Some of the changes in CL golang.org/cl/38071/ assumed that / and %
could always be combined to use only one DIV instruction. However,
this is not the case for 64bit operands on a 32bit platform which use
seperate runtime functions to calculate division and modulo.
This CL restores the original optimizations that help on 32bit platforms
with negligible impact on 64bit platforms.
386:
name old time/op new time/op delta
FormatInt-2 6.06µs ± 0% 6.02µs ± 0% -0.70% (p=0.000 n=20+20)
AppendInt-2 4.98µs ± 0% 4.98µs ± 0% ~ (p=0.747 n=18+18)
FormatUint-2 1.93µs ± 0% 1.85µs ± 0% -4.19% (p=0.000 n=20+20)
AppendUint-2 1.71µs ± 0% 1.64µs ± 0% -3.68% (p=0.000 n=20+20)
amd64:
name old time/op new time/op delta
FormatInt-2 2.41µs ± 0% 2.41µs ± 0% -0.09% (p=0.010 n=18+18)
AppendInt-2 1.77µs ± 0% 1.77µs ± 0% +0.08% (p=0.000 n=18+18)
FormatUint-2 653ns ± 1% 653ns ± 0% ~ (p=0.178 n=20+20)
AppendUint-2 514ns ± 0% 513ns ± 0% -0.13% (p=0.000 n=20+17)
Change-Id: I574a18e54fb41b25fbe51ce696e7a8765abc79a6
Reviewed-on: https://go-review.googlesource.com/38051
Run-TryBot: Martin Möhrmann <moehrmann@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
This appears to be leftover from when instruction selection happened
in the linker. Many of the morestackX functions listed don't even
exist anymore.
Now that we select instructions within the compiler and assembler,
normal deadcode elimination mechanisms should suffice for these
symbols.
Change-Id: I2cb1e435101392e7c983957c4acfbbcc87a5ca7d
Reviewed-on: https://go-review.googlesource.com/38077
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Determine int, uint and uintptr bit sizes from GOARCH environment
variable if it is set. Otherwise use host-specific sizes.
Fixes#19321
Change-Id: I494b8e4b49b59d32794f50ff2ce06ba040cb8460
Reviewed-on: https://go-review.googlesource.com/37950
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
This avoids a problem that occurs on FreeBSD 11, in which the clang
3.8 assembler issues a pointless warning when invoked with -g on a
file that contains an empty .note.GNU-stack section.
No test because there is no reasonable way to write one, but should
fix the build on FreeBSD 11.
Fixes#14705.
Change-Id: I8c49bbf79a2c715c0e75495da19045fc92280e81
Reviewed-on: https://go-review.googlesource.com/38072
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
nat.setUint64 is nicely generic.
By assuming 32- or 64-bit words, however,
we can write simpler code,
and eliminate some shifts
in dead code that vet complains about.
Generated code for 64 bit systems is unaltered.
Generated code for 32 bit systems is much better.
For 386, the routine length drops from 325
bytes of code to 271 bytes of code, with fewer loops.
Change-Id: I1bc14c06272dee37a7fcb48d33dd1e621eba945d
Reviewed-on: https://go-review.googlesource.com/38070
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
The compiler recognizes that in a sequence q = x/y; r = x%y only
one division is required. Remove prior work-arounds and write
more readable straight-line code (this also results in fewer
instructions, though it doesn't appear to affect the benchmarks
significantly).
name old time/op new time/op delta
FormatInt-8 2.95µs ± 1% 2.92µs ± 5% ~ (p=0.952 n=5+5)
AppendInt-8 1.91µs ± 1% 1.89µs ± 2% ~ (p=0.421 n=5+5)
FormatUint-8 795ns ± 2% 782ns ± 4% ~ (p=0.444 n=5+5)
AppendUint-8 557ns ± 1% 557ns ± 2% ~ (p=0.548 n=5+5)
https://perf.golang.org/search?q=upload:20170310.1
Also:
- use uint instead of uintptr where we want to guarantee single-
register operations
- remove some unnecessary conversions (before indexing)
- add more comments and fix some comments
Change-Id: I04858dc2d798a6495879d9c7cfec2fdc2957b704
Reviewed-on: https://go-review.googlesource.com/38071
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
In FreeBSD when run Go proc under a given sub-list of
processors(e.g. 'cpuset -l 0 ./a.out' in multi-core system),
runtime.NumCPU() still return all physical CPUs from sysctl
hw.ncpu instead of account from sub-list.
Fix by use syscall cpuset_getaffinity to account the number of sub-list.
Fixes#15206
Change-Id: If87c4b620e870486efa100685db5debbf1210a5b
Reviewed-on: https://go-review.googlesource.com/29341
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
We are vendoring pprof from github.com/google/pprof, which comes with
a main package. If we don't explicitly skip that main package, then
`go install cmd` will install the compiled program in $GOROOT/bin.
Fixes#19441.
Change-Id: Ib268ffd16d4be65f7d80e4f8d9dc6e71523a94de
Reviewed-on: https://go-review.googlesource.com/38007
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Raul Silvera <rsilvera@google.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Found by github.com/mvdan/unparam.
Change-Id: Iabcdfec2ae42c735aa23210b7183080d750682ca
Reviewed-on: https://go-review.googlesource.com/38030
Reviewed-by: Peter Weinberger <pjw@google.com>
Run-TryBot: Peter Weinberger <pjw@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Change-Id: I0fa68ca9812fe5e82ffb9d0b9598e95b47183eb8
Reviewed-on: https://go-review.googlesource.com/38011
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
It appears that this test is particularly
sensitive to resource starvation.
Returning it to non-parallel should reduce flakiness,
by giving it the full system resources to run.
Fixes#19161
Change-Id: I6e8906516629badaa0cffeb5712af649dc197f39
Reviewed-on: https://go-review.googlesource.com/38005
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Minor fix, because it's the right thing to do.
No significant impact.
Change-Id: I2138285d397494daa9a88c414149c2a7860edd7e
Reviewed-on: https://go-review.googlesource.com/38001
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Values have an Aux and an AuxInt.
We're setting AuxInt, not Aux.
Say so.
Change-Id: I41aa783273bb7e1ba47c941aa4233f818e37dadd
Reviewed-on: https://go-review.googlesource.com/37997
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Compiler errors now show the exact line and line byte offset (sometimes
called "column") of where an error occured. For `go tool compile x.go`:
package p
const c int = false
//line foo.go:123
type t intg
reports
x.go:2:7: cannot convert false to type int
foo.go:123[x.go:4:8]: undefined: intg
(Some errors use the "wrong" position for the error message; arguably
the byte offset for the first error should be 15, the position of 'false',
rathen than 7, the position of 'c'. But that is an indepedent issue.)
The byte offset (column) values are measured in bytes; they start at 1,
matching the convention used by editors and IDEs.
Positions modified by //line directives show the line offset only for the
actual source location (in square brackets), not for the "virtual" file and
line number because that code is likely generated and the //line directive
only provides line information.
Because the new format might break existing tools or scripts, printing
of line offsets can be disabled with the new compiler flag -C. We plan
to remove this flag eventually.
Fixes#10324.
Change-Id: I493f5ee6e78457cf7b00025aba6b6e28e50bb740
Reviewed-on: https://go-review.googlesource.com/37970
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
We could leave it alone and fix line offset (column) numbers when
reporting errors, but that is likely to cause confusion (internal
numbers don't match reported numbers). Instead, switch to default
numbering starting at 1.
For package syntax-internal use only, introduced constants defining
the line and column bases, and use them throughout the code and its
tests. It is possible to change these constants and package syntax
will continue to work. But changing them is going to break any client
that makes explicit assumptions about line and column numbers (which
is "all of them").
