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
cgo[1-4].go, go1.go couldn't be tested now
(cgo[1-4].go can only be tested when cgo is enabled, go1.go
contain a list of filenames in the current directory)
R=golang-dev, alex.brainman, rsc
CC=golang-dev
https://golang.org/cl/6218048
Accomplished by synchronizing the formatting of conversion errors between typecheck.c and subr.c
Fixes#3984.
R=golang-dev, remyoudompheng, rsc
CC=golang-dev
https://golang.org/cl/6500064
There was mail on golang-nuts a few weeks ago
from someone who understood the message perfectly
and knew he had a cyclic dependency but assumed
that Go, like Python or Java, was supposed to handle it.
R=golang-dev, bradfitz, dave
CC=golang-dev
https://golang.org/cl/6488069
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
This fixes a spurious 'invalid recursive type' error, and stops the compiler from emitting errors on uses of the invalid type.
Fixes#3766.
R=golang-dev, dave, minux.ma, rsc
CC=golang-dev
https://golang.org/cl/6443100
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
In the example "units" program for goyacc, the exchange rates were
reciprocals of the correct amounts. Turn them right-side-up
and update them to current figures.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6495053
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
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
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
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
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
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
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
- 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
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
This CL adds a step to the build procedure for cgo programs. It uses 'ld -r'
to combine all gcc compiled object file and generate a relocatable object file
for our ld. Additionally, this linking step will combine some static linking
gcc library into the relocatable object file, so that we can use libgcc,
libmingwex and libmingw32 without problem.
Fixes#3261.
Fixes#1741.
Added a testcase for linking in libgcc.
TODO:
1. still need to fix the INDIRECT_SYMBOL_LOCAL problem on Darwin/386.
2. still need to enable the libgcc test on Linux/ARM, because 5l can't deal
with thumb libgcc.
Tested on Darwin/amd64, Darwin/386, FreeBSD/amd64, FreeBSD/386, Linux/amd64,
Linux/386, Linux/ARM, Windows/amd64, Windows/386
R=iant, rsc, bradfitz, coldredlemur
CC=golang-dev
https://golang.org/cl/5822049
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
When generating replacement elements for an <isindex> tag, the old
addSyntheticElement method was producing the wrong nesting. Replace
it with parseImpliedToken.
Pass the one remaining test in the test suite.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6453114
If a tag doesn't have a closing '>', it isn't considered a tag;
it is just ignored and EOF is returned instead.
Pass one additional test in the test suite.
Change tokenizer tests to match correct behavior.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6454131
In HTML content, having a self-closing tag is a parse error unless
the tag would be self-closing anyway (like <img>). The only place a
self-closing tag actually makes a difference is in XML-based foreign
content.
Pass 1 additional test.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6450109
Also simplified parsing of interface
types since they can only contain
methods (and no embedded interfaces)
in the export data.
R=rsc
CC=golang-dev
https://golang.org/cl/6446084
The keyword reprents an untyped nil and is useful for
passing nil values to methods and functions. The
nil will be promoted to the appropriate type when
used; if a type cannot be assigned, an error results.
R=rsc, dsymonds
CC=golang-dev
https://golang.org/cl/6459056
Although I don't use PAX enabled ARM kernels, PAX
does have support for ARM, so we're better off add
PT_PAX_FLAGS now in case people use PAX kernels.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6453092
Surrogate halves are part of UTF-16 and should never appear in UTF-8.
(The rune that two combined halves represent in UTF-16 should
be encoded directly.)
Encoding: encode as RuneError.
Decoding: convert to RuneError, consume one byte.
This requires changing:
package unicode/utf8
runtime for range over string
Also added utf8.ValidRune and fixed bug in utf.RuneLen.
Fixes#3927.
R=golang-dev, rsc, bsiegert
CC=golang-dev
https://golang.org/cl/6458099
If a table contained whitespace, text nodes would not get foster parented
correctly.
Pass 1 additional test.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6459054
Depends on CL 6197045.
Result obtained on Core i7 620M, Darwin/amd64:
benchmark old ns/op new ns/op delta
BenchmarkComplex128DivNormal 57 28 -50.78%
BenchmarkComplex128DivNisNaN 49 15 -68.90%
BenchmarkComplex128DivDisNaN 49 15 -67.88%
BenchmarkComplex128DivNisInf 40 12 -68.50%
BenchmarkComplex128DivDisInf 33 13 -61.06%
Result obtained on Core i7 620M, Darwin/386:
benchmark old ns/op new ns/op delta
BenchmarkComplex128DivNormal 89 50 -44.05%
BenchmarkComplex128DivNisNaN 307 802 +161.24%
BenchmarkComplex128DivDisNaN 309 788 +155.02%
BenchmarkComplex128DivNisInf 278 237 -14.75%
BenchmarkComplex128DivDisInf 46 22 -52.46%
Result obtained on 700MHz OMAP4460, Linux/ARM:
benchmark old ns/op new ns/op delta
BenchmarkComplex128DivNormal 1557 465 -70.13%
BenchmarkComplex128DivNisNaN 1443 220 -84.75%
BenchmarkComplex128DivDisNaN 1481 218 -85.28%
BenchmarkComplex128DivNisInf 952 216 -77.31%
BenchmarkComplex128DivDisInf 861 231 -73.17%
The 386 version has a performance regression, but as we have
decided to use SSE2 instead of x87 FPU for 386 too (issue 3912),
I won't address this issue.
R=dsymonds, mchaten, iant, dave, mtj, rsc, r
CC=golang-dev
https://golang.org/cl/6024045
The <title> element was getting removed from the stack of open elements,
when its parent, the <head> element should have been removed instead.
Pass 2 additional tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6449101
This is an attempt to fix windows build. Will restore
once we know what the problem is.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6461046
our old choice is not working properly at least on VFPv2 in
ARM1136JF-S (it's not preserved across float64->float32 conversions).
Fixes#3745.
R=dave, rsc
CC=golang-dev
https://golang.org/cl/6344078
Enhances test/run.go to support testing other directories
Will enable stdio tests on Windows in a follow-up CL.
R=golang-dev, alex.brainman, rsc
CC=golang-dev
https://golang.org/cl/6220049
When an element (like <nobr> or <p>) was implicitly closed by another
start tag, it would keep foster parenting from working because
the check for what was on top of the stack of open elements was
in the wrong place.
Move the check to addChild.
Pass 2 additional tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6460045
Hold ForkLock during dup of fd + cloexec in the net pkg,
per the locking policy documented in syscall/exec_unix.go.
R=golang-dev, dsymonds, adg
CC=golang-dev
https://golang.org/cl/6457080
Since NUL usually terminates strings in underlying syscalls, allowing
it when converting string arguments is a security risk, especially
when dealing with filenames. For example, a program might reason that
filename like "/root/..\x00/" is a subdirectory or "/root/" and allow
access to it, while underlying syscall will treat "\x00" as an end of
that string and the actual filename will be "/root/..", which might
be unexpected. Returning EINVAL when string arguments have NUL in
them makes sure this attack vector is unusable.
R=golang-dev, r, bradfitz, fullung, rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6458050
Too many people use it without reading what it does.
Those people want ReadBytes or ReadString.
Fixes#3906.
R=golang-dev, iant, r
CC=golang-dev
https://golang.org/cl/6442087
The improvement is obtained by eliminating the zero
initialization of a large structure that is only
needed when the fast path fails.
Also add a missing roundtrip test for float32s.
benchmark old ns/op new ns/op delta
BenchmarkAppendFloatDecimal 301 180 -40.20%
BenchmarkAppendFloat 486 388 -20.16%
BenchmarkAppendFloatExp 492 383 -22.15%
BenchmarkAppendFloatNegExp 478 370 -22.59%
BenchmarkAppendFloatBig 650 541 -16.77%
BenchmarkAppendFloat32Integer 308 180 -41.56%
BenchmarkAppendFloat32ExactFraction 449 333 -25.84%
BenchmarkAppendFloat32Point 494 390 -21.05%
BenchmarkAppendFloat32Exp 488 387 -20.70%
BenchmarkAppendFloat32NegExp 488 378 -22.54%
R=r, rsc
CC=golang-dev, remy
https://golang.org/cl/6346081
The content of an HTML <title> element is RCDATA, but the content of an SVG
<title> element is parsed as tags. Now the parser doesn't go into RCDATA
mode in foreign content.
Pass 4 additional tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6448111
I have C functions implemented in .syso file (rather than .so or inlined in .go file).
W/o this change the gcc invocation fails with undefined symbols.
R=minux.ma, rsc
CC=golang-dev
https://golang.org/cl/6352076
The compiler is incorrectly rejecting switches on arrays of
comparable types. It also doesn't catch incomparable structs
when typechecking the switch, leading to unreadable errors
during typechecking of the generated code.
