Skip to content

G
golang
Switch branch/tag
  1. 17 Apr, 2017 17 commits
  3. 16 Apr, 2017 3 commits
    • Takuto Ikuta's avatar
      regexp: Use []bool instead of map[uint32]bool in makeOnePass · a306a851
      Takuto Ikuta authored 
      Variable m is accessed like variable onePassRunes.
So, we can use array instead of map for variable m.

Onepass compile performance is improved 10~20%.

name                                      old time/op    new time/op    delta
CompileOnepass/^(?:(?:(?:.(?:$))?))...-4    4.60µs ± 1%    4.10µs ± 1%  -10.98%  (p=0.008 n=5+5)
CompileOnepass/^abcd$-4                     3.76µs ± 2%    3.33µs ± 1%  -11.43%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a{0,})*?)$-4          4.47µs ± 1%    4.14µs ± 1%   -7.49%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a+)*)$-4              4.07µs ± 1%    3.81µs ± 2%   -6.60%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a|(?:aa)))$-4         5.21µs ± 1%    4.80µs ± 2%   -7.90%  (p=0.008 n=5+5)
CompileOnepass/^(?:[^\s\S])$-4              3.26µs ± 1%    3.06µs ± 1%   -6.15%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:(?:a*)+))$-4          4.33µs ± 0%    4.04µs ± 3%   -6.78%  (p=0.008 n=5+5)
CompileOnepass/^[a-c]+$-4                   3.33µs ± 1%    3.06µs ± 4%   -8.24%  (p=0.008 n=5+5)
CompileOnepass/^[a-c]*$-4                   3.69µs ± 1%    3.39µs ± 4%   -8.26%  (p=0.008 n=5+5)
CompileOnepass/^(?:a*)$-4                   3.78µs ± 2%    3.36µs ± 0%  -11.09%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:aa)|a)$-4             5.11µs ± 1%    4.60µs ± 1%   -9.85%  (p=0.008 n=5+5)
CompileOnepass/^...$-4                      3.51µs ± 1%    3.25µs ± 1%   -7.37%  (p=0.008 n=5+5)
CompileOnepass/^(?:a|(?:aa))$-4             5.05µs ± 1%    4.59µs ± 1%   -9.13%  (p=0.008 n=5+5)
CompileOnepass/^a((b))c$-4                  5.24µs ± 1%    4.66µs ± 1%  -11.16%  (p=0.008 n=5+5)
CompileOnepass/^a.[l-nA-Cg-j]?e$-4          5.94µs ± 7%    5.02µs ± 1%  -15.62%  (p=0.008 n=5+5)
CompileOnepass/^a((b))$-4                   5.34µs ± 7%    4.21µs ± 2%  -21.05%  (p=0.008 n=5+5)
CompileOnepass/^a(?:(b)|(c))c$-4            7.25µs ± 7%    5.86µs ± 1%  -19.17%  (p=0.008 n=5+5)
CompileOnepass/^a(?:b|c)$-4                 4.13µs ± 9%    3.35µs ± 1%  -18.99%  (p=0.008 n=5+5)
CompileOnepass/^a(?:b?|c)$-4                6.48µs ± 4%    5.05µs ± 1%  -22.16%  (p=0.008 n=5+5)
CompileOnepass/^a(?:b?|c+)$-4               7.05µs ± 7%    5.48µs ± 1%  -22.21%  (p=0.008 n=5+5)
CompileOnepass/^a(?:bc)+$-4                 4.71µs ± 4%    3.85µs ± 0%  -18.14%  (p=0.008 n=5+5)
CompileOnepass/^a(?:[bcd])+$-4              4.10µs ± 2%    3.69µs ± 1%   -9.98%  (p=0.008 n=5+5)
CompileOnepass/^a((?:[bcd])+)$-4            5.20µs ± 1%    4.59µs ± 0%  -11.72%  (p=0.008 n=5+5)
CompileOnepass/^a(:?b|c)*d$-4               9.29µs ± 2%    8.23µs ± 1%  -11.43%  (p=0.008 n=5+5)
CompileOnepass/^.bc(d|e)*$-4                6.33µs ± 3%    5.51µs ± 5%  -12.95%  (p=0.008 n=5+5)
CompileOnepass/^loooooooooooooooooo...-4     162µs ± 0%     138µs ± 2%  -14.70%  (p=0.008 n=5+5)

