mirror of
https://github.com/golang/go
synced 2024-10-04 16:21:22 -06:00
408238e20b
Changing from 4 kB to 8 kB brings significant improvement
on a variety of the Go 1 benchmarks, on both amd64
and 386 systems.
Significant runtime reductions:
amd64 386
GoParse -14% -1%
GobDecode -12% -20%
GobEncode -64% -1%
JSONDecode -9% -4%
JSONEncode -15% -5%
Template -17% -14%
In the longer term, khr's new stacks will avoid needing to
make this decision at all, but for Go 1.2 this is a reasonable
stopgap that makes performance significantly better.
Demand paging should mean that if the second 4 kB is not
used, it will not be brought into memory, so the change
should not adversely affect resident set size.
The same argument could justify bumping as high as 64 kB
on 64-bit machines, but there are diminishing returns
after 8 kB, and using 8 kB limits the possible unintended
memory overheads we are not aware of.
Benchmark graphs at
http://swtch.com/~rsc/gostackamd64.html
http://swtch.com/~rsc/gostack386.html
Full data at
http://swtch.com/~rsc/gostack.zip
R=golang-dev, khr, dave, bradfitz, dvyukov
CC=golang-dev
https://golang.org/cl/14317043
113 lines
3.9 KiB
C
113 lines
3.9 KiB
C
// Copyright 2011 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.
|
|
|
|
/*
|
|
Stack layout parameters.
|
|
Included both by runtime (compiled via 6c) and linkers (compiled via gcc).
|
|
|
|
The per-goroutine g->stackguard is set to point StackGuard bytes
|
|
above the bottom of the stack. Each function compares its stack
|
|
pointer against g->stackguard to check for overflow. To cut one
|
|
instruction from the check sequence for functions with tiny frames,
|
|
the stack is allowed to protrude StackSmall bytes below the stack
|
|
guard. Functions with large frames don't bother with the check and
|
|
always call morestack. The sequences are (for amd64, others are
|
|
similar):
|
|
|
|
guard = g->stackguard
|
|
frame = function's stack frame size
|
|
argsize = size of function arguments (call + return)
|
|
|
|
stack frame size <= StackSmall:
|
|
CMPQ guard, SP
|
|
JHI 3(PC)
|
|
MOVQ m->morearg, $(argsize << 32)
|
|
CALL morestack(SB)
|
|
|
|
stack frame size > StackSmall but < StackBig
|
|
LEAQ (frame-StackSmall)(SP), R0
|
|
CMPQ guard, R0
|
|
JHI 3(PC)
|
|
MOVQ m->morearg, $(argsize << 32)
|
|
CALL morestack(SB)
|
|
|
|
stack frame size >= StackBig:
|
|
MOVQ m->morearg, $((argsize << 32) | frame)
|
|
CALL morestack(SB)
|
|
|
|
The bottom StackGuard - StackSmall bytes are important: there has
|
|
to be enough room to execute functions that refuse to check for
|
|
stack overflow, either because they need to be adjacent to the
|
|
actual caller's frame (deferproc) or because they handle the imminent
|
|
stack overflow (morestack).
|
|
|
|
For example, deferproc might call malloc, which does one of the
|
|
above checks (without allocating a full frame), which might trigger
|
|
a call to morestack. This sequence needs to fit in the bottom
|
|
section of the stack. On amd64, morestack's frame is 40 bytes, and
|
|
deferproc's frame is 56 bytes. That fits well within the
|
|
StackGuard - StackSmall = 128 bytes at the bottom.
|
|
The linkers explore all possible call traces involving non-splitting
|
|
functions to make sure that this limit cannot be violated.
|
|
*/
|
|
|
|
enum {
|
|
// StackSystem is a number of additional bytes to add
|
|
// to each stack below the usual guard area for OS-specific
|
|
// purposes like signal handling. Used on Windows and on
|
|
// Plan 9 because they do not use a separate stack.
|
|
#ifdef GOOS_windows
|
|
StackSystem = 512 * sizeof(uintptr),
|
|
#else
|
|
#ifdef GOOS_plan9
|
|
// The size of the note handler frame varies among architectures,
|
|
// but 512 bytes should be enough for every implementation.
|
|
StackSystem = 512,
|
|
#else
|
|
StackSystem = 0,
|
|
#endif // Plan 9
|
|
#endif // Windows
|
|
|
|
// The amount of extra stack to allocate beyond the size
|
|
// needed for the single frame that triggered the split.
|
|
StackExtra = 2048,
|
|
|
|
// The minimum stack segment size to allocate.
|
|
// If the amount needed for the splitting frame + StackExtra
|
|
// is less than this number, the stack will have this size instead.
|
|
StackMin = 8192,
|
|
FixedStack = StackMin + StackSystem,
|
|
|
|
// Functions that need frames bigger than this use an extra
|
|
// instruction to do the stack split check, to avoid overflow
|
|
// in case SP - framesize wraps below zero.
|
|
// This value can be no bigger than the size of the unmapped
|
|
// space at zero.
|
|
StackBig = 4096,
|
|
|
|
// The stack guard is a pointer this many bytes above the
|
|
// bottom of the stack.
|
|
StackGuard = 256 + StackSystem,
|
|
|
|
// After a stack split check the SP is allowed to be this
|
|
// many bytes below the stack guard. This saves an instruction
|
|
// in the checking sequence for tiny frames.
|
|
StackSmall = 128,
|
|
|
|
// The maximum number of bytes that a chain of NOSPLIT
|
|
// functions can use.
|
|
StackLimit = StackGuard - StackSystem - StackSmall,
|
|
|
|
// The assumed size of the top-of-stack data block.
|
|
// The actual size can be smaller than this but cannot be larger.
|
|
// Checked in proc.c's runtime.malg.
|
|
StackTop = 96,
|
|
};
|
|
|
|
// Goroutine preemption request.
|
|
// Stored into g->stackguard0 to cause split stack check failure.
|
|
// Must be greater than any real sp.
|
|
// 0xfffffade in hex.
|
|
#define StackPreempt ((uint64)-1314)
|