fix compile error on the std.c.EXC.MASK field on darwin

[srijan-paul]

The backing integer for the MASK struct is a u32, but the bool fields sum up to a u13.
I've added 19 bits of padding.

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?
EDIT: Looks using std.c.darwing.exception_mask_t would be a better idea here.
Is there a reason why the std library doesn't use C types in these places?

Reference: #21094

P.S: Should we consider having a test block for this just to see if it compiles? Something like:

test {
std.testing.expectEqual(32, @bitSizeOf(MASK));
}

[FnControlOption]

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?

I would use exception_mask_t instead (or whichever type has the same size as c_uint)

Reference: https://github.com/apple-oss-distributions/xnu/blob/main/osfmk/mach/exception_types.h#L207

[srijan-paul]

or whichever type has the same size as c_uint

In that case, why not use c_uint directly?

[FnControlOption]

exception_mask_t is already defined as c_uint in std.c.darwin, but packed structs in the standard library don’t use C types for some reason

[srijan-paul]

Interesting. I wonder why that decision was taken.
I could change the u32 to exception_mask_t in this PR.

[srijan-paul]

The backing integer for the MASK struct is a u32, but the bool fields sum up to a u13.
I've added 19 bits of padding.

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?
EDIT: Looks using std.c.darwing.exception_mask_t would be a better idea here.
Is there a reason why the std library doesn't use C types in these places?

Reference: #21094

P.S: Should we consider having a test block for this just to see if it compiles? Something like:

test {
std.testing.expectEqual(32, @bitSizeOf(MASK));
}

[FnControlOption]

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?

I would use exception_mask_t instead (or whichever type has the same size as c_uint)

Reference: https://github.com/apple-oss-distributions/xnu/blob/main/osfmk/mach/exception_types.h#L207

[srijan-paul]

or whichever type has the same size as c_uint

In that case, why not use c_uint directly?

[FnControlOption]

exception_mask_t is already defined as c_uint in std.c.darwin, but packed structs in the standard library don’t use C types for some reason

[srijan-paul]

Interesting. I wonder why that decision was taken.
I could change the u32 to exception_mask_t in this PR.

[srijan-paul]

The backing integer for the MASK struct is a u32, but the bool fields sum up to a u13.
I've added 19 bits of padding.

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?
EDIT: Looks using std.c.darwing.exception_mask_t would be a better idea here.
Is there a reason why the std library doesn't use C types in these places?

Reference: #21094

P.S: Should we consider having a test block for this just to see if it compiles? Something like:

test {
std.testing.expectEqual(32, @bitSizeOf(MASK));
}

[FnControlOption]

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?

I would use exception_mask_t instead (or whichever type has the same size as c_uint)

Reference: https://github.com/apple-oss-distributions/xnu/blob/main/osfmk/mach/exception_types.h#L207

[srijan-paul]

or whichever type has the same size as c_uint

In that case, why not use c_uint directly?

[FnControlOption]

exception_mask_t is already defined as c_uint in std.c.darwin, but packed structs in the standard library don’t use C types for some reason

[srijan-paul]

Interesting. I wonder why that decision was taken.
I could change the u32 to exception_mask_t in this PR.

[srijan-paul]

The backing integer for the MASK struct is a u32, but the bool fields sum up to a u13.
I've added 19 bits of padding.

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?
EDIT: Looks using std.c.darwing.exception_mask_t would be a better idea here.
Is there a reason why the std library doesn't use C types in these places?

Reference: #21094

P.S: Should we consider having a test block for this just to see if it compiles? Something like:

test {
std.testing.expectEqual(32, @bitSizeOf(MASK));
}

[FnControlOption]

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?

I would use exception_mask_t instead (or whichever type has the same size as c_uint)

Reference: https://github.com/apple-oss-distributions/xnu/blob/main/osfmk/mach/exception_types.h#L207

[srijan-paul]

or whichever type has the same size as c_uint

In that case, why not use c_uint directly?

[FnControlOption]

exception_mask_t is already defined as c_uint in std.c.darwin, but packed structs in the standard library don’t use C types for some reason

[srijan-paul]

Interesting. I wonder why that decision was taken.
I could change the u32 to exception_mask_t in this PR.

