1. Use n->alloc, not n->left, to hold the allocated temp being
passed from orderstmt/orderexpr to walk.
2. Treat method values the same as closures.
3. Use killed temporary for composite literal passed to
non-escaping function argument.
4. Clean temporaries promptly in if and for statements.
5. Clean temporaries promptly in select statements.
As part of this, move all the temporary-generating logic
out of select.c into order.c, so that the temporaries can
be reclaimed.
With the new temporaries, can re-enable the 1-entry
select optimization. Fixes issue 7672.
While we're here, fix a 1-line bug in select processing
turned up by the new liveness test (but unrelated; select.c:72).
Fixes#7686.
6. Clean temporaries (but not particularly promptly) in switch
and range statements.
7. Clean temporary used during convT2E/convT2I.
8. Clean temporaries promptly during && and || expressions.
---
CL 81940043 reduced the number of ambiguously live temps
in the godoc binary from 860 to 711.
CL 83090046 reduced the number from 711 to 121.
This CL reduces the number from 121 to 23.
15 the 23 that remain are in fact ambiguously live.
The final 8 could be fixed but are not trivial and
not common enough to warrant work at this point
in the release cycle.
These numbers only count ambiguously live temps,
not ambiguously live user-declared variables.
There are 18 such variables in the godoc binary after this CL,
so a total of 41 ambiguously live temps or user-declared
variables.
The net effect is that zeroing anything on entry to a function
should now be a rare event, whereas earlier it was the
common case.
This is good enough for Go 1.3, and probably good
enough for future releases too.
Fixes#7345.
LGTM=khr
R=khr
CC=golang-codereviews
https://golang.org/cl/83000048
1. In functions with heap-allocated result variables or with
defer statements, the return sequence requires more than
just a single RET instruction. There is an optimization that
arranges for all returns to jump to a single copy of the return
epilogue in this case. Unfortunately, that optimization is
fundamentally incompatible with PC-based liveness information:
it takes PCs at many different points in the function and makes
them all land at one PC, making the combined liveness information
at that target PC a mess. Disable this optimization, so that each
return site gets its own copy of the 'call deferreturn' and the
copying of result variables back from the heap.
This removes quite a few spurious 'ambiguously live' variables.
2. Let orderexpr allocate temporaries that are passed by address
to a function call and then die on return, so that we can arrange
an appropriate VARKILL.
2a. Do this for ... slices.
2b. Do this for closure structs.
2c. Do this for runtime.concatstring, which is the implementation
of large string additions. Change representation of OADDSTR to
an explicit list in typecheck to avoid reconstructing list in both
walk and order.
3. Let orderexpr allocate the temporary variable copies used for
range loops, so that they can be killed when the loop is over.
Similarly, let it allocate the temporary holding the map iterator.
CL 81940043 reduced the number of ambiguously live temps
in the godoc binary from 860 to 711.
This CL reduces the number to 121. Still more to do, but another
good checkpoint.
Update #7345
LGTM=khr
R=khr
CC=golang-codereviews
https://golang.org/cl/83090046
Make the loop nesting depth of &x depend on where x is declared,
not on where the &x appears. The latter is only a conservative
estimate of the former. Being more careful can avoid some
variables escaping, and it is easier to reason about.
It would have avoided issue 7313, although that was still a bug
worth fixing.
Not much effect in the tree: one variable in the whole tree
is saved from a heap allocation (something in x509 parsing).
LGTM=daniel.morsing
R=daniel.morsing
CC=golang-codereviews
https://golang.org/cl/62380043
Logically, the init statement is in the enclosing scopes loopdepth, not inside the for loop.
Fixes#7313.
LGTM=rsc
R=golang-codereviews, gobot, rsc
CC=golang-codereviews
https://golang.org/cl/62430043
The escape analysis works by tracing assignment paths from
variables that start with pointer type, or addresses of variables
(addresses are always pointers). It does allow non-pointers
in the path, so that in this code it sees x's value escape into y:
var x *[10]int
y := (*int)(unsafe.Pointer(uintptr(unsafe.Pointer(x))+32))
It must allow uintptr in order to see through this kind of
"pointer arithmetic".
It also traces such values if they end up as uintptrs passed to
functions. This used to be important because packages like
encoding/gob passed around uintptrs holding real pointers.
The introduction of precise collection of stacks has forced
code to be more honest about which declared stack variables
hold pointers and which do not. In particular, the garbage
collector no longer sees pointers stored in uintptr variables.
