Separate the built-in intrinsic feature implementations into different C files and then amalgamate them back together #216
Comments
|
It is possible that the buitins will be moved to a separate translation unit someday in order to speed up the compilation. However, I see no advantage in splitting with several #include. Having large source files is not a problem for me. |
chenzd123456
pushed a commit
to Micro-Application-Runtime/quickjs
that referenced
this issue
Feb 5, 2024
This JS_ReadObject() flag no longer works for bytecode. The IC opcodes are patched during execution. Fixes: quickjs-ng/quickjs#206 Refs: quickjs-ng/quickjs#120
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
The source code for quickjs.c is currently at 55K lines, and it is expected the grow. We should consider splitting the intrinsic implementations into its own C files, and then we can
#includethem back again.For example, we can split the atomic intrinsic support into
intrinsic/atomics.cand then#include "intrinsic/atomics.c"back again in the quickjs.c file.This way, hacking the actual QJS engine source code itself would be cleaner and easier. We can do more by formally splitting the intrinsic code into its own translation unit (that means into its own object file), but I think that is not going to be easy at the moment.
The text was updated successfully, but these errors were encountered: