1 files changed, 60 insertions, 0 deletions
diff --git a/compiler/rustc_target/src/spec/thumb_base.rs b/compiler/rustc_target/src/spec/thumb_base.rs
new file mode 100644
index 000000000..049142b89
--- /dev/null
+++ b/compiler/rustc_target/src/spec/thumb_base.rs
@@ -0,0 +1,60 @@
+// These `thumbv*` targets cover the ARM Cortex-M family of processors which are widely used in
+// microcontrollers. Namely, all these processors:
+//
+// - Cortex-M0
+// - Cortex-M0+
+// - Cortex-M1
+// - Cortex-M3
+// - Cortex-M4(F)
+// - Cortex-M7(F)
+// - Cortex-M23
+// - Cortex-M33
+//
+// We have opted for these instead of one target per processor (e.g., `cortex-m0`, `cortex-m3`,
+// etc) because the differences between some processors like the cortex-m0 and cortex-m1 are almost
+// non-existent from the POV of codegen so it doesn't make sense to have separate targets for them.
+// And if differences exist between two processors under the same target, rustc flags can be used to
+// optimize for one processor or the other.
+//
+// Also, we have not chosen a single target (`arm-none-eabi`) like GCC does because this makes
+// difficult to integrate Rust code and C code. Targeting the Cortex-M4 requires different gcc flags
+// than the ones you would use for the Cortex-M0 and with a single target it'd be impossible to
+// differentiate one processor from the other.
+//
+// About arm vs thumb in the name. The Cortex-M devices only support the Thumb instruction set,
+// which is more compact (higher code density), and not the ARM instruction set. That's why LLVM
+// triples use thumb instead of arm. We follow suit because having thumb in the name let us
+// differentiate these targets from our other `arm(v7)-*-*-gnueabi(hf)` targets in the context of
+// build scripts / gcc flags.
+
+use crate::spec::TargetOptions;
+use crate::spec::{FramePointer, LinkerFlavor, LldFlavor, PanicStrategy, RelocModel};
+
+pub fn opts() -> TargetOptions {
+    // See rust-lang/rfcs#1645 for a discussion about these defaults
+    TargetOptions {
+        linker_flavor: LinkerFlavor::Lld(LldFlavor::Ld),
+        // In most cases, LLD is good enough
+        linker: Some("rust-lld".into()),
+        // Because these devices have very little resources having an unwinder is too onerous so we
+        // default to "abort" because the "unwind" strategy is very rare.
+        panic_strategy: PanicStrategy::Abort,
+        // Similarly, one almost always never wants to use relocatable code because of the extra
+        // costs it involves.
+        relocation_model: RelocModel::Static,
+        // When this section is added a volatile load to its start address is also generated. This
+        // volatile load is a footgun as it can end up loading an invalid memory address, depending
+        // on how the user set up their linker scripts. This section adds pretty printer for stuff
+        // like std::Vec, which is not that used in no-std context, so it's best to left it out
+        // until we figure a way to add the pretty printers without requiring a volatile load cf.
+        // rust-lang/rust#44993.
+        emit_debug_gdb_scripts: false,
+        // LLVM is eager to trash the link register when calling `noreturn` functions, which
+        // breaks debugging. Preserve LR by default to prevent that from happening.
+        frame_pointer: FramePointer::Always,
+        // ARM supports multiple ABIs for enums, the linux one matches the default of 32 here
+        // but any arm-none or thumb-none target will be defaulted to 8 on GCC and clang
+        c_enum_min_bits: 8,
+        ..Default::default()
+    }
+}