Because of this, packages like encoding/gob have been fixed.
There is not much point in the escape analysis accepting
uintptrs as holding pointers at call boundaries if the garbage
collector does not.
Excluding uintptr-valued arguments brings the escape
analysis in line with the garbage collector and has the
useful side effect of making arguments to syscall.Syscall
not appear to escape.
That is, this CL should yield the same benefits as
CL 45930043 (rolled back in CL 53870043), but it does
so by making uintptrs less special, not more.
R=golang-codereviews, iant
CC=golang-codereviews
https://golang.org/cl/53940043
The compiler change is an ugly hack.
We can do better.
««« original CL description
syscall: mark arguments to Syscall as noescape
Heap arguments to "async" syscalls will break when/if we have moving GC anyway.
With this change is must not break until moving GC, because a user must
reference the object in Go to preserve liveness. Otherwise the code is broken already.
Reduces number of leaked params from 125 to 36 on linux.
R=golang-codereviews, mikioh.mikioh, bradfitz
CC=cshapiro, golang-codereviews, khr, rsc
https://golang.org/cl/45930043
»»»
R=golang-codereviews, r
CC=bradfitz, dvyukov, golang-codereviews
https://golang.org/cl/53870043
Heap arguments to "async" syscalls will break when/if we have moving GC anyway.
With this change is must not break until moving GC, because a user must
reference the object in Go to preserve liveness. Otherwise the code is broken already.
Reduces number of leaked params from 125 to 36 on linux.
R=golang-codereviews, mikioh.mikioh, bradfitz
CC=cshapiro, golang-codereviews, khr, rsc
https://golang.org/cl/45930043
Design doc at golang.org/s/go12slice.
This is an experimental feature and may not be included in the release.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/10743046
Functions without bodies were excluded from the ordering logic,
because when I wrote the ordering logic there was no reason to
analyze them.
But then we added //go:noescape tags that need analysis, and we
didn't update the ordering logic.
So in the absence of good ordering, //go:noescape only worked
if it appeared before the use in the source code.
Fixes#5773.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/10570043
It used to not mark parameters as escaping if only one of the
fields it points to leaks out of the function. This causes
problems when importing from another package.
Fixes#4964.
R=rsc, lvd, dvyukov, daniel.morsing
CC=golang-dev
https://golang.org/cl/7648045
A new comment directive //go:noescape instructs the compiler
that the following external (no body) func declaration should be
treated as if none of its arguments escape to the heap.
Fixes#4099.
R=golang-dev, dave, minux.ma, daniel.morsing, remyoudompheng, adg, agl, iant
CC=golang-dev
https://golang.org/cl/7289048
If the analysis reached a node twice, then the analysis was cut off.
However, if the second arrival is at a lower depth (closer to escaping)
then it is important to repeat the traversal.
The repeating must be cut off at some point to avoid the occasional
infinite recursion. This CL cuts it off as soon as possible while still
passing all tests.
Fixes#4751.
R=ken2
CC=golang-dev, lvd
https://golang.org/cl/7303043
A fatal error used to happen when escassign-ing a multiple
function return to a single node. However, the situation
naturally appears when using "go f(g())" or "defer f(g())",
because g() is escassign-ed to sink.
Fixes#4529.
R=golang-dev, lvd, minux.ma, rsc
CC=golang-dev
https://golang.org/cl/6920060
The code assumed that the only choices were EscNone, EscScope, and EscHeap,
so that it makes sense to set EscScope only if the current setting is EscNone.
Now that we have the many variants of EscReturn, this logic is false, and it was
causing important EscScopes to be ignored in favor of EscReturn.
Fixes#4360.
R=ken2
CC=golang-dev, lvd
https://golang.org/cl/6816103
Avoids problems with local declarations shadowing other names.
We write a more explicit form than the incoming program, so there
may be additional type annotations. For example:
int := "hello"
j := 2
would normally turn into
var int string = "hello"
var j int = 2
but the int variable shadows the int type in the second line.
This CL marks all local variables with a per-function sequence number,
so that this would instead be:
var int·1 string = "hello"
var j·2 int = 2
Fixes#4326.
R=ken2
CC=golang-dev
https://golang.org/cl/6816100
Plan 9 and Go's lib9/fmt disagree on whether %#x includes the 0x prefix
when printing 0, because ANSI C gave bad advice long ago.
