Robert Griesemer [Fri, 18 Nov 2016 18:49:29 +0000 (10:49 -0800)]
spec: add subtitles to section on "for" statements
This matches what we already do for switch statements and makes
this large section more visibly organized. No other changes besides
introducing the titles.
Fixes #4486.
Change-Id: I73f274e4fdd27c6cfeaed79090b4553e57a9c479
Reviewed-on: https://go-review.googlesource.com/33410 Reviewed-by: Rob Pike <r@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Austin Clements [Fri, 4 Nov 2016 15:13:27 +0000 (11:13 -0400)]
cmd/trace: fix goroutine view
Currently, trace processing interleaves state/statistics updates and
emitting trace viewer objects. As a result, if events are being
filtered, either by time or by goroutines, we'll miss those
state/statistics updates. At best, this leads to bad statistics;
however, since we're now strictly checking G state transitions, it
usually leads to a failure to process the trace if there is any
filtering.
Fix this by separating state updates from emitting trace object. State
updates are done before filtering, so we always have correct state
information and statistics. Trace objects are only emitted if we pass
the filter. To determine when we need to emit trace counters, rather
than duplicating the knowledge of which events might modify
statistics, we keep track of the previously emitted counters and emit
a trace counter object whenever these have changed.
Robert Griesemer [Fri, 18 Nov 2016 17:27:18 +0000 (09:27 -0800)]
spec: remove => (alias) operator from Operators and Delimiters section
(Revert of https://go-review.googlesource.com/#/c/32310/)
For #16339.
Fixes #17975.
Change-Id: I36062703c423a81ea1c5b00f4429a4faf00b3782
Reviewed-on: https://go-review.googlesource.com/33365 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Robert Griesemer [Fri, 18 Nov 2016 00:39:11 +0000 (16:39 -0800)]
spec: clarify type elision rules for composite literals
- organize examples better
- add an example illustrating behavior if element type is a named pointer type
- both compilers and go/types (per https://go-review.googlesource.com/33358)
follow this now
See the issue for detailed discussion.
Fixes #17954.
Change-Id: I8d90507ff2347d9493813f75b73233819880d2b4
Reviewed-on: https://go-review.googlesource.com/33361 Reviewed-by: Rob Pike <r@golang.org>
The existing implementation of text/template handles the option
"missingkey=error" in an inconsitent manner: If the provided data is
a nil-interface, no error is returned (despite the fact that no key
can be found in it).
This patch makes text/template return an error if "missingkey=error"
is set and the provided data is a not a valid reflect.Value.
Fixes #15356
Change-Id: Ia0a83da48652ecfaf31f18bdbd78cb21dbca1164
Reviewed-on: https://go-review.googlesource.com/31638 Reviewed-by: Rob Pike <r@golang.org>
Run-TryBot: Rob Pike <r@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Cherry Zhang [Sat, 29 Oct 2016 03:11:04 +0000 (23:11 -0400)]
cmd/compile: make a copy of Phi input if it is still live
Register of Phi input is allocated to the Phi. So if the Phi
input is still live after Phi, we may need to use a spill. In
this case, copy the Phi input to a spare register to avoid a
spill.
Originally targeted the code in issue #16187, and this CL
indeed removes the spill, but doesn't seem to help on benchmark
result. It may help in general, though.
Elias Naur [Sun, 6 Nov 2016 20:40:57 +0000 (21:40 +0100)]
runtime: handle SIGPIPE in c-archive and c-shared programs
Before this CL, Go programs in c-archive or c-shared buildmodes
would not handle SIGPIPE. That leads to surprising behaviour where
writes on a closed pipe or socket would raise SIGPIPE and terminate
the program. This CL changes the Go runtime to handle
SIGPIPE regardless of buildmode. In addition, SIGPIPE from non-Go
code is forwarded.
Fixes #17393
Updates #16760
Change-Id: I155e82020a03a5cdc627a147c27da395662c3fe8
Reviewed-on: https://go-review.googlesource.com/32796
Run-TryBot: Elias Naur <elias.naur@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Adam Langley [Thu, 17 Nov 2016 20:15:19 +0000 (12:15 -0800)]
crypto/tls: reject zero-length SCTs.
The SignedCertificateTimestampList[1] specifies that both the list and
each element must not be empty. Checking that the list is not empty was
handled in [2] and this change checks that the SCTs themselves are not
zero-length.
woodsaj [Thu, 17 Nov 2016 12:14:32 +0000 (20:14 +0800)]
crypto/tls: reject CT extension with no SCTs included
When the CT extension is enabled but no SCTs are present, the existing
code calls "continue" which causes resizing the data byte slice to be
skipped. In fact, such extensions should be rejected.
Ian Lance Taylor [Wed, 16 Nov 2016 17:55:24 +0000 (09:55 -0800)]
cmd/cgo: ignore top-level qualifiers in function args/results
The top-level qualifiers are unimportant for our purposes. If a C
function is defined as `const int f(const int i)`, the `const`s are
meaningless to C, and we want to avoid using them in the struct we
create where the `const` has a completely different meaning.
This unwinds https://golang.org/cl/33097 with regard to top-level
qualifiers.
Change-Id: I3d66b0eb43b6d9a586d9cdedfae5a2306b46d96c
Reviewed-on: https://go-review.googlesource.com/33325
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Daniel Theophanes [Thu, 17 Nov 2016 17:33:31 +0000 (09:33 -0800)]
database/sql: ensure all driver Stmt are closed once
Previously driver.Stmt could could be closed multiple times in
edge cases that drivers may not test for initially. Make their
job easier by ensuring the driver is only closed a single time.
Brad Fitzpatrick [Thu, 17 Nov 2016 16:28:45 +0000 (16:28 +0000)]
runtime/pprof: fix typo in test
Not sure what I was thinking.
Change-Id: I143cdf7c5ef8e7b2394afeca6b30c46bb2c19a55
Reviewed-on: https://go-review.googlesource.com/33340 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Elias Naur [Thu, 17 Nov 2016 12:52:00 +0000 (13:52 +0100)]
misc/cgo: decrease test failure timeouts
CL 33239 changed the polling loops from using sched_yield to a sleep
for 1/1000 of a second. The loop counters were not updated, so failing
tests now take 100 seconds to complete. Lower the loop counts to 5
seconds instead.
Change-Id: I7c9a343dacc8188603ecf7e58bd00b535cfc87f5
Reviewed-on: https://go-review.googlesource.com/33280
Run-TryBot: Elias Naur <elias.naur@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Alex Brainman [Sat, 12 Nov 2016 06:57:35 +0000 (17:57 +1100)]
debug/pe: do not create symbol table if FileHeader.PointerToSymbolTable is 0
https://github.com/tpn/pdfs/raw/master/Microsoft Portable Executable and Common Object File Format Specification - 1999 (pecoff).doc
says this about PointerToSymbolTable:
File offset of the COFF symbol table or 0 if none is present.
Do as it says.
Fixes #17809.
Change-Id: Ib1ad83532f36a3e56c7e058dc9b2acfbf60c4e72
Reviewed-on: https://go-review.googlesource.com/33170 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Alex Brainman <alex.brainman@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Alex Brainman [Thu, 10 Nov 2016 02:00:35 +0000 (13:00 +1100)]
os: add more tests in TestReadStdin
TestReadStdin always fill up buffer provided by ReadFile caller full.
But we do not know if real ReadFile does the same. Add tests where
buffer is only filled with limited data.
Change-Id: I0fc776325c2b1fe60511126c439f4b0560e9d653
Reviewed-on: https://go-review.googlesource.com/33030 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Mikio Hara [Thu, 17 Nov 2016 03:18:39 +0000 (12:18 +0900)]
net: tweak comment on ExampleCIDRMask
CIDRMask just returns a mask which corresponds to an address
prefix in CIDR nonation. A subnet for an IPv6 mask sounds a bit
confusing.
Change-Id: Ic7859ce992bc2de4043d3b25caf9a1051d118b0e
Reviewed-on: https://go-review.googlesource.com/33262 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Lynn Boger [Wed, 9 Nov 2016 15:09:37 +0000 (09:09 -0600)]
runtime: handle bad ftab index in symtab.go
If a program has had its text section split into multiple
sections then the ftab that is built is based on addresses
prior to splitting. That means all the function addresses
are there and correct because of relocation but the
but the computed idx won't always match up quite right and
in some cases go beyond the end of the table, causing a panic.
To resolve this, determine if the idx is too large and if it is,
set it to the last index in ftab. Then search backward to find the
matching function address.
Fixes #17854
Change-Id: I6940e76a5238727b0a9ac23dc80000996db2579a
Reviewed-on: https://go-review.googlesource.com/32972 Reviewed-by: David Chase <drchase@google.com>
Ian Lance Taylor [Thu, 17 Nov 2016 00:13:22 +0000 (16:13 -0800)]
cmd/compile, reflect: use field pkgPath if needed
It's possible for the pkgPath of a field to be different than that of
the struct type as a whole. In that case, store the field's pkgPath in
the name field. Use the field's pkgPath when setting PkgPath and when
checking for type identity.
Fixes #17952.
Change-Id: Iebaf92f0054b11427c8f6e4158c3bebcfff06f45
Reviewed-on: https://go-review.googlesource.com/33333
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Crawshaw <crawshaw@golang.org>
Daniel Martí [Wed, 16 Nov 2016 22:39:25 +0000 (22:39 +0000)]
cmd/cover: don't ignore os.Create error
Failing to create the output file would give confusing errors such as:
cover: invalid argument
Also do out.Close() even if Execute() errored.
Fixes #17951.
Change-Id: I897e1d31f7996871c54fde7cb09614cafbf6c3fc
Reviewed-on: https://go-review.googlesource.com/33278 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Scott Bell [Wed, 18 May 2016 16:56:51 +0000 (09:56 -0700)]
database/sql: additional underlying types in DefaultValueConverter
The previous documentation purported to convert underlying strings to
[]byte, which it did not do. This adds support for underlying bool,
string, and []byte, which convert directly to their underlying type.
Fixes #15174.
Change-Id: I7fc4e2520577f097a48f39c9ff6c8160fdfb7be4
Reviewed-on: https://go-review.googlesource.com/27812 Reviewed-by: Daniel Theophanes <kardianos@gmail.com> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Kevin Burke [Thu, 3 Nov 2016 22:34:05 +0000 (15:34 -0700)]
net: add example for CIDRMask
I had trouble translating the documentation language into a subnet
- e.g. whether /31 was CIDRMask(1, 31) or CIDRMask(1, 32) or
CIDRMask(31, 32) so I thought I'd add a short example showing how to
create the right masks.
Change-Id: Ia6a6de08c5c30b6d2249b3194cced2d3c383e317
Reviewed-on: https://go-review.googlesource.com/32677 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
David Chase [Mon, 14 Nov 2016 23:00:17 +0000 (18:00 -0500)]
cmd/compile: ensure necessary types appear in .debug_info
Autotmp filtering was too aggressive and excluded types
necessary to make debuggers work properly. Restore the
"late filter" in dwarf.go based on names to exclude autotmps,
and remove the "early filter" in pgen.go based on how the
name was introduced. However, the updated naming scheme
with a dot prefix is retained to prevent accidental clashes
with legal Go identifier names.
Includes test (grouped with runtime gdb tests),
verified to fail without the fix.
Updates #17644.
Fixes #17830.
Change-Id: I7ec3f7230083889660236e5f6bc77ba5fe434e93
Reviewed-on: https://go-review.googlesource.com/33233
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Mikio Hara [Wed, 16 Nov 2016 07:34:02 +0000 (16:34 +0900)]
net: deflake TestAcceptTimeout
This change makes use of synchronization primitive instead of
context-based canceling not to depend on defer execution scheduling.
Fixes #17927.
Change-Id: I5ca9287a48bb5cdda6845a7f12757f95175c5db8
Reviewed-on: https://go-review.googlesource.com/33257 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Daniel Martí [Wed, 16 Nov 2016 15:26:41 +0000 (15:26 +0000)]
all: call flag.Parse from TestMain only if used
These don't use any flags in TestMain itself, so the call is redundant
as M.Run will do it.
Change-Id: I00f2ac7f846dc2c3ad3535eb8177616b2d900149
Reviewed-on: https://go-review.googlesource.com/33275 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
David Crawshaw [Mon, 14 Nov 2016 02:20:58 +0000 (21:20 -0500)]
cmd/link: handle R_GOTPCREL separately on darwin
To generate the correct section offset the shared code path for
R_CALL, R_PCREL, and R_GOTPCREL on darwin when externally linking
walks up the symbol heirarchy adding the differences. This is fine,
except in the case where we are generating a GOT lookup, because
the topmost symbol is left in r.Xsym instead of the symbol we are
looking up. So all funcsym GOT lookups were looking up the outer
"go.func.*" symbol.
