setAddr was showing up in profiles due to string concatenation construction the os.File name field. netFD.sysfile's Name() is never used, except in dup() so I believe it is safe to avoid this allocation.
R=mikioh.mikioh, rsc
CC=golang-dev
https://golang.org/cl/6742058
- Allow secondary values below the default value in second form. This is
to support before tags for secondary values, as used by Chinese.
- Eliminate collation elements that are guaranteed to be immaterial
after a weight increment.
R=r
CC=golang-dev
https://golang.org/cl/6739051
Previously, multi-byte characters were not allowed. Also certain single-byte
characters, such as '-', were disallowed.
Fixes#3813.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6641052
I just realized that there is no good place for adding
exposed function or method tests because server, unicast
and multicast_test.go do test complicated multiple test
objects, platform behaviros, protocol behaviors and API,
at the same time. Perhaps splitting them into per test
object might be better, so this CL provides tests focused
on API.
R=rsc
CC=gobot, golang-dev
https://golang.org/cl/6501057
PauseNs is a circular buffer of recent pause times, and the
most recent one is at [((NumGC-1)+256)%256].
Also fix comments cross-linking the Go and C definition of
various structs.
R=golang-dev, rsc, bradfitz
CC=golang-dev
https://golang.org/cl/6657047
many small writes to a network may be less efficient that a few
large writes.
This fixes net/http's TestClientWrites, broken by 6565056 that
introduced Writer.ReadFrom. That test implicitly assumed that
calling io.Copy on a *bufio.Writer wouldn't write to the
underlying network until the buffer was full.
R=dsymonds
CC=bradfitz, golang-dev, mchaten, mikioh.mikioh
https://golang.org/cl/6743044
This is part 2 of 2 for issue 4028.
benchmark old ns/op new ns/op delta
BenchmarkWriterCopyOptimal 53293 28326 -46.85%
BenchmarkWriterCopyUnoptimal 53757 30537 -43.19%
BenchmarkWriterCopyNoReadFrom 53192 36642 -31.11%
Fixes#4028.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6565056
The value of cosines are cached in a global array
instead of being recomputed each time.
The test was terribly slow on arm.
R=golang-dev, dave, nigeltao
CC=golang-dev
https://golang.org/cl/6733046
Since this patch changes the way complex literals are written
in export data, there are a few other glitches.
Fixes#4159.
R=golang-dev, rsc
CC=golang-dev, remy
https://golang.org/cl/6674047
A 4-bit window is convenient because 4 divides both 32 and 64,
therefore we never have a window spanning words of the exponent.
Additionaly, the benefit of a 5-bit window is only 2.6% at 1024-bits
and 3.3% at 2048-bits.
This code is still not constant time, however.
benchmark old ns/op new ns/op delta
BenchmarkRSA2048Decrypt 17108590 11180370 -34.65%
Benchmark3PrimeRSA2048Decrypt 13003720 7680390 -40.94%
R=gri
CC=golang-dev
https://golang.org/cl/6716048
Define the properties of the arguments better. In particular,
explain that the path is (sort of) relative to the argument to
Walk.
Fixes#4119.
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6721048
- always return 1 for y <= 0
- document that the sign of m is ignored
- protect against div-0 panics by treating
m == 0 the same way as m == nil
- added extra tests
Fixes#4239.
R=agl, remyoudompheng, agl
CC=golang-dev
https://golang.org/cl/6724046
The RFC doesn't actually have an opinion on whether this is a fatal or
warning level alert, but common practice suggests that it should be a
warning.
This involves rebasing most of the tests.
Fixes#3413.
R=golang-dev, shanemhansen, rsc
CC=golang-dev
https://golang.org/cl/6654050
To be clear, this supports decoding the bytes on the wire into an
in-memory image. There is no API change: jpeg.Decode will still not
return until the entire image is decoded.
The code is obviously more complicated, and costs around 10% in
performance on baseline JPEGs. The processSOS code could be cleaned up a
bit, and maybe some of that loss can be reclaimed, but I'll leave that
for follow-up CLs, to keep the diff for this one as small as possible.
Before:
BenchmarkDecode 1000 2855637 ns/op 21.64 MB/s
After:
BenchmarkDecodeBaseline 500 3178960 ns/op 19.44 MB/s
BenchmarkDecodeProgressive 500 4082640 ns/op 15.14 MB/s
Fixes#3976.
