The default behaviour for fatal errors and runtime panics is to dump
the goroutine stack traces and exit with code 2. However, when the process is
owned by foreign code, it is suprising and inappropriate to suddenly exit
the whole process, even on fatal errors. Instead, re-use the crash behaviour
from GOTRACEBACK=crash and abort.

The motivating use case is issue #11382, where an Android crash reporter
is confused by an exiting process, but I believe the aborting behaviour
is appropriate for all cases where Go does not own the process.

The change is simple and contained and will enable reliable crash reporting
for Android apps in Go 1.5, but I'll leave it to others to judge whether it
is too late for Go 1.5.

Fixes #11382

Change-Id: I477328e1092f483591c99da1fbb8bc4411911785
Reviewed-on: https://go-review.googlesource.com/12032Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

Recent change (CL 10370) unexpectedly broke TestRaiseException on
Windows XP amd64. I still do not know why. But reverting old
CL 8165 fixes the problem.

This effectively makes Windows XP amd64 use AddVectoredContinueHandler
instead of SetUnhandledExceptionFilter for exception handling. That is
what we do for all recent Windows versions too.

Fixes #11481

Change-Id: If2e8037711f05bf97e3c69f5a8d86af67c58f6fc
Reviewed-on: https://go-review.googlesource.com/11888
Run-TryBot: Alex Brainman <alex.brainman@gmail.com>
Reviewed-by: Daniel Theophanes <kardianos@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

Fixes #11442

Change-Id: I2053fe752c6a122924d28565f1338f73e00ed417
Reviewed-on: https://go-review.googlesource.com/11791Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

Instead of silently truncating integers to their expected range, check
that they're within range and emit errors if not.  Intended to help
narrow down the cause of issue #11617.

Change-Id: Ia7b577270f8438ca7479262702371e26277f1ea7
Reviewed-on: https://go-review.googlesource.com/12050Reviewed-by: Ian Lance Taylor <iant@golang.org>

When GOROOT_FINAL is set when running all.bash, the tests are run
before the files are copied to GOROOT_FINAL.  The tests are run with
GOROOT set, so most work fine.  This fixes two cases that do not.

In cmd/go/go_test.go we were explicitly removing GOROOT from the
environment, causing tests that did not themselves explicitly set
GOROOT to fail.  There was no need to explicitly remove GOROOT, so
don't do it.  If people choose to run "go test cmd/go" with a bad
GOROOT, that is their own lookout.

In the runtime GDB test, the linker has told gdb to find the support
script in GOROOT_FINAL, which will fail.  Check for that case, and
skip the test when we see it.

Fixes #11652.

Change-Id: I4d3a32311e3973c30fd8a79551aaeab6789d0451
Reviewed-on: https://go-review.googlesource.com/12021
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Update #11609.

Change-Id: Ie363facf13f5e62f1af4a8bdc42a18fb36e16ebf
Reviewed-on: https://go-review.googlesource.com/12022Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Don't treat IPv4-mapped link-local IP addresses as IPv6 link-local
addresses, an IPv4 broadcast address as a global unicast IP address.

Fixes #11585.

Change-Id: I6a7a0c0601f18638f5c624ab63e12ee40f77b182
Reviewed-on: https://go-review.googlesource.com/11883Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Updates #6465.

Change-Id: I5babbcf4c92dae47c823a41628b01e4ceb2332eb
Reviewed-on: https://go-review.googlesource.com/11951
Run-TryBot: Mikio Hara <mikioh.mikioh@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

People use 80-column terminals because their grandparents used
punched cards. When I last used a punched card, in 1978, it seemed
antiquated even then. But today, people still set their terminal
widths to 80 to honor the struggles their fallen ancestors made to
endure this painful technology.

We must all stand and salute the 80 column flag, or risk the opprobium
of our peers.

For Pete's sake, I don't even use a fixed-width font. I don't even
believe in columns.

Fixes #11639 with extreme reluctance.

P.S. To avoid the horror of an automatically folded line of text, this commit message has been formatted to fit on an 80-column line, except for this postscript.

Change-Id: Ia2eb2dcf293dabe804c22ee5abb4bbb703f45c33
Reviewed-on: https://go-review.googlesource.com/12011Reviewed-by: Ian Lance Taylor <iant@golang.org>

Section.Data returns disk section data, but those are rounded up to
some predefined value. Processing these as is confuses dwarf parser
because of garbage at the end. Truncate Section.Data as per memory
section description.

Sometimes dwarf sections have memory section size of 0
(for pe object files). Keep those to their disk size.

Fixes #11608

Change-Id: I8de0a2271201a24aa9ac8dac44f1e9c8a9285183
Reviewed-on: https://go-review.googlesource.com/11950Reviewed-by: Ian Lance Taylor <iant@golang.org>

GODEBUG=netdns=1 prints a one-time strategy decision. (cgo or go DNS lookups)
GODEBUG=netdns=2 prints the per-lookup strategy as a function of the hostname.

The new "netcgo" build tag forces cgo DNS lookups.

GODEBUG=netdns=go (or existing build tag "netgo") forces Go DNS resolution.
GODEBUG=netdns=cgo (or new build tag "netcgo") forces libc DNS resolution.

Options can be combined with e.g. GODEBUG=netdns=go+1 or GODEBUG=netdns=2+cgo.

Fixes #11322
Fixes #11450

Change-Id: I7a67e9f759fd0a02320e7803f9ded1638b19e861
Reviewed-on: https://go-review.googlesource.com/11584Reviewed-by: Russ Cox <rsc@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>

Change-Id: Ide5b0c81405fbf6b83a6f02527d629898e0fcf02
Reviewed-on: https://go-review.googlesource.com/12000Reviewed-by: Ian Lance Taylor <iant@golang.org>

sysmon triggers a GC if there has been no GC for two minutes.
Currently, this is a STW GC. There is no reason for this to be STW, so
make it concurrent.

Fixes #10261.

Change-Id: I92f3ac37272d5c2a31480ff1fa897ebad08775a9
Reviewed-on: https://go-review.googlesource.com/11955Reviewed-by: Rob Pike <r@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>

The architecture-specific details will be updated and expanded in
a subsequent CL (or series thereof).

Update #10096

Change-Id: I59c6be1fcc123fe8626ce2130e6ffe71152c87af
Reviewed-on: https://go-review.googlesource.com/11954Reviewed-by: Russ Cox <rsc@golang.org>

Fix typo.

TBR=rsc

Change-Id: I85d1b46744a9a4524e7949e452cdebb53afe0740
Reviewed-on: https://go-review.googlesource.com/11959Reviewed-by: Rob Pike <r@golang.org>

Change-Id: Ie3d603a95826b9b6a7acd4825991f24c3c61408b
Reviewed-on: https://go-review.googlesource.com/11956Reviewed-by: Russ Cox <rsc@golang.org>

Change the default behavior when showing the package docs
for a command to elide the symbols. This makes

	go doc somecommand

show the top-level package docs only and hide the symbols,
which are probably irrelevant to the user. This has no effect
on explicit requests for internals, such as

	go doc somecommand.sometype

The new -cmd flag restores the old behavior.

Fixes #10733.

Change-Id: I4d363081fe7dabf76ec8e5315770ac3609592f80
Reviewed-on: https://go-review.googlesource.com/11953Reviewed-by: Russ Cox <rsc@golang.org>

They were missing from the inputs.
Unfortunately this means the .out files all have wrong line numbers,
but they are easy to update.

Change-Id: I254742f24ab803421f34d52d13b9afa93674edd6
Reviewed-on: https://go-review.googlesource.com/11958Reviewed-by: Russ Cox <rsc@golang.org>

This was added in commit 8b4278ff.

Change-Id: I95279f2779c2bab2767e34389fb4324900c01e6c
Reviewed-on: https://go-review.googlesource.com/11952Reviewed-by: Rob Pike <r@golang.org>

It was crashing.
This fixes the build for
GO15VENDOREXPERIMENT=1 go test -short runtime

Fixes #11416.

Change-Id: I74a9114cdd8ebafcc9d2a6f40bf500db19c6e825
Reviewed-on: https://go-review.googlesource.com/11964Reviewed-by: Russ Cox <rsc@golang.org>

This avoids both a write barrier and then dynamic initialization
globals of the form

	var x something
	var xp = unsafe.Pointer(&x)

Using static initialization avoids emitting a relocation for &x,
which helps cgo.

Fixes #9411.

Change-Id: I0dbf480859cce6ab57ab805d1b8609c45b48f156
Reviewed-on: https://go-review.googlesource.com/11693Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>

Fixes #10884

Change-Id: I7cab3c96548867612f579d2cd4ec736309787443
Reviewed-on: https://go-review.googlesource.com/11961Reviewed-by: Andrew Gerrand <adg@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

The recent https://golang.org/cl/11810 is reportedly a bit too
aggressive.

Apparently some HTTP requests in the wild do contain both a
Transfer-Encoding along with a bogus Content-Length. Instead of
returning a 400 Bad Request error, we should just ignore the
Content-Length like we did before.

Change-Id: I0001be90d09f8293a34f04691f608342875ff5c4
Reviewed-on: https://go-review.googlesource.com/11962Reviewed-by: Andrew Gerrand <adg@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

The expansion of structure, array, slice, and map literals
does not use the right line number in its introduced assignments
to temporaries, which leads to incorrect line number attribution
for expressions in those literals.

Inlining also incorrectly replaced the line numbers of args to
inlined functions.

This was revealed in CL 9721 because a now-avoided temporary
assignment introduced the correct line number.
I.e. before CL 9721
  "tmp_wrongline := expr"
was transformed to
  "tmp_rightline := expr; tmp_wrongline := tmp_rightline"

Also includes a repair to CL 10334 involving line numbers
where a spurious -1 remained (should have been 0, now is 0).

Fixes #11400.

Change-Id: I3a4687efe463977fa1e2c996606f4d91aaf22722
Reviewed-on: https://go-review.googlesource.com/11730
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Sameer Ajmani <sameer@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>

Basic randomization of goroutine scheduling for -race mode.
It is probably possible to do much better (there's a paper linked
in the issue that I haven't read, for example), but this suffices
to introduce at least some unpredictability into the scheduling order.
The goal here is to have _something_ for Go 1.5, so that we don't
start hitting more of these scheduling order-dependent bugs
if we change the scheduler order again in Go 1.6.

For #11372.

Change-Id: Idf1154123fbd5b7a1ee4d339e93f97635cc2bacb
Reviewed-on: https://go-review.googlesource.com/11795Reviewed-by: Austin Clements <austin@google.com>

Fixes #11409

Change-Id: Ic1610e124b2d8b2b12310fc9538d5078cc7302a0
Reviewed-on: https://go-review.googlesource.com/11316Reviewed-by: Russ Cox <rsc@golang.org>

Fixes #11065.

Change-Id: Idd854facd5fa78c0334f86740f351d404f9a5b2d
Reviewed-on: https://go-review.googlesource.com/11511Reviewed-by: Russ Cox <rsc@golang.org>

Fixes #11414

Change-Id: I45a41b98554f00362d9222e9c68a441dbfc23cb8
Reviewed-on: https://go-review.googlesource.com/11700Reviewed-by: Russ Cox <rsc@golang.org>

Go's continuous build system depends on HTTP trailers for the buildlet
interface.

