Like AMD64, don't issue NilCheck instruction if the subsequent block
has a load or store at the same address.

Pass test/nilptr3_ssa.go.

Updates #15365.

Change-Id: Ic88780dab8c4893c57d1c95f663760cc185fe51e
Reviewed-on: https://go-review.googlesource.com/24451Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: David Chase <drchase@google.com>

CSE may substitute a tuple generator with another one in a different
block. In this case, since we want tuple selectors to stay together
with the tuple generator, copy the selector to the new generator's
block and rewrite its use.

Op.isTupleGenerator and Op.isTupleSelector are introduced to assert
tuple ops. Use it in tighten as well.

Updates #15365.

Change-Id: Ia9e8c734b9cc3bc9fca4a2750041eef9cdfac5a5
Reviewed-on: https://go-review.googlesource.com/24137Reviewed-by: David Chase <drchase@google.com>

Minor code cleanup. Done as part of understanding
OpARMMOVWaddr, since other architectures will
need to do something similar.

Change-Id: Iea2ecf3defb4f884e63902c369cd55e4647bce7a
Reviewed-on: https://go-review.googlesource.com/24157Reviewed-by: Cherry Zhang <cherryyz@google.com>

This will be used verbatim in other architectures.

Change-Id: I307891ae597d797fd45f296b6a38ffe9fac6b975
Reviewed-on: https://go-review.googlesource.com/24155
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

Provide better diagnostic messages.

Use an int for numRegs comparisons,
to avoid asking whether a uint8 is > 255.

Change-Id: I33ae193ce292b24b369865abda3902c3207d7d3f
Reviewed-on: https://go-review.googlesource.com/24135Reviewed-by: Keith Randall <khr@golang.org>

This will be needed by other architectures as well.
Put a cleaner encapsulation around it.

Change-Id: I0ac25d600378042b2233301678e9d037e20701d8
Reviewed-on: https://go-review.googlesource.com/24154
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Cherry Zhang <cherryyz@google.com>

Use hardware g register (R10) for GetG, allow g to appear at LHS of
some ops.

Progress on SSA backend for ARM. Now everything compiles and runs.

Updates #15365.

Change-Id: Icdf93585579faa86cc29b1e17ab7c90f0119fc4e
Reviewed-on: https://go-review.googlesource.com/23952Reviewed-by: David Chase <drchase@google.com>

Change-Id: Idd150294aaeced0176b53d6b95852f5d21ff4fdc

I don't see how the call could fail, so, no test. Just a code cleanup in
case it can fail in the future.

Fixes #15987.

Change-Id: If4af0d5e7d19cc8b13fb6a4f7661c37fb0015e83
Reviewed-on: https://go-review.googlesource.com/23860
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Mikio Hara <mikioh.mikioh@gmail.com>

A .syso file may include information that should go into the object file
that is not object code, and should be included even if not using cgo.
The example in the issue is a Windows manifest file.

Fixes #16050.

Change-Id: I1f4f3f80bb007e84d153ca2d26e5919213ea4f8d
Reviewed-on: https://go-review.googlesource.com/24032
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Alex Brainman <alex.brainman@gmail.com>

The dir parameter was removed in https://golang.org/cl/5732045.

Fixes #15503.

Change-Id: I02a6d8317233bea08633715a095ea2514822032b
Reviewed-on: https://go-review.googlesource.com/24011
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <shurcool@gmail.com>
Reviewed-by: Andrew Gerrand <adg@golang.org>

Change-Id: I21fad94ff94e342ada18e0e41ca90296d030115f
Reviewed-on: https://go-review.googlesource.com/24061
Run-TryBot: Mikio Hara <mikioh.mikioh@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

This change is a followup to https://go-review.googlesource.com/23967
for FreeBSD.

Updates #13987.
Updates #16028.