Change-Id: Ia3d136a8ec8d9372ed9c05ca47d3dff222cf030e
Reviewed-on: https://go-review.googlesource.com/37996
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
When LookupIP is performing multiple subqueries, this option causes a
timeout/servfail affecting a single query to abort the whole operation,
instead of returning a partial (IPv4/IPv6-only) result.
Similarly, operations that walk the DNS search list will also abort when
encountering one of these errors.
Fixes#17448
Change-Id: Ice22e4aceb555c5a80d19bd1fde8b8fe87ac9517
Reviewed-on: https://go-review.googlesource.com/32572
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
This change adds a method to replace expressions
of the form
v.Args[len(v.Args)-1]
so that the code's intention to walk memory arguments
is explicit.
Passes toolstash-check.
Change-Id: I0c80d73bc00989dd3cdf72b4f2c8e1075a2515e0
Reviewed-on: https://go-review.googlesource.com/37757
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
make.bash used mostly tabs and buildall.bash used mostly spaces, but
they were both mixing them. Be consistent and use tabs, as that's what's
more common and what the Go code uses.
Change-Id: Ia6affbfccfe64fda800c1ac400965df364d2c545
Reviewed-on: https://go-review.googlesource.com/37967
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
The type of the OffPtr for the first field was incorrect. It should
have been a pointer to the field type, rather than the field
type itself.
Fixes#19475.
Change-Id: I3960b404da0f4bee759331126cce6140d2ce1df7
Reviewed-on: https://go-review.googlesource.com/37869
Run-TryBot: Michael Munday <munday@ca.ibm.com>
Reviewed-by: Keith Randall <khr@golang.org>
A typo in the previous revision ("act" instead of "oldact") caused us
to return the sa_flags from the new (or zeroed) sigaction rather than
the old one.
In the presence of a signal handler registered before
runtime.libpreinit, this caused setsigstack to erroneously zero out
important sa_flags (such as SA_SIGINFO) in its attempt to re-register
the existing handler with SA_ONSTACK.
Change-Id: I3cd5152a38ec0d44ae611f183bc1651d65b8a115
Reviewed-on: https://go-review.googlesource.com/37852
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
There are a few problems from change 35494, discovered during testing
of change 37852.
1. I was confused about the usage of n.key in the sema variant, so we
were looping on the wrong condition. The error was not caught by
the TryBots (presumably due to missing TSAN coverage in the BSD and
darwin builders?).
2. The sysmon goroutine sometimes skips notetsleep entirely, using
direct usleep syscalls instead. In that case, we were not calling
_cgo_yield, leading to missed signals under TSAN.
3. Some notetsleep calls have long finite timeouts. They should be
broken up into smaller chunks with a yield at the end of each
chunk.
updates #18717
Change-Id: I91175af5dea3857deebc686f51a8a40f9d690bcc
Reviewed-on: https://go-review.googlesource.com/37867
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
The tzdata 2017a update (2017-02-28) changed the abbreviation of the
Asia/Baghdad time zone (used in TestParseInLocation) from 'AST' to the
numeric '+03'.
Update the test so that it skips the checks if we're using a recent
tzdata release.
Fixes#19457
Change-Id: I45d705a5520743a611bdd194dc8f8d618679980c
Reviewed-on: https://go-review.googlesource.com/37964
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
write(2) is defined in unistd.h.
For the iOS builder.
Change-Id: I411ffe81988d8fbafffde89e4732a20af1a63325
Reviewed-on: https://go-review.googlesource.com/37962
Run-TryBot: Elias Naur <elias.naur@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Previously the base register was unset, which lead to the disassembler
using "FP" instead of "SP" as the base register. That lead to some
confusion as to what the difference betweeen the two was.
Be consistent and always use SP.
