This helps reduce confusion with cmd/internal/pgo, which performs
compilation-independent analysis. pgoir associates that data with the
IR from the current package compilation.
For #58102.
Change-Id: I9ef1c8bc41db466d3340f41f6d071b95c09566de
Reviewed-on: https://go-review.googlesource.com/c/go/+/569338
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
CL 543657 dedup'd the go/defer statement recognition between the
inliner and static devirtualizer. This CL extends that for PGO-based
devirtualization too.
Change-Id: I998753132af1ef17329676f4e17515f16e0acb03
Reviewed-on: https://go-review.googlesource.com/c/go/+/543775
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Extend the pgodevirtualize debug flag to distinguish interface and
function devirtualization. Setting 1 keeps interface devirtualization
enabled but disables function value devirtualization.
For #64209.
Change-Id: I33aa7eb95ca0bdb215256d8c7cc8f9dac53ae30e
Reviewed-on: https://go-review.googlesource.com/c/go/+/543115
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
runtime.memhash_varlen is defined as a normal function, but it is
actually a closure. All references are generated by
cmd/compile/internal/reflectdata.genhash, which creates a closure
containing the size of the type, which memhash_varlen accesses with
runtime.getclosureptr.
Since this doesn't look like a normal closure, ir.Func.OClosure is not
set, thus PGO function value devirtualization is willing to devirtualize
it, generating a call that completely ignores the closure context. This
causes memhash_varlen to either crash or generate incorrect results.
Skip this function, which is the only caller of getclosureptr.
Unfortunately there isn't a good way to detect these ineligible
functions more generally.
Fixes#64209.
Change-Id: Ibf509406667c6d4e5d431f10e5b1d1f926ecd7dc
Reviewed-on: https://go-review.googlesource.com/c/go/+/543195
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Today, PGO-based devirtualization only applies to interface calls. This
CL extends initial support to function values (i.e., function/closure
pointers passed as arguments or stored in a struct).
This CL is a minimal implementation with several limitations.
* Export data lookup of function value callees not implemented
(equivalent of CL 497175; done in CL 540258).
* Callees must be standard static functions. Callees that are closures
(requiring closure context) are not supported.
For #61577.
Change-Id: I7d328859035249e176294cd0d9885b2d08c853f6
Reviewed-on: https://go-review.googlesource.com/c/go/+/539699
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
When selecting the hottest edge to use for PGO-based devirtualization,
edges are order by:
1. Edge weight
2. If weights are equal, prefer the edge with IR available in the
package.
3. Otherwise, simply sort lexicographically.
The existing logic for (2) is incomplete.
If the hottest edge so far is missing IR, but the new edge has IR, then
it works as expected and selects the new edge.
But if the hottest edge so far has IR and the new edge is missing IR, we
want to always keep the hottest edge so far, but this logic will fall
through and use lexicographical ordering instead.
Adjust the check to always make an explicit choice when IR availability
differs.
Change-Id: Ia7fcc286aa9a62ac209fd978cfce60463505f4cd
Reviewed-on: https://go-review.googlesource.com/c/go/+/539475
Auto-Submit: Michael Pratt <mpratt@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
CallExpr.X -> CallExpr.Fun
This consistent with go/ast and cmd/compile/internal/syntax.
OPRINTN -> OPRINTLN
This op represents the "println" builtin; might as well spell it the
same way.
Change-Id: Iead1b007776658c717879cf0997b3c48028428f4
Reviewed-on: https://go-review.googlesource.com/c/go/+/532795
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Keith Randall <khr@google.com>
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
When a PGO build fails or produces incorrect program, it is often
unclear what the problem is. Add pgo hash so we can bisect to
individual optimization decisions, which often helps debugging.
Related to #58153.
Change-Id: I651ffd9c53bad60f2f28c8ec2a90a3f532982712
Reviewed-on: https://go-review.googlesource.com/c/go/+/528400
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
When two callees have equal weight, we need to sort by another criteria
to ensure that we get stable output.
Note that this is only for the CallStat debug JSON output. The actual
callee selection already does this secondary sort in
findHotConcreteCallee.
Change-Id: I0de105623c5ccc793ca6f5799ea25e57bc286722
Reviewed-on: https://go-review.googlesource.com/c/go/+/527796
Auto-Submit: Michael Pratt <mpratt@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This CL removes a lot of the redundant methods for accessing struct
fields and signature parameters. In particular, users never have to
write ".Slice()" or ".FieldSlice()" anymore; the exported APIs just do
what you want.
Further internal refactorings to follow.