name                                      old alloc/op   new alloc/op   delta
CompileOnepass/^(?:(?:(?:.(?:$))?))...-4    3.38kB ± 0%    3.27kB ± 0%   -3.08%  (p=0.008 n=5+5)
CompileOnepass/^abcd$-4                     2.74kB ± 0%    2.59kB ± 0%   -5.54%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a{0,})*?)$-4          3.34kB ± 0%    3.23kB ± 0%   -3.12%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a+)*)$-4              2.95kB ± 0%    2.85kB ± 0%   -3.52%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a|(?:aa)))$-4         3.74kB ± 0%    3.58kB ± 0%   -4.07%  (p=0.008 n=5+5)
CompileOnepass/^(?:[^\s\S])$-4              2.45kB ± 0%    2.35kB ± 0%   -4.20%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:(?:a*)+))$-4          3.13kB ± 0%    3.02kB ± 0%   -3.32%  (p=0.008 n=5+5)
CompileOnepass/^[a-c]+$-4                   2.48kB ± 0%    2.38kB ± 0%   -4.19%  (p=0.008 n=5+5)
CompileOnepass/^[a-c]*$-4                   2.52kB ± 0%    2.42kB ± 0%   -4.13%  (p=0.008 n=5+5)
CompileOnepass/^(?:a*)$-4                   2.63kB ± 0%    2.53kB ± 0%   -3.95%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:aa)|a)$-4             3.62kB ± 0%    3.47kB ± 0%   -4.19%  (p=0.008 n=5+5)
CompileOnepass/^...$-4                      2.87kB ± 0%    2.78kB ± 0%   -3.34%  (p=0.008 n=5+5)
CompileOnepass/^(?:a|(?:aa))$-4             3.62kB ± 0%    3.47kB ± 0%   -4.19%  (p=0.008 n=5+5)
CompileOnepass/^a((b))c$-4                  4.38kB ± 0%    4.23kB ± 0%   -3.33%  (p=0.008 n=5+5)
CompileOnepass/^a.[l-nA-Cg-j]?e$-4          4.30kB ± 0%    4.15kB ± 0%   -3.35%  (p=0.008 n=5+5)
CompileOnepass/^a((b))$-4                   4.05kB ± 0%    3.90kB ± 0%   -3.56%  (p=0.008 n=5+5)
CompileOnepass/^a(?:(b)|(c))c$-4            5.30kB ± 0%    5.05kB ± 0%   -4.83%  (p=0.008 n=5+5)
CompileOnepass/^a(?:b|c)$-4                 2.87kB ± 0%    2.77kB ± 0%   -3.62%  (p=0.008 n=5+5)
CompileOnepass/^a(?:b?|c)$-4                4.35kB ± 0%    4.21kB ± 0%   -3.31%  (p=0.008 n=5+5)
CompileOnepass/^a(?:b?|c+)$-4               4.58kB ± 0%    4.44kB ± 0%   -3.14%  (p=0.008 n=5+5)
CompileOnepass/^a(?:bc)+$-4                 3.15kB ± 0%    3.00kB ± 0%   -4.82%  (p=0.008 n=5+5)
CompileOnepass/^a(?:[bcd])+$-4              2.94kB ± 0%    2.84kB ± 0%   -3.53%  (p=0.008 n=5+5)
CompileOnepass/^a((?:[bcd])+)$-4            4.08kB ± 0%    3.94kB ± 0%   -3.53%  (p=0.008 n=5+5)
CompileOnepass/^a(:?b|c)*d$-4               6.10kB ± 0%    5.85kB ± 0%   -4.19%  (p=0.008 n=5+5)
CompileOnepass/^.bc(d|e)*$-4                4.46kB ± 0%    4.31kB ± 0%   -3.28%  (p=0.008 n=5+5)
CompileOnepass/^loooooooooooooooooo...-4     135kB ± 0%     127kB ± 0%   -5.88%  (p=0.008 n=5+5)