Fixes#3894.
R=rsc
CC=gobot, golang-dev, r, remy
https://golang.org/cl/6442074
These aren't needed for scalar multiplication, but since we export a
generic Add function we should handle it.
This change also corrects two bugs in p224Contract that it turned up.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6458076
PAX systems are Linux systems that are more paranoid about memory permissions.
These flags tell them to relax when running Go binaries.
Fixes#47.
R=iant
CC=golang-dev
https://golang.org/cl/6326054
This removes some (non-short only) tests that involved connections to
the outside world and thus were flakey. This does remove some test
coverage of the root fetchers on non-Windows platforms, but the right
place for that is crypto/x509.
R=golang-dev, krautz, rsc
CC=golang-dev
https://golang.org/cl/6455085
We already support reading ECDSA certificates and this change adds
write support.
R=golang-dev, bradfitz, rsc
CC=golang-dev
https://golang.org/cl/6422046
for dealing with CLDR files:
- Add now taxes a list of indexes of colelems that are variables. Checking and
handling is now done by the Builder. VariableTop is now also properly generated
using the Build method.
- Introduced separate Builder, called Tailoring, for creating tailorings of root
table. This clearly separates the functionality for building a table based on
weights (the allkeys* files) versus tables based on LDML XML files.
- Tailorings are now added by two calls instead of one: SetAnchor and Insert.
This more closely reflects the structure of LDML side and simplifies the
implementation of both the client and library side. It also preserves
some information that is otherwise hard to recover for the Builder.
- Allow the LDML XML element extend to be passed to Insert. This simplifies
both client and library implementation.
R=r
CC=golang-dev
https://golang.org/cl/6454061
Don't print unexported struct fields; their values are
not accessible via reflection.
Fixes#3898.
Also:
- added support for arrays
- print empty maps, arrays, slices, structs on one line
for a denser output
- added respective test cases
R=r
CC=golang-dev
https://golang.org/cl/6454089
Process a package's object in a reproducible
order (rather then in map order) so that we
get error messages in reproducible order.
R=r
CC=golang-dev
https://golang.org/cl/6449076
The text inside <script> tags is not ordinary raw text; there are all sorts
of other complications. This CL implements those complications.
Pass 76 additional tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6443070
This is more in sync with the rest of the package;
for instance, we have functions (not methods) to
deref or find the underlying type of a Type.
In the process use a single bytes.Buffer to create
the string representation for a type rather than
the (occasional) string concatenation.
R=r
CC=golang-dev
https://golang.org/cl/6458057
If an end tag has an attribute that is a quoted string containing '>',
the tokenizer would end the tag prematurely. Now it reads the attributes
on end tags just as it does on start tags, but the high-level interface
still doesn't return them, because their presence is a parse error.
Pass 1 additional test.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6457060
The receive operator was given incorrect precedence
resulting in incorrect deletion of parentheses.
Fixes#3843.
R=rsc
CC=golang-dev, remy
https://golang.org/cl/6442049
LLVM-based gcc will place all-zero data in a zero-filled
section, but our debug/macho can't handle that.
Fixes#3821.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6444049
If the X-Forwarded-For header already exists on a request, we
should append our client's IP to it after a comma+space instead
of overwriting it.
Fixes#3846.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6448053
This syscall was inadvertently exported when fixing
Getrlimit/Setrlimit on 32-bit platforms.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6458051
The situation only affects diagnostics but is easy to fix.
When computing lineNumber, use the position of the last item
returned by nextItem rather than the current state of the lexer.
This is internal only and does not affect the API.
Fixes#3886.
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6445061
They were previously ignored when deciding order and
detecting dependency loops.
Fixes#3824.
R=rsc, golang-dev
CC=golang-dev, remy
https://golang.org/cl/6455055
Should probably be fixed in the syscall package, either
additional or instead of this CL.
Fixes#3842
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6442061
This fixes a data race (usually just harmlessly updating
"GET" to "GET"), but also follows RFC 2616 Sec 5.1.1 which
says that the request method is case-sensitive.
Fixes#3881
R=golang-dev, rsc, dsymonds
CC=golang-dev
https://golang.org/cl/6446063
Darwin kernels have a bug in accept() where error result from
an internal call is not checked and socket is accepted instead
of ECONNABORTED error. However, such sockets have no sockaddr,
which results in EAFNOSUPPORT error from anyToSockaddr, making
Go http servers running on Mac OS X easily susceptible to
denial of service from simple port scans with nmap.
Fixes#3849.
R=golang-dev, adg, mikioh.mikioh
CC=golang-dev
https://golang.org/cl/6456045
- Allow handles into the trie for different locales. Multiple tables share the same
try to allow for reuse of blocks.
- Significantly improved memory footprint and reduced allocations of trieNodes.
This speeds up generation by about 30% and allows keeping trieNodes around
for multiple locales during generation.
- Renamed print method to fprint.
R=r
CC=golang-dev
https://golang.org/cl/6408052
pkg log already appends a linefeed to the log message,
so log/syslog doesn't need to append another.
R=golang-dev, bradfitz, r
CC=golang-dev
https://golang.org/cl/6441048
When a cgo program calls setuid, setgid, etc., the GNU/Linux
pthread library sends signal SIGSETXID to each thread to tell
it to update its UID info. If Go is permitted to intercept
the default SIGSETXID signal handler, the program will hang.
This patch tells the runtime package to not try to intercept
SIGSETXID on GNU/Linux. This will be odd if a Go program
wants to try to use that signal, but it means that cgo
programs that call setuid, etc., won't hang.
Fixes#3871.
R=rsc, r, minux.ma, bradfitz
CC=golang-dev
https://golang.org/cl/6455050
The go/ast comment for FuncType.Params says that the field may be nil.
Make sure the printer accepts such a value. The go/parser always sets
the field (to provide parenthesis position information), but a program
creating a Go AST from scatch may not.
Added corresponding test case.
Fixes#3870.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6448060
If NUL bytes occur inside certain elements, convert them to U+FFFD
replacement character.
Pass 1 additional test.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6452047
Fixed creation of Func's, taking IsVariadic from parameter list rather
than results.
Updated checkObj to process ast.Fun objects.
R=gri
CC=golang-dev
https://golang.org/cl/6402046
Declare crosscall2. Declare the functions passed to it as
returning void, rather than relying on implicit return type.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6432060
If the body of an HTML document contains text, the <frameset> tag is
ignored. But not if the text is only whitespace.
Pass 4 additional tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6442043
If the key is not present, return value of the type of the element
not the type of the key. Also fix a test that should have caught this case.
Fixes#3850.
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6405078
Add commentary to explain better what's going on, but the
code change is a simple one-line reversal to the previous
form.
R=rsc
CC=golang-dev
https://golang.org/cl/6428072
Don't unescape entities in attributes when they don't end with
a semicolon and they are followed by '=', a letter, or a digit.
Pass 6 more tests from the WebKit test suite, plus one that was
commented out in token_test.go.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6405073
Surrogates are still admitted, but I have sent mail to golang-dev on that topic.
Fixes#3785.
R=golang-dev, rogpeppe, iant
CC=golang-dev
https://golang.org/cl/6398049
If a Client's CheckRedirect function returns an error, we
again return both a non-nil *Response and a non-nil error.
Fixes#3795
R=golang-dev, n13m3y3r
CC=golang-dev
https://golang.org/cl/6429044
By default, all are still run, but a particular test can be
selected with the new flags.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6395053
Some class of bugs (data races, runtime bugs) can be found
only with real parallelism.
Note that GOMAXPROCS=32 is somewhat different from go test -cpu=32,
this intentionally uses GOMAXPROCS to stress program bootstrap,
testing code, garbage collections, etc.
Package selection is mostly random.
R=golang-dev, dave, r
CC=golang-dev
https://golang.org/cl/6346070
They can generate huge amounts of memory, causing failure on
small machines. Also they can be very slow. So slow that one test
was commented out! We uncomment it and use a flag.
Fixes#3742.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6373044
- Changed the representation of colElem to support a few cases
for some languages not supported by the current format.
- Changed offsets for implicit primary values. This makes the
values both easier to read and debug (last 4 nibbles are identical to
implicit primary value) and also results in better packing.
- Fixed bug in weight conversion code that did not pop up yet by
sheer luck.
Note that tables.go also includes changes to the contraction trie
from CL 6346092.
R=r, mpvl
CC=golang-dev
https://golang.org/cl/6392060
which has a rather large contraction table. The value of the next state
offset now starts after the current block, instead of before. This is
slightly less efficient (on extra addition per state change), but gives
some extra range for the offsets.
