chore(main): release 4.0.0 #327

Open

gabrielfrasantos mentioned this pull request Oct 5, 2026

perf(math): build matrix results without zero-filling them first #349

Merged

Navigation Menu

Repository navigation

perf!: stop setting optimization options per header and per function - #348

perf!: stop setting optimization options per header and per function#348
gabrielfrasantos merged 2 commits into
mainfrom
ccr-90fcd432-u9pqe7

gabrielfrasantos commented Oct 4, 2026

Uh oh!

github-actions Bot commented Oct 4, 2026

Uh oh!

sonarqubecloud Bot commented Oct 4, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

gabrielfrasantos commented Oct 4, 2026

Summary

QEMU test lockup

Optimization options

Effect in e-foc (embedded-pro/e-foc#333)

Validation (local)

Uh oh!

github-actions Bot commented Oct 4, 2026

✅⚠️MegaLinter analysis: Success with warnings

Detailed Issues

Notices

Uh oh!

sonarqubecloud Bot commented Oct 4, 2026

Quality Gate passed

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants