Commits · b72636cbde8e241ae216748ad915a9f9cf620988 · Kirill Smelkov / go

28 Mar, 2017 11 commits

cmd/compile/internal/gc: cleanup selecttype · b72636cb

Matthew Dempsky authored Mar 28, 2017

Use namedfield consistently.

Passes toolstash-check.

Change-Id: Ic5a3acb4bfaa1f60dd2eac94612160509e8d7f94
Reviewed-on: https://go-review.googlesource.com/38741
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

b72636cb

cmd/compile/internal/gc: use anonfield and namedfield · 3c0072e0

Matthew Dempsky authored Mar 28, 2017

Automated refactoring using gofmt.

Passes toolstash-check.

Change-Id: I8624e1c231dc736e1bb4cc800acaf629a0af91d7
Reviewed-on: https://go-review.googlesource.com/38740
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

3c0072e0

cmd/compile/internal/gc: npos(p, nod(o, l, r)) -> nodl(p, o, l, r) · 3f4f119b

Matthew Dempsky authored Mar 28, 2017

Prepared with gofmt -r.

Passes toolstash-check.

Change-Id: I8a4f7a8c365dfe464dfc5868a18c67d56af37749
Reviewed-on: https://go-review.googlesource.com/38739
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

3f4f119b

test: add test for gccgo compiler crash · 835b17c8

Ian Lance Taylor authored Mar 28, 2017

Gccgo crashed compiling a function that returned multiple zero-sized values.

Change-Id: I499112cc310e4a4f649962f4d2bc9fee95dee1b6
Reviewed-on: https://go-review.googlesource.com/38772Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

835b17c8

test/fixedbugs: add a test for 19201 · 4b50c813

Ilya Tocar authored Feb 21, 2017

This was cherry-picked to 1.8 as CL 38587, but on master issue was fixed
by CL 37661. Add still relevant part (test) and close issue, since test passes.

Fixes #19201

Change-Id: I6415792e2c465dc6d9bd6583ba1e54b107bcf5cc
Reviewed-on: https://go-review.googlesource.com/37376
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Keith Randall <khr@golang.org>

4b50c813

cmd/compile: fix two instances of { lineno = ...; yyerror } · 95bfd927

Josh Bleecher Snyder authored Mar 28, 2017

Updates #19683

Change-Id: Ic00d5a9807200791cf37553f4f802dbf27beea19
Reviewed-on: https://go-review.googlesource.com/38770
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

95bfd927

cmd/compile: number autotmps per-func, not per-package · b87fcc6e

Josh Bleecher Snyder authored Mar 28, 2017

Prior to this CL, autotmps were global to a package.
They also shared numbering with static variables.
Switch autotmp numbering to be per-function instead,
and do implicit numbering based on len(Func.Dcl).
This eliminates a dependency on a global variable
from the backend without adding to the Func struct.
While we're here, move statuniqgen closer to its
sole remaining user.

This actually improves compiler performance,
because the autotmp_* names can now be
reused across functions.

name       old alloc/op    new alloc/op    delta
Template      40.6MB ± 0%     40.1MB ± 0%  -1.38%  (p=0.000 n=10+10)
Unicode       29.9MB ± 0%     29.9MB ± 0%    ~     (p=0.912 n=10+10)
GoTypes        116MB ± 0%      114MB ± 0%  -1.53%  (p=0.000 n=10+10)
SSA            865MB ± 0%      856MB ± 0%  -1.04%  (p=0.000 n=10+10)
Flate         25.8MB ± 0%     25.4MB ± 0%  -1.36%  (p=0.000 n=10+10)
GoParser      32.2MB ± 0%     32.0MB ± 0%  -0.72%  (p=0.000 n=10+10)
Reflect       80.3MB ± 0%     79.0MB ± 0%  -1.65%  (p=0.000 n=9+10)
Tar           27.0MB ± 0%     26.7MB ± 0%  -0.86%  (p=0.000 n=10+9)
XML           42.8MB ± 0%     42.4MB ± 0%  -0.95%  (p=0.000 n=10+10)

name       old allocs/op   new allocs/op   delta
Template        398k ± 1%       396k ± 1%  -0.59%  (p=0.002 n=10+10)
Unicode         321k ± 1%       321k ± 0%    ~     (p=0.912 n=10+10)
GoTypes        1.17M ± 0%      1.16M ± 0%  -0.77%  (p=0.000 n=10+10)
SSA            7.65M ± 0%      7.62M ± 0%  -0.40%  (p=0.000 n=10+10)
Flate           240k ± 1%       238k ± 1%  -0.56%  (p=0.001 n=10+10)
GoParser        323k ± 1%       320k ± 1%  -0.65%  (p=0.002 n=10+10)
Reflect        1.01M ± 0%      1.00M ± 0%  -0.37%  (p=0.001 n=9+10)
Tar             256k ± 1%       255k ± 0%    ~     (p=0.101 n=10+8)
XML             400k ± 1%       398k ± 1%    ~     (p=0.063 n=10+10)


Change-Id: I3c23ca98129137d373106990b1a3e1507bbe0cc3
Reviewed-on: https://go-review.googlesource.com/38729
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>

b87fcc6e

cmd/compile: strengthen SetFields/Width safety guarantee · 8ee2d5bc

Josh Bleecher Snyder authored Mar 23, 2017

It is currently possible in the compiler to create a struct type,
calculate the widths of types that depend on it,
and then alter the struct type.

transformclosure has local protection against this.
Protect against it at a deeper level.

This is preparation to call dowidth automatically,
rather than explicitly.

This is a re-roll of CL 38469.

Change-Id: Ic5b4baa250618504611fc57cbf51ab01d1eddf80
Reviewed-on: https://go-review.googlesource.com/38534
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

8ee2d5bc

cmd/compile: add Type.WidthCalculated · 8dafdb1b

Josh Bleecher Snyder authored Mar 23, 2017

Prior to this CL, Type.Width != 0 was the mark
of a Type whose Width had been calculated.
As a result, dowidth always recalculated
the width of struct{}.
This, combined with the prohibition on calculating
the width of a FuncArgsStruct and the use of
struct{} as a function argument,
meant that there were circumstances in which
it was forbidden to call dowidth on a type.
This inhibits refactoring to call dowidth automatically,
rather than explicitly.
Instead add a helper method, Type.WidthCalculated,
and implement as Type.Align > 0.
Type.Width is not a good candidate for tracking
whether the width has been calculated;
0 is a value type width, and Width is subject to
too much magic value game-playing.

For good measure, add a test for #11354.

Change-Id: Ie9a9fb5d924e7a2010c1904ae5e38ed4a38eaeb2
Reviewed-on: https://go-review.googlesource.com/38468
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>

8dafdb1b

regexp: reduce allocations at onePassCopy · f5c1926e

haya14busa authored Mar 28, 2017

It reduces needless allocations on compiling onepass regex.

name old time/op new time/op delta
CompileOnepass/^(?:(?:(?:.(?:$))?))...-4 6.31µs ± 3% 6.11µs ± 3% ~ (p=0.056 n=5+5)
CompileOnepass/^abcd$-4 5.69µs ±12% 4.93µs ± 4% -13.42% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a{0,})*?)$-4 7.10µs ±12% 5.82µs ± 5% -17.95% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a+)*)$-4 5.99µs ±10% 6.07µs ±11% ~ (p=1.000 n=5+5)
CompileOnepass/^(?:(?:a|(?:aa)))$-4 7.36µs ± 4% 7.81µs ±19% ~ (p=0.310 n=5+5)
CompileOnepass/^(?:[^\s\S])$-4 4.71µs ± 3% 4.71µs ± 5% ~ (p=1.000 n=5+5)
CompileOnepass/^(?:(?:(?:a*)+))$-4 6.06µs ± 2% 6.23µs ± 9% ~ (p=0.310 n=5+5)
CompileOnepass/^[a-c]+$-4 4.74µs ± 4% 4.64µs ± 6% ~ (p=0.421 n=5+5)
CompileOnepass/^[a-c]*$-4 5.17µs ± 2% 4.68µs ± 0% -9.57% (p=0.016 n=5+4)
CompileOnepass/^(?:a*)$-4 5.34µs ± 3% 5.08µs ±12% ~ (p=0.151 n=5+5)
CompileOnepass/^(?:(?:aa)|a)$-4 7.24µs ± 5% 7.33µs ±12% ~ (p=0.841 n=5+5)
CompileOnepass/^...$-4 5.28µs ± 3% 4.99µs ± 9% ~ (p=0.095 n=5+5)
CompileOnepass/^(?:a|(?:aa))$-4 7.20µs ± 4% 7.24µs ±10% ~ (p=0.841 n=5+5)
CompileOnepass/^a((b))c$-4 7.99µs ± 3% 7.76µs ± 8% ~ (p=0.151 n=5+5)
CompileOnepass/^a.[l-nA-Cg-j]?e$-4 8.30µs ± 5% 7.29µs ± 4% -12.08% (p=0.008 n=5+5)
CompileOnepass/^a((b))$-4 7.34µs ± 4% 7.24µs ±19% ~ (p=0.690 n=5+5)
CompileOnepass/^a(?:(b)|(c))c$-4 9.80µs ± 6% 9.49µs ±18% ~ (p=0.151 n=5+5)
CompileOnepass/^a(?:b|c)$-4 5.23µs ± 3% 4.80µs ±10% ~ (p=0.056 n=5+5)
CompileOnepass/^a(?:b?|c)$-4 8.26µs ± 3% 7.30µs ± 3% -11.62% (p=0.008 n=5+5)
CompileOnepass/^a(?:b?|c+)$-4 9.18µs ± 2% 8.16µs ± 2% -11.06% (p=0.008 n=5+5)
CompileOnepass/^a(?:bc)+$-4 6.16µs ± 3% 6.41µs ±13% ~ (p=0.548 n=5+5)
CompileOnepass/^a(?:[bcd])+$-4 5.75µs ± 5% 5.50µs ±12% ~ (p=0.151 n=5+5)
CompileOnepass/^a((?:[bcd])+)$-4 7.65µs ± 5% 6.93µs ± 9% ~ (p=0.056 n=5+5)
CompileOnepass/^a(:?b|c)*d$-4 13.0µs ± 1% 12.1µs ± 2% -6.91% (p=0.008 n=5+5)
CompileOnepass/^.bc(d|e)*$-4 9.20µs ± 4% 8.25µs ± 3% -10.38% (p=0.008 n=5+5)
CompileOnepass/^loooooooooooooooooo...-4 254µs ± 2% 220µs ± 6% -13.47% (p=0.008 n=5+5)

name old alloc/op new alloc/op delta
CompileOnepass/^(?:(?:(?:.(?:$))?))...-4 3.92kB ± 0% 3.41kB ± 0% -13.06% (p=0.008 n=5+5)
CompileOnepass/^abcd$-4 3.20kB ± 0% 2.75kB ± 0% -14.00% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a{0,})*?)$-4 3.85kB ± 0% 3.34kB ± 0% -13.31% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a+)*)$-4 3.46kB ± 0% 2.95kB ± 0% -14.78% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a|(?:aa)))$-4 4.20kB ± 0% 3.75kB ± 0% -10.67% (p=0.008 n=5+5)
CompileOnepass/^(?:[^\s\S])$-4 3.10kB ± 0% 2.46kB ± 0% -20.62% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:(?:a*)+))$-4 3.64kB ± 0% 3.13kB ± 0% -14.07% (p=0.008 n=5+5)
CompileOnepass/^[a-c]+$-4 3.06kB ± 0% 2.48kB ± 0% -18.85% (p=0.008 n=5+5)
CompileOnepass/^[a-c]*$-4 3.10kB ± 0% 2.52kB ± 0% -18.60% (p=0.008 n=5+5)
CompileOnepass/^(?:a*)$-4 3.21kB ± 0% 2.63kB ± 0% -17.96% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:aa)|a)$-4 4.09kB ± 0% 3.64kB ± 0% -10.96% (p=0.008 n=5+5)
CompileOnepass/^...$-4 3.42kB ± 0% 2.91kB ± 0% -14.95% (p=0.008 n=5+5)
CompileOnepass/^(?:a|(?:aa))$-4 4.09kB ± 0% 3.64kB ± 0% -10.96% (p=0.008 n=5+5)
CompileOnepass/^a((b))c$-4 5.67kB ± 0% 4.39kB ± 0% -22.59% (p=0.008 n=5+5)
CompileOnepass/^a.[l-nA-Cg-j]?e$-4 5.73kB ± 0% 4.32kB ± 0% -24.58% (p=0.008 n=5+5)
CompileOnepass/^a((b))$-4 5.41kB ± 0% 4.06kB ± 0% -24.85% (p=0.008 n=5+5)
CompileOnepass/^a(?:(b)|(c))c$-4 6.40kB ± 0% 5.31kB ± 0% -17.00% (p=0.008 n=5+5)
CompileOnepass/^a(?:b|c)$-4 3.46kB ± 0% 2.88kB ± 0% -16.67% (p=0.008 n=5+5)
CompileOnepass/^a(?:b?|c)$-4 5.77kB ± 0% 4.36kB ± 0% -24.41% (p=0.008 n=5+5)
CompileOnepass/^a(?:b?|c+)$-4 5.94kB ± 0% 4.59kB ± 0% -22.64% (p=0.008 n=5+5)
CompileOnepass/^a(?:bc)+$-4 3.60kB ± 0% 3.15kB ± 0% -12.44% (p=0.008 n=5+5)
CompileOnepass/^a(?:[bcd])+$-4 3.46kB ± 0% 2.94kB ± 0% -14.81% (p=0.008 n=5+5)
CompileOnepass/^a((?:[bcd])+)$-4 5.50kB ± 0% 4.09kB ± 0% -25.62% (p=0.008 n=5+5)
CompileOnepass/^a(:?b|c)*d$-4 7.24kB ± 0% 6.15kB ± 0% -15.03% (p=0.008 n=5+5)
CompileOnepass/^.bc(d|e)*$-4 5.75kB ± 0% 4.47kB ± 0% -22.25% (p=0.008 n=5+5)
CompileOnepass/^loooooooooooooooooo...-4 225kB ± 0% 135kB ± 0% -39.99% (p=0.008 n=5+5)

name old allocs/op new allocs/op delta
CompileOnepass/^(?:(?:(?:.(?:$))?))...-4 52.0 ± 0% 49.0 ± 0% -5.77% (p=0.008 n=5+5)
CompileOnepass/^abcd$-4 44.0 ± 0% 41.0 ± 0% -6.82% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a{0,})*?)$-4 52.0 ± 0% 49.0 ± 0% -5.77% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a+)*)$-4 47.0 ± 0% 44.0 ± 0% -6.38% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:a|(?:aa)))$-4 57.0 ± 0% 54.0 ± 0% -5.26% (p=0.008 n=5+5)
CompileOnepass/^(?:[^\s\S])$-4 36.0 ± 0% 33.0 ± 0% -8.33% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:(?:a*)+))$-4 49.0 ± 0% 46.0 ± 0% -6.12% (p=0.008 n=5+5)
CompileOnepass/^[a-c]+$-4 39.0 ± 0% 36.0 ± 0% -7.69% (p=0.008 n=5+5)
CompileOnepass/^[a-c]*$-4 44.0 ± 0% 41.0 ± 0% -6.82% (p=0.008 n=5+5)
CompileOnepass/^(?:a*)$-4 45.0 ± 0% 42.0 ± 0% -6.67% (p=0.008 n=5+5)
CompileOnepass/^(?:(?:aa)|a)$-4 56.0 ± 0% 53.0 ± 0% -5.36% (p=0.008 n=5+5)
CompileOnepass/^...$-4 46.0 ± 0% 43.0 ± 0% -6.52% (p=0.008 n=5+5)
CompileOnepass/^(?:a|(?:aa))$-4 56.0 ± 0% 53.0 ± 0% -5.36% (p=0.008 n=5+5)
CompileOnepass/^a((b))c$-4 57.0 ± 0% 53.0 ± 0% -7.02% (p=0.008 n=5+5)
CompileOnepass/^a.[l-nA-Cg-j]?e$-4 62.0 ± 0% 58.0 ± 0% -6.45% (p=0.008 n=5+5)
CompileOnepass/^a((b))$-4 51.0 ± 0% 47.0 ± 0% -7.84% (p=0.008 n=5+5)
CompileOnepass/^a(?:(b)|(c))c$-4 69.0 ± 0% 65.0 ± 0% -5.80% (p=0.008 n=5+5)
CompileOnepass/^a(?:b|c)$-4 43.0 ± 0% 40.0 ± 0% -6.98% (p=0.008 n=5+5)
CompileOnepass/^a(?:b?|c)$-4 61.0 ± 0% 57.0 ± 0% -6.56% (p=0.008 n=5+5)
CompileOnepass/^a(?:b?|c+)$-4 67.0 ± 0% 63.0 ± 0% -5.97% (p=0.008 n=5+5)
CompileOnepass/^a(?:bc)+$-4 49.0 ± 0% 46.0 ± 0% -6.12% (p=0.008 n=5+5)
CompileOnepass/^a(?:[bcd])+$-4 46.0 ± 0% 43.0 ± 0% -6.52% (p=0.008 n=5+5)
CompileOnepass/^a((?:[bcd])+)$-4 53.0 ± 0% 49.0 ± 0% -7.55% (p=0.008 n=5+5)
CompileOnepass/^a(:?b|c)*d$-4 109 ± 0% 105 ± 0% -3.67% (p=0.008 n=5+5)
CompileOnepass/^.bc(d|e)*$-4 66.0 ± 0% 62.0 ± 0% -6.06% (p=0.008 n=5+5)
CompileOnepass/^loooooooooooooooooo...-4 1.10k ± 0% 1.09k ± 0% -0.91% (p=0.008 n=5+5)

Fixes #19735

Change-Id: Ic68503aaa08e42fafcf7e11cf1f584d674f5ea7b
Reviewed-on: https://go-review.googlesource.com/38750Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

f5c1926e

cmd/internal/obj: eliminate stray ctxt.Cursym write · aca58647

Josh Bleecher Snyder authored Mar 27, 2017

It is explicitly assigned in each of the
assemblers as needed.
I plan to remove Cursym entirely eventually,
but this is a helpful intermediate step.

Passes toolstash-check -all.

Updates #15756

Change-Id: Id7ddefae2def439af44d03053886ca8cc935731f
Reviewed-on: https://go-review.googlesource.com/38727
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

aca58647

27 Mar, 2017 11 commits

cmd/vet: remove Peek from list of canonical methods · 73912a1b

Josh Bleecher Snyder authored Mar 27, 2017

It is insufficiently canonical;
see the discussion at issue 19719.

Fixes #19719

Change-Id: I0559ff3b1b39d7bc4b446d104f36fdf8ce3da50e
Reviewed-on: https://go-review.googlesource.com/38722
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Rob Pike <r@golang.org>

73912a1b

cmd/compile: construct typename in walk instead of SSA conversion · a0d6d385

Josh Bleecher Snyder authored Mar 27, 2017

This eliminates references to lineno and
other globals from ssa conversion.

Passes toolstash-check.

Updates #15756

Change-Id: I9792074fab0036b42f454b79139d0b27db913fb5
Reviewed-on: https://go-review.googlesource.com/38721
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>

a0d6d385

cmd/asm: add support to shift operands on arm64 · 579297e1

wei xiao authored Nov 28, 2016

Fixes: #18070
Also added a test in: cmd/asm/internal/asm/testdata/arm64.s

Change-Id: Icc43ff7383cc06b8eaccabd9ff0aefa61c4ecb88
Reviewed-on: https://go-review.googlesource.com/33595
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>

579297e1

cmd/internal/obj: eliminate Curp · 7e817859

Josh Bleecher Snyder authored Mar 27, 2017

Remove the global obj.Link.Curp.

In asmz.go, replace the only use by passing it as an argument.
In asm0.go and asm9.go, it was written but never read.
In asm5.go and asm7.go, thread it through as an argument.

Passes toolstash-check -all.

Updates #15756

Change-Id: I1a0faa89e768820f35d73a8b37ec8088d78d15f7
Reviewed-on: https://go-review.googlesource.com/38715
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

7e817859

cmd/internal/obj: remove prasm · 1acba7d4

Josh Bleecher Snyder authored Mar 27, 2017

Fold the printing of the offending instruction
into the neighboring Diag call, if it is not
already present.

Change-Id: I310f1479e16a4d2a24ff3c2f7e2c60e5e2015c1b
Reviewed-on: https://go-review.googlesource.com/38714
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

1acba7d4

runtime/cgo: CFRelease result from CFBundleCopyResourceURL · 2b4274d6

Elias Naur authored Mar 27, 2017

The result from CFBundleCopyResourceURL is owned by the caller. This
CL adds the necessary CFRelease to release it after use.

Fixes #19722

Change-Id: I7afe22ef241d21922a7f5cef6498017e6269a5c3
Reviewed-on: https://go-review.googlesource.com/38639
Run-TryBot: Elias Naur <elias.naur@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>

2b4274d6

cmd/internal/dwarf: remove global encbuf · f9783011

Josh Bleecher Snyder authored Mar 26, 2017

The global encbuf helped avoid allocations.
It is incompatible with a concurrent backend.
To avoid a performance regression while removing it,
introduce two optimizations.
First, re-use a buffer in dwarf.PutFunc.
Second, avoid a buffer entirely when the int
being encoded fits in seven bits, which is about 75%
of the time.

Passes toolstash-check.

Updates #15756


name       old alloc/op    new alloc/op    delta
Template      40.6MB ± 0%     40.6MB ± 0%  -0.08%  (p=0.001 n=8+9)
Unicode       29.9MB ± 0%     29.9MB ± 0%    ~     (p=0.068 n=8+10)
GoTypes        116MB ± 0%      116MB ± 0%  +0.05%  (p=0.043 n=10+9)
SSA            864MB ± 0%      864MB ± 0%  +0.01%  (p=0.010 n=10+9)
Flate         25.8MB ± 0%     25.8MB ± 0%    ~     (p=0.353 n=10+10)
GoParser      32.2MB ± 0%     32.2MB ± 0%    ~     (p=0.353 n=10+10)
Reflect       80.2MB ± 0%     80.2MB ± 0%    ~     (p=0.165 n=10+10)
Tar           27.0MB ± 0%     26.9MB ± 0%    ~     (p=0.143 n=10+10)
XML           42.8MB ± 0%     42.8MB ± 0%    ~     (p=0.400 n=10+9)

name       old allocs/op   new allocs/op   delta
Template        398k ± 0%       397k ± 0%  -0.20%  (p=0.002 n=8+9)
Unicode         320k ± 0%       321k ± 1%    ~     (p=0.122 n=8+10)
GoTypes        1.16M ± 0%      1.17M ± 0%    ~     (p=0.053 n=10+9)
SSA            7.65M ± 0%      7.65M ± 0%    ~     (p=0.122 n=10+8)
Flate           240k ± 1%       240k ± 1%    ~     (p=0.243 n=10+9)
GoParser        322k ± 1%       322k ± 1%    ~     (p=0.481 n=10+10)
Reflect        1.00M ± 0%      1.00M ± 0%    ~     (p=0.211 n=9+10)
Tar             256k ± 0%       255k ± 1%    ~     (p=0.052 n=10+10)
XML             400k ± 1%       400k ± 0%    ~     (p=0.631 n=10+10)


Change-Id: Ia39d9de09232fdbfc9c9cec14587bbf6939c9492
Reviewed-on: https://go-review.googlesource.com/38713
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

f9783011

net/http/httptest: don't panic on Close of user-constructed Server value · 89ebe5bb

Brad Fitzpatrick authored Mar 27, 2017

If the user created an httptest.Server directly without using a
constructor it won't have the new unexported 'client' field. So don't
assume it's non-nil.

Fixes #19729

Change-Id: Ie92e5da66cf4e7fb8d95f3ad0f4e3987d3ae8b77
Reviewed-on: https://go-review.googlesource.com/38710
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Kevin Burke <kev@inburke.com>

89ebe5bb

runtime: improve systemstack-on-Go stack message · 4234d1de

Austin Clements authored Mar 23, 2017

We reused the old C stack check mechanism for the implementation of
//go:systemstack, so when we execute a //go:systemstack function on a
user stack, the system fails by calling morestackc. However,
morestackc's message still talks about "executing C code".

Fix morestackc's message to reflect its modern usage.

Change-Id: I7e70e7980eab761c0520f675d3ce89486496030f
Reviewed-on: https://go-review.googlesource.com/38572
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

4234d1de

runtime/cgo: raise the thread-local storage slot search limit on Android · 0476c7a7

Elias Naur authored Mar 25, 2017

On Android, the thread local offset is found by looping through memory
starting at the TLS base address. The search is limited to
PTHREAD_KEYS_MAX, but issue 19472 made it clear that in some cases, the
slot is located further from the TLS base.

The limit is merely a sanity check in case our assumptions about the
thread-local storage layout are wrong, so this CL raises it to 384, which
is enough for the test case in issue 19472.

Fixes #19472

Change-Id: I89d1db3e9739d3a7fff5548ae487a7483c0a278a
Reviewed-on: https://go-review.googlesource.com/38636
Run-TryBot: Elias Naur <elias.naur@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

0476c7a7

runtime/pprof: fix proto tests on NetBSD · aa4c2ca3

Elias Naur authored Mar 25, 2017

The proto_test tests are failing on NetBSD:

https://build.golang.org/log/a3a577144ac48c6ef8e384ce6a700ad30549fb78

the failures seem similar to previous failures on Android:

https://build.golang.org/log/b5786e0cd6d5941dc37b6a50be5172f6b99e22f0

The Android failures where fixed by CL 37896. This CL is an attempt
to fix the NetBSD failures with a similar fix.

Change-Id: I3834afa5b32303ca226e6a31f0f321f66fef9a3f
Reviewed-on: https://go-review.googlesource.com/38637Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

aa4c2ca3

26 Mar, 2017 7 commits

cmd/internal/obj/x86: change AsmBuf.Lock to bool · 4909ecc4

Josh Bleecher Snyder authored Mar 26, 2017

Follow-up to CL 38668.

Passes toolstash-check -all.

Change-Id: I78a62509c610b5184b5e7ef2c4aa146fc8038840
Reviewed-on: https://go-review.googlesource.com/38670
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Dave Cheney <dave@cheney.net>

4909ecc4

cmd/internal/obj: eliminate Prog.Mode · cf2b32e7

Josh Bleecher Snyder authored Mar 26, 2017

Follow-up to CL 38446.

Passes toolstash-check -all.

Change-Id: I04cadc058cbaa5f396136502c574e5a395a33276
Reviewed-on: https://go-review.googlesource.com/38669
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

cf2b32e7

cmd/internal/obj: move fields from obj.Link to x86.AsmBuf · 4f122e82

Josh Bleecher Snyder authored Mar 26, 2017

These fields are used to encode a single instruction.
Add them to AsmBuf, which is also per-instruction,
and which is not global.

Updates #15756

Change-Id: I0e5ea22ffa641b07291e27de6e2ff23b6dc534bd
Reviewed-on: https://go-review.googlesource.com/38668
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

4f122e82

cmd/internal/obj/x86: make ctxt.Cursym local · 166160b4

Josh Bleecher Snyder authored Mar 26, 2017

Thread it through as an argument instead of using a global.

Passes toolstash-check -all.

Updates #15756

Change-Id: Ia8c6ce09b43dbb2e6c7d889ded8dbaeb5366048d
Reviewed-on: https://go-review.googlesource.com/38667
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

166160b4

cmd/internal/obj/x86: remove ctxt.Curp references · c06679f7

Josh Bleecher Snyder authored Mar 25, 2017

Empirically, p == ctxt.Curp here.
A scan of (the thousands of lines of) asm6.go
shows no clear opportunity for them to diverge.

Passes toolstash-check -all.

Updates #15756

Change-Id: I9f5ee9585a850fbe24be3b851d8fdc2c966c65ce
Reviewed-on: https://go-review.googlesource.com/38665
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>

c06679f7

cmd/compile: remove likely bits from generated assembly · 214be5b3

Keith Randall authored Mar 25, 2017

We don't need them any more since #15837 was fixed.

Fixes #19718

Change-Id: I13e46c62b321b2c9265f44c977b63bfb23163ca2
Reviewed-on: https://go-review.googlesource.com/38664
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>

214be5b3

cmd/internal/obj: make x86's asmbuf a local variable · 2ebe1bdd

Josh Bleecher Snyder authored Mar 25, 2017

The x86 assembler requires a buffer to build
variable-length instructions.
It used to be an obj.Link field.
That doesn't play nicely with concurrent assembly.
Move the AsmBuf type to the x86 package,
where it belongs anyway,
and make it a local variable.

Passes toolstash-check -all.
No compiler performance impact.

Updates #15756

Change-Id: I8014e52145380bfd378ee374a0c971ee5bada917
Reviewed-on: https://go-review.googlesource.com/38663
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

2ebe1bdd

25 Mar, 2017 7 commits

cmd/internal/obj: eagerly initialize x86 assembler · 3b251e60

Josh Bleecher Snyder authored Mar 25, 2017

Prior to this CL, instinit was called as needed.
This does not work well in a concurrent backend.
Initialization is very cheap; do it on startup instead.

Passes toolstash-check -all.
No compiler performance impact.

Updates #15756

Change-Id: Ifa5e82e8abf4504435e1b28766f5703a0555f42d
Reviewed-on: https://go-review.googlesource.com/38662
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

3b251e60

cmd/compile: add test for non interface type switch · e76d6a45

Emmanuel Odeke authored Mar 25, 2017

Ensure that we have a test for when the compiler
encounters a type switch on a non-interface value.

Change-Id: Icb222f986894d0190e1241ca65396b4950e7d14f
Reviewed-on: https://go-review.googlesource.com/38661Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

e76d6a45

cmd/compile/internal/syntax: remove need for missing_statement (fixed TODO) · 783f166e

Robert Griesemer authored Mar 24, 2017

Now that we have consistent use of xOrNil parse methods, we don't
need a special missing_statement singleton to distinguish between
missing actually statements and other errors (which have returned
nil before).

For #19663.

Change-Id: I8364f1441bdf8dd966bcd6d8219b2a42d6b88abd
Reviewed-on: https://go-review.googlesource.com/38656
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>

783f166e

cmd/compile/internal/syntax: always construct a correct syntax tree · d1f5e5f4

Robert Griesemer authored Mar 24, 2017

- parser creates sensible nodes in case of syntax errors instead of nil
- a new BadExpr node is used in places where we can't do better
- fixed error message for incorrect type switch guard
- minor cleanups

Fixes #19663.

Change-Id: I028394c6db9cba7371f0e417ebf93f594659786a
Reviewed-on: https://go-review.googlesource.com/38653
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>

d1f5e5f4

cmd/compile: prevent modification of ONAME/OLITERAL/OTYPES nodes in walkexpr · ecc6a816

Josh Bleecher Snyder authored Mar 24, 2017

ONAME, OLITERAL, and OTYPE nodes can be shared between functions.
In a concurrent backend, such nodes might be walked concurrently
with being read in other functions.
Arrange for them to be unmodified by walk.

This is a follow-up to CL 38609.

Passes toolstash-check.

Updates #15756

Change-Id: I03ff1d2c0ad81dafac3fd55caa218939cf7c0565
Reviewed-on: https://go-review.googlesource.com/38655
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>

ecc6a816

cmd/compile: enforce no uses of Curfn in backend · ef1ab0f0

Josh Bleecher Snyder authored Mar 23, 2017

Updates #15756

Change-Id: Id8d65ca9a3f1a7f9ea43e26cdd5e7d3befef8ba0
Reviewed-on: https://go-review.googlesource.com/38593
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

ef1ab0f0

cmd/compile: provide pos and curfn to temp · 34975095

Josh Bleecher Snyder authored Mar 22, 2017

Concurrent compilation requires providing an
explicit position and curfn to temp.
This implementation of tempAt temporarily
continues to use the globals lineno and Curfn,
so as not to collide with mdempsky's
work for #19683 eliminating the Curfn dependency
from func nod.

Updates #15756
Updates #19683

Change-Id: Ib3149ca4b0740e9f6eea44babc6f34cdd63028a9
Reviewed-on: https://go-review.googlesource.com/38592
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>

34975095

24 Mar, 2017 4 commits

net/http: Fix TestLinuxSendfile without strace permissions · 34396ada

Kenny Grant authored Mar 17, 2017

If go doesn't have permission to run strace, this test hangs while
waiting for strace to run. Instead try invoking strace with
Run() first - on fail skip and report error, otherwise run
the test normally using strace.

Also fix link to open mips64 issue in same test.

Fixes #9711

Change-Id: Ibbc5fbb143ea6d0f8b6cfdca4b385ef4c8960b3d
Reviewed-on: https://go-review.googlesource.com/38633Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

34396ada

cmd/compile: combine walkexpr cases · 0b647ffb

Josh Bleecher Snyder authored Mar 24, 2017

The type switch in walkexpr is giant.
Shrink it a little by coalescing identical cases
and removing some vertical whitespace.

No functional changes.

Passes toolstash-check.

Change-Id: I7f7efb4faae1f8657dfafac04585172f99d8b37d
Reviewed-on: https://go-review.googlesource.com/38652
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

0b647ffb

cmd/compile: simplify funcsyms · 2c50bffe

Josh Bleecher Snyder authored Mar 24, 2017

Sym.Fsym is used only to avoid adding duplicate
entries to funcsyms, but that is easily
accomplished by detecting the first lookup
vs subsequent lookups of the func sym name.

This avoids creating an unnecessary ONAME node
during funcsym, which eliminates a dependency
in the backend on Curfn and lineno.

It also makes the code a lot simpler and clearer.

Updates #15756

Passes toolstash-check -all.
No compiler performance changes.
funcsymname does generate garbage via string
concatenation, but it is not called very much,
and this CL also eliminates allocation of several
Nodes and Names.

Change-Id: I7116c78fa39d975b7bd2c65a1d228749cf0dd46b
Reviewed-on: https://go-review.googlesource.com/38605Reviewed-by: Matthew Dempsky <mdempsky@google.com>

2c50bffe

cmd/compile: replace nod(ONAME) with newname · 3a89065c

Matthew Dempsky authored Mar 24, 2017

Passes toolstash-check -all.

Change-Id: Ib9f969e5ecc1537b7eab186dc4fd504a50f800f2
Reviewed-on: https://go-review.googlesource.com/38586
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

3a89065c