Fix this by separating out the R_GOTPCREL code path.
For #17828 (and may fix it).
Change-Id: I2c9f4d135e77c17270aa064d8c876dc6d485d659
Reviewed-on: https://go-review.googlesource.com/33211
Run-TryBot: David Crawshaw <crawshaw@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Mikio Hara [Wed, 16 Nov 2016 04:35:59 +0000 (13:35 +0900)]
net: don't run TestTCPBig unconditionally
The test requires tons of memory and results various failures, mainly
runtime errors and process termination by SIGKILL, caused by resource
exhaustion when the node under test doesn't have much resources.
This change makes use of -tcpbig flag to enable the test.
Change-Id: Id53fa5d88543e2e60ca9bb4f55a1914ccca844e1
Reviewed-on: https://go-review.googlesource.com/33254 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Mikio Hara <mikioh.mikioh@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Bryan C. Mills [Wed, 9 Nov 2016 20:28:24 +0000 (15:28 -0500)]
runtime/cgo: use libc for sigaction syscalls when possible
This ensures that runtime's signal handlers pass through the TSAN and
MSAN libc interceptors and subsequent calls to the intercepted
sigaction function from C will correctly see them.
Fixes #17753.
Change-Id: I9798bb50291a4b8fa20caa39c02a4465ec40bb8d
Reviewed-on: https://go-review.googlesource.com/33142 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Mikio Hara [Wed, 16 Nov 2016 04:51:45 +0000 (13:51 +0900)]
net/http: fix a typo in test
Change-Id: I897237667ffe9e9b2a5f92251a6f665d29479fd2
Reviewed-on: https://go-review.googlesource.com/33255 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Ian Lance Taylor [Wed, 16 Nov 2016 01:55:28 +0000 (17:55 -0800)]
os/exec: add closeOnce.WriteString method
Add an explicit WriteString method to closeOnce that acquires the
writers lock. This overrides the one promoted from the
embedded *os.File field. The promoted one naturally does not acquire
the lock, and can therefore race with the Close method.
Fixes #17647.
Change-Id: I3460f2a0d503449481cfb2fd4628b4855ab0ecdf
Reviewed-on: https://go-review.googlesource.com/33298
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Ian Lance Taylor [Mon, 14 Nov 2016 23:51:30 +0000 (15:51 -0800)]
cmd/cgo: run cgo pointer checks for pointer to union
If a C union type (or a C++ class type) can contain a pointer field,
then run the cgo checks on pointers to that type. This will test the
pointer as though it were an unsafe.Pointer, and will crash if it points
to Go memory that contains a pointer.
Fixes #15942.
Change-Id: Ic2d07ed9648d4b27078ae7683e26196bcbc59fc9
Reviewed-on: https://go-review.googlesource.com/33237
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
David Crawshaw [Sat, 12 Nov 2016 11:24:36 +0000 (06:24 -0500)]
cmd/link, runtime, plugin: versioning
In plugins and every program that opens a plugin, include a hash of
every imported package.
There are two versions of each hash: one local and one exported.
As the program starts and plugins are loaded, the first exported
symbol for each package becomes the canonical version.
Any subsequent plugin's local package hash symbol has to match the
canonical version.
Fixes #17832
Change-Id: I4e62c8e1729d322e14b1673bada40fa7a74ea8bc
Reviewed-on: https://go-review.googlesource.com/33161 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Ian Lance Taylor [Tue, 15 Nov 2016 00:50:15 +0000 (16:50 -0800)]
misc/cgo/testcarchive, misc/cgo/testcshared: sleep instead of sched_yield
Apparently when GOMAXPROCS == 1 a simple sched_yield in a tight loop is
not necessarily sufficient to permit a signal handler to run. Instead,
sleep for 1/1000 of a second.
Fixes #16649.
