Add a new function attribute, nonlazybind, which inhibits lazy-loading

optimizations when emitting calls to the function;  instead those calls may
use faster relocations which require the function to be immediately resolved
upon loading the dynamic object featuring the call.  This is useful when it
is known that the function will be called frequently and pervasively and
therefore there is no merit in delaying binding of the function.

Currently only implemented for x86-64, where it turns into a call through
the global offset table.

Patch by Dan Gohman, who assures me that he's going to add LangRef documentation
for this once it's committed.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@133080 91177308-0d34-0410-b5e6-96231b3b80d8

1 files changed, 27 insertions, 0 deletions

diff --git a/test/CodeGen/X86/non-lazy-bind.ll b/test/CodeGen/X86/non-lazy-bind.ll
new file mode 100644
index 0000000000..f72965877d
--- /dev/null
+++ b/test/CodeGen/X86/non-lazy-bind.ll
@@ -0,0 +1,27 @@
+; RUN: llc -mtriple=x86_64-apple-darwin < %s | FileCheck %s
+
+declare void @lazy() nonlazybind
+declare void @not()
+
+; CHECK: foo:
+; CHECK:  callq _not
+; CHECK:  callq *_lazy@GOTPCREL(%rip)
+define void @foo() nounwind {
+  call void @not()
+  call void @lazy()
+  ret void
+}
+
+; CHECK: tail_call_regular:
+; CHECK:   jmp _not
+define void @tail_call_regular() nounwind {
+  tail call void @not()
+  ret void
+}
+
+; CHECK: tail_call_eager:
+; CHECK:   jmpq *_lazy@GOTPCREL(%rip)
+define void @tail_call_eager() nounwind {
+  tail call void @lazy()
+  ret void
+}