name                                      old allocs/op  new allocs/op  delta
CompileOnepass/^(?:(?:(?:.(?:$))?))...-4      47.0 ± 0%      46.0 ± 0%   -2.13%  (p=0.008 n=5+5)
CompileOnepass/^abcd$-4                       41.0 ± 0%      40.0 ± 0%   -2.44%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a{0,})*?)$-4            49.0 ± 0%      48.0 ± 0%   -2.04%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a+)*)$-4                44.0 ± 0%      43.0 ± 0%   -2.27%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a|(?:aa)))$-4           54.0 ± 0%      53.0 ± 0%   -1.85%  (p=0.008 n=5+5)
CompileOnepass/^(?:[^\s\S])$-4                33.0 ± 0%      32.0 ± 0%   -3.03%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:(?:a*)+))$-4            46.0 ± 0%      45.0 ± 0%   -2.17%  (p=0.008 n=5+5)
CompileOnepass/^[a-c]+$-4                     36.0 ± 0%      35.0 ± 0%   -2.78%  (p=0.008 n=5+5)
CompileOnepass/^[a-c]*$-4                     41.0 ± 0%      40.0 ± 0%   -2.44%  (p=0.008 n=5+5)
CompileOnepass/^(?:a*)$-4                     42.0 ± 0%      41.0 ± 0%   -2.38%  (p=0.008 n=5+5)
CompileOnepass/^(?:(?:aa)|a)$-4               53.0 ± 0%      52.0 ± 0%   -1.89%  (p=0.008 n=5+5)
CompileOnepass/^...$-4                        39.0 ± 0%      38.0 ± 0%   -2.56%  (p=0.008 n=5+5)
CompileOnepass/^(?:a|(?:aa))$-4               53.0 ± 0%      52.0 ± 0%   -1.89%  (p=0.008 n=5+5)
CompileOnepass/^a((b))c$-4                    53.0 ± 0%      52.0 ± 0%   -1.89%  (p=0.008 n=5+5)
CompileOnepass/^a.[l-nA-Cg-j]?e$-4            56.0 ± 0%      55.0 ± 0%   -1.79%  (p=0.008 n=5+5)
CompileOnepass/^a((b))$-4                     47.0 ± 0%      46.0 ± 0%   -2.13%  (p=0.008 n=5+5)
CompileOnepass/^a(?:(b)|(c))c$-4              65.0 ± 0%      64.0 ± 0%   -1.54%  (p=0.008 n=5+5)
CompileOnepass/^a(?:b|c)$-4                   40.0 ± 0%      39.0 ± 0%   -2.50%  (p=0.008 n=5+5)
CompileOnepass/^a(?:b?|c)$-4                  57.0 ± 0%      56.0 ± 0%   -1.75%  (p=0.008 n=5+5)
CompileOnepass/^a(?:b?|c+)$-4                 63.0 ± 0%      62.0 ± 0%   -1.59%  (p=0.008 n=5+5)
CompileOnepass/^a(?:bc)+$-4                   46.0 ± 0%      45.0 ± 0%   -2.17%  (p=0.008 n=5+5)
CompileOnepass/^a(?:[bcd])+$-4                43.0 ± 0%      42.0 ± 0%   -2.33%  (p=0.008 n=5+5)
CompileOnepass/^a((?:[bcd])+)$-4              49.0 ± 0%      48.0 ± 0%   -2.04%  (p=0.008 n=5+5)
CompileOnepass/^a(:?b|c)*d$-4                  101 ± 0%       100 ± 0%   -0.99%  (p=0.008 n=5+5)
CompileOnepass/^.bc(d|e)*$-4                  60.0 ± 0%      59.0 ± 0%   -1.67%  (p=0.008 n=5+5)
CompileOnepass/^loooooooooooooooooo...-4     1.09k ± 0%     1.08k ± 0%   -0.74%  (p=0.008 n=5+5)

Fixes #19984

Change-Id: I2268b28d068926a057c62751528de15b6de61a7b
Reviewed-on: https://go-review.googlesource.com/40890Reviewed-by: 's avatarRalph Corderoy <ralph@inputplus.co.uk>
Reviewed-by: 's avatarBrad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
      a306a851
    • Andreas Auernhammer's avatar
      vendor: update golang_org/x/net/route package · d71863ad
      Andreas Auernhammer authored 
      Update the route package to git rev 6b27048a.

Introduce the following changes:
 - 6b27048 route: drop support for go1.5
 - b7fd658 route: fix typo
 - 41bba8d route: add support for the manipulation of routing informaion

Updates #19967

Change-Id: Id2bb93df97a45254a2df2b048db0143e3e52bbdf
Reviewed-on: https://go-review.googlesource.com/40830
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarBrad Fitzpatrick <bradfitz@golang.org>
      d71863ad
    • Josh Bleecher Snyder's avatar
      cmd/compile: remove flag_largemodel · 0e4824aa
      Josh Bleecher Snyder authored 
      It was added in 2013 in CL 7064048.
All uses of it in the compiler disappeared with
(or possibly before) the SSA backend.
Several releases have gone by without it,
from which I conclude that it is now not needed.

Change-Id: I2095f4ac05d4d7ab998168993a7fd5d954aeee88
Reviewed-on: https://go-review.googlesource.com/40856
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarBrad Fitzpatrick <bradfitz@golang.org>
      0e4824aa
  5. 15 Apr, 2017 4 commits
  7. 14 Apr, 2017 8 commits
    • Caleb Spare's avatar
      testing: add TB.Helper to better support test helpers · bc293137
      Caleb Spare authored 
      This CL implements the proposal at
https://github.com/golang/proposal/blob/master/design/4899-testing-helper.md.

It's based on Josh's CL 79890043 from a few years ago:
https://codereview.appspot.com/79890043 but makes several changes,
most notably by using the new CallersFrames API so that it works with
mid-stack inlining.

Another detail came up while I was working on this: I didn't want the
user to be able to call t.Helper from inside their TestXxx function
directly (which would mean we'd print a file:line from inside the
testing package itself), so I explicitly prevented this from working.

Fixes #4899.

Change-Id: I37493edcfb63307f950442bbaf993d1589515310
Reviewed-on: https://go-review.googlesource.com/38796
Run-TryBot: Caleb Spare <cespare@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarIan Lance Taylor <iant@golang.org>
      bc293137
    • Alberto Donizetti's avatar
      cmd/compile: add test for Issue 12536 · 6266b0f0
      Alberto Donizetti authored 
      The fixedbugs/issue12536.go file was erroneously deleted just before
committing the patch that fixed the issue (CL 14400).

That's an easy test and there's a small reproducer in the issue, add
it back.

Updates #12536

Change-Id: Ib7b0cd245588299e9a5469e1d75805fd0261ce1a
Reviewed-on: https://go-review.googlesource.com/40712Reviewed-by: 's avatarBrad Fitzpatrick <bradfitz@golang.org>
      6266b0f0
    • Andreas Auernhammer's avatar
      vendor: update golang_org/x/crypto packages · 3b6ecdf9
      Andreas Auernhammer authored 
      Update the poly1305 and curve25519 packages to the current state of /x/crypto.

Updates #19967

Change-Id: Ib71534f78040f31bfd5debb06f3c4a54a77955b3
Reviewed-on: https://go-review.googlesource.com/40711Reviewed-by: 's avatarIan Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
      3b6ecdf9
    • David Lazar's avatar
      runtime/trace: iterate over frames instead of PCs · 3249cb0a
      David Lazar authored 
      Now the runtime/trace tests pass with -l=4.

This also gets rid of the frames cache for multiple reasons:

1) The frames cache was used to avoid repeated calls to funcname and
funcline. Now these calls happen inside the CallersFrames iterator.

2) Maintaining a frames cache is harder: map[uintptr]traceFrame
doesn't work since each PC can map to multiple traceFrames.

3) It's not clear that the cache is important.

Change-Id: I2914ac0b3ba08e39b60149d99a98f9f532b35bbb
Reviewed-on: https://go-review.googlesource.com/40591
Run-TryBot: David Lazar <lazard@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarAustin Clements <austin@google.com>
      3249cb0a
    • David Lazar's avatar
      runtime/trace: better output when test fails · a7276742
      David Lazar authored 
      Change-Id: I108d15eb4cd25904bb76de4ed7548c039c69d1a3
Reviewed-on: https://go-review.googlesource.com/40590
Run-TryBot: David Lazar <lazard@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarAustin Clements <austin@google.com>
      a7276742
    • Mikio Hara's avatar
      net: validate network in Dial{,IP} and Listen{Packet,IP} for IP networks · 7337181d
      Mikio Hara authored 
      The argument of the first parameter for connection setup functions on
IP networks must contain a protocol name or number. This change adds
validation for arguments of IP networks to connection setup functions.

Fixes #18185.

