gtker

`tcc_1_7` calls functions using a format which breaks the function call operator detection. ```c int (*t)(); return (*t)(argc - optind, argv + optind); ```

The logic to call a function pointer is entirely inside `variable_load` which makes global function pointers not work. Duplicating the logic would work initially, but it would probably be better...

The file ```c #if 0 #endif ``` results in `impossible #if block` but the file (same content but with a starting newline) ```c #if 0 #endif ``` compiles correctly. There...

Because of how we currently implement `typedef`s as a copy of the original object the `typedef` isn't correctly updated to reflect when a declaration gets a definition. So ```c struct...

`fopen` with "a" doesn't work for UEFI

2

[M2-Mesoplanet has functionality that requires](https://github.com/oriansj/M2-Mesoplanet/pull/17) append mode. This [wasn't implemented for UEFI](https://github.com/oriansj/M2libc/pull/85) since there wasn't a straightforward way to implement it. As far as I can tell UEFI doesn't natively...

This isn't a shopping list of things we want implemented right now this is primarily to summarize what we have and what's missing, as well as categorize things that are...

This improves performance slightly and leads to significantly cleaner assembly.

```c int main() { char buf[4] = {1, 2, 3, 4}; } ``` and ```c int main() { struct { int a; } s = { 10 }; } ```...

From `known_issues.org` file. Pointer arithmetic only works for compound (+=, -=) and postfix/prefix (++, --) operators but is broken for regular operators (+, -). ## Example of failing code ```c...

From `known_issues.org`: Both sides of `&&` evaluate because it hasn't been shown to be worth the effort of implementation of short-circuit logic ## Example of failing code ```c #include #include...