Change-Id: I83910144228556e742b7a92a441732ef61aa49d9
Reviewed-on: https://go-review.googlesource.com/33239
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Caleb Spare [Mon, 14 Nov 2016 02:06:16 +0000 (18:06 -0800)]
html/template: fix multiple Clones of redefined template
This change redoes the fix for #16101 (CL 31092) in a different way by
making t.Clone return the template associated with the t.Name() while
allowing for the case that a template of the same name is define-d.
Mikio Hara [Mon, 14 Nov 2016 00:17:15 +0000 (09:17 +0900)]
os: gofmt -w -s
Change-Id: I9a42cb55544185ade20b2a4a9de5d39a6cfc6fc6
Reviewed-on: https://go-review.googlesource.com/33172 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Jesse Szwedko [Sun, 13 Nov 2016 21:29:19 +0000 (13:29 -0800)]
syscall: Clearenv now unsets env vars on Windows
Previously, `os.Clearenv()` (by way of `syscall.Clearenv`) would simply
set all environment variables' values to `""` rather than actually
unsetting them causing subsequent `os.LookupEnv` calls to return that
they were still set.
net/error_test.go:254: unrecognized printf flag for verb 'T': '#'
os/os_test.go:1067: arg mt for printf verb %d of wrong type: time.Time
runtime/debug/garbage_test.go:83: arg dt for printf verb %d of wrong type: time.Time
Brad Fitzpatrick [Fri, 11 Nov 2016 23:40:25 +0000 (23:40 +0000)]
context: document appropriate WithValue key type more
Fixes #17826
Updates #17302
Change-Id: I7c1ebd965e679e7169a97e62d27ae3ede2473aa1
Reviewed-on: https://go-review.googlesource.com/33152 Reviewed-by: Ian Lance Taylor <iant@golang.org>
David du Colombier [Fri, 11 Nov 2016 19:49:11 +0000 (20:49 +0100)]
net: implement asynchonous cancelable I/O on Plan 9
This change is an experimental implementation of asynchronous
cancelable I/O operations on Plan 9, which are required to
implement deadlines.
There are no asynchronous syscalls on Plan 9. I/O operations
are performed with blocking pread and pwrite syscalls.
Implementing deadlines in Go requires a way to interrupt
I/O operations.
It is possible to interrupt reads and writes on a TCP connection
by forcing the closure of the TCP connection. This approach
has been used successfully in CL 31390.
However, we can't implement deadlines with this method, since
we require to be able to reuse the connection after the timeout.
On Plan 9, I/O operations are interrupted when the process
receives a note. We can rely on this behavior to implement
a more generic approach.
When doing an I/O operation (read or write), we start the I/O in
its own process, then wait for the result asynchronously. The
process is able to handle the "hangup" note. When receiving the
"hangup" note, the currently running I/O operation is canceled
and the process returns.
This way, deadlines can be implemented by sending an "hangup"
note to the process running the blocking I/O operation, after
the expiration of a timer.
Quentin Smith [Thu, 10 Nov 2016 16:45:56 +0000 (11:45 -0500)]
cmd/go: skip TestCgoPkgConfig if pkg-config is too old
pkg-config 0.24 adds support for quoting and escaping whitespace;
distros like CentOS 6 are still shipping pkg-config 0.23. Skip the test
there since there's no way to get whitespace into the pkg-config output.
Rhys Hiltner [Tue, 25 Oct 2016 18:56:29 +0000 (11:56 -0700)]
runtime: include pre-panic/throw logs in core dumps
When a Go program crashes with GOTRACEBACK=crash, the OS creates a
core dump. Include the text-formatted output of some of the cause of
that crash in the core dump.
Output printed by the runtime before crashing is maintained in a
circular buffer to allow access to messages that may be printed
immediately before calling runtime.throw.
The stack traces printed by the runtime as it crashes are not stored.
The information required to recreate them should be included in the
core file.
Updates #16893
There are no tests covering the generation of core dumps; this change
has not added any.
This adds (reentrant) locking to runtime.gwrite, which may have an
undesired performance impact.
Brad Fitzpatrick [Fri, 11 Nov 2016 20:35:26 +0000 (20:35 +0000)]
net/smtp: make Client.Auth trim final space if Auth.Start toServer is empty
Users can implement the smtp.Auth interface and return zero bytes in
the "toServer []byte" return value from the Auth.Start method. People
apparently do this to implement the SMTP "LOGIN" method.