Change-Id: I6aaedd7806e3ed1043d4b1c834024f350b99361d
Reviewed-on: https://go-review.googlesource.com/40512
Run-TryBot: Mikio Hara <mikioh.mikioh@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarIan Lance Taylor <iant@golang.org>
      7337181d
    • Mikio Hara's avatar
      net: don't enclose non-literal IPv6 addresses in square brackets · dc74f51c
      Mikio Hara authored 
      The net package uses various textual representations for network
identifiers and locators on the Internet protocol suite as API.
In fact, the representations are the composition of subset of multple
RFCs: RFC 3986, RFC 4007, RFC 4632, RFC 4291 and RFC 5952.

RFC 4007 describes guidelines for the use of textual representation of
IPv6 addressing/routing scope zone and doesn't prohibit the format for
implementation dependent purposes, as in, specifying a literal IPv6
address and its connected region of routing topology as application
user interface. However, a non-literal IPv6 address, for example, a
host name, with a zone enclosed in square brackets confuses us because
a zone is basically for non-global IPv6 addresses and a pair of square
brackets is used as a set of delimiters between a literal IPv6 address
and a service name or transport port number.

To mitigate such confusion, this change makes JoinHostPort not enclose
non-literal IPv6 addresses in square brackets and SplitHostPort accept
the form "host%zone:port" to recommend that anything enclosed in
square brackets should be a literal IPv6 address.

Before this change:
	JoinHostPort("name%zone", "80") = "[name%zone]:80"
	JoinHostPort("[::1%zone]", "80") = "[::1%zone]:80"
	SplitHostPort("name%zone:80") = "", "", "address name%zone:80: missing brackets in address"
	SplitHostPort("[name%zone]:80") = "name%zone", "80", nil
	SplitHostPort("[::1%zone]:80") = "::1%zone", "80", nil

After this change:
	JoinHostPort("name%zone", "80") = "name%zone:80"
	JoinHostPort("[::1%zone]", "80") = "[::1%zone]:80"
	SplitHostPort("name%zone:80") = "name%zone", "80", nil
	SplitHostPort("[name%zone]:80") = "name%zone", "80", nil // for backwards compatibility
	SplitHostPort("[::1%zone]:80") = "::1%zone", "80", nil

Also updates docs and test cases on SplitHostPort and JoinHostPort for
clarification.

Fixes #18059.
Fixes #18060.

Change-Id: I5c3ccce4fa0fbdd58f698fc280635ea4a14d2a37
Reviewed-on: https://go-review.googlesource.com/40510
Run-TryBot: Mikio Hara <mikioh.mikioh@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarIan Lance Taylor <iant@golang.org>
      dc74f51c
    • Wei Xiao's avatar
      cmd/dist: fix incorrect platform string shared by all tests · 1cf6d474
      Wei Xiao authored 
      all tests currently share the same platform string and fail to
vet expected platforms

Fixes #19958

Change-Id: I2801e1e84958e31975769581e27ea5ca6a0edf5b
Reviewed-on: https://go-review.googlesource.com/40511
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarBrad Fitzpatrick <bradfitz@golang.org>
      1cf6d474
  9. 13 Apr, 2017 8 commits
    • Jess Frazelle's avatar
      syscall: fix TestGroupCleanup{UserNamespace} on Alpine · 75176947
      Jess Frazelle authored 
      This updates TestGroupCleanup and TestGroupCleanupUserNamespace to pass in the
Alpine builder.

Updates #19938

Change-Id: Iacbfd73782eccd57f872f9e85726c6024529c277
Reviewed-on: https://go-review.googlesource.com/40692Reviewed-by: 's avatarDaniel Martí <mvdan@mvdan.cc>
Reviewed-by: 's avatarBrad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Gobot Gobot <gobot@golang.org>
      75176947
    • Michael Munday's avatar
      cmd/compile: fix s390x unsigned comparison constant merging rules · f75b5e99
      Michael Munday authored 
      On s390x unsigned integer comparisons with immediates require the immediate
to be an unsigned 32-bit integer. The rule was checking that the immediate
was a signed 32-bit integer.

This CL also adds a test for comparisons that could be turned into compare
with immediate or equivalent instructions (depending on architecture and
optimizations applied).

Fixes #19940.