The test data was generated by:
# Create intermediate files; cjpeg on Ubuntu 10.04 can't read PNG.
convert video-001.png video-001.bmp
convert video-005.gray.png video-005.gray.pgm
# Create new test files.
cjpeg -quality 100 -sample 1x1,1x1,1x1 -progressive video-001.bmp > video-001.progressive.jpeg
cjpeg -quality 50 -sample 2x2,1x1,1x1 video-001.bmp > video-001.q50.420.jpeg
cjpeg -quality 50 -sample 2x1,1x1,1x1 video-001.bmp > video-001.q50.422.jpeg
cjpeg -quality 50 -sample 1x1,1x1,1x1 video-001.bmp > video-001.q50.444.jpeg
cjpeg -quality 50 -sample 2x2,1x1,1x1 -progressive video-001.bmp > video-001.q50.420.progressive.jpeg
cjpeg -quality 50 -sample 2x1,1x1,1x1 -progressive video-001.bmp > video-001.q50.422.progressive.jpeg
cjpeg -quality 50 -sample 1x1,1x1,1x1 -progressive video-001.bmp > video-001.q50.444.progressive.jpeg
cjpeg -quality 50 video-005.gray.pgm > video-005.gray.q50.jpeg
cjpeg -quality 50 -progressive video-005.gray.pgm > video-005.gray.q50.progressive.jpeg
# Delete intermediate files.
rm video-001.bmp video-005.gray.pgm
R=r
CC=golang-dev
https://golang.org/cl/6684046
caller of ioutil.TempFile() can use f.Name() to get "pathname"
of the temporary file, instead of just the "name" of the file.
Also remove an out-of-date comment about random number state.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6649054
Better explanation of width for floating-point values.
Explain that scanning does not handle %#q etc.
Fixes#4202.
Fixes#4206.
R=golang-dev, adg, rsc, iant
CC=golang-dev
https://golang.org/cl/6620074
I was an idiot and was thinking that a small base didn't matter
because the exponentiation would quickly make the number the same size
as the modulus. But, of course, the small base continues to make
multiplications unrealistically cheap throughout the computation.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6649048
It was suggested that it's too easy to use crypto/hmac insecurely and
I think that has some merit. This change adds a Equal function to
make it obvious that MAC values should be compared in constant time.
R=rsc, max
CC=golang-dev
https://golang.org/cl/6632044
If source are not available, then the stack looks like:
stack_test.go:40: /tmp/gobuilder/linux-amd64-race-72b15c5d6f65/go/src/pkg/runtime/debug/bla-bla-bla/src/pkg/runtime/debug/stack_test.go:15 (0x43fb11)
stack_test.go:40: /tmp/gobuilder/linux-amd64-race-72b15c5d6f65/go/src/pkg/runtime/debug/bla-bla-bla/src/pkg/runtime/debug/stack_test.go:18 (0x43fb7a)
stack_test.go:40: /tmp/gobuilder/linux-amd64-race-72b15c5d6f65/go/src/pkg/runtime/debug/bla-bla-bla/src/pkg/runtime/debug/stack_test.go:37 (0x43fbf4)
stack_test.go:40: /tmp/gobuilder/linux-amd64-race-72b15c5d6f65/go/src/pkg/testing/bla-bla-bla/src/pkg/testing/testing.go:301 (0x43b5ba)
stack_test.go:40: /tmp/gobuilder/linux-amd64-race-72b15c5d6f65/go/src/pkg/runtime/bla-bla-bla/src/pkg/runtime/proc.c:276 (0x410670)
stack_test.go:40:
which is 6 lines.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6637060
Also:
- Refactored testing framework to permit easier
idempotency testing.
- Applied gofmt -w src misc
This CL depends on CL 6639044 being applied first.
Formatting is not idempotent for all files: In those
files the comment position has changed (due to missing
precise location information) and/or the comment formatting
cannot/is not aware of independent code re-formatting.
In general it is very hard to make format idempotent when
running it in one pass only. Leaving that aside for now.
Fixes#1835.
R=r, rsc
CC=golang-dev
https://golang.org/cl/6624051
Preparation for forthcoming CL 6624051: Will make it
easier to see if/what changes are incurred by it.
The alignment changes in this CL are due to CL 6610051
(fix to alignment heuristic) where it appears that an
old version of gofmt was run (and thus the correct
alignment updates were not done).
R=r
CC=golang-dev
https://golang.org/cl/6639044
I typoed the code and tried to parse all the way to the end of the
message. Therefore it fails when NPN is not the last extension in the
ServerHello.
Fixes#4088.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6637052
This is a part of a bigger change that adds data race detection feature:
https://golang.org/cl/6456044
The purpose of this patch is to provide coarse-grained synchronization
between all Read() and Write() calls.
R=rsc, bradfitz, alex.brainman
CC=golang-dev
https://golang.org/cl/6610064
Check for specific, important misalignment in garbage collector.
Not a complete fix for issue 599 but an important workaround.
Update #599.
R=golang-dev, iant, dvyukov
CC=golang-dev
https://golang.org/cl/6641049
No longer assume that e_lfanew (in the IMAGE_DOS_HEADER strcuture) is always one byte. It is now regarded as a 4 byte uint32.
Fixes#4177.
R=golang-dev, alex.brainman, dave, minux.ma
CC=golang-dev
https://golang.org/cl/6587048
- 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
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 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
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
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
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
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
elimination) in idct.go.
benchmark old ns/op new ns/op delta
BenchmarkIDCT 5649 5610 -0.69%
BenchmarkDecodeRGBOpaque 2948607 2941051 -0.26%
The "type block" declaration moved so that idct.go is compilable
as a stand-alone file: "go tool 6g -S idct.go" works.
R=r
CC=golang-dev
https://golang.org/cl/6619056
The profiler collects goroutine blocking information similar to Google Perf Tools.
You may see an example of the profile (converted to svg) attached to
http://code.google.com/p/go/issues/detail?id=3946
The public API changes are:
+pkg runtime, func BlockProfile([]BlockProfileRecord) (int, bool)
+pkg runtime, func SetBlockProfileRate(int)
+pkg runtime, method (*BlockProfileRecord) Stack() []uintptr
+pkg runtime, type BlockProfileRecord struct
+pkg runtime, type BlockProfileRecord struct, Count int64
+pkg runtime, type BlockProfileRecord struct, Cycles int64
+pkg runtime, type BlockProfileRecord struct, embedded StackRecord
R=rsc, dave, minux.ma, r
CC=gobot, golang-dev, r, remyoudompheng
https://golang.org/cl/6443115
The Go run-time assumes that all SSE floating-point exceptions
are masked so that Go programs are not broken by such invalid
operations. By default, the 64-bit version of the Plan 9 kernel
masks only some SSE floating-point exceptions. Here, we mask
them all on a per-thread basis.
R=rsc, rminnich, minux.ma
CC=golang-dev
https://golang.org/cl/6592056
The AST representation is already identical. Making the
code (nearly) identical in the parser reduces code size
and ensures that the ast.ValueSpec nodes have the same
values (specifically, iota). This in turn permits the
sharing of much of the respective code in the typechecker.
While at it: type functions work now, so use them.
R=r
CC=golang-dev
https://golang.org/cl/6624047
Code for parsing email addresses was already partially part of the public API with "func (Header) AddressList". This CL adds a trivial implementation for two public methods to parse address and lists from a string. With tests.
R=dsymonds
CC=golang-dev
https://golang.org/cl/5676067
Adds a DecryptBlock function which takes a password and a *pem.Block and
returns the decrypted DER bytes suitable for passing into other crypto/x509
functions.
R=golang-dev, agl, leterip
CC=golang-dev
https://golang.org/cl/6555052
They now show the correct name, the byte offset on the line, and context for the failed evaluation.
Before:
template: three:7: error calling index: index out of range: 5
After:
template: top:7:20: executing "three" at <index "hi" $>: error calling index: index out of range: 5
Here top is the template that was parsed to create the set, and the error appears with the action
starting at byte 20 of line 7 of "top", inside the template called "three", evaluating the expression
<index "hi" $>.
Also fix a bug in index: it didn't work on strings. Ouch.
Also fix bug in error for index: was showing type of index not slice.
The real previous error was:
template: three:7: error calling index: can't index item of type int
The html/template package's errors can be improved by building on this;
I'll do that in a separate pass.
Extends the API for text/template/parse but only by addition of a field and method. The
old API still works.
Fixes#3188.
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6576058
This is a backward-compatible API change.
Without the correct <- position information,
certain channel types have incorrect position
information.
R=iant, iant
CC=golang-dev
https://golang.org/cl/6585063
This fixes some example code in the tar package documentation, which
first refers to tar.NewWriter and then to Header, which is inconsistent
because NewWriter and Header are both in the tar namespace.
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6595050
This change updates CL 6576057 for exceptional cases where
return values from Syscall/RawSyscall functions are used.
The system calls return 32-bit integers. With the recent change
in size of `int' in Go for amd64, the type conversion was not
catching `-1' return values. This change makes the conversion
explicitly `int32'.
R=rsc, r
CC=golang-dev
https://golang.org/cl/6590047
The previous implementation was a mess with invariants
maintained inconsistently. Essentially reimplemented
the package:
- used a circular list as internal representation for
significantly simpler implementation with fewer
special cases while maintaining the illusion of
a nil-terminated doubly linked list externally
- more precise documentation
- cleaned up and simplified tests, added test case
for issue 4103.
No changes to the API or documented semantics.
