R=r
CC=golang-dev
https://golang.org/cl/6612066

- Changed Check signature to take function parameters for
  more flexibility: Now a client can interrupt type checking
  early (via panic in one the upcalls) once the desired
  type information or number of errors is reached. Default
  use is still simple.

- Cleaned up main typechecking loops. Now does not neglect
  _ declarations anymore.

- Various other cleanups.

R=golang-dev, r, rsc
CC=golang-dev
https://golang.org/cl/6612049

Still lots of pieces missing, but basic framework working.
Lots of tests.

R=rsc
CC=golang-dev
https://golang.org/cl/6594054

This code has been reviewed before. The most significant
change is to check_test which now can handle more than
one error at the same error position (due to spurious
errors - should not happen in praxis once error handling
has been fine-tuned). This change makes check_test easier
to use during development.

R=rsc
CC=golang-dev
https://golang.org/cl/6584057

Also: apply gofmt -w -s to existing tables.

R=mpvl, minux.ma, rsc
CC=golang-dev
https://golang.org/cl/6611051

Bug introduced with CL 6624047.

R=r
CC=golang-dev
https://golang.org/cl/6620073

No JavaScript in HTML anymore.

R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6619066

R=rsc, bradfitz
CC=golang-dev
https://golang.org/cl/6614064

Address several instances of unneeded temporaries when using gcmp.

func M(m map[int]bool) int {
        return len(m)
}

--- prog list "M" ---
0000 (/home/dfc/src/map.go:3) TEXT      M+0(SB),$0-8
0001 (/home/dfc/src/map.go:4) MOVW      m+0(FP),R0
0002 (/home/dfc/src/map.go:4) MOVW      $0,R1
0003 (/home/dfc/src/map.go:4) CMP       R1,R0,
0004 (/home/dfc/src/map.go:4) BEQ       ,6(APC)
0005 (/home/dfc/src/map.go:4) MOVW      0(R0),R0
0006 (/home/dfc/src/map.go:4) MOVW      R0,.noname+4(FP)
0007 (/home/dfc/src/map.go:4) RET       ,

after:

--- prog list "M" ---
0000 (/home/dfc/src/map.go:3) TEXT      M+0(SB),$0-8
0001 (/home/dfc/src/map.go:4) MOVW      m+0(FP),R0
0002 (/home/dfc/src/map.go:4) CMP       $0,R0,
0003 (/home/dfc/src/map.go:4) BEQ       ,5(APC)
0004 (/home/dfc/src/map.go:4) MOVW      0(R0),R0
0005 (/home/dfc/src/map.go:4) MOVW      R0,.noname+4(FP)
0006 (/home/dfc/src/map.go:4) RET       ,

func C(c chan int) int {
        return cap(c)
}

--- prog list "C" ---
0000 (/home/dfc/src/map.go:3) TEXT      C+0(SB),$0-8
0001 (/home/dfc/src/map.go:4) MOVW      c+0(FP),R0
0002 (/home/dfc/src/map.go:4) MOVW      $0,R1
0003 (/home/dfc/src/map.go:4) CMP       R1,R0,
0004 (/home/dfc/src/map.go:4) BEQ       ,6(APC)
0005 (/home/dfc/src/map.go:4) MOVW      4(R0),R0
0006 (/home/dfc/src/map.go:4) MOVW      R0,.noname+4(FP)
0007 (/home/dfc/src/map.go:4) RET       ,

after:

--- prog list "C" ---
0000 (/home/dfc/src/map.go:3) TEXT      C+0(SB),$0-8
0001 (/home/dfc/src/map.go:4) MOVW      c+0(FP),R0
0002 (/home/dfc/src/map.go:4) CMP       $0,R0,
0003 (/home/dfc/src/map.go:4) BEQ       ,5(APC)
0004 (/home/dfc/src/map.go:4) MOVW      4(R0),R0
0005 (/home/dfc/src/map.go:4) MOVW      R0,.noname+4(FP)
0006 (/home/dfc/src/map.go:4) RET       ,

R=rsc, minux.ma, remyoudompheng
CC=golang-dev
https://golang.org/cl/6618054

Update #4139.

