34489eb2af
The current implementation of Gogo::pkgpath_for_symbol was written in
a way that allowed two distinct package paths to map to the same
symbol, which could cause collisions at link- time or compile-time.
Switch to a better mangling scheme to insure that we get a unique
packagepath symbol for each package. In the new scheme instead of having
separate mangling schemes for identifiers and package paths, the
main identifier mangler ("go_encode_id") now handles mangling of
both packagepath characters and identifier characters.
The new mangling scheme is more intrusive: "foo/bar.Baz" is mangled as
"foo..z2fbar.Baz" instead of "foo_bar.Baz". To mitigate this, this
patch also adds a demangling capability so that function names
returned from runtime.CallersFrames are converted back to their
original unmangled form.
Changing the pkgpath_for_symbol scheme requires updating a number of
//go:linkname directives and C "__asm__" directives to match the new
scheme, as well as updating the 'gotest' driver (which makes
assumptions about the correct mapping from pkgpath symbol to package
name).
Fixes golang/go#27534.
Reviewed-on: https://go-review.googlesource.com/c/135455
From-SVN: r265510
41 lines
1.5 KiB
Go
41 lines
1.5 KiB
Go
// Copyright 2017 The Go Authors. All rights reserved.
|
|
// Use of this source code is governed by a BSD-style
|
|
// license that can be found in the LICENSE file.
|
|
|
|
package runtime
|
|
|
|
import "unsafe"
|
|
|
|
var labelSync uintptr
|
|
|
|
//go:linkname runtime_setProfLabel runtime..z2fpprof.runtime_setProfLabel
|
|
func runtime_setProfLabel(labels unsafe.Pointer) {
|
|
// Introduce race edge for read-back via profile.
|
|
// This would more properly use &getg().labels as the sync address,
|
|
// but we do the read in a signal handler and can't call the race runtime then.
|
|
//
|
|
// This uses racereleasemerge rather than just racerelease so
|
|
// the acquire in profBuf.read synchronizes with *all* prior
|
|
// setProfLabel operations, not just the most recent one. This
|
|
// is important because profBuf.read will observe different
|
|
// labels set by different setProfLabel operations on
|
|
// different goroutines, so it needs to synchronize with all
|
|
// of them (this wouldn't be an issue if we could synchronize
|
|
// on &getg().labels since we would synchronize with each
|
|
// most-recent labels write separately.)
|
|
//
|
|
// racereleasemerge is like a full read-modify-write on
|
|
// labelSync, rather than just a store-release, so it carries
|
|
// a dependency on the previous racereleasemerge, which
|
|
// ultimately carries forward to the acquire in profBuf.read.
|
|
if raceenabled {
|
|
racereleasemerge(unsafe.Pointer(&labelSync))
|
|
}
|
|
getg().labels = labels
|
|
}
|
|
|
|
//go:linkname runtime_getProfLabel runtime..z2fpprof.runtime_getProfLabel
|
|
func runtime_getProfLabel() unsafe.Pointer {
|
|
return getg().labels
|
|
}
|