All this said, I would be in favor of removing
this package eventually. container/ring provides
a faster implementation and a simpler and more
powerful API.
Fixes#4103.
R=r
CC=golang-dev
https://golang.org/cl/6569072
The string searching is implemented separately so other functions
may make use of it in the future.
benchmark old ns/op new ns/op delta
BenchmarkSingleMaxSkipping 125889 2474 -98.03%
BenchmarkSingleLongSuffixFail 16252 1996 -87.72%
BenchmarkSingleMatch 260793 136266 -47.75%
benchmark old MB/s new MB/s speedup
BenchmarkSingleMaxSkipping 79.43 4041.57 50.88x
BenchmarkSingleLongSuffixFail 61.65 501.81 8.14x
BenchmarkSingleMatch 57.52 110.08 1.91x
R=nigeltao
CC=golang-dev
https://golang.org/cl/6545049
The system calls return 32-bit integers. With the recent change
in size of `int' in Go for amd64, the type conversion was not
catching `-1' return values. This change makes the conversion
explicitly `int32'.
R=rsc, rminnich, npe, r
CC=golang-dev
https://golang.org/cl/6576057
Silly and small but easy to be consistent.
To make it worthwhile, I eliminated an allocation when using
%x on a byte slice.
Fixes#4149.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6574046
The previous version was created by an idiot. This time, Rog Peppe
wrote the text. Thanks, Rog.
(== doesn't work on slices in general, so it makes no sense to
talk about in the context of DeepEqual.)
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6566054
This code relies on some functions that are not yet in staging,
but it get's harder to keep all this in sync in a piece-meal
fashion.
R=rsc
CC=golang-dev
https://golang.org/cl/6492124
More pieces of the typechecker code:
- Operands are temporary objects representing an expressions's
type and value (for constants). An operand is the equivalent of
an "attribute" in attribute grammars except that it's not stored
but only passed around during type checking.
- Constant operations are implemented in const.go. Constants are
represented as bool (booleans), int64 and *big.Int (integers),
*big.Rat (floats), complex (complex numbers), and string (strings).
- Error reporting is consolidated in errors.go. Only the first
dozen of lines is new code, the rest of the file contains the
exprString and typeString functions formerly in two separate
files (which have been removed).
This is a replacement CL for 6492101 (which was created without
proper use of hg).
R=rsc, r
CC=golang-dev
https://golang.org/cl/6500114
The assembly offsets were converted mechanically using
code.google.com/p/rsc/cmd/asmlint. The instruction
changes were done by hand.
Fixes#2188.
R=iant, r, bradfitz, remyoudompheng
CC=golang-dev
https://golang.org/cl/6550058
Both methods allow to access the IP ancillary data through
socket control messages.
This CL is required for CL 6482044; go.net/ipv4: new package.
R=rsc, r, dave
CC=golang-dev
https://golang.org/cl/6426047
Session resumption saves a round trip and removes the need to perform
the public-key operations of a TLS handshake when both the client and
server support it (which is true of Firefox and Chrome, at least).
R=golang-dev, bradfitz, rsc
CC=golang-dev
https://golang.org/cl/6555051
This CL makes the runtime understand that the type of
the len or cap of a map, slice, or string is 'int', not 'int32',
and it is also careful to distinguish between function arguments
and results of type 'int' vs type 'int32'.
In the runtime, the new typedefs 'intgo' and 'uintgo' refer
to Go int and uint. The C types int and uint continue to be
unavailable (cause intentional compile errors).
This CL does not change the meaning of int, but it should make
the eventual change of the meaning of int on amd64 a bit
smoother.
Update #2188.
R=iant, r, dave, remyoudompheng
CC=golang-dev
https://golang.org/cl/6551067
Using offsets from Tos is cumbersome and we've had problems
in the past. Since it's only being used to grab the PID, we'll just
get that from the default TLS instead.
R=rsc, rminnich, npe
CC=golang-dev
https://golang.org/cl/6543049
various implementation of collation. The tool provides commands for soring,
regressing one implementation against another, and benchmarking.
Currently it includes collation implementations for the Go collator, ICU,
and one using Darwin's CoreFoundation framework.
To avoid building this tool in the default build, the colcmp tag has been
added to all files. This allows other tools/colcmp in this directory (e.g. it may make
sense to move maketables here) to be put in this directory as well.
R=r, rsc, mpvl
CC=golang-dev
https://golang.org/cl/6496118
This CL fixes code that incorrectly assumes that int is 32 bits wide.
Specifically, the socketpair system call expects a pointer to a pair
of int32s, not a pair of ints. Fix this inside the wrappers without
changing the APIs.
Update #2188.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6552063
Change the grammar so that field access is a proper operator.
