Process a slice of equivalent values by setting replaced values to nil
instead of removing them from the slice to eliminate copying.  Also take
advantage of the entry number sort to break early once we reach a value
in a block that is not dominated.

For the code in issue #15112:

Before:
real    0m52.603s
user    0m56.957s
sys     0m1.213s

After:
real    0m22.048s
user    0m26.445s
sys     0m0.939s

Updates #15112

Change-Id: I06d9e1e1f1ad85d7fa196c5d51f0dc163907376d
Reviewed-on: https://go-review.googlesource.com/22068Reviewed-by: David Chase <drchase@google.com>

RFC 1952, section 3.2.3 says:
>>>
If FHCRC is set, a CRC16 for the gzip header is present,
immediately before the compressed data. The CRC16 consists of the two
least significant bytes of the CRC32 for all bytes of the
gzip header up to and not including the CRC16.
<<<

Thus, instead of computing the CRC only over the first 10 bytes
of the header, we compute it over the whole header (minus CRC16).

Fixes #15070

Change-Id: I55703fd30b535b12abeb5e3962d4da0a86ed615a
Reviewed-on: https://go-review.googlesource.com/21466Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

This will allow us to mechanically substitute these strings
using javascript (in a forthcoming change to x/tools/godoc).

Updates #14371

Change-Id: I96e876283060ffbc9f3eabaf55d6b880685453e1
Reviewed-on: https://go-review.googlesource.com/22055Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

- position info for all exported globals, plus methods and fields
- use delta-encoded line number info in most cases
- canonicalize all strings: each filename appears only once,
  but will also compact other strings (names) to at most one
  occurence in encoding
- positions not yet hooked up when reading in

Also:
- adjusted go/importer (gcimporter)
- some refactoring for better symmetry

Stats:
- comparison of export data size w/o and w/ position info (bytes).
- delta is increase in %
- overall (see bottom of table): 14% increase
- however, the current binary format decreased from
  the original binary format last week by 14%
- compared to original textual format: 65% decrease
  (increase by 14% after decrease by 14% still leads
  to a decrease from original textual format)

(caveat: we used the textual size from last week, assuming
it has not changed - there may be a small error here).

package				w/o pos	w/ pos	delta

archive/tar			4234	4902	16%
archive/zip			6387	7340	15%
bufio				3106	3419	10%
bytes				4362	4757	9%
cmd/addr2line			27	70	159%
cmd/api				12065	13590	13%
cmd/asm				27	64	137%
cmd/asm/internal/arch		9957	11529	16%
cmd/asm/internal/asm		11788	13385	14%
cmd/asm/internal/flags		239	311	30%
cmd/asm/internal/lex		13415	15358	14%
cmd/cgo				13064	15006	15%
cmd/compile			27	67	148%
cmd/compile/internal/amd64	461	869	89%
cmd/compile/internal/arm	5963	7273	22%
cmd/compile/internal/arm64	363	657	81%
cmd/compile/internal/big	7186	8590	20%
cmd/compile/internal/gc		48242	56234	17%
cmd/compile/internal/mips64	367	666	81%
cmd/compile/internal/ppc64	372	721	94%
cmd/compile/internal/s390x	330	569	72%
cmd/compile/internal/ssa	30464	35058	15%
cmd/compile/internal/x86	429	770	79%
cmd/cover			3984	4731	19%
cmd/dist			74	154	108%
cmd/doc				7272	8591	18%
cmd/expdump			27	71	163%
cmd/fix				342	419	23%
cmd/go				8126	9520	17%
cmd/gofmt			27	70	159%
cmd/gofmt2			27	69	156%
cmd/gofmt2/internal/format	702	856	22%
cmd/gofmt2/internal/lexical	2954	3509	19%
cmd/gofmt2/internal/parse	6185	7295	18%
cmd/gofmt2/internal/syntax	3533	4738	34%
cmd/gofmt2/internal/test	540	615	14%
cmd/internal/bio		5395	6060	12%
cmd/internal/gcprog		533	663	24%
cmd/internal/goobj		1022	1277	25%
cmd/internal/obj		10951	12825	17%
cmd/internal/obj/arm		8612	9985	16%
cmd/internal/obj/arm64		15814	17638	12%
cmd/internal/obj/mips		10928	12487	14%
cmd/internal/obj/ppc64		13576	15277	13%
cmd/internal/obj/s390x		16513	18708	13%
cmd/internal/obj/x86		21152	23482	11%
cmd/internal/objfile		14442	16505	14%
cmd/internal/pprof/commands	1663	1885	13%
cmd/internal/pprof/driver	9517	10789	13%
cmd/internal/pprof/fetch	7632	8635	13%
cmd/internal/pprof/plugin	13150	14809	13%
cmd/internal/pprof/profile	7004	8248	18%
cmd/internal/pprof/report	7763	8942	15%
cmd/internal/pprof/svg		1332	1534	15%
cmd/internal/pprof/symbolizer	7376	8439	14%
cmd/internal/pprof/symbolz	6970	7976	14%
cmd/internal/pprof/tempfile	3645	4093	12%
cmd/internal/sys		505	619	23%
cmd/internal/unvendor/golang.org/x/arch/arm/armasm	73951	79188	7%
cmd/internal/unvendor/golang.org/x/arch/x86/x86asm	10140	11738	16%
cmd/link			27	64	137%
cmd/link/internal/amd64		9317	11034	18%
cmd/link/internal/arm		110	213	94%
cmd/link/internal/arm64		112	219	96%
cmd/link/internal/ld		53524	60149	12%
cmd/link/internal/mips64	113	222	96%
cmd/link/internal/ppc64		113	220	95%
cmd/link/internal/s390x		112	219	96%
cmd/link/internal/x86		110	212	93%
cmd/nm				27	61	126%
cmd/objdump			27	68	152%
cmd/pack			4141	4688	13%
cmd/pprof			27	67	148%
cmd/trace			624	842	35%
cmd/vet				11194	13140	17%
cmd/vet/internal/whitelist	52	113	117%
cmd/yacc			1141	1317	15%
compress/bzip2			2101	2484	18%
compress/flate			3619	4336	20%
compress/gzip			6261	7111	14%
compress/lzw			276	401	45%
compress/zlib			3630	4158	15%
container/heap			187	250	34%
container/list			1370	1506	10%
container/ring			466	546	17%
context				3005	3338	11%
crypto				728	856	18%
crypto/aes			181	321	77%
crypto/cipher			744	1163	56%
crypto/des			220	320	45%
crypto/dsa			4526	4990	10%
crypto/ecdsa			5341	5982	12%
crypto/elliptic			4969	5593	13%
crypto/hmac			188	250	33%
crypto/md5			560	706	26%
crypto/rand			4218	4746	13%
crypto/rc4			214	321	50%
crypto/rsa			5648	6355	13%
crypto/sha1			597	751	26%
crypto/sha256			228	351	54%
crypto/sha512			354	484	37%
crypto/subtle			586	621	6%
crypto/tls			20909	23438	12%
