compiler: do not perform nil checking when indexing slices

The x/tools/go/ssa package splits slice loads/stores into two
operations. So for code like this:

    x = p[3]

It has two instructions:

    x_ptr = &p[3]
    x = *x_ptr

This makes the IR simpler, but also means we're accidentally inserting
more nil checks than necessary: the slice index operation has
effectively already checked for nil by performing a bounds check.
Therefore, omit nil pointer checks for pointers created by
*ssa.IndexAddr.

This change is necessary to make sure a future removal of runtime.isnil
will not cause the escape analysis pass to regress. Apart from that, it
reduces code size slightly in many smoke tests (with no increases in
code size).
This commit is contained in:
Ayke van Laethem
2020-03-25 19:23:13 +01:00
committed by Ron Evans
parent 85854cd58b
commit 19f8874764
3 changed files with 17 additions and 8 deletions
+10 -1
View File
@@ -8,6 +8,7 @@ import (
"go/token"
"go/types"
"golang.org/x/tools/go/ssa"
"tinygo.org/x/go-llvm"
)
@@ -151,12 +152,20 @@ func (b *builder) createChanBoundsCheck(elementSize uint64, bufSize llvm.Value,
// createNilCheck checks whether the given pointer is nil, and panics if it is.
// It has no effect in well-behaved programs, but makes sure no uncaught nil
// pointer dereferences exist in valid Go code.
func (b *builder) createNilCheck(ptr llvm.Value, blockPrefix string) {
func (b *builder) createNilCheck(inst ssa.Value, ptr llvm.Value, blockPrefix string) {
// Check whether we need to emit this check at all.
if !ptr.IsAGlobalValue().IsNil() {
return
}
switch inst.(type) {
case *ssa.IndexAddr:
// This pointer is the result of an index operation into a slice or
// array. Such slices/arrays are already bounds checked so the pointer
// must be a valid (non-nil) pointer. No nil checking is necessary.
return
}
// Compare against nil.
var isnil llvm.Value
if ptr.Type().PointerAddressSpace() == 0 {