Andrew rewrote the makerelease tool to work in terms of Go's builder
system (now at x/build/cmd/release), but it previously could only
create GCE-based buildlets, which meant x/build/cmd/release couldn't
build the release for Darwin.

https://golang.org/cl/11901 added support for proxying buildlet
connections via the coordinator, but that exposed the fact that
httputil.ReverseProxy couldn't proxy Trailers. A fork of that code
also wasn't possible because net/http needlessly deleted the "Trailer"
response header in the Transport code.  This mistake goes back to
"release-branch.r56" and earlier but was never noticed because nobody
ever uses Trailers, and servers via ResponseWriter never had the
ability to even set trailers before this Go 1.5. Note that setting
trailers requires pre-declaring (in the response header) which
trailers you'll set later (after the response body). Because you could
never set them, before this release you could also never proxy them.

Change-Id: I2410a099921790dcd391675ae8610300efa19108
Reviewed-on: https://go-review.googlesource.com/11940Reviewed-by: Andrew Gerrand <adg@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>

Don't know why, but git deleted the previous version of this change.
This is the same change as https://go-review.googlesource.com/11884,
which I will now abandon, with a couple of fixes.

Almost all done now. Could use help with the TODOs.

Major missing piece is the trace command. Vendoring
section is also weak, but it's also undocumented elsewhere.

Change-Id: I5d8556b23aa6628eb7bf0e330d4dd8d4ac2157c5
Reviewed-on: https://go-review.googlesource.com/11887Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Change-Id: If33ffe1da23a8463fc5479f453422d280b736372
Reviewed-on: https://go-review.googlesource.com/11865Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Don't talk about commands that no longer exist.

There are still references throughout the tree, mostly in comments,
but they provide a charming historical backdrop for the idle tourist.

Change-Id: I637ebdce05bbc7df5addcc46cb772d2bb9f3e073
Reviewed-on: https://go-review.googlesource.com/11885Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Change-Id: I78f0f9ac33db153c4a0d7f9866da20ebcd3d3639
Reviewed-on: https://go-review.googlesource.com/11872Reviewed-by: Rob Pike <r@golang.org>

The second (fallback) draw is a no-op, but it's a non-trivial amount of work.

Fixes #11550.

benchmark               old ns/op     new ns/op     delta
BenchmarkPaletted-4     16301219      7309568       -55.16%

Change-Id: Ic88c537b2b0c710cf517888f3dd15cb702dd142f
Reviewed-on: https://go-review.googlesource.com/11858Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Also add words about the assembler.

Change-Id: I9bd8cc88076f06b0eef36a07f57d1ad5d9261d8d
Reviewed-on: https://go-review.googlesource.com/11853Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Fixes #11508

Change-Id: I72e83893b76f75685d6edfe65ca6691d97539226
Reviewed-on: https://go-review.googlesource.com/11864Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

This reverts commit 6f7961da.

Russ suggests changing the frozon syscall package and obviously it's a
better solution. Perhaps he will also let me know the way how to get the
project owners to agree later.

Fixes #11492.

Change-Id: I98f9f366b72b85db54b4acfc3a604b62fb6d783c
Reviewed-on: https://go-review.googlesource.com/11854
Run-TryBot: Mikio Hara <mikioh.mikioh@gmail.com>
Reviewed-by: Rob Pike <r@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Fixes #10616

Change-Id: I4ef25eb0be6ccf474976fdb5087dd2c62c66c510
Reviewed-on: https://go-review.googlesource.com/11811Reviewed-by: Andrew Gerrand <adg@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>

As per comments in cl/11834.

Change-Id: I285536b882fa9496e15d77d0d4c16ee913aca581
Reviewed-on: https://go-review.googlesource.com/11861Reviewed-by: Daniel Theophanes <kardianos@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Comment change only.

Change-Id: I2e32c2b34d5a5659ead6d6082b06e1b039bf1147
Reviewed-on: https://go-review.googlesource.com/11852Reviewed-by: Russ Cox <rsc@golang.org>