Change-Id: Ifd6aa989fd3d50e282f7d30fec9db462c28422b1
Reviewed-on: https://go-review.googlesource.com/40433
Run-TryBot: Michael Munday <munday@ca.ibm.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarDavid Chase <drchase@google.com>
      f75b5e99
    • Brad Fitzpatrick's avatar
      test: disable flaky test/fixedbugs/issue10958.go · dc09a7af
      Brad Fitzpatrick authored 
      Updates #18589

Change-Id: I2c3bbc8257c68295051bd2e63e1e11794d0609c3
Reviewed-on: https://go-review.googlesource.com/40651
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: 's avatarJosh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: 's avatarDavid Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
      dc09a7af
    • Austin Clements's avatar
      runtime: free workbufs during sweeping · 051809e3
      Austin Clements authored 
      This extends the sweeper to free workbufs back to the heap between GC
cycles, allowing this memory to be reused for GC'd allocations or
eventually returned to the OS.

This helps for applications that have high peak heap usage relative to
their regular heap usage (for example, a high-memory initialization
phase). Workbuf memory is roughly proportional to heap size and since
we currently never free workbufs, it's proportional to *peak* heap
size. By freeing workbufs, we can release and reuse this memory for
other purposes when the heap shrinks.

This is somewhat complicated because this costs ~1–2 µs per workbuf
span, so for large heaps it's too expensive to just do synchronously
after mark termination between starting the world and dropping the
worldsema. Hence, we do it asynchronously in the sweeper. This adds a
list of "free" workbuf spans that can be returned to the heap. GC
moves all workbuf spans to this list after mark termination and the
background sweeper drains this list back to the heap. If the sweeper
doesn't finish, that's fine, since getempty can directly reuse any
remaining spans to allocate more workbufs.

Performance impact is negligible. On the x/benchmarks, this reduces
GC-bytes-from-system by 6–11%.

Fixes #19325.

Change-Id: Icb92da2196f0c39ee984faf92d52f29fd9ded7a8
Reviewed-on: https://go-review.googlesource.com/38582
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarRick Hudson <rlh@golang.org>
      051809e3
    • Austin Clements's avatar
      runtime: allocate GC workbufs from manually-managed spans · 9cc883a4
      Austin Clements authored 
      Currently the runtime allocates workbufs from persistent memory, which
means they can never be freed.

Switch to allocating them from manually-managed heap spans. This
doesn't free them yet, but it puts us in a position to do so.

For #19325.

Change-Id: I94b2512a2f2bbbb456cd9347761b9412e80d2da9
Reviewed-on: https://go-review.googlesource.com/38581
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarRick Hudson <rlh@golang.org>
      9cc883a4
    • Austin Clements's avatar
      runtime: eliminate write barriers from alloc/mark bitmaps · 42c12147
      Austin Clements authored 
      This introduces a new type, *gcBits, to use for alloc/mark bitmap
allocations instead of *uint8. This type is marked go:notinheap, so
uses of it correctly eliminate write barriers. Since we now have a
type, this also extracts some common operations to methods both for
convenience and to avoid (*uint8) casts at most use sites.

For #19325.

Change-Id: Id51f734fb2e96b8b7715caa348c8dcd4aef0696a
Reviewed-on: https://go-review.googlesource.com/38580
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarRick Hudson <rlh@golang.org>
      42c12147
    • Austin Clements's avatar
      runtime: rename gcBits -> gcBitsArena · 9d1b2f88
      Austin Clements authored 
      This clarifies that the gcBits type is actually an arena of gcBits and
will let us introduce a new gcBits type representing a single
mark/alloc bitmap allocated from the arena.

For #19325.

Change-Id: Idedf76d202d9174a17c61bcca9d5539e042e2445
Reviewed-on: https://go-review.googlesource.com/38579
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: 's avatarRick Hudson <rlh@golang.org>
      9d1b2f88
    • Austin Clements's avatar
      runtime: don't count manually-managed spans from heap_{inuse,sys} · dc0f0ab7
      Austin Clements authored 
      Currently, manually-managed spans are included in memstats.heap_inuse
and memstats.heap_sys, but when we export these stats to the user, we
subtract out how much has been allocated for stack spans from both.
This works for now because stacks are the only manually-managed spans
we have.

However, we're about to use manually-managed spans for more things
that don't necessarily have obvious stats we can use to adjust the
user-presented numbers. Prepare for this by changing the accounting so
manually-managed spans don't count toward heap_inuse or heap_sys. This
makes these fields align with the fields presented to the user and
means we don't have to track more statistics just so we can adjust
these statistics.

For #19325.

Change-Id: I5cb35527fd65587ff23339276ba2c3969e2ad98f
Reviewed-on: https://go-review.googlesource.com/38577
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: 's avatarRick Hudson <rlh@golang.org>
      dc0f0ab7