R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6609051

R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6620071

R=ken
CC=golang-dev
https://golang.org/cl/6614063

R=ken
CC=golang-dev
https://golang.org/cl/6610065

This keeps the BOM runes from causing other errors.

R=golang-dev, remyoudompheng
CC=golang-dev
https://golang.org/cl/6625062

1. R_ARM_CALL can also be used to call a PLT entry
2. add support for R_ARM_PC24 and R_ARM_JUMP24
3. refactor, remove D_PLT32 in favor of D_CALL

Fixes #4006.

R=rsc, dave
CC=fullung, golang-dev
https://golang.org/cl/6622057

        Fixes #4131.

R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6614061

The test is not about type mismatches and it causes
an extra error to appear.

R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6614062

If we're benchmarking 8g, use gcc -m32.
If we're benchmarking 6g, use gcc -m64.

R=golang-dev, bradfitz, minux.ma, remyoudompheng
CC=golang-dev
https://golang.org/cl/6625061

The protection against segfaults does not completely solve
crashes and breaks test/fixedbugs/bug365.go

R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6615058

R=ken2
CC=golang-dev
https://golang.org/cl/6609052

See issue 3970 (it's already marked as Fixed).

R=rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6624059

This is a part of a bigger change that adds data race detection feature:
https://golang.org/cl/6456044

R=rsc, remyoudompheng
CC=gobot, golang-dev
https://golang.org/cl/6536059

This is a part of a bigger change that adds data race detection feature:
https://golang.org/cl/6456044

R=rsc, minux.ma
CC=gobot, golang-dev
https://golang.org/cl/6529053

This is a part of a bigger change that adds data race detection feature:
https://golang.org/cl/6456044

R=rsc
CC=gobot, golang-dev
https://golang.org/cl/6535050

Also add call to GC() to make it easier to re-enable the test.

Update #4155.
When we have precise GC merged, re-enable this test.

R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6622058

R=golang-dev, daniel.morsing
CC=golang-dev
https://golang.org/cl/6624060

Don't assume the test has a clean environment within /tmp.
Use an actual new tempdir for its tests.

Fixes FreeBSD build failure as seen at:
http://build.golang.org/log/396738676356d7fb6bab6eaf1b97cac820f8a90f

--- FAIL: TestMkdirAllWithSymlink (0.00 seconds)
path_test.go:178:                 Mkdir /tmp/dir: mkdir /tmp/dir: file exists
FAIL
FAIL    os      1.091s

R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6615057

Also adds tests, which didn't exist before.

Fixes #4188

R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6613062

Update #4139.

R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6620070

t.Log("line 1\nline 2\nline 3")

Old output:
=== RUN TestLine3
--- PASS: TestLine3 (0.00 seconds)
testing_test.go:25: 	line 1
		line 2
		line 3
		PASS

New output:
=== RUN TestLine3
--- PASS: TestLine3 (0.00 seconds)
testing_test.go:24: 	line 1
		line 2
		line 3
PASS

R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6613069

R=golang-dev, dave, rsc
CC=golang-dev
https://golang.org/cl/6615055

For issue 3757:
BEFORE:  test/fixedbugs/bug463.go:12: typechecking loop involving a
             test/fixedbugs/bug463.go:12 a
             test/fixedbugs/bug463.go:12 <node DCLCONST>
AFTER:   test/fixedbugs/bug463.go:12: constant definition loop
             test/fixedbugs/bug463.go:12: a uses a

For issue 3937:
BEFORE: test/fixedbugs/bug464.go:12: typechecking loop involving foo
            test/fixedbugs/bug464.go:12 <T>
            test/fixedbugs/bug464.go:12 foo
            test/fixedbugs/bug464.go:12 <node DCLFUNC>
AFTER:  test/fixedbugs/bug464.go:12: foo is not a type

Fixes #3757.
Fixes #3937.

