[casaroli]

Summary

The TIME function of bas (fn_timei() in interpreters/bas/bas_global.c) divides clock() by CLK_TCK. CLK_TCK is the system tick rate, but clock() counts in CLOCKS_PER_SEC units. In NuttX both are the tick rate today, so the result is correct, but the code uses the wrong constant. This PR divides by CLOCKS_PER_SEC.

This is the bas part of the CLOCKS_PER_SEC series of the dev@ thread "[DISCUSS] Breaking change: make CLOCKS_PER_SEC 1000000" (https://lists.apache.org/thread/bqxry570g9t63j63zmk4c5l82lybtx4s). #3904 (step 1, the other apps) and apache/nuttx#20923 (step 2) are merged. This PR does not need either of them.

Impact

None today. The object code does not change (see Testing). The result stays correct if CLOCKS_PER_SEC changes. Not a breaking change.

Testing

qemu-armv8a:nsh with CONFIG_INTERPRETERS_BAS=y, built with apps master and with this PR. bas_global.o compared with objdump -dr (code and relocations) and objdump -s (all sections except debug information): the only difference is the line number of three assert() calls after the change, because the new line break adds one line. fn_timei() and all data are the same. (The build needs -Wno-error on master too: bas_statement.c has older -Werror=array-bounds errors with GCC 14.)

Style check: tools/checkpatch.sh -c -u -m -g checks the whole of bas_global.c and reports 1045 errors, the same number as on master. 1020 of them are mixed-case identifiers of the third-party bas code (for example Value_new_INTEGER()), which a whitespace commit cannot fix; the others are older alignment and brace errors. The changed lines add no error.

[xiaoxiang781216]

let's ignore style error since the foreign code base don't follow nuttx style and ci error which is fixed by apache/nuttx#20949 instead wasting ci resource by retrigger.

[cederom]

@xiaoxiang781216: let's ignore style error since the foreign code base don't follow nuttx style and ci error which is fixed by apache/nuttx#20949 instead wasting ci resource by retrigger.

@jerpelea what do you think can we do that or would that mess up backporting later on? :-)