Change-Id: I45212f6772fe16aad39d0e68b82d71b0796e5639
Reviewed-on: https://go-review.googlesource.com/c/go/+/521295
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
This CL refactors common patterns for constructing field and method
selector expressions. Notably, XDotField and XDotMethod are now the
only two functions where a SelecterExpr with OXDOT is constructed.
Change-Id: I4c087225d8b295c4a6a92281ffcbcabafe2dc94d
Reviewed-on: https://go-review.googlesource.com/c/go/+/520979
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This CL updates several frontend passes to stop relying on
ir.CurFunc (at least directly).
Change-Id: I3c3529e81e27fb05d54a828f081f7c7efc31af67
Reviewed-on: https://go-review.googlesource.com/c/go/+/520606
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
Steps towards eliminating implicit dependencies on base.Pos and
ir.CurFunc. Mechanical CL produced with gofmt -r.
Change-Id: I070015513cb955cbe87f9a148d81db8c0d4b0dc5
Reviewed-on: https://go-review.googlesource.com/c/go/+/520605
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
Currently, we devirtualize an interface call if the profile
indicates a concrete callee is hot on the same line, and the
concrete receiver implements the interface. But it is possible
that (likely due to another call on the same line, or possibly a
stale profile) the concrete call is to a different method.
With the current AST construction we generate correct code, as we
extract the method name from the interface call and use that to
create the concrete call. But the devirtualization decision is
based on an unrelated call in the profile.
Check the method name when finding the hottest callee, so we won't
use unrelated calls to different methods.
Change-Id: I75c026997926f21bd6cc5266d3ffe99649a9b2d9
Reviewed-on: https://go-review.googlesource.com/c/go/+/500961
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Make it more consistent with the static devirtualization
diagnostic message. Keep the print of concrete callee's method
name, as it is clearer.
Change-Id: Ibe9b40253eaff2c0071353a2b388177213488822
Reviewed-on: https://go-review.googlesource.com/c/go/+/500960
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
The new PGO-driven indirect call specialization from CL 492436
in theory should allow for devirtualization on methods
in another package when those methods are directly referenced
in the current package.
However, inline.InlineImpossible was checking for a zero-length
fn.Body and would cause devirtualization to fail
with a debug log message like:
"should not PGO devirtualize (*Speaker1).Speak: no function body"
Previously, the logic in inline.InlineImpossible was only
called on local functions, but with PGO-based devirtualization,
it can now be called on imported functions, where inlinable
imported functions will have a zero-length fn.Body but a
non-nil fn.Inl.
We update inline.InlineImpossible to handle imported functions
by adding a call to typecheck.HaveInlineBody in the check
that was previously failing.
For the test, we need to have a hopefully temporary workaround
of adding explicit references to the callees in another package
for devirtualization to work. CL 497175 or similar should
enable removing this workaround.
Fixes#60561
Updates #59959
Change-Id: I48449b7d8b329d84151bd3b506b8093c262eb2a3
GitHub-Last-Rev: 2d53c55fd8
GitHub-Pull-Request: golang/go#60565
Reviewed-on: https://go-review.googlesource.com/c/go/+/500155
Run-TryBot: thepudds <thepudds1460@gmail.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This CL is originally based on CL 484838 from rajbarik@uber.com.
Add a new PGO-based devirtualize pass. This pass conditionally
devirtualizes interface calls for the hottest callee. That is, it
performs a transformation like:
type Iface interface {
Foo()
}
type Concrete struct{}
func (Concrete) Foo() {}
func foo(i Iface) {
i.Foo()
}
to:
func foo(i Iface) {
if c, ok := i.(Concrete); ok {
c.Foo()
} else {
i.Foo()
}
}
The primary benefit of this transformation is enabling inlining of the
direct calls.
Today this change has no impact on the escape behavior, as the fallback
interface always forces an escape. But improving escape analysis to take
advantage of this is an area of potential work.
This CL is the bare minimum of a devirtualization implementation. There
are still numerous limitations:
* Callees not directly referenced in the current package can be missed
(even if they are in the transitive dependences).
* Callees not in the transitive dependencies of the current package are
missed.
* Only interface method calls are supported, not other indirect function
calls.
* Multiple calls to compatible interfaces on the same line cannot be
distinguished and will use the same callee target.
* Callees that only partially implement an interface (they are embedded
in another type that completes the interface) cannot be devirtualized.
* Others, mentioned in TODOs.
Fixes#59959
Change-Id: I8bedb516139695ee4069650b099d05957b7ce5ee
Reviewed-on: https://go-review.googlesource.com/c/go/+/492436
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