R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6614058

Fixes #4067.

R=golang-dev, minux.ma, dave, rsc
CC=golang-dev
https://golang.org/cl/6622056

decoder struct, inside the unmappedzero limit, to eliminate some
TESTB instructions in the inner decoding loop.

benchmark          old ns/op    new ns/op    delta
BenchmarkDecode      2943204      2746360   -6.69%

R=r, dave
CC=golang-dev
https://golang.org/cl/6625058

R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6610060

The nil dereference in the next few lines doesn't seem
to cause a segmentation fault on Unix, but does seem
to halt the Go compiler.

The following is a test case:

>>>
package main

func mine(int b) int {
        return b + 2
}

func main() {
        mine()

        c = mine()
}
<<<

Without this change only the following is caught:

typecheck.go:3: undefined: b
typecheck.go:4: undefined: b

with it, we catch all the errors:

typecheck.go:3: undefined: b
typecheck.go:4: undefined: b
typecheck.go:10: undefined: c
typecheck.go:10: cannot assign to c .

R=rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6542060

Fixes #4124.

R=golang-dev, dave, minux.ma, remyoudompheng, rsc
CC=golang-dev
https://golang.org/cl/6543057

before

func addr(s[]int) *int {
        return &s[2]
   10c1c:       e28d0008        add     r0, sp, #8
   10c20:       e5901004        ldr     r1, [r0, #4]
   10c24:       e3a02002        mov     r2, #2
   10c28:       e1510002        cmp     r1, r2
   10c2c:       8a000000        bhi     10c34 <main.addr+0x34>
   10c30:       eb0035e6        bl      1e3d0 <runtime.panicindex>
   10c34:       e5900000        ldr     r0, [r0]
   10c38:       e2800008        add     r0, r0, #8
   10c3c:       e58d0014        str     r0, [sp, #20]
   10c40:       e49df004        pop     {pc}            ; (ldr pc, [sp], #4)

after

func addr(s[]int) *int {
	return &s[2]
   10c1c:       e28d0008        add     r0, sp, #8
   10c20:       e5901004        ldr     r1, [r0, #4]
   10c24:       e3510002        cmp     r1, #2
   10c28:       8a000000        bhi     10c30 <main.addr+0x30>
   10c2c:       eb0035e6        bl      1e3cc <runtime.panicindex>
   10c30:       e5900000        ldr     r0, [r0]
   10c34:       e2800008        add     r0, r0, #8
   10c38:       e58d0014        str     r0, [sp, #20]
   10c3c:       e49df004        pop     {pc}            ; (ldr pc, [sp], #4)

Also, relax gcmp restriction that 2nd operand must be a register. A followup
CL will address the remaining TODO items.

R=rsc, remyoudompheng, minux.ma
CC=golang-dev
https://golang.org/cl/6620064

to be consistent with the fdct function, and to ease any future
idct rewrites in assembly.

The BenchmarkIDCT delta is obviously just an accounting change and not
a real saving, but it does give an indication of what proportion of
time was spent in the actual IDCT and what proportion was in shift and
clip. The idct time taken is now comparable to fdct.

The BenchmarkFDCT delta is an estimate of benchmark noise.

benchmark                   old ns/op    new ns/op    delta
BenchmarkFDCT                    3842         3837   -0.13%
BenchmarkIDCT                    5611         3478  -38.01%
BenchmarkDecodeRGBOpaque      2932785      2929751   -0.10%

R=r
CC=golang-dev
https://golang.org/cl/6625057

refer to opacity. Those references were copy/pasted from the
image/png encoding benchmarks, which cares whether or not the
source image is opaque, but the JPEG encoder does not care.

R=r
CC=golang-dev
https://golang.org/cl/6623052