This introduces a new node, ChainNode, into the public (but
actually internal) API of text/template/parse. For
compatibility, we only use the new node type for the specific
construct, which was not parseable before. Therefore this
should be backward-compatible.
Before, .X.Y was a token in the lexer; this CL breaks it out
into .Y applied to .X. But for compatibility we mush them
back together before delivering. One day we might remove
that hack; it's the simple TODO in parse.go/operand.
This change also provides grammatical distinction between
f
and
(f)
which might permit function values later, but not now.
Fixes#3999.
R=golang-dev, dsymonds, gri, rsc, mikesamuel
CC=golang-dev
https://golang.org/cl/6494119
The code was trying to interpret a uint32 n as a signed quantity
and then store it in an int. For this, int(n) currently works, but when
int becomes 64 bits one must write int(int32(n)) instead, to get
the 32-bit sign extension.
Update #2188.
R=golang-dev, nigeltao
CC=golang-dev
https://golang.org/cl/6551068
After each line, ast.Print would do a zero-length write,
which would hit the boundary condition on Plan 9 when
reading over pipes (since message boundaries are
preserved). This change makes sure we only do positive-
length writes.
R=rsc, rminnich, dave, r
CC=golang-dev
https://golang.org/cl/6558046
*** There is an API change here: the introduction of the
LatinOffset int in the RangeTable struct. ***
* Avoid checking Latin range multiple times for non-Latin runes.
* Use linear search when it is faster than binary search.
go test -calibrate runs the calibration for where the linear/binary
crossover should be.
benchmark old MB/s new MB/s speedup
BenchmarkFields 36.27 41.43 1.14x
BenchmarkFieldsFunc 36.23 41.38 1.14x
The speedup here is evenly split between the linear scans
and the LatinOffset change. Both are about 1.07x.
R=r
CC=bradfitz, gobot, golang-dev
https://golang.org/cl/6526048
Previously we tried to parse email constraints as the maximum base
distance (which is unused and must be omitted according to RFC 5280)
because they share the same tag number. Rather than complicate the
code further, this CL just ignores the unused values.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6552044
This is the first part of a bigger change that adds data race detection feature:
https://golang.org/cl/6456044
Adds -race flag to go command.
API change:
+pkg go/build, type Context struct, InstallTag string
R=rsc
CC=golang-dev
https://golang.org/cl/6488075
The change is a preparation for the new scheduler.
It introduces runtime.park() function,
that will atomically unlock the mutex and park the goroutine.
It will allow to remove the racy readyonstop flag
that is difficult to implement w/o the global scheduler mutex.
R=rsc, remyoudompheng, dave
CC=golang-dev
https://golang.org/cl/6501077
Fixes#3456.
This proposal is a reformulation of CL 5987063. This CL resets
the default GOARM value to 6 and allows the use of the VFPv3
optimisation if GOARM=7. Binaries built with this CL in place
will abort if GOARM=7 was used and the target host does not
support VFPv3.
R=minux.ma, rsc, ajstarks
CC=golang-dev
https://golang.org/cl/6501099
This also fixes the semantics of some corner cases with the empty
match. TODOs for genericReplacer in the tests are fixed.
benchmark old ns/op new ns/op delta
BenchmarkGenericNoMatch 71395 3132 -95.61%
BenchmarkGenericMatch1 75610 20280 -73.18%
BenchmarkGenericMatch2 837995 86725 -89.65%
R=nigeltao, rsc
CC=golang-dev
https://golang.org/cl/6492076
Give the right name for errors, and add a test to check we're
getting the errors we expect.
Also fix an ordering bug (calling add after stopParse) that
caused a nil indirection rather than a helpful error.
Fixes#3280.
R=golang-dev, adg
CC=golang-dev
https://golang.org/cl/6520043
Command.Start could crash before if no fds were available
because a nil *os.File of /dev/null was added to the cleanup
list, which crashed before returning the proper error.
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6514043
When reading Go files, read through import block.
When reading non-Go files, read only leading comments.
R=nigeltao, adg, r
CC=golang-dev
https://golang.org/cl/6493068
instead of variables. Several reasons:
- Encourage users of the API to minimize the number of creations and reuse Collate objects.
- Don't rule out the possibility of using initialization code for collators. For some locales
it will be possible to have very compact representations that can be quickly expanded
into a proper table on demand.
Other changes:
- Change name of root* vars to main*, as the tables are shared between locales.
- Added Locales() method to get a list of supported locales.
R=r
CC=golang-dev
https://golang.org/cl/6498107
X509KeyPair wasn't really supposed to allow the certificate and
key to be in the same file, but it did work if you put the key
first. Since some HTTPS servers support loading keys and certs
like this, this change makes it work in either order.
Fixes#3986.
R=golang-dev, dave, rsc
CC=golang-dev
https://golang.org/cl/6499103
This is the easy part of issue 3528.
(What to do about "noescape" is the hard part, left open.)
Update #3528.
R=mikesamuel, r, dsymonds
CC=golang-dev
https://golang.org/cl/6493113
Right now we only have 32-bit ints so that's a no-op.
Took the opportunity to check for some other invalid values too.
Suggestions for additions or modifications welcome.
R=agl
CC=golang-dev
https://golang.org/cl/6493112
This fixes a problem with ELF tools thinking they know the
format of the symbol table, as we do not use any of the
standard formats for that table.
This change will probably annoy the Plan 9 users, but I
believe there are other incompatibilities already that mean
they have to use a Go-specific nm.
Fixes#3473.
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6500117
SO_PEERCRED on unix domain socket will fill a Ucred struct,
thus linux needs a custom Getsockopt variant.
Fixes#3836.
R=golang-dev, rsc, iant
CC=golang-dev
https://golang.org/cl/6445104
Rename the first argument of CompareAndSwapT and AddT s/val/addr/
for consistency with LoadT and StoreT.
R=rsc, r, dvyukov
CC=golang-dev
https://golang.org/cl/6494112
This verifies existing behavior. Some replacements are arguably wrong
(these are marked with TODO) but changing behavior is left for a
follow-up CL.
Also fix that BenchmarkGenericMatch wasn't actually matching anything.
R=rsc, eric.d.eisner
CC=bradfitz, golang-dev
https://golang.org/cl/6488110
First set of type checker files for review.
The primary concern here is the typechecker
API (types.go).
R=rsc, adonovan, r, rogpeppe
CC=golang-dev
https://golang.org/cl/6490089
Also, clear Content-Type and Content-Length on Not Modified
responses before server.go strips them and spams the logs with
warnings.
R=rsc
CC=golang-dev
https://golang.org/cl/6503090
Refactored build + buildTrie into build + buildOrdering.
Note that since the tailoring code is not checked in yet, all tailorings are identical
to root. The table therefore should not and does not grow at this point.
R=r
CC=golang-dev
https://golang.org/cl/6500087
Fixes#3911.
Requires CL 6449127.
dfc@qnap:~$ ./runtime.test
runtime: this CPU has no floating point hardware, so it cannot run
this GOARM=7 binary. Recompile using GOARM=5.
R=rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6442109
Fixes#3862.
There were many areas where conn.err was being accessed
outside the mutex. This proposal moves the err value to
an embedded struct to make it more obvious when the error
value is being accessed.
As there are no Benchmark tests in this package I cannot
feel confident of the impact of this additional locking,
although most will be uncontended.
R=dvyukov, agl
CC=golang-dev
https://golang.org/cl/6497070
order.
JPEG images generated prior to this CL are still valid JPEGs, as the
quantization tables used are encoded in the wire format. Such JPEGs just
don't use the recommended quantization tables.
R=r, dsymonds, raph, adg
CC=golang-dev, tuom.larsen
https://golang.org/cl/6497083
The old code was a depth first graph traversal that could, under the
right conditions, end up re-exploring the same subgraphs multiple
times, once for each way to arrive at that subgraph at a given depth.
The new code uses a breadth first search to make sure that it only
visits each reachable embedded struct once.
Also add fast path for the trivial case.
benchmark old ns/op new ns/op delta
BenchmarkFieldByName1 1321 187 -85.84%
BenchmarkFieldByName2 6118 5186 -15.23%
BenchmarkFieldByName3 8218553 42112 -99.49%
R=gri, r
CC=golang-dev
https://golang.org/cl/6458090
All of them call `newFileFD' which must properly restore close-on-exec on
duplicated fds.
R=golang-dev, bradfitz, mikioh.mikioh
CC=golang-dev
https://golang.org/cl/6445081
Signal handlers are global resources but many language
environments (Go, C++ at Google, etc) assume they have sole
ownership of a particular handler. Signal handlers in
mixed-language applications must therefore be robust against
unexpected delivery of certain signals, such as SIGPROF.
The default Go signal handler runtime·sigtramp assumes that it
will never be called on a non-Go thread, but this assumption
is violated by when linking in C++ code that spawns threads.
Specifically, the handler asserts the thread has an associated
"m" (Go scheduler).
This CL is a very simple workaround: discard SIGPROF delivered to non-Go threads. runtime.badsignal(int32) now receives the signal number; if it returns without panicking (e.g. sig==SIGPROF) the signal is discarded.
I don't think there is any really satisfactory solution to the
problem of signal-based profiling in a mixed-language
application. It's not only the issue of handler clobbering,
but also that a C++ SIGPROF handler called in a Go thread
can't unwind the Go stack (and vice versa). The best we can
hope for is not crashing.
Note:
- I've ported this to all POSIX platforms, except ARM-linux which already ignores unexpected signals on m-less threads.
- I've avoided tail-calling runtime.badsignal because AFAICT the 6a/6l don't support it.
- I've avoided hoisting 'push sig' (common to both function calls) because it makes the code harder to read.
- Fixed an (apparently incorrect?) docstring.
R=iant, rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6498057
Added instructions for starting an http server
to the godoc header for this package. With the old
instructions, the example "go tool pprof..." commands
wouldn't work unless you happen to be running an http
server on port 6060 in your application.
R=golang-dev, minux.ma, adg, giacomo.tartari
CC=golang-dev
https://golang.org/cl/6483049
for both locale-specific exemplar characters and tailorings to
the collation table.
Some specifices:
- Moved stringSet to the beginning of the file and added some functionality
to parse command line files.
- openReader now modifies the input URL for localFiles to guarantee that
any http source listed in the generated file is indeed this source.
- Note that the implementation of the Tailoring API used by maketables.go
is not yet checked in. So for now adding tailorings are simply no-ops.
- The generated file of exemplar characters will be used somewhere else.
Here is a snippet of how the body of the generated file looks like:
type exemplarType int
const (
exCharacters exemplarType = iota
exContractions
exPunctuation
exAuxiliary
exCurrency
exIndex
exN
)
var exemplarCharacters = map[string][exN]string{
"af": [exN]string{
0: "a á â b c d e é è ê ë f g h i î ï j k l m n o ô ö p q r s t u û v w x y z",
3: "á à â ä ã æ ç é è ê ë í ì î ï ó ò ô ö ú ù û ü ý",
4: "a b c d e f g h i j k l m n o p q r s t u v w x y z",
},
...
}
R=r
CC=golang-dev
https://golang.org/cl/6501070
- Elements in the array are now sorted as a linked list. This makes it easier to
apply tailorings.
- Added code to sort entries by collation elements.
- Added logical reset points. This is used for tailoring relative to certain
properties, rather than characters.
NOTE: all code for type entry should now be in order.go. To keep the diffs for
this CL reasonable, though, the existing code is left in builder.go. I'll move
this in a separate CL.
R=r
CC=golang-dev
https://golang.org/cl/6493063
Fixes#3525.
PTRACE_SYSCALL behaves like PTRACE_CONT and can deliver
a signal to the process. Ideally PtraceSingleStep should
support the signal argument, but its interface is frozen
by Go1.
R=golang-dev, r, rsc
CC=golang-dev
https://golang.org/cl/6353051
Parses a time header value into a time.Time according to rfc2616 sec 3.3.
R=golang-dev, dave, rsc, r
CC=bradfitz, golang-dev
https://golang.org/cl/6344046
A crash happens in the first request in a connection
if "params" field is missing because c.req.Params is Nil.
Fixes#3848.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6446051
Reverts part of CL 6460082.
If a doc comment describes a type by explaining the
meaning of one instance of the type, a leading article
is fine and makes the text less awkward.
Compare:
// A dog is a kind of animal.
// Dog is a kind of animal.
R=golang-dev, dsymonds, dvyukov, r
CC=golang-dev
https://golang.org/cl/6494066
This set of changes extends the Plan 9 support
to include the AMD64 architecture and should
work on all versions of Plan 9.
R=golang-dev, rminnich, noah.evans, rsc, minux.ma, npe
CC=akskuma, golang-dev, jfflore, noah.evans
https://golang.org/cl/6479052
The Handler method makes the ServeMux dispatch logic
available to wrappers that enforce additional constraints
on requests.
R=golang-dev, bradfitz, dsymonds
CC=golang-dev
https://golang.org/cl/6450165
In serve mux, if pattern contains a host name, pass only the path to
the redirect handler.
Add tests for serve mux redirections.
R=rsc
CC=bradfitz, gobot, golang-dev
https://golang.org/cl/6329045
Also rename Node.{Add,Remove} to Node.{AppendChild,RemoveChild} to
be consistent with the DOM.
benchmark old ns/op new ns/op delta
BenchmarkParser 4042040 3749618 -7.23%
benchmark old MB/s new MB/s speedup
BenchmarkParser 19.34 20.85 1.08x
BenchmarkParser mallocs per iteration is also:
10495 before / 7992 after
R=andybalholm, r, adg
CC=golang-dev
https://golang.org/cl/6495061
This is required by the spec to produce the replacement char.
The fix lies in lib9's rune code.
R=golang-dev, nigeltao, rsc
CC=golang-dev
https://golang.org/cl/6443109
Other than catching an error case that was missed before, this
CL introduces no changes to the template language or API.
For simplicity, templates use spaces as argument separators.
This means that spaces are significant: .x .y is not the same as .x.y.
In the existing code, these cases are discriminated by the lexer,
but that means for instance that (a b).x cannot be distinguished
from (a b) .x, which is lousy. Although that syntax is not
supported yet, we want to support it and this CL is a necessary
step.
This CL emits a "space" token (actually a run of spaces) from
the lexer so the parser can discriminate these cases. It therefore
fixes a couple of undisclosed bugs ("hi".x is now an error) but
doesn't otherwise change the language. Later CLs will amend
the grammar to make .X a proper operator.
There is one unpleasantness: With space a token, three-token
lookahead is now required when parsing variable declarations
to discriminate them from plain variable references. Otherwise
the change isn't bad.
The CL also moves the debugging print code out of the lexer
into the test, which is the only place it's needed or useful.
Step towards resolving issue 3999.
It still remains to move field chaining out of the lexer
and into the parser and make field access an operator.
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6492054
Fixes#3892.
Swapping the order of the writers inside the MultiWriter ensures
the request will be written to buf before http.ReadRequest completes.
The fencedBuffer is not required to make the test pass on
any machine that I have access too, but as the buf is shared
across goroutines, I think it is necessary for correctness.
R=bradfitz, fullung, franciscossouza
CC=golang-dev
https://golang.org/cl/6483061
Use version 2 of the NetBSD signal ABI - both version 2 and version 3
are supported by the kernel, with near identical behaviour. However,
the netbsd32 compat code does not allow version 3 to be used, which
prevents Go netbsd/386 binaries from running in compat mode on a
NetBSD amd64 kernel. Switch to version 2 of the ABI, which is the
same version currently used by NetBSD's libc.
R=minux.ma
CC=golang-dev
https://golang.org/cl/6476068
This shouldn't be an error (see issue 3999), but until it's handled
correctly, treat it as one to avoid confusion. Without this CL,
(A).X parses as two arguments.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6473059
Based on work by Russ Cox. From his CL:
This is generally useful but especially helpful when trying
to use the built-in boolean operators. It lets you write:
{{if not (f 1)}} foo {{end}}
{{if and (f 1) (g 2)}} bar {{end}}
{{if or (f 1) (g 2)}} quux {{end}}
instead of
{{if f 1 | not}} foo {{end}}
{{if f 1}}{{if g 2}} bar {{end}}{{end}}
{{$do := 0}}{{if f 1}}{{$do := 1}}{{else if g 2}}{{$do := 1}}{{end}}{{if $do}} quux {{end}}
The result can be a bit LISPy but the benefit in expressiveness and readability
for such a small change justifies it.
I believe no changes are required to html/template.
Fixes#3276.
R=golang-dev, adg, rogpeppe, minux.ma
CC=golang-dev
https://golang.org/cl/6482056
In the regtest data, surrogates are assigned primary weights based on
the surrogate code point value. Go now converts surrogates to FFFD, however,
meaning that the primary weight is based on this code point instead.
This change drops tests with surrogates and lets the tests pass.
R=r
CC=golang-dev
https://golang.org/cl/6461100
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
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
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
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
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
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
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
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
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
Now that the parser passes all tests in the test suite,
it is no longer necessary to keep track of which tests
pass and which don't. So remove the testlogs directory
and the code that uses it.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6453124
This currently fails on NetBSD due to the cloned file descriptors
that result from opening /dev/urandom. Disable the additional checking
until this is investigated and properly fixed.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6443129
Disable the crash handler test on NetBSD until I can figure out why
it triggers failures in later tests.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6460090
All of the remaining tests that had as status of PARSE rather than PASS had
good reasons for not passing the render-and-reparse step: the correct parse tree is
badly formed, so when it is rendered out as HTML, the result doesn't parse into the
same tree. So add them to the list of tests where that step is skipped.
Also, I discovered that it is possible to end up with HTML elements (not just text)
inside a raw text element through reparenting. So change the rendering routines to
handle that situation as sensibly as possible (which still isn't very sensible, but
this is HTML5).
R=nigeltao
CC=golang-dev
https://golang.org/cl/6446137
Optimize some common cases.
benchmark old ns/op new ns/op delta
BenchmarkScanFile 718907 667960 -7.09%
benchmark old MB/s new MB/s speedup
BenchmarkScanFile 23.03 25.51 1.11x
R=r
CC=golang-dev
https://golang.org/cl/6454150