Fix handling of FFI arguments

[nagisa]

r? @eddyb @nikomatsakis or whoever else.

cc @alexcrichton @rust-lang/core

The strategy employed here was to essentially change code we generate from

  %s = alloca %S ; potentially smaller than argument, but never larger
  %1 = bitcast %S* %s to { i64, i64 }*
  store { i64, i64 } %0, { i64, i64 }* %1, align 4

to

  %1 = alloca { i64, i64 } ; the copy of argument itself
  store { i64, i64 } %0, { i64, i64 }* %1, align 4
  %s = bitcast { i64, i64 }* %1 to %S* ; potentially truncate by casting to a pointer of smaller type.

[nagisa]

Hmm, it clashes with debuginfo.

[eddyb]

This won't work with debuginfo, which wants an alloca with no casts on top.
I would suggest giving it the alloca before the cast and hope that it doesn't care about the type (debuginfo metadata should describe the actual value in the alloca all by itself).
cc @michaelwoerister

[alexcrichton]

Thanks for the quick fix @nagisa!

[nagisa]

Added a check for debuginfo thing. All should work now.

If there’s any hard-pressing nits, feel to fork and make another PR as I’m likely to be in bed for a while now.

[eddyb]

@bors r+ p=100

[bors]

📌 Commit e0e50a4 has been approved by eddyb

[bors]

⌛️ Testing commit e0e50a4 with merge 73fdf78...

[bors]

💔 Test failed - auto-linux-64-debug-opt

[eddyb]

Miscompiling the compiler, resulting in this for stage2 libcore:

thread 'rustc' panicked at 'RefCell<T> already borrowed', ../src/libcore/cell.rs:492

@alexcrichton Serious question: why is RUST_BACKTRACE=1 not set on the bots? I assume it has to do with some tests not taking that very well?

[alexcrichton]

No particular reason, it just hasn't been done yet. We'd probably benefit from just setting it in the build system directly.

[nagisa]

@bors r=eddyb 2f0da79

[nagisa]

@bors r=eddyb 5b40452

[bors]

⌛️ Testing commit 5b40452 with merge 3c795e0...

[bors]

💔 Test failed - auto-win-gnu-32-opt-rustbuild

[nagisa]

@bors retry force

[bors]

⚡️ Previous build results for auto-linux-64-cargotest, auto-linux-cross-opt, auto-win-msvc-64-cargotest are reusable. Rebuilding only auto-linux-32-nopt-t, auto-linux-32-opt, auto-linux-32cross-opt, auto-linux-64-cross-armhf, auto-linux-64-cross-armsf, auto-linux-64-cross-freebsd, auto-linux-64-cross-netbsd, auto-linux-64-debug-opt, auto-linux-64-nopt-t, auto-linux-64-opt, auto-linux-64-opt-mir, auto-linux-64-opt-rustbuild, auto-linux-64-x-android-t, auto-linux-musl-64-opt, auto-mac-32-opt, auto-mac-64-nopt-t, auto-mac-64-opt, auto-mac-64-opt-rustbuild, auto-mac-cross-ios-opt, auto-win-gnu-32-nopt-t, auto-win-gnu-32-opt, auto-win-gnu-32-opt-rustbuild, auto-win-gnu-64-nopt-t, auto-win-gnu-64-opt, auto-win-msvc-32-cross-opt, auto-win-msvc-32-opt, auto-win-msvc-64-opt, auto-win-msvc-64-opt-mir, auto-win-msvc-64-opt-rustbuild...

[sanxiyn]

Another case of #33844.

[bors]

☀️ Test successful - auto-linux-32-nopt-t, auto-linux-32-opt, auto-linux-32cross-opt, auto-linux-64-cargotest, auto-linux-64-cross-armhf, auto-linux-64-cross-armsf, auto-linux-64-cross-freebsd, auto-linux-64-cross-netbsd, auto-linux-64-debug-opt, auto-linux-64-nopt-t, auto-linux-64-opt, auto-linux-64-opt-mir, auto-linux-64-opt-rustbuild, auto-linux-64-x-android-t, auto-linux-cross-opt, auto-linux-musl-64-opt, auto-mac-32-opt, auto-mac-64-nopt-t, auto-mac-64-opt, auto-mac-64-opt-rustbuild, auto-mac-cross-ios-opt, auto-win-gnu-32-nopt-t, auto-win-gnu-32-opt, auto-win-gnu-32-opt-rustbuild, auto-win-gnu-64-nopt-t, auto-win-gnu-64-opt, auto-win-msvc-32-cross-opt, auto-win-msvc-32-opt, auto-win-msvc-64-cargotest, auto-win-msvc-64-opt, auto-win-msvc-64-opt-mir, auto-win-msvc-64-opt-rustbuild

[alexcrichton]

Discussed on IRC and accepted for a beta backport