Also introduced constants for final (0) and noIndex (0xFF).
tables.go is updated in a separate CL.
R=r
CC=golang-dev
https://golang.org/cl/6346092
Gob decoding reads a whole message into memory and then
copies it into a bytes.Buffer. For large messages this wastes
an entire copy of the message. In this CL, we use a staging
buffer to avoid the large temporary.
Update #2539
RSS drops to 775MB from 1GB.
Active memory drops to 858317048 from 1027878136,
essentially the size of one copy of the input file.
R=dsymonds, nigeltao
CC=golang-dev
https://golang.org/cl/6392057
Allows a client to pre-allocate buffer space that is known to be necessary,
avoiding expensive reallocations.
R=gri, gri, adg
CC=golang-dev
https://golang.org/cl/6392061
The error was caused by a call to implements() even when
the type switch variable was not an interface.
Fixes#3786.
R=golang-dev, r
CC=golang-dev, remy
https://golang.org/cl/6354102
The previous implementation used to have a O(n) recursion
depth for unbalanced inputs. A test is added to check that a
reasonable amount of bytes is allocated in this case.
Fixes#3807.
R=golang-dev, dsymonds, gri
CC=golang-dev, remy
https://golang.org/cl/6345075
Also adds a new test for GOMAXPROCS=16 explicitly, which now passes
reliably in a stress loop like:
$ go test -c
$ (while ./http.test -test.v -test.run=Concurrency; do echo pass; done ) 2>&1 | tee foo; less foo
(It used to fail very quickly and reliably on at least Linux/amd64)
Fixes#3793
R=golang-dev, adg, r
CC=golang-dev
https://golang.org/cl/6347061
OpenSSL requires that RSA signatures be exactly the same byte-length
as the modulus. Currently it'll reject ~1/256 of our signatures: those
that end up a byte shorter.
Fixes#3796.
R=golang-dev, edsrzf, r
CC=golang-dev
https://golang.org/cl/6352093
Previously, strings that didn't have an explicit ASN.1 string type
were taken to be ASN.1 PrintableStrings. This resulted in an error if
a unrepresentable charactor was included.
For compatibility reasons, I'm too afraid to switch the default string
type to UTF8String, but this patch causes untyped strings to become
UTF8Strings if they contain a charactor that's not valid in a
PrintableString.
Fixes#3791.
R=golang-dev, bradfitz, r, r
CC=golang-dev
https://golang.org/cl/6348074
Entrust have issued a root certificate that's not marked as valid for
signing certificates.
This results in Go programs failing to validate certificates that
chain up to this root (i.e. gateway.push.apple.com:2195).
Although this is clearly a mistake on Entrust's part, it seems that we
will have to bodge around it.
R=golang-dev, dave, r
CC=golang-dev
https://golang.org/cl/6346064
Ran 'double.pl' on the pkg tree to identify doubled words.
One change to an error string return in x509; the rest are in comments.
Thanks to Matt Jibson for the idea.
R=golang-dev, bsiegert
CC=golang-dev
https://golang.org/cl/6344089
Made the following changes:
- Export errprintf() from all three OS-specific modules
- Added errprintf() to a.h
- Moved errprintf() in windows.c under xprintf(), since they are so similar
- Replaced all instances of xprintf() with errprintf() where a vflag check is done
Fixes#3788.
R=golang-dev, alex.brainman
CC=golang-dev
https://golang.org/cl/6346056
Move panic/defer/recover-related stuff from proc.c/runtime.c to a new file panic.c.
No semantic changes.
proc.c is 1800+ LOC and is a bit difficult to work with.
R=golang-dev, dave, r
CC=golang-dev
https://golang.org/cl/6343071
The use of math.Pow for mere squaring can be extremely
slow on soft-float ARM. Even on systems with hardware
floating-point, a speedup in test duration is observed.
On amd64
Before: ok math/rand 2.009s
After: ok math/rand 0.340s
Fixes#3740.
R=dave, golang-dev, r, r
CC=golang-dev
https://golang.org/cl/6348061
The previous version was not handling an off_t (64 bit)
parameter on 32 bit i386 systems. This patch splits sendfile
into two implementations in their respective arch-specific files.
Tested on FreeBSD amd64 and i386.
R=bradfitz
CC=golang-dev
https://golang.org/cl/6356048
1. Rename 'g' and 'm' local vars to 'gp' and 'mp' (convention already used in some functions)
'g' and 'm' are global vars that mean current goroutine and current machine,
when they are shadowed by local vars, it's confusing, no ability to debug log both, etc.
