Meghan Denny
Meghan Denny
can u import this in a relevant file so that it is parsed in tests and doesnt bitrot?
- the types used are very wrong - the constants should be in a namespace - the functions should not use @bitCast (was going to enumerate the way the types...
`std.c` should use `c_int`/`c_uint` `shmdt` should accept `*const anyopaque` `shmctl`'s buf accepts a struct ptr, not a byte one `shmid` should be unsigned returning `usize` then accepting `u32` feels sus
yeah this is more like https://github.com/nektro/gimme/blob/master/src/FailingAllocator.zig
yeah doesn't seem like it'd be your fault at first glance but its UB in the compiler causing a crash https://zigbin.io/544ee8
I'm not totally sure, was more just documenting the finding to ease anyone else looking into it as well
letting the human do the diff would be better imo ``` test.zig:4:19: error: expected type 'T0', found 'T1' test.zig:4:19: note: 'T0' = '*array_hash_map.ArrayHashMapUnmanaged(u32,u32,array_hash_map.AutoContext(u32),false)' test.zig:4:19: note: 'T1' = 'array_hash_map.ArrayHashMapUnmanaged(u32,u32,array_hash_map.AutoContext(u32),false)' ```
not sure this is worth is worth the complexity for the common case given that unless the whole array fits in less than 8 bytes there will be a massive...
oh the change i thought this was, it seems was made 5 months ago. my suggestion would be to make the `.len` field itself a usize
note that this patch should additionally be sent upstream so its not lost in zig when we next upgrade glibc