Fixes#19458
Change-Id: Ie8f8ee54653bd202c0cf6fbf1d350e3c8c8b67a0
Reviewed-on: https://go-review.googlesource.com/37971
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
The iOS test harness only includes the current test directory in its
app bundles, but the tests need access to all source code.
Change-Id: I8a902b183bc2745b4fbfffef867002d573abb1f5
Reviewed-on: https://go-review.googlesource.com/37961
Run-TryBot: Elias Naur <elias.naur@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
On Android devices where the stub fallback for Current fails to
extract a User from the environment, return a dummy fallback instead
of failing.
While we're here, use / instead of /home/nacl for the NaCL fallback.
Hopefully fixes the Android builder.
Change-Id: Ia29304fbc224ee5f9c0f4e706d1756f765a7eae5
Reviewed-on: https://go-review.googlesource.com/37960
Run-TryBot: Elias Naur <elias.naur@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
After benchmarking with a compiler modified to have better
spill location, it became clear that this method of checking
was actually faster on (at least) two different architectures
(ppc64 and amd64) and it also provides more timely interruption
of loops.
This change adds a modified FOR loop node "FORUNTIL" that
checks after executing the loop body instead of before (i.e.,
always at least once). This ensures that a pointer past the
end of a slice or array is not made visible to the garbage
collector.
Without the rescheduling checks inserted, the restructured
loop from this change apparently provides a 1% geomean
improvement on PPC64 running the go1 benchmarks; the
improvement on AMD64 is only 0.12%.
Inserting the rescheduling check exposed some peculiar bug
with the ssa test code for s390x; this was updated based on
initial code actually generated for GOARCH=s390x to use
appropriate OpArg, OpAddr, and OpVarDef.
NaCl is disabled in testing.
Change-Id: Ieafaa9a61d2a583ad00968110ef3e7a441abca50
Reviewed-on: https://go-review.googlesource.com/36206
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
The position information recorded now consists of the line-
directive relative filename and line number. It would be
relatively easy to also encode absolute position information
as necessary (by serializing src.PosBase data).
For example, given $GOROOT/src/tmp/x.go:
package p
const C0 = 0
//line c.go:10
const C1 = 1
//line t.go:20
type T int
//line v.go:30
var V T
//line f.go:40
func F() {}
The recorded positions for the exported entities are:
C0 $GOROOT/src/tmp/x.go 3
C1 c.go 10
T t.go 20
V v.go 30
F f.go 40
Fix verified by manual inspection. There's currently no easy way
to test this, but it will eventually be tested when we fix#7311.
Fixes#19391.
Change-Id: I6269067ea58358250fe6dd1f73bdf9e5d2adfe3d
Reviewed-on: https://go-review.googlesource.com/37936
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
This change was originally written by Marko Tiikkaja <marko@joh.to>.
https://go-review.googlesource.com/#/c/2035/
Previously *Tx.Stmt always prepared a new statement, even if an
existing one was available on the connection the transaction was on.
Now we first see if the statement is already available on the
connection and only prepare if it isn't. Additionally, when we do
need to prepare one, we store it in the parent *Stmt to allow it to be
later reused by other calls to *Tx.Stmt on that statement or just
straight up by *Stmt.Exec et al.
To make sure that the statement doesn't disappear unexpectedly, we
record a dependency from the statement returned by *Tx.Stmt to the
*Stmt it came from and set a new field, parentStmt, to point to the
originating *Stmt. When the transaction's *Stmt is closed, we remove
the dependency. This way the "parent" *Stmt can be closed by the user
without her having to know whether any transactions are still using it
or not.
Fixes#15606
Change-Id: I41b5056847e117ac61130328b0239d1e000a4a08
Reviewed-on: https://go-review.googlesource.com/35476
Run-TryBot: Daniel Theophanes <kardianos@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Daniel Theophanes <kardianos@gmail.com>
After merging https://go-review.googlesource.com/c/34639/,
it was pointed out to me that a lot of tests under net/http
could use the new functionality to simplify and unify testing.
Using the httptest.Server provided Client removes the need to
call CloseIdleConnections() on all Transports created, as it
is automatically called on the Transport associated with the
client when Server.Close() is called.
Change the transport used by the non-TLS
httptest.Server to a new *http.Transport rather than using
http.DefaultTransport implicitly. The TLS version already
used its own *http.Transport. This change is to prevent
concurrency problems with using DefaultTransport implicitly
across several httptest.Server's.
Add tests to ensure the httptest.Server.Client().Transport
RoundTripper interface is implemented by a *http.Transport,
as is now assumed across large parts of net/http tests.
Change-Id: I9f9d15f59d72893deead5678d314388718c91821
Reviewed-on: https://go-review.googlesource.com/37771
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
This helps systems that maintain an external database mapping
build ID to symbol information for the given binary, especially
in the case where /proc/self/maps lists many different files
(for example, many shared libraries).
Avoid importing debug/elf to avoid dragging in that whole
package (and its dependencies like debug/dwarf) into the
build of every program that generates a profile.
Fixes#19431.
Change-Id: I6d4362a79fe23e4f1726dffb0661d20bb57f766f
Reviewed-on: https://go-review.googlesource.com/37855
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
The old code simply printed the position of the line directive in
square brackets for a position modified by a line directive. Now
we print the corresponding actual source file position instead.
Fixes#19392.
Change-Id: I933f3e435d03a6ee8269df36ae35f9202b7b2e76
Reviewed-on: https://go-review.googlesource.com/37932
Run-TryBot: Robert Griesemer <gri@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Currently selectgo is just a wrapper around selectgoImpl. This keeps
the hard-coded frame skip counts for tracing the same between the
channel implementation and the select implementation.
However, this is fragile and confusing, so pass a skip parameter to
send and recv, join selectgo and selectgoImpl into one function, and
use decrease all of the skips in selectgo by one.
Change-Id: I11b8cbb7d805b55f5dc6ab4875ac7dde79412ff2
Reviewed-on: https://go-review.googlesource.com/37860
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Several SSA ops will always behave identically regardless of target
architecture, so handle those within gc/ssa.go instead.
Passes toolstash-check -all.
Change-Id: I54d514e80ab86723e44332a5a38e3054cbca8c5d
Reviewed-on: https://go-review.googlesource.com/37931
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
If the bad pointer is on a stack, this makes it possible to find the
frame containing the bad pointer.
Change-Id: Ieda44e054aa9ebf22d15d184457c7610b056dded
Reviewed-on: https://go-review.googlesource.com/37858
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
This commit reworks multiway select statements to use normal control
flow primitives instead of the previous setjmp/longjmp-like behavior.
This simplifies liveness analysis and should prevent issues around
"returns twice" function calls within SSA passes.
test/live.go is updated because liveness analysis's CFG is more
representative of actual control flow. The case bodies are the only
real successors of the selectgo call, but previously the selectsend,
selectrecv, etc. calls were included in the successors list too.
Updates #19331.
Change-Id: I7f879b103a4b85e62fc36a270d812f54c0aa3e83
Reviewed-on: https://go-review.googlesource.com/37661
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
It's only ever called with the value it was using, but the code was
counterintuitive. Use the parameter instead, like the other funcs near
it.
Found by github.com/mvdan/unparam.
Change-Id: I45855e11d749380b9b2a28e6dd1d5dedf119a19b
Reviewed-on: https://go-review.googlesource.com/37893
Reviewed-by: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Using the stubs, user.Current will no longer fail on Android, fixing
the os/exec.TestCredentialNoSetGroups test.
Change-Id: I8b9842aa6704c0cde383c549a614bab0a0ed7695
Reviewed-on: https://go-review.googlesource.com/37765
Run-TryBot: Elias Naur <elias.naur@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>