crypto/x509			14862	16857	13%
crypto/x509/pkix		8384	9278	11%
database/sql			6721	7715	15%
database/sql/driver		1243	1535	23%
debug/dwarf			7867	9153	16%
debug/elf			25479	28025	10%
debug/gosym			1887	2267	20%
debug/macho			7222	8846	22%
debug/pe			6921	8081	17%
debug/plan9obj			1084	1319	22%
encoding			217	280	29%
encoding/ascii85		587	722	23%
encoding/asn1			1043	1268	22%
encoding/base32			929	1112	20%
encoding/base64			1166	1368	17%
encoding/binary			2168	2410	11%
encoding/csv			3761	4203	12%
encoding/gob			11304	12936	14%
encoding/hex			510	606	19%
encoding/json			9965	11395	14%
encoding/pem			202	266	32%
encoding/xml			11817	13361	13%
errors				126	170	35%
expvar				930	1142	23%
flag				5905	6519	10%
fmt				1027	1190	16%
go/ast				12910	15541	20%
go/build			5460	6173	13%
go/constant			1645	1816	10%
go/doc				3107	3882	25%
go/format			1416	1729	22%
go/importer			1426	1668	17%
go/internal/gccgoimporter	1624	2028	25%
go/internal/gcimporter		2650	3095	17%
go/parser			6220	7073	14%
go/printer			1924	2306	20%
go/scanner			3137	3602	15%
go/token			3053	3474	14%
go/types			21793	25561	17%
hash				234	327	40%
hash/adler32			465	553	19%
hash/crc32			668	817	22%
hash/crc64			630	727	15%
hash/fnv			1413	1582	12%
html				76	114	50%
html/template			14382	16457	14%
image				10248	11409	11%
image/color			2247	2562	14%
image/color/palette		107	169	58%
image/draw			2313	2494	8%
image/gif			3079	3450	12%
image/internal/imageutil	3136	3456	10%
image/jpeg			2349	2735	16%
image/png			2404	2695	12%
index/suffixarray		4978	5596	12%
internal/race			225	278	24%
internal/singleflight		551	697	26%
internal/syscall/windows/sysdll	97	166	71%
internal/testenv		4488	5052	13%
internal/trace			1392	1680	21%
io				2811	3318	18%
io/ioutil			3988	4467	12%
log				3532	3907	11%
log/syslog			4247	4775	12%
math				3021	4499	49%
math/big			7250	8456	17%
math/cmplx			1034	1617	56%
math/rand			734	885	21%
mime				1889	2194	16%
mime/multipart			4313	4849	12%
mime/quotedprintable		1758	1996	14%
net				15686	18617	19%
net/http			42182	47848	13%
net/http/cgi			19496	21768	12%
net/http/cookiejar		4615	5248	14%
net/http/fcgi			17758	19771	11%
net/http/httptest		26108	29350	12%
net/http/httputil		20732	23286	12%
net/http/internal		2195	2497	14%
net/http/pprof			17596	19545	11%
net/internal/socktest		1689	2153	27%
net/mail			4328	4810	11%
net/rpc				24328	27249	12%
net/rpc/jsonrpc			11052	12438	13%
net/smtp			17127	19174	12%
net/textproto			3705	4329	17%
net/url				1193	1371	15%
os				8493	10113	19%
os/exec				6625	7532	14%
os/signal			137	236	72%
os/user				529	761	44%
path				295	372	26%
path/filepath			3452	3952	14%
reflect				5091	6028	18%
regexp				4848	5585	15%
regexp/syntax			2590	3076	19%
runtime				8721	11598	33%
runtime/cgo			17	17	0%
runtime/debug			2721	3130	15%
runtime/internal/atomic		569	704	24%
runtime/internal/sys		1874	2318	24%
runtime/pprof			478	582	22%
runtime/race			18	18	0%
runtime/trace			95	146	54%
sort				1052	1215	15%
strconv				1389	1667	20%
strings				3372	3772	12%
sync				946	1371	45%
sync/atomic			962	1079	12%
syscall				41574	45613	10%
testing				6184	7243	17%
testing/iotest			883	1116	26%
testing/quick			4659	5443	17%
text/scanner			2930	3269	12%
text/tabwriter			2333	2607	12%
text/template			13335	15274	15%
text/template/parse		8270	9285	12%
time				4687	5313	13%
unicode				3831	4355	14%
unicode/utf16			530	584	10%
unicode/utf8			872	946	8%
vendor/golang.org/x/net/http2/hpack	3386	3970	17%

				1295440	1481566	14%
orig. textual			4253585	1481566	-65%
orig. binary			1724071	1481566 -14%

Change-Id: I4177c6511cc57ebe5eb80c89bf3aefc83376ce86
Reviewed-on: https://go-review.googlesource.com/22096Reviewed-by: Matthew Dempsky <mdempsky@google.com>

The result of ODOTPTR, as well as a bunch of other ops,
should be the type of the result, not always a pointer type.

This fixes an amd64p32 bug where we were incorrectly truncating
a 64-bit slice index to 32 bits, and then barfing on a weird
load-64-bits-but-then-truncate-to-32-bits op that doesn't exist.

Fixes #15252

Change-Id: Ie62f4315fffd79f233e5449324ccc0879f5ac343
Reviewed-on: https://go-review.googlesource.com/22094
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

sync/atomic.StorePointer (which is implemented in
runtime/atomic_pointer.go) writes the pointer twice (through two
completely different code paths, no less). Fix it to only write once.

Change-Id: Id3b2aef9aa9081c2cf096833e001b93d3dd1f5da
Reviewed-on: https://go-review.googlesource.com/21999
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
Reviewed-by: Rick Hudson <rlh@golang.org>

SwapPointer is declared as

  func SwapPointer(addr *unsafe.Pointer, new unsafe.Pointer) (old unsafe.Pointer)

in sync/atomic, but defined in the runtime (where it's actually
implemented) as

  func sync_atomic_SwapPointer(ptr unsafe.Pointer, new unsafe.Pointer) unsafe.Pointer

Make ptr a *unsafe.Pointer in the runtime definition to match the type
in sync/atomic.

Change-Id: I99bab651b995001bbe54f9e790fdef2417ef0e9e
Reviewed-on: https://go-review.googlesource.com/21998
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>

Use deBruijn sequences to count low-order zeros.
Reorg bswap to not use &^, it takes another instruction on x86.

Change-Id: I4a5ed9fd16ee6a279d88c067e8a2ba11de821156
Reviewed-on: https://go-review.googlesource.com/22084Reviewed-by: David Chase <drchase@google.com>

This code was fixed a while ago to ensure that xtest and fake packages came
first on the link line, but golang.org/cl/16775 added --whole-archive ...
--no-whole-archive around all the .a files and rendered this fix useless.

So, take a different approach and only put one .a file on the linker command
line for each ImportPath we see while traversing the action graph, not for each
*Package we see. The way we walk the graph ensures that we'll see the .a files
that need to be first first.

Change-Id: I137f00f129ccc9fc99f40eee885cc04cc358a62e
Reviewed-on: https://go-review.googlesource.com/21692Reviewed-by: Ian Lance Taylor <iant@golang.org>

The unique difficulty of #cgo pkg-config is that the linker flags are recorded
when the package is compiled but (obviously) must be used when the package is
linked into an executable -- so the flags need to be stored on disk somewhere.
As it happens cgo already writes out a _cgo_flags file: nothing uses it
currently, but this change adds it to the lib$pkg.a file when compiling a
package, reads it out when linking (and passes a version of the .a file with
_cgo_flags stripped out of it to the linker). It's all fairly ugly but it works
and I can't really think of any way of reducing the essential level of
ugliness.

Fixes #11739

Change-Id: I35621878014e1e107eda77a5b0b23d0240ec5750
Reviewed-on: https://go-review.googlesource.com/18790
Run-TryBot: Michael Hudson-Doyle <michael.hudson@canonical.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

Notably, this fixes two incorrect argument sizes.

Update #11041

Change-Id: Ie4a3b1a59cd6a6707f6d2f4d3be978fc70322b46
Reviewed-on: https://go-review.googlesource.com/22091
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Update #14759.

Change-Id: I8a174aad721beb62380e10071d9648b6b1c21b8c
Reviewed-on: https://go-review.googlesource.com/22072
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Andrew Gerrand <adg@golang.org>

This ensures that importpath symbols are treated like other type data
and end up in the same section under all build modes.

Fixes: go test -buildmode=pie reflect

Change-Id: Ibb8348648e8dcc850f2424d206990a06090ce4c6
Reviewed-on: https://go-review.googlesource.com/22081
Run-TryBot: David Crawshaw <crawshaw@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

While we’re here, clean up a few comments.

Fixes #15281

Change-Id: Ia6173e9941133db08f57bc80bdd3c5722122bfdb
Reviewed-on: https://go-review.googlesource.com/22082
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>

Print numGC followed by numGC1, rather than printing numGC twice.

Change-Id: I8e7144b6a11d4ae9be0d82d88b86fed04b906e2f
Reviewed-on: https://go-review.googlesource.com/22087Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

These comments were left behind after runtime.h was converted
from C to Go. I examined the original code and tried to move these
to the places that the most sense.

Change-Id: I8769d60234c0113d682f9de3bd8d6c34c450c188
Reviewed-on: https://go-review.googlesource.com/21969Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

s390x can handle unaligned loads and stores of 64-bit values.

Change-Id: Iae5621781e3ba56e27b4a1f4788772c86e4f6475
Reviewed-on: https://go-review.googlesource.com/22086Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

This is a fix for the ssacheck builder
http://build.golang.org/log/baa00f70c34e41186051cfe90568de3d91f115d7
after CL 21307 for sinking spills down loop exits
https://go-review.googlesource.com/#/c/21037/

The fix is to reuse (move) the original spill, thus preserving
the definition of the variable and its use count. Original and
copy both use the same stack slot, but ssacheck needs to see
a definition for the variable itself.

Fixes #15279.

Change-Id: I286285490193dc211b312d64dbc5a54867730bd6
Reviewed-on: https://go-review.googlesource.com/21995Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

+ test

Fixes #15305

Change-Id: Ica657c00c92f0b19f0df7452cdbe5a95d23cc8a4
Reviewed-on: https://go-review.googlesource.com/22085Reviewed-by: Robert Griesemer <gri@golang.org>

Fixes #15271

Change-Id: I28e3fb5bde1e6fd5b263b1434873b8ce051aee97
Reviewed-on: https://go-review.googlesource.com/22083
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Rob Pike <r@golang.org>

Also add MustClose and MustWriter to cmd/internal/bio, and use them in
cmd/asm.

Change-Id: I07f5df3b66c17bc5b2e6ec9c4357d9b653e354e0
Reviewed-on: https://go-review.googlesource.com/21938
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

The AuthorityKeyId is optional for self-signed certificates, generally
useless, and takes up space. This change causes an AuthorityKeyId not to
be added to self-signed certificates, although it can still be set in
the template if the caller really wants to include it.

Fixes #15194.

Change-Id: If5d3c3d9ca9ae5fe67458291510ec7140829756e
Reviewed-on: https://go-review.googlesource.com/21895Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Adam Langley <agl@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

Error strings in this package were all over the place: some were
prefixed with “tls:”, some with “crypto/tls:” and some didn't have a
prefix.

This change makes everything use the prefix “tls:”.

Change-Id: Ie8b073c897764b691140412ecd6613da8c4e33a2
Reviewed-on: https://go-review.googlesource.com/21893Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>

We can trust that untyped composite literals are part of a slice literal
and not emit a vet warning for those.

Fixes #9171

Change-Id: Ia7c081e543b850f8be1fd1f9e711520061e70bed
Reviewed-on: https://go-review.googlesource.com/22000
Run-TryBot: Rob Pike <r@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Rob Pike <r@golang.org>

The old trace-viewer is broken since Chrome 49:
https://bugs.chromium.org/p/chromium/issues/detail?id=569417
It was fixed in:
https://github.com/catapult-project/catapult/commit/506457cbd726324f327b80ae11f46c1dfeb8710d

This change updates trace-viewer to the latest version
(now it is called catapult).

This version has a bug in the lean config that we use, though:
https://github.com/catapult-project/catapult/issues/2247
So use full config for now (it works, but leads to larger html).
When the bug is fixed we need to switch back to lean config (issue #15302).

Change-Id: Ifb8d782ced66e3292d81c5604039fe18eaf267c5
Reviewed-on: https://go-review.googlesource.com/22013Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

The synchronization in this test is a bit complicated and likely
incorrect, judging from the sporadically hanging trybots.
Most of what this is supposed to test is already tested in
TestTestContext, so I'll just remove it.

Fixes #15170

Change-Id: If54db977503caa109cec4516974eda9191051888
Reviewed-on: https://go-review.googlesource.com/22080
Run-TryBot: Marcel van Lohuizen <mpvl@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Fixes #15291.

Change-Id: I563140c2acd37d4989a940488b217414cf73f6c2
Reviewed-on: https://go-review.googlesource.com/22077Reviewed-by: Alex Brainman <alex.brainman@gmail.com>

Change-Id: I6644081df495cb92b3d208f867066f9acb08946f
Reviewed-on: https://go-review.googlesource.com/22074
Run-TryBot: Mikio Hara <mikioh.mikioh@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

Change-Id: I41691134770d01805c19c0f84f8828b00b85de0c
Reviewed-on: https://go-review.googlesource.com/22058
Run-TryBot: Mikio Hara <mikioh.mikioh@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

Change-Id: I9bda2ce6f45fb8292503f86d8f9f161601f222b7
Reviewed-on: https://go-review.googlesource.com/22053Reviewed-by: Michael Hudson-Doyle <michael.hudson@canonical.com>

Some of the Debug[x] flags are actually boolean too, but not all, so
they need to be handled separately.

While here, change some obj.Flagstr and obj.Flagint64 calls to
directly use flag.StringVar and flag.Int64Var instead.

Change-Id: Iccedf6fed4328240ee2257f57fe6d66688f237c4
Reviewed-on: https://go-review.googlesource.com/22052Reviewed-by: Michael Hudson-Doyle <michael.hudson@canonical.com>

Tested with debugFormat enabled and running
(export GO_GCFLAGS=-newexport; sh all.bash).

Change-Id: If7d43e1e594ea43c644232b89e670f7abb6b003e
Reviewed-on: https://go-review.googlesource.com/22033Reviewed-by: Matthew Dempsky <mdempsky@google.com>

Documentation change only.

Fixes #15178.

Change-Id: I3c7d80ce9e668ac7515f7ebb9da80f3bd8e534d6
Reviewed-on: https://go-review.googlesource.com/22006Reviewed-by: Ian Lance Taylor <iant@golang.org>

Also:
- "rewrite" node Op in exporter for some nodes instead of importer
- more comments

Change-Id: I809e6754d14987b28f1da9379951ffa2e690c2a7
Reviewed-on: https://go-review.googlesource.com/22008Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

The existing epoll_event structure used by many of
the epoll_* syscalls was defined incorrectly
for use with ppc64le & ppc64 in the syscall
directory.  This resulted in the caller getting
incorrect information on return from these
syscalls.  This caused failures in fsnotify as
well as builds with upstream Docker.  The
structure is defined correctly in gccgo.

This adds a pad field that is expected for
these syscalls on ppc64le, ppc64.
Fixes #15135

Change-Id: If7e8ea9eb1d1ca5182c8dc0f935b334127341ffd
Reviewed-on: https://go-review.googlesource.com/21582Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>

By replacing the *string used to represent pkgPath with a
reflect.name everywhere, the embedded *string for package paths
inside the reflect.name can be replaced by an offset, nameOff.
This reduces the number of pointers in the type information.

This also moves all reflect.name types into the same section, making
it possible to use nameOff more widely in later CLs.

No significant binary size change for normal binaries, but:

linux/amd64 PIE:
	cmd/go: -440KB (3.7%)
	jujud:  -2.6MB (3.2%)

For #6853.

Change-Id: I3890b132a784a1090b1b72b32febfe0bea77eaee
Reviewed-on: https://go-review.googlesource.com/21395
Run-TryBot: David Crawshaw <crawshaw@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

Change-Id: If0e0bc5a85101db1e70faaab168fc2d12024eb93
Reviewed-on: https://go-review.googlesource.com/22005Reviewed-by: Ian Lance Taylor <iant@golang.org>

Merge them together into os1_darwin.go. A future CL will rename it.

Change-Id: Ia4380d3296ebd5ce210908ce3582ff184566f692
Reviewed-on: https://go-review.googlesource.com/22004
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

Fixes #15104.

Change-Id: I9ddfbbf39ef0a873b703ee3e04fbb7d1192f5f39
Reviewed-on: https://go-review.googlesource.com/21581
Run-TryBot: David Crawshaw <crawshaw@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

We do two O(n) scans of all values in an eqclass when computing
substitutions for CSE.

In unfortunate cases, like those found in #15112, we can have a large
eqclass composed of values found in blocks none of whom dominate the
other.  This leads to O(n^2) behavior. The elements are removed one at a
time, with O(n) scans each time.

This CL removes the linear scan by sorting the eqclass so that dominant
values will be sorted first.  As long as we also ensure we don't disturb
the sort order, then we no longer need to scan for the maximally
dominant value.

For the code in issue #15112:

Before:
real    1m26.094s
user    1m30.776s
sys     0m1.125s

Aefter:
real    0m52.099s
user    0m56.829s
sys     0m1.092s

Updates #15112

Change-Id: Ic4f8680ed172e716232436d31963209c146ef850
Reviewed-on: https://go-review.googlesource.com/21981Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
Run-TryBot: Todd Neal <todd@tneal.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>