[srijan-paul]

The backing integer for the MASK struct is a u32, but the bool fields sum up to a u13.
I've added 19 bits of padding.

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?
EDIT: Looks using std.c.darwing.exception_mask_t would be a better idea here.
Is there a reason why the std library doesn't use C types in these places?

Reference: #21094

P.S: Should we consider having a test block for this just to see if it compiles? Something like:

test {
std.testing.expectEqual(32, @bitSizeOf(MASK));
}

[FnControlOption]

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?

I would use exception_mask_t instead (or whichever type has the same size as c_uint)

Reference: https://github.com/apple-oss-distributions/xnu/blob/main/osfmk/mach/exception_types.h#L207

[srijan-paul]

or whichever type has the same size as c_uint

In that case, why not use c_uint directly?

[FnControlOption]

exception_mask_t is already defined as c_uint in std.c.darwin, but packed structs in the standard library don’t use C types for some reason

[srijan-paul]

Interesting. I wonder why that decision was taken.
I could change the u32 to exception_mask_t in this PR.

[srijan-paul]

The backing integer for the MASK struct is a u32, but the bool fields sum up to a u13.
I've added 19 bits of padding.

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?
EDIT: Looks using std.c.darwing.exception_mask_t would be a better idea here.
Is there a reason why the std library doesn't use C types in these places?

Reference: #21094

P.S: Should we consider having a test block for this just to see if it compiles? Something like:

test {
std.testing.expectEqual(32, @bitSizeOf(MASK));
}

[FnControlOption]

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?

I would use exception_mask_t instead (or whichever type has the same size as c_uint)

Reference: https://github.com/apple-oss-distributions/xnu/blob/main/osfmk/mach/exception_types.h#L207

[srijan-paul]

or whichever type has the same size as c_uint

In that case, why not use c_uint directly?

[FnControlOption]

exception_mask_t is already defined as c_uint in std.c.darwin, but packed structs in the standard library don’t use C types for some reason

[srijan-paul]

Interesting. I wonder why that decision was taken.
I could change the u32 to exception_mask_t in this PR.

[srijan-paul]

The backing integer for the MASK struct is a u32, but the bool fields sum up to a u13.
I've added 19 bits of padding.

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?
EDIT: Looks using std.c.darwing.exception_mask_t would be a better idea here.
Is there a reason why the std library doesn't use C types in these places?

Reference: #21094

P.S: Should we consider having a test block for this just to see if it compiles? Something like:

test {
std.testing.expectEqual(32, @bitSizeOf(MASK));
}

[FnControlOption]

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?

I would use exception_mask_t instead (or whichever type has the same size as c_uint)

Reference: https://github.com/apple-oss-distributions/xnu/blob/main/osfmk/mach/exception_types.h#L207

[srijan-paul]

or whichever type has the same size as c_uint

In that case, why not use c_uint directly?

[FnControlOption]

exception_mask_t is already defined as c_uint in std.c.darwin, but packed structs in the standard library don’t use C types for some reason

[srijan-paul]

Interesting. I wonder why that decision was taken.
I could change the u32 to exception_mask_t in this PR.

[srijan-paul]

The backing integer for the MASK struct is a u32, but the bool fields sum up to a u13.
I've added 19 bits of padding.

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?
EDIT: Looks using std.c.darwing.exception_mask_t would be a better idea here.
Is there a reason why the std library doesn't use C types in these places?

Reference: #21094

P.S: Should we consider having a test block for this just to see if it compiles? Something like:

test {
std.testing.expectEqual(32, @bitSizeOf(MASK));
}

[FnControlOption]

Alternatively, we could add 3 bits of padding and make it a u16 backed packed struct?

I would use exception_mask_t instead (or whichever type has the same size as c_uint)

Reference: https://github.com/apple-oss-distributions/xnu/blob/main/osfmk/mach/exception_types.h#L207

[srijan-paul]

or whichever type has the same size as c_uint

In that case, why not use c_uint directly?

[FnControlOption]

exception_mask_t is already defined as c_uint in std.c.darwin, but packed structs in the standard library don’t use C types for some reason

[srijan-paul]

Interesting. I wonder why that decision was taken.
I could change the u32 to exception_mask_t in this PR.