2. White-space shuffling.
No semantic changes.
In preparation to bigger changes.
R=golang-dev, dave
CC=golang-dev
https://golang.org/cl/6355061
There may be further savings if convT2I can avoid the function call
if the cache is good and T is uintptr-shaped, a la convT2E, but that
will be a follow-up CL.
src/pkg/runtime:
benchmark old ns/op new ns/op delta
BenchmarkConvT2ISmall 43 15 -64.01%
BenchmarkConvT2IUintptr 45 14 -67.48%
BenchmarkConvT2ILarge 130 101 -22.31%
test/bench/go1:
benchmark old ns/op new ns/op delta
BenchmarkBinaryTree17 8588997000 8499058000 -1.05%
BenchmarkFannkuch11 5300392000 5358093000 +1.09%
BenchmarkGobDecode 30295580 31040190 +2.46%
BenchmarkGobEncode 18102070 17675650 -2.36%
BenchmarkGzip 774191400 771591400 -0.34%
BenchmarkGunzip 245915100 247464100 +0.63%
BenchmarkJSONEncode 123577000 121423050 -1.74%
BenchmarkJSONDecode 451969800 596256200 +31.92%
BenchmarkMandelbrot200 10060050 10072880 +0.13%
BenchmarkParse 10989840 11037710 +0.44%
BenchmarkRevcomp 1782666000 1716864000 -3.69%
BenchmarkTemplate 798286600 723234400 -9.40%
R=rsc, bradfitz, go.peter.90, daniel.morsing, dave, uriel
CC=golang-dev
https://golang.org/cl/6337058
The switch at the beginning of the function already ensures n > 1,
so testing for n < 2 is redundant.
R=golang-dev, gri
CC=golang-dev
https://golang.org/cl/6350051
Occasionally I see:
--- FAIL: TestAllocations-15 (0.00 seconds)
all_test.go:1575: 6 mallocs after 100 iterations
Tested:
$ go test -cpu=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20 reflect
R=golang-dev, dsymonds, r
CC=golang-dev
https://golang.org/cl/6354063
Generating env.bat using dist env -wp > env.bat failed silently
if case of an error, because the message was redirected to env.bat.
Verbose messages still go to stdout, causing problems, but that's
a seperate change.
Made errprintf() identical to xprintf(), except for the output handle.
Yes, it's duplicate code, but most of the function is unpacking
the argument list and preparing it for WriteFile(), which has to be
done anyway.
R=golang-dev, alex.brainman
CC=golang-dev
https://golang.org/cl/6343047
Number represents the actual JSON text,
preserving the precision and
formatting of the original input.
R=rsc, iant
CC=golang-dev
https://golang.org/cl/6202068
go/build section "Build Constraints", first paragraph said:
"... they must be appear near the top of the file ..."
fixed to:
"... they must appear near the top of the file ..."
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6329060
Removes an incorrect code comment and some superfluous variables.
The comment I removed says that struct fields which implement
Unmarshaler must be pointers, even if they're in an addressable
struct. That's not the case, and there's already a test in decode_test.go
that demonstrates as much.
Encoding/json has quite a few assignments of reflect.Values to extra
variables – things like "iv := v" when there's no need to make a copy. I
think these are left over from a previous version of the reflect API. If they
aren't wanted, I wouldn't mind going through the package and getting
rid of the rest of them.
R=rsc
CC=golang-dev
https://golang.org/cl/6318047
This is a new, not yet committed API.
- Changed NewCommentMap to be independent of
*File nodes and more symmetric with the
Filter and Comments methods.
- Implemented Update method for use in
AST modifications.
- Implemented String method for debugging
R=rsc
CC=golang-dev
https://golang.org/cl/6303086
A few performance improvements, but without the stack sorting
change to avoid allocating, which is instead waiting on better
escape analysis.
R=rsc
CC=golang-dev
https://golang.org/cl/6265047
The issue seems to not be triggered right now,
but I've seen the deadlock after some other legal
modifications to runtime.
So I think we are safer this way.
R=rsc, r
CC=golang-dev
https://golang.org/cl/6339051
This can only happen if the hash function we're using is getting
far more than it's fair share of collisions, but that has happened
to us repeatedly as we've expanded the allowed use cases for
hash tables (issue 1544, issue 2609, issue 2630, issue 2883, issue 3695).
