Repository navigation
apm32: link newlib-nano in make builds - #4378
Conversation
family.mk linked full newlib while family.cmake passes --specs=nano.specs. On apm32f103_dev_board (20 KB RAM) that costs ~2.8 KB of RAM (mallocr, locale, impure, findfp), so the make build of net_lwip_webserver overflowed RAM by 744 bytes while the CMake build fit. Pass --specs=nano.specs in both apm32 families, as family.cmake and the stm32 families do.
|
No actionable comments were generated in the recent review.
 |
| Check name | Status | Explanation |
|---|---|---|
| Title check |  |
The title "apm32: link newlib-nano in make builds" matches the main change. The PR adds --specs=nano.specs to the Make LDFLAGS for apm32f10x and apm32f0xx, so the build links newlib-nano instead of fu… |
| Linked Issues check | |
Check skipped because no linked issues were found for this pull request. |
| Out of Scope Changes check | |
Check skipped because no linked issues were found for this pull request. |
| Description check | |
The description directly explains the linker flag changes, the RAM overflow issue, the intended newlib-nano alignment, and the reported verification results. |
Full details: Title check
-
Explanation
The title "apm32: link newlib-nano in make builds" matches the main change. The PR adds --specs=nano.specs to the Make LDFLAGS for apm32f10x and apm32f0xx, so the build links newlib-nano instead of full newlib. The title is specific and concise, and it names the platform and the build system.
Finishing Touches
-
Generate unit tests (beta)
-
- Commit to this branch
- Create a new PR
Simplify code
-
- Commit to this branch
- Create a new PR
-
- Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.
 [Medium impact] This PR appears safe to merge; no actionable issues were found.
|
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Code sizeNothing to measure: no code change, or no snapshot build leg selected. |
Hardware-in-the-loop (HIL) Test ReportNo HIL run for this push (no affected boards, or hardware testing did not run). |
Summary
The apm32
family.mkfiles linked full newlib, whilefamily.cmakepasses--specs=nano.specs. Onapm32f103_dev_board(20 KB RAM), full newlib adds about 2.8 KB of RAM (mallocr,locale,impure,findfp). The make build ofnet_lwip_webserverthen overflowed RAM by 744 bytes, while the CMake build fit.This adds
--specs=nano.specsto the make LDFLAGS ofapm32f10xandapm32f0xx, asfamily.cmakeand the stm32 families already do.Found through the random make job
build---one-random-make-apm32f10x-large-arm-gccon #4373. It fails on master too.Fixes #4330
Verification
net_lwip_webserveronapm32f103_dev_board, make: master overflows RAM by 744 B. With this change it links at 19520 B of 20480 B, the same as the CMake build, and the map shows newlib-nano.tools/build.py -s make, all examples:apm32f103_dev_board: 31 OK;apm32f107_dev_board: 42 OK;apm32f072_dev_board: 32 OK.