But we were then sending "AUTH LOGIN \r\n" to the server, which some
servers apparently choke on. So, trim it when the toServer value is
empty.
Fixes #17794
Change-Id: I83662dba9e0f61b1c5000396c096cf7110f78361
Reviewed-on: https://go-review.googlesource.com/33143
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
func sigprof(pc, sp, lr uintptr, gp *g, mp *m) {
if prof.hz == 0 {
return
}
// Profiling runs concurrently with GC, so it must not allocate.
mp.mallocing++
... lots of code ...
mp.mallocing--
}
A borrowed M may migrate between threads. Between the
atomic.Loaduintptr(&mp.thread) and the SuspendThread, mp may have
moved to a new thread, so that it's in active use. In particular
it might be calling malloc, as in the crash stack trace. If so, the
mp.mallocing++ in sigprof would provoke the crash.
Those lines are trying to guard against allocation during sigprof.
But on Windows, mp is the thread being traced, not the current
thread. Those lines should really be using getg().m.mallocing, which
is the same on Unix but not on Windows. With that change, it's
possible the race on the actual thread is not a problem: the traceback
would get confused and eventually return an error, but that's fine.
The code expects that possibility.
Fixes #17165.
Change-Id: If6619731910d65ca4b1a6e7de761fa2518ef339e
Reviewed-on: https://go-review.googlesource.com/33132
Run-TryBot: Russ Cox <rsc@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Bill O'Farrell [Sun, 30 Oct 2016 04:11:37 +0000 (00:11 -0400)]
math: use SIMD to accelerate some scalar math functions on s390x
Note, most math functions are structured to use stubs, so that they can
be accelerated with assembly on any platform.
Sinh, cosh, and tanh were not structued with stubs, so this CL does
that. This set of routines was chosen as likely to produce good speedups
with assembly on any platform.
Technique used was minimax polynomial approximation using tables of
polynomial coefficients, with argument range reduction.
A table of scaling factors was also used for cosh and log10.
Accuracy was tested against a high precision
reference function to determine maximum error.
Approximately 4,000,000 points were tested for each function,
producing the following result.
Note: ulperr is error in "units in the last place"
max
ulperr
sin 1.43 (returns NaN beyond +-2^50)
cos 1.79 (returns NaN beyond +-2^50)
cosh 1.05
sinh 3.02
tanh 3.69
log10 1.75
Also includes a set of tests to test non-vector functions even
when SIMD is enabled
Change-Id: Icb45f14d00864ee19ed973d209c3af21e4df4edc
Reviewed-on: https://go-review.googlesource.com/32352
Run-TryBot: Michael Munday <munday@ca.ibm.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Michael Munday <munday@ca.ibm.com>
Brad Fitzpatrick [Fri, 11 Nov 2016 18:43:39 +0000 (18:43 +0000)]
net/http: make Server respect shutdown state after handler finishes
If the Server's Shutdown (or SetKeepAlivesEnabled) method was called
while a connection was in a Handler, but after the headers had been
written, the connection was not later closed.
Fixes #9478
Updates #17754 (reverts that workaround)
Change-Id: I65324ab8217373fbb38e12e2b8bffd0a91806072
Reviewed-on: https://go-review.googlesource.com/33141 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Russ Cox [Fri, 11 Nov 2016 16:54:18 +0000 (11:54 -0500)]
text/template: reintroduce implicit indirect of interface values in builtin funcs
CL 31462 made it possible to operate directly on reflect.Values
instead of always forcing a round trip to interface{} and back.
The round trip was losing addressability, which hurt users.
The round trip was also losing "interface-ness", which helped users.
That is, using reflect.ValueOf(v.Interface()) instead of v was doing
an implicit indirect any time v was itself an interface{} value: the result
was the reflect.Value for the underlying concrete value contained in the
interface, not the interface itself.
CL 31462 eliminated some "unnecessary" reflect.Value round trips
in order to preserve addressability, but in doing so it lost this implicit
indirection. This CL adds the indirection back.
It may help to compare the changes in this CL against funcs.go from CL 31462:
https://go-review.googlesource.com/#/c/31462/4/src/text/template/funcs.go
Everywhere CL 31462 changed 'v := reflect.ValueOf(x)' to 'v := x',
this CL changes 'v := x' to 'v := indirectInterface(x)'.