bug fix in named-return

R=r
OCL=16377
CL=16377

prints that insert spaces and new line

R=r
OCL=16370
CL=16370

delete unused variables

R=ken
OCL=16360
CL=16360

R=gri
DELTA=18  (8 added, 2 deleted, 8 changed)
OCL=16356
CL=16356

R=r
OCL=16345
CL=16345

- typename cannot refer to interface types
- clarified syntax
- added TODO re: type name issue with qualified identifiers

Changed/fixed order of operators/keywords.

R=r
DELTA=10  (3 added, 0 deleted, 7 changed)
OCL=16337
CL=16339

invalid.  The receiver type is not permitted to be a pointer
type.  Since 6g currently accepts it, this moves the test case
back to the bugs directory.

R=gri
DELTA=22  (11 added, 11 deleted, 0 changed)
OCL=16284
CL=16336

- fixed parsing of parameter lists (sigh)

R=r
DELTA=48  (22 added, 7 deleted, 19 changed)
OCL=16319
CL=16321

var a,b int;

R=r
OCL=16318
CL=16318

R=r
DELTA=41  (41 added, 0 deleted, 0 changed)
OCL=16290
CL=16293

R=r
OCL=16274
CL=16274

- some scanner cleanup
- new pretty-printed code can be compiled again (for some files)

R=r
OCL=16272
CL=16272

R=r
OCL=16250
CL=16250

so that they don't look like functions.

R=ken
OCL=16244
CL=16244

R=ken
OCL=16240
CL=16242

R=r
DELTA=41  (41 added, 0 deleted, 0 changed)
OCL=16212
CL=16223

and still have a variable named "net"

R=r
DELTA=18  (0 added, 0 deleted, 18 changed)
OCL=16220
CL=16222

a few missing FLUSH.

R=ken
OCL=16221
CL=16221

R=ken
OCL=16219
CL=16219

R=ken
OCL=16218
CL=16218

R=rsc
OCL=16213
CL=16213

- "nil" for interfaces, comparison against "nil"

R=r
DELTA=38  (24 added, 5 deleted, 9 changed)
OCL=16207
CL=16211

type t1 int;
type t2 int;
type t3 int;
func f1(t1, t2, t3);
func f2(t1, t2, t3 bool);
func f3(t1, t2, x t3);
func f4(*t2, x t3);	// error: cannot mix
func f5(t1, *t3);
func (x *t1) f6(y *[]t2) (t1, *t3);
func f7() (int, *string);
func f8(t1, *t2, x t3);	// error: cannot mix
func f9() (x int, *string);
func f10(*t2, t3);

R=ken
OCL=16202
CL=16210

update golden
add the "goroutines" test

R=iant
DELTA=74  (36 added, 23 deleted, 15 changed)
OCL=16194
CL=16206

- changed terminology from "embedded type" to "anonymous field"

R=r
DELTA=38  (7 added, 2 deleted, 29 changed)
OCL=16193
CL=16196

R=r
DELTA=35  (12 added, 13 deleted, 10 changed)
OCL=16162
CL=16164

R=r
OCL=16163
CL=16163

R=gri
OCL=16161
CL=16161

More to come.

R=r
DELTA=74  (47 added, 8 deleted, 19 changed)
OCL=16156
CL=16159

Currently 6g seems to resolve them to type "double".

R=ken,r
DELTA=30  (30 added, 0 deleted, 0 changed)
OCL=16132
CL=16152

signed integer overflow when negating the most negative
integer.

R=rsc
DELTA=11  (0 added, 7 deleted, 4 changed)
OCL=16105
CL=16120

* isolate OS-specific polling goop in Pollster type
* move generic poll loop out of fd_darwin.go into fd.go

R=r
DELTA=782  (448 added, 281 deleted, 53 changed)
OCL=16108
CL=16119

R=r
DELTA=494  (0 added, 494 deleted, 0 changed)
OCL=16109
CL=16115

R=r
DELTA=6  (0 added, 0 deleted, 6 changed)
OCL=16110
CL=16114

- added "..." to list of delimiters (slightly re-arranged that table)
- rename of 2 productions for more consistent naming

R=r
DELTA=20  (7 added, 1 deleted, 12 changed)
OCL=16101
CL=16103

ideally cause the compiler to give an error.  Right now 6g
warns about large shifts but does not give an error.  This CL
removes the out of range shift from shift.go, so that it will
work with gccgo, and adds a test case in bugs/ to be fixed at
a later date.

R=ken,r
DELTA=23  (9 added, 14 deleted, 0 changed)
OCL=16085
CL=16088

R=r
OCL=16068
CL=16068

R=r
OCL=16053
CL=16053

- temporary work-around for 6g bug

R=r
OCL=16052
CL=16052

- changed FieldDecl production in preparation for inlined types

R=r
DELTA=6  (4 added, 0 deleted, 2 changed)
OCL=16046
CL=16046