Change-Id: I0f0737372fce6df89d090fe9847305749b79eb4c
Reviewed-on: https://go-review.googlesource.com/24021Reviewed-by: Ian Lance Taylor <iant@golang.org>

This change is a followup to https://go-review.googlesource.com/23967
for Darwin.

Updates #13987.
Updates #16028.

Change-Id: Ib1fb9f957fafd0f91da6fceea56620e29ad82b00
Reviewed-on: https://go-review.googlesource.com/24020Reviewed-by: Ian Lance Taylor <iant@golang.org>

Fixes #15994.

Change-Id: I5aca91ab53985ac7dcb07ce094ec15eb8ec341f8
Reviewed-on: https://go-review.googlesource.com/23891
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

Functions should be declared to end after the last real instruction, not
after the last padding byte. We achieve this by adding the padding while
assembling the text section in the linker instead of adding the padding
to the function symbol in the compiler. This change makes dtrace happy.

TODO: check that this works with external linking

Fixes #15969

Change-Id: I973e478d0cd34b61be1ddc55410552cbd645ad62
Reviewed-on: https://go-review.googlesource.com/24040
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

Introduce an op MOVWaddr for addresses on ARM, instead of overuse
ADDconst.

Mark MOVWaddr as rematerializable. This fixes a liveness problem: if
it were not rematerializable, the address of a variable may be spilled
and later use of the address may just load the spilled value without
mentioning the variable, and the liveness code may think it is dead
prematurely.

Update #15365.

Change-Id: Ib0b0fa826bdb75c9e6bb362b95c6cf132cc6b1c0
Reviewed-on: https://go-review.googlesource.com/23942Reviewed-by: David Chase <drchase@google.com>

SSA treats SP as constant throughout a function, so as OffPtr [off] SP.
When the stack moves, spilled OffPtr values become invalid, if they are
not pointer-typed.

(Currently it is fine because of the optimization rules that folds OffPtr
into Load/Store. But it'd better be "optimization", not requirement.)

Updates #15365.

Change-Id: I76cf4008dfdc169e1cb5a55a2605b6678efc915d
Reviewed-on: https://go-review.googlesource.com/23941
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>

Special case for rewriting OAS inits omitted OASWB, added
that and OAS2FUNC.  The special case cannot be default case,
that causes racewalk to fail in horrible ways.

Fixes #16008.

Change-Id: Ie0d2f5735fe9d8255a109597b36d196d4f86703a
Reviewed-on: https://go-review.googlesource.com/23954Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Dmitry Vyukov <dvyukov@google.com>
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

Generate package comment in alldocs.go using line comments rather than
general comments. This scales better, general comments cannot contain the
"*/" character sequence. Line comments do not have any restrictions on
the comment text that can be contained.

Remove the dependency on sed, which is not cross-platform, not go-gettable
external command.

Remove trailing whitespace from usage string in test.go. It's unnecessary.

Fixes #16030.

Change-Id: I3c0bc9955e7c7603c3d1fb4878218b0719d02e04
Reviewed-on: https://go-review.googlesource.com/23968Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

Instead of doing:

x = input
one round of aes on x
x ^= seed
two rounds of aes on x

Do:

x = input
x ^= seed
three rounds of aes on x

This change provides some additional seed-dependent scrambling
which should help prevent collisions.

Change-Id: I02c774d09c2eb6917cf861513816a1024a9b65d7
Reviewed-on: https://go-review.googlesource.com/23577Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

On systems that support the POSIX.1-2008 waitid function, we can use it
to block until a wait will succeed. This avoids a possible race
condition: if a program calls p.Kill/p.Signal and p.Wait from two
different goroutines, then it is possible for the wait to complete just
before the signal is sent. In that case, it is possible that the system
will start a new process using the same PID between the wait and the
signal, causing the signal to be sent to the wrong process. The
Process.isdone field attempts to avoid that race, but there is a small
gap of time between when wait returns and isdone is set when the race
can occur.

This CL avoids that race by using waitid to wait until the process has
exited without actually collecting the PID. Then it sets isdone, then
waits for any active signals to complete, and only then collects the PID.

No test because any plausible test would require starting enough
processes to recycle all the process IDs.

Update #13987.
Update #16028.

Change-Id: Id2939431991d3b355dfb22f08793585fc0568ce8
Reviewed-on: https://go-review.googlesource.com/23967
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

+ panic with explicit error if no file set it provided

(Not providing a file set is invalid use of the API; panic
is the appropriate action rather than returning an error.)

Fixes #16018.

Change-Id: I207f5b2a2e318d65826bdd9522fce46d614c24ee
Reviewed-on: https://go-review.googlesource.com/24010
Run-TryBot: Robert Griesemer <gri@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

Make the documentation for `go get` match the documentation for `go
install`, since `go get` essentially invokes `go install`.

Update #15825.

Change-Id: I374d80efd301814b6d98b86b7a4a68dd09704c92
Reviewed-on: https://go-review.googlesource.com/23925Reviewed-by: Andrew Gerrand <adg@golang.org>

Fixes #16021.

Change-Id: I55df38bbccd2641abcb54704115002a9aa04325d
Reviewed-on: https://go-review.googlesource.com/23962
Run-TryBot: Michael Munday <munday@ca.ibm.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

This patch updates the doc about comments whitespace for the
encoding/csv package to reflect that leading whitespace before
the hash will treat the line as not a comment.

Fixes #13775.

Change-Id: Ia468c75b242a487b4b2b4cd3d342bfb8e07720ba
Reviewed-on: https://go-review.googlesource.com/23302Reviewed-by: Ian Lance Taylor <iant@golang.org>

When setting $pc, gdb does a backtrace using the current value of $sp,
and it may complain if $sp does not match that $pc (although the
assignment went through successfully).

This happens with ARM SSA backend: when setting $pc it prints
> Cannot access memory at address 0x0

As well as occasionally on MIPS64:
> warning: GDB can't find the start of the function at 0xc82003fe07.
> ...

Setting $sp before setting $pc makes it happy.

Change-Id: Idd96dbef3e9b698829da553c6d71d5b4c6d492db
Reviewed-on: https://go-review.googlesource.com/23940Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Austin Clements <austin@google.com>

The generated code for interface stubs sometimes just messes
with a few of the args and then tail-calls to the target routine.
The args that aren't explicitly modified appear to not be used.
But they are used, by the thing we're tail calling.

Fixes #16016

Change-Id: Ib9b3a8311bb714a201daee002885fcb59e0463fa
Reviewed-on: https://go-review.googlesource.com/23960
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>

The combination of https://golang.org/cl/23650 and
https://golang.org/cl/23675 did not work--they were tested separately
but not together.

The problem was that 23650 introduced deferred argument checking, and
the deferred function loses the type that 23675 started requiring. The
fix is to go back to using an empty interface type in a deferred
argument check.

No new test required--fixes broken build.

Change-Id: I5ea023c5aed71d70e57b11c4551242d3ef25986d
Reviewed-on: https://go-review.googlesource.com/23961
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>

When cgo writes a _cgoCheckPointerN function to handle unsafe.Pointer,
use the function's argument type rather than interface{}. This permits
type errors to be detected at build time rather than run time.

Fixes #13830.

Change-Id: Ic7090905e16b977e2379670e0f83640dc192b565
Reviewed-on: https://go-review.googlesource.com/23675
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>

Fixes #14505.

Change-Id: I46196b26c9339609e6e3ef9159de38c5b50c2a1b
Reviewed-on: https://go-review.googlesource.com/23922
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Emmanuel Odeke <emm.odeke@gmail.com>
Reviewed-by: Andrew Gerrand <adg@golang.org>

getAbbrs looks like it is checking each month looking for a change
in the time zone abbreviation, but starts in Dec of the previous year
and skips the month of February because of the overflow rules for
AddDate. Changing the day to 1 starts at Jan 1 and tries all months
in the current year. This isn't very important or likely to change
output as zones usually span several months. Discovered when
looking into time.AddDate behavior when adding months.

Change-Id: I685254c8d21c402ba82cc4176e9a86b64ce8f7f7
Reviewed-on: https://go-review.googlesource.com/23322Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

Change-Id: I68d66fccf9cc8f7137c92b94820ce7d6f478a185
Reviewed-on: https://go-review.googlesource.com/23310Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

Fixes #15970.

Change-Id: I3f7d8316069a69d0e3859aaa96bc1414487fead0
Reviewed-on: https://go-review.googlesource.com/23921
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Andrew Gerrand <adg@golang.org>

The test TestGoGetHTTPS404 downloads a package that does not build on
every OS, so change it to only run where the package builds. It's not
great for the test to depend on an external package, but this is an
improvement on the current situation.

Fixes #15644.

Change-Id: I1679cee5ab1e61a5b26f4ad39dc8a397fbc0da69
Reviewed-on: https://go-review.googlesource.com/23920
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Andrew Gerrand <adg@golang.org>

- 64x signed right shift was wrong for shift larger than 0x80000000.
- for Lsh-followed-by-Rsh, the intermediate value should be full int
  width, so when it is spilled MOVW should be used.
- use RET for RetJmp, so the assembler can take case of restoring LR
  for non-leaf case.
- reserve R9 in dynlink mode. R9 is used for GOT by the assembler.

Progress on SSA backend for ARM. Still not complete.

Updates #15365.

Change-Id: I3caca256b92ff7cf96469da2feaf4868a592efc5
Reviewed-on: https://go-review.googlesource.com/23793Reviewed-by: David Chase <drchase@google.com>

We want tuple-reading ops immediately follow tuple-generating op, so
that tuple values will not be spilled/copied.

The mechanism introduced in the previous CL cannot really avoid tuples
interleaving. In this CL we always emit tuple and their selectors together.
Maybe remove the tuple scores if it does not help on performance (todo).

Also let tighten not move tuple-reading ops across blocks.

In the previous CL a special case of regenerating flags with tuple-reading
pseudo-op is added, but it did not cover end-of-block case. This is fixed
in this CL and the condition is generalized.

Progress on SSA backend for ARM. Still not complete.

Updates #15365.

Change-Id: I8980b34e7a64eb98153540e9e19a3782e20406ff
Reviewed-on: https://go-review.googlesource.com/23792Reviewed-by: David Chase <drchase@google.com>

This matches the behavior of the legacy backend.

Fixes #15975 (if this is the intended behavior)

Change-Id: Id277959069b8b8bf9958fa8f2cbc762c752a1a19
Reviewed-on: https://go-review.googlesource.com/23820Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

R13 needs to be set to g because C code may have clobbered R13.

Fixes #16006.

Change-Id: I66311fe28440e85e589a1695fa1c42416583b4c6
Reviewed-on: https://go-review.googlesource.com/23910Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

Before, the Android exec wrapper expected the trailing exit code
output on its own line, like this:

PASS
exitcode=0

However, some tests can sometimes squeeze in some output after
the test harness outputs "PASS" and the newline. The
TestWriteHeapDumpFinalizers test is particularly prone to this,
since its finalizers println to standard out. When it happens, the
output looks like this:

PASS
finalizedexitcode=0

Two recent failures caused by this race:

https://build.golang.org/log/185605e1b936142c22350eef22d20e982be53c29
https://build.golang.org/log/e61cf6a050551d10360bd90be3c5f58c3eb07605

Since the "exitcode=" string is always echoed after the test output,
the fix is simple: instead of looking for the last newline in the
output, look for the last exitcode string instead.

Change-Id: Icd6e53855eeba60b982ad3108289d92549328b86
Reviewed-on: https://go-review.googlesource.com/23750
Run-TryBot: Elias Naur <elias.naur@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>