Avoiding that case makes binaries compiled on different systems compatible.
R=ken2
CC=akumar, golang-dev
https://golang.org/cl/6814066
includes step 0: synthesize outparams, from 6600044
includes step 1,2: give outparams loopdepth 0 and verify unchanged results
generate esc:$mask tags, but still tie to sink if a param has mask != 0
from 6610054
adds final steps:
- have esccall generate n->escretval, a list of nodes the function results flow to
- use these in esccall and ORETURN/OAS2FUNC/and f(g())
- only tie parameters to sink if tag is absent, otherwise according to mask, tie them to escretval
R=rsc, bradfitz
CC=dave, gobot, golang-dev, iant, rsc
https://golang.org/cl/6741044
includes step 0: synthesize outparams, from 6600044
step 1: give outparams loopdepth 0 and verify unchanged results
step 2: generate esc:$mask tags, but still tie to sink if a param has mask != 0
next step: use in esccall (and ORETURN with implicit OAS2FUNC) to avoid tying to sink
R=rsc
CC=golang-dev
https://golang.org/cl/6610054
Was not handling &x.y[0] and &x.y.z correctly where
y is an array or struct-valued field (not a pointer).
R=ken2
CC=golang-dev
https://golang.org/cl/6551059
If there are mutually recursive functions, there is a cycle in
the dependency graph, so the order is actually dependency order
among the strongly connected components: mutually recursive
functions get put into the same batch and analyzed together.
(Until now the entire package was put in one batch.)
The non-recursive case (single function, maybe with some
closures inside) will be able to be more precise about inputs
that escape only back to outputs, but that is not implemented yet.
R=ken2
CC=golang-dev, lvd
https://golang.org/cl/6304050
The original implementation of closures created the
underlying top-level function during walk, which is fairly
late in the compilation process and caused ordering-based
complications due to earlier stages that had to be repeated
any number of times.
Create the underlying function during typecheck, much
earlier, so that later stages can be run just once.
The result is a simpler compilation sequence.
R=ken2
CC=golang-dev
https://golang.org/cl/6279049
The escape analysis code does not make a distinction between
scalar and pointers fields in structs. Non-pointer fields
that escape should not make the whole struct escape.
R=lvd, rsc
CC=golang-dev, remy
https://golang.org/cl/5489128
This avoids degraded performance caused by extra labels
emitted by inlining (breaking strconv ftoa alloc count unittest) and is better in any case.
R=rsc
CC=golang-dev
https://golang.org/cl/5483071
Got rid of all the magic mystery globals. Now
for %N, %T, and %S, the flags +,- and # set a sticky
debug, sym and export mode, only visible in the new fmt.c.
Default is error mode. Handle h and l flags consistently with
the least side effects, so we can now change
things without worrying about unrelated things
breaking.
fixes#2361
R=rsc
CC=golang-dev
https://golang.org/cl/5316043
string literals used as package qualifiers are now prefixed with '@'
which obviates the need for the extra ':' before tags.
R=rsc, gri, lvd
CC=golang-dev
https://golang.org/cl/5129057
Escape analysis was incorrectly assuming that
functions without bodies don't leak their
parameters. This meant that sync/atomic's
TestAddInt64 was allocating x on its stack,
and then x was not properly aligned for use
with the atomic 64-bit instructions. Obviously
we should figure out the alignment story on 5g
too, but this fix is correct and should restore the
build to 'ok'.
TBR=lvd
CC=golang-dev
https://golang.org/cl/4964047
-s now means *disable* escape analysis.
Fix escape leaks for struct/slice/map literals.
Add ... tracking.
Rewrite new(T) and slice literal into stack allocation when safe.
Add annotations to reflect.
Reflect is too chummy with the compiler,
so changes like these affect it more than they should.
R=lvd, dave, gustavo
CC=golang-dev
https://golang.org/cl/4954043
#include "go.h" (or "gg.h")
becomes
#include <u.h>
#include <libc.h>
#include "go.h"
so that go.y can #include <stdio.h>
after <u.h> but before "go.h".
This is necessary on Plan 9.
R=ken2
CC=golang-dev
https://golang.org/cl/4971041
For now it's switch-on-and-offable with -s, and the effects can be inspected
with -m. Defaults are the old codepaths.
R=rsc
CC=golang-dev
https://golang.org/cl/4634073
