When using llvm opaque pointers, typed pointers and pointer bitcasts are
no longer needed. This also avoids needing packed struct layouts that
are nested inside pointers, letting us avoid computing struct layouts
in Sema that could cause unnecessary dependency loops.
Normally when we want a pointer to the end of a struct we just add 1 to
the struct pointer. However, when it is a zero-bit struct, the pointer
type being used during lowering is often a dummy pointer type that
actually points to a non-zero-bit type, so we actually want to add 0
instead, since a zero-bit struct begins and ends at the same address.
* Initialize `big_align` with 1 as 0 is not a valid alignment.
* Add an assert to `alignForwardGeneric` to catch this issue earlier.
* Refactor valid alignment checks to call a more descriptive function.
* Handle a `null` return from `llvmFieldIndex`.
* Add a behavior test to test this code path.
* Reword this test name, which incorrectly described how pointers to
zero-bit fields behave, and instead describe the actual test.
* Handle a `null` return from `llvmFieldIndex`.
* Add a behavior test to test this code path.
* Reword this test name, which incorrectly described how pointers to
zero-bit fields behave, and instead describe the actual test.
These bugs were triggered in the C backend by aarch64-specific code in
os/windows.zig. Intentionally not updating zig1.wasm yet because of
upcoming changes and since aarch64-windows is not tested on master yet.
This will make stack traces and debugging experience more consistent
in the sense that the presence of source lines in stack traces will
not be dependent on the current working directory of the running process.
There are still a few occurrences of "stage1" in the standard library
and self-hosted compiler source, however, these instances need a bit
more careful inspection to ensure no breakage.
When a local defined outside a loop dies inside the loop, it can still
be needed on subsequent loop iterations, so reuse of the local must be
deferred until after the loop ends. This causes behavior tests to pass.
In both backends they did not observe the Liveness information for try
instructions. Now they do. For the C backend this is necessary for
correctness; for the LLVM backend, it improves code generation.
