The second parameter for sql.putConn() (err) is always nil. As a result bad
connections are reused, even if the driver returns an driver.ErrBadConn.
Unsing a pointer to err instead achievs the desired behavior.
See http://code.google.com/p/go/issues/detail?id=3777 for more details.
Fixes #3777.

R=golang-dev, dave, bradfitz, jameshuachow, BlakeSGentry
CC=golang-dev
https://golang.org/cl/6348069

R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6459050

R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6480058

Bug introduced with CL 6454150.

Fixes #4000.

R=r
CC=golang-dev
https://golang.org/cl/6474061

BenchmarkAddVV_1          500000000        7.24 ns/op     8844.11 MB/s
BenchmarkAddVV_2          100000000       10.4 ns/op     12290.41 MB/s
BenchmarkAddVV_3          100000000       10.7 ns/op     17966.58 MB/s
BenchmarkAddVV_4          100000000       12.3 ns/op     20848.67 MB/s
BenchmarkAddVV_5          100000000       14.5 ns/op     21993.82 MB/s
BenchmarkAddVV_1e1        100000000       24.0 ns/op     26720.65 MB/s
BenchmarkAddVV_1e2         10000000      246 ns/op       26014.58 MB/s
BenchmarkAddVV_1e3          1000000     2416 ns/op       26485.06 MB/s
BenchmarkAddVV_1e4           100000    23874 ns/op       26806.36 MB/s
BenchmarkAddVV_1e5            10000   241155 ns/op       26538.87 MB/s
BenchmarkAddVW_1          500000000        6.12 ns/op    10461.91 MB/s
BenchmarkAddVW_2          200000000       11.0 ns/op     11596.63 MB/s
BenchmarkAddVW_3          200000000        8.97 ns/op    21409.82 MB/s
BenchmarkAddVW_4          100000000       10.8 ns/op     23696.72 MB/s
BenchmarkAddVW_5          100000000       12.5 ns/op     25524.88 MB/s
BenchmarkAddVW_1e1        100000000       21.5 ns/op     29786.32 MB/s
BenchmarkAddVW_1e2         10000000      168 ns/op       37925.36 MB/s
BenchmarkAddVW_1e3          1000000     1658 ns/op       38579.15 MB/s
BenchmarkAddVW_1e4           100000    16492 ns/op       38805.85 MB/s
BenchmarkAddVW_1e5            10000   172155 ns/op       37175.69 MB/s
BenchmarkAddMulVVW_1      100000000       12.9 ns/op      4968.49 MB/s
BenchmarkAddMulVVW_2      100000000       15.5 ns/op      8279.42 MB/s
BenchmarkAddMulVVW_3      100000000       13.4 ns/op     14340.53 MB/s
BenchmarkAddMulVVW_4      100000000       15.8 ns/op     16194.94 MB/s
BenchmarkAddMulVVW_5      100000000       18.9 ns/op     16906.61 MB/s
BenchmarkAddMulVVW_1e1     50000000       32.3 ns/op     19838.35 MB/s
BenchmarkAddMulVVW_1e2     10000000      285 ns/op       22427.28 MB/s
BenchmarkAddMulVVW_1e3      1000000     2777 ns/op       23040.42 MB/s
BenchmarkAddMulVVW_1e4       100000    27901 ns/op       22938.01 MB/s
BenchmarkAddMulVVW_1e5        10000   281087 ns/op       22768.73 MB/s

R=r
CC=golang-dev
https://golang.org/cl/6478055

Update #3930
Return a better error message in this situation.

R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6474055

Fixes #3721.

R=dave, rsc
CC=golang-dev
https://golang.org/cl/6395055

    CVTSS2SQ's rounding mode is controlled by the RC field of MXCSR;
as we specifically need truncate semantic, we should use CVTTSS2SQ.

    Fixes #3804.

R=rsc, r
CC=golang-dev
https://golang.org/cl/6352079

To make it more compliant.
This won't affect the behavior of running on OABI-only kernels.

R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6475044

in 13416:67c0b8c8fb29 "faster code, mainly for rotate" [1]. The codegen
can run out of registers if there are too many small-int arithmetic ops.

An alternative approach is to copy 6g's sbop/abop codegen to 8g, but
this change is less risky.

Fixes #3835.

[1] http://code.google.com/p/go/source/diff?spec=svn67c0b8c8fb29b1b7b6221977af6b89cae787b941&name=67c0b8c8fb29&r=67c0b8c8fb29b1b7b6221977af6b89cae787b941&format=side&path=/src/cmd/8g/cgen.c

R=rsc, remyoudompheng, r
CC=golang-dev
https://golang.org/cl/6450163

No semantic change.

R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6475050

YearDay provides the day in the year represented by a given time.Time
object. This value is normally computed as part of other date calculations,
but not exported.

Fixes #3932.

R=golang-dev, r, remyoudompheng
CC=golang-dev, rsc
https://golang.org/cl/6460069

R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6479045

Fixes #3967.

R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6460115

This makes SNI "just work" for callers using the standard http.Client.

Since we now have a test that depends on the httptest.Server cert, change
the cert to be a CA (keeping all other fields the same).

R=bradfitz
CC=agl, dsymonds, gobot, golang-dev
https://golang.org/cl/6448154

R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6484046

NetBSD's built-in linker script for 'ld -r' does not provide a
SEARCH_DIR. As a result libgcc.a is not found when -lgcc is used.

Work around this by determining the path to libgcc (by invoking
gcc with the -print-libgcc-file-name option) and explicitly
referencing the resulting library.

R=golang-dev, iant, aram, lucio.dere, minux.ma
CC=golang-dev
https://golang.org/cl/6470044

R=golang-dev, r, adg
CC=golang-dev
https://golang.org/cl/6463050

R=golang-dev, nigeltao
CC=golang-dev
https://golang.org/cl/6476043

R=andybalholm
CC=golang-dev
https://golang.org/cl/6463070

Moves the defer (again).

Also, correct the example documentation to match.

R=r, robert.hencke, iant, dsymonds, bradfitz
CC=golang-dev
https://golang.org/cl/6458158

Fixes #3793.

Tested using GOMAXPROCS=81 which was able to trigger a panic
in TestStressSurpriseServerCloses continually on a Core i5.

R=fullung, bradfitz
CC=golang-dev
https://golang.org/cl/6445069

Fixes #3969.

R=dvyukov, r, alex.brainman, minux.ma
CC=golang-dev
https://golang.org/cl/6462081

Current results on linux_amd64, 8 HT cores @2.4GHz:
BenchmarkTCPOneShot	   10000	    194037 ns/op
BenchmarkTCPOneShot-2	   20000	     93641 ns/op
BenchmarkTCPOneShot-4	   20000	     94039 ns/op
BenchmarkTCPOneShot-8	   20000	     94667 ns/op
BenchmarkTCPOneShot-16	   10000	    301924 ns/op
BenchmarkTCPOneShotTimeout	   10000	    193264 ns/op
BenchmarkTCPOneShotTimeout-2	   20000	     98247 ns/op
BenchmarkTCPOneShotTimeout-4	   20000	     94442 ns/op
BenchmarkTCPOneShotTimeout-8	   20000	     95297 ns/op
BenchmarkTCPOneShotTimeout-16	   10000	    307970 ns/op
BenchmarkTCPPersistent	   50000	     52050 ns/op
BenchmarkTCPPersistent-2	  100000	     29452 ns/op
BenchmarkTCPPersistent-4	  100000	     28823 ns/op
BenchmarkTCPPersistent-8	   50000	     30473 ns/op
BenchmarkTCPPersistent-16	   10000	    311777 ns/op
BenchmarkTCPPersistentTimeout	   50000	     32574 ns/op
BenchmarkTCPPersistentTimeout-2	   50000	     29723 ns/op
BenchmarkTCPPersistentTimeout-4	  100000	     28592 ns/op
BenchmarkTCPPersistentTimeout-8	  100000	     28997 ns/op
BenchmarkTCPPersistentTimeout-16	   10000	    314354 ns/op

R=golang-dev, alex.brainman, dave, mikioh.mikioh, r, iant, bradfitz, iant
CC=golang-dev
https://golang.org/cl/6458128

The last fix was wrong w.r.t C's operator precedence,
and it also failed to really skip the NONE relocation.

The offending R_386_NONE relocation is a absolute
relocation in section .eh_frame.

TBR=golang-dev
CC=golang-dev
https://golang.org/cl/6463058

     Fixes #3973.

R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6464078

If ParseUint is like ParseInt it should have the same parameter
names, thus rename b to to base as in ParseInt's documentation.

R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6453152

benchmark                           old ns/op    new ns/op    delta
BenchmarkClientServerParallel          155909       154454   -0.93%
BenchmarkClientServerParallel-2         86012        82986   -3.52%
BenchmarkClientServerParallel-4         70211        55168  -21.43%
BenchmarkClientServerParallel-8         80755        47862  -40.73%
BenchmarkClientServerParallel-12        77753        51478  -33.79%
BenchmarkClientServerParallel-16        77920        50278  -35.47%
The benchmark is https://golang.org/cl/6441134
The machine is 2 x 4 HT cores (16 HW threads total).
Fixes #3946.
Now contention moves to net.pollServer.AddFD().

R=bradfitz
CC=bradfitz, dave, dsymonds, gobot, golang-dev, remyoudompheng
https://golang.org/cl/6454142

with table changes.
NOTE: there is no test for this, but 1) the code has now the same
control flow as scan in exp/locale/collate/contract.go, which is
tested and 2) Builder verifies the generated table so bugs in this
code are quickly and easily found (which is how this bug was discovered).

R=r
CC=golang-dev
https://golang.org/cl/6461082

The main table will need to get a slightly different collation table as the one
used by regtest, as the regtest is based on the standard UCA DUCET, while
the locale-specific tables are all based on a CLDR root table.
This change allows changing the table without affecting the regression test.

R=r
CC=golang-dev
https://golang.org/cl/6453089

default as ICU.

R=r
CC=golang-dev
https://golang.org/cl/6445080

against previously generated tables.

R=r
CC=golang-dev
https://golang.org/cl/6441098

instead of touching the tokenizer's internal state.

R=andybalholm
CC=golang-dev
https://golang.org/cl/6446153

If we aren't explicit about the KeyUsages, the verifier
will treat the certificate as a server certificate and require
it to have a ExtKeyUsageServerAuth key usage.

R=golang-dev
CC=golang-dev
https://golang.org/cl/6453148

- pprof is a perl script, so go command should invoke
  perl instead of trying to run pprof directly;
- pprof should use "go tool nm" unconditionally on windows,
  no one else can extract symbols from Go program;
- pprof should use "go tool nm" instead of "6nm".

Fixes #3879.

R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6445082

Before, pointers always appeared as 0x1234ABCD. This CL
keeps that as the default for %p and %v, but lets explicit
numeric verbs override the default.
Fixes #3936.

R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6441152

Only affects some tests and none seem likely to be problematic, but let's fix them.
Fixes #3971.

R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6463060

R=r, rsc, iant, ken, iant
CC=golang-dev
https://golang.org/cl/6450150

When manipulating the stack pointer use the UESP register instead
of the ESP register, since the UESP register is the one that gets
restored from the machine context. Fixes broken tests on netbsd/386.

R=golang-dev, minux.ma, r, bsiegert
CC=golang-dev
https://golang.org/cl/6465054

see issus 3358 for similar problem

R=golang-dev
CC=golang-dev, minux.ma
https://golang.org/cl/6464072