Maybe this will help the next time we try something new.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6306083
pkg/math/all_test.go tests Atan (and therefore Asin and Acos) to a
relative accuracy of 4e-16, but the test vector misses values where
the old algorithm was in error by more than that. For example:
x newError oldError
0.414215746 1.41e-16 -4.24e-16
0.414216076 1.41e-16 -4.24e-16
0.414217632 1.41e-16 -4.24e-16
0.414218770 1.41e-16 -4.24e-16
0.414225466 0 -5.65e-16
0.414226244 1.41e-16 -4.24e-16
0.414228756 0 -5.65e-16
0.414235089 0 -5.65e-16
0.414237070 0 -5.65e-16
R=rsc, golang-dev
CC=golang-dev
https://golang.org/cl/6302093
Several of my students were confused by trying to use both the error
return and a reply return, so I figured it was worth explicitly clarifying
that returning an error overrides the reply.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6327051
Flame motivated me to get around to adding extended key usage support
so that code signing certificates can't be used for TLS server
authentication and vice versa.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6304065
if we were to use sizeof(sa.sa_mask) instead of 8 as the last argument
to rt_sigaction, we would have already fixed this bug, so also updated
Linux/386 and Linux/amd64 files to use that; also test the return value
of rt_sigaction.
R=dave, rsc
CC=golang-dev
https://golang.org/cl/6297087
If the server replies with an HTTP response before we're done
writing our body (for instance "401 Unauthorized" response), we
were previously ignoring that, since we returned our write
error ("broken pipe", etc) before ever reading the response.
Now we read and write at the same time.
Fixes#3595
R=rsc, adg
CC=golang-dev
https://golang.org/cl/6238043
Also, fixes one violation found during testing where both
response and error could be non-nil when a CheckRedirect test
failed. This is arguably a minor API (behavior, not
signature) change, but it wasn't documented either way and was
inconsistent & non-Go like. Any code depending on the old
behavior was wrong anyway.
R=adg, rsc
CC=golang-dev
https://golang.org/cl/6307088
A comment to that effect was introduced
with rev d332f4b9cef5 but the respective
code wasn't deleted.
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6304086
Fixes a situation where a nested bad type would still
permit the outer type to install a working engine, leading
to inconsistent behavior.
Fixes#3273.
R=bsiegert, rsc
CC=golang-dev
https://golang.org/cl/6294067
A comment map associates comments with AST nodes
and permits correct updating of the AST's comment
list when the AST is manipulated.
R=rsc
CC=golang-dev
https://golang.org/cl/6281044
This slipped in with the implementation of Getpid in CL 5909043.
I'd exclude that CL entirely but it is tangled up in the Win32finddata changes.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6297065
Preserve old API by using correct struct in system call
and then copying the results, as we did for SetsockoptLinger.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6307065
This new error is the only API change in the current draft of
Go 1.0.2 CLs. I'd like to include the CL that introduced it,
because it replaces a mysterious 'internal error' with a
useful error message, but I don't want any API changes,
so unexport the error constant for now. It can be
re-exported for Go 1.1.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6294055
Fixes#3708.
The fix to allow 5{c,g,l} to compile under clang 3.1 broke cross
compilation on darwin using the Apple default compiler on 10.7.3.
This failure was introduced in 9b455eb64690.
This has been tested by cross compiling on darwin/amd64 to linux/arm using
* gcc version 4.2.1 (Based on Apple Inc. build 5658) (LLVM build 2336.1.00)
* clang version 3.1 (branches/release_31)
As well as on linux/arm using
* gcc version 4.6.3 (Ubuntu/Linaro 4.6.3-1ubuntu5)
* Ubuntu clang version 3.0-6ubuntu3 (tags/RELEASE_30/final) (based on LLVM 3.0)
* Debian clang version 3.1-4 (branches/release_31) (based on LLVM 3.1)
R=consalus, rsc
CC=golang-dev
https://golang.org/cl/6307058
The type declarations were being generated using
a range over a map, which meant that successive
runs produced different orders. This will make sure
successive runs produce the same files.
Fixes#3707.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6300062
If there are mutually recursive functions, there is a cycle in
the dependency graph, so the order is actually dependency order
among the strongly connected components: mutually recursive
functions get put into the same batch and analyzed together.
(Until now the entire package was put in one batch.)
The non-recursive case (single function, