R=rsc
DELTA=4  (3 added, 0 deleted, 1 changed)
OCL=29314
CL=29314

R=r
DELTA=2  (0 added, 0 deleted, 2 changed)
OCL=29305
CL=29308

R=r
DELTA=1  (0 added, 0 deleted, 1 changed)
OCL=29307
CL=29307

take 2

R=ken
OCL=29304
CL=29306

*** Reason for rollback ***

too many files included

*** Original change description ***

simplifying grammar: delete LBASETYPE and LACONST

R=ken
OCL=29303
CL=29303

R=ken
OCL=29300
CL=29302

fix Reset bug in sha1.
add hmac, sha1, md5 tests.
document hmac.

R=r
DELTA=146  (111 added, 6 deleted, 29 changed)
OCL=29294
CL=29299

R=r
DELTA=53  (52 added, 0 deleted, 1 changed)
OCL=29295
CL=29298

extract comment text code out of go/doc/doc.go into comment.go.
no code changes, just rearrangement.

first step so i can write tests.

R=gri
DELTA=633  (318 added, 301 deleted, 14 changed)
OCL=29269
CL=29293

plan9 libmach.

R=rsc
APPROVED=rsc
DELTA=142  (139 added, 3 deleted, 0 changed)
OCL=29281
CL=29281

TBR=r
OCL=29279
CL=29279

R=rsc
APPROVED=rsc
DELTA=1  (1 added, 0 deleted, 0 changed)
OCL=29275
CL=29278

  comment markers must only be stripped if they are
  at the very beginning otherwise comments that contain
  code with comments get screwed up (the ast delivers clean
  comments with no junk before or after)

- fix indentation in google/net/rpc/rpc.go which screwed up
  godoc formatting

R=rsc
DELTA=3  (0 added, 0 deleted, 3 changed)
OCL=29223
CL=29267

R=rsc
DELTA=106  (69 added, 9 deleted, 28 changed)
OCL=29203
CL=29254

R=rsc
DELTA=184  (69 added, 73 deleted, 42 changed)
OCL=29178
CL=29253

R=r
DELTA=21  (21 added, 0 deleted, 0 changed)
OCL=29212
CL=29249

(will submit with compiler fixes).

R=r
DELTA=44  (44 added, 0 deleted, 0 changed)
OCL=29211
CL=29248

R=rsc
DELTA=1  (0 added, 0 deleted, 1 changed)
OCL=29221
CL=29221

R=r
DELTA=41  (0 added, 21 deleted, 20 changed)
OCL=29219
CL=29219

R=ken
OCL=29209
CL=29216

R=dsymonds
DELTA=2  (0 added, 0 deleted, 2 changed)
OCL=29210
CL=29215

update tests.

R=ken
OCL=29207
CL=29207

renamed 6g/(gen|align|obj).c

R=r
OCL=29205
CL=29205

R=r
DELTA=79  (79 added, 0 deleted, 0 changed)
OCL=29196
CL=29202

so that callers cannot edit large values inside interfaces.

R=r
DELTA=52  (42 added, 1 deleted, 9 changed)
OCL=29180
CL=29195

when assigning a multifield object
(structs or arrays of structs) they
must not contain any fields that could
not be assigned individually.

R=ken
OCL=29192
CL=29194

move interface tests to subdirectory.

R=r
DELTA=1632  (827 added, 804 deleted, 1 changed)
OCL=29181
CL=29191

structure and array literals

R=r
OCL=29190
CL=29190

R=r
DELTA=3  (0 added, 1 deleted, 2 changed)
OCL=29151
CL=29176

delete unnecessary reflect. throughout

R=r
DELTA=124  (18 added, 0 deleted, 106 changed)
OCL=29173
CL=29175

(this time i ran g4 nothave)

TBR=r
OCL=29156
CL=29156

TBR=dsymonds
OCL=29155
CL=29155

must have non-nil i.

R=ken
OCL=29136
CL=29136

* new package hash defining interfaces Hash and Hash32.
* adler32 and crc32 return Hash32 instead of specific types.
* adler32 and crc32 provide non-allocating methods for single slices.
* sha1 and md5 move to crypto, return Hash.
* sum.go, a simple test program, moves to /usr/rsc.
* refresh Make.deps

R=r
DELTA=1908  (935 added, 923 deleted, 50 changed)
OCL=29095
CL=29135

in x.(T) x cannot be a nil (uninitialized) interface variable.

remove TODO per discussion.

DELTA=3  (1 added, 1 deleted, 1 changed)
OCL=29123
CL=29134

- format.go implementation complete
- lots of documentation added (not quite complete)

TBR=r
OCL=29130
CL=29130

R=r
DELTA=250  (250 added, 0 deleted, 0 changed)
OCL=29120
CL=29127

as far as I can tell there's no reason not to.

the Nillable test was succeeding because NewZeroValue
returned the nil interface value and the type guard
was letting it through.  the only change in the test is
more detail in the print.

R=r
DELTA=8  (0 added, 7 deleted, 1 changed)
OCL=29124
CL=29126

that came from two different enums.
spilt into ifacecvt and ifaceop depending
on which enum the argument is.

R=ken
OCL=29122
CL=29122

this is not a user-visible change.

before, all interface values were

	struct Itype {
		Sigt *type;
		Sigi *inter;
		void *method[n];
	}

	struct Iface {
		void *addr;
		Itype *itype;
	}

the itype is basically a vtable, but it's unnecessary
if the static type is interface{ }.
for interface values with static type empty, the
new representation is

	struct Eface {
		void *addr;
		Sigt *type;
	}

this complicates the code somewhat, but
it reduces the number of Itypes that
have to be computed and cached,
it opens up opportunities to avoid function
calls in a few common cases,
and it will make it possible to lay out
interface{} values at compile time,
which i think i'll need for the new reflection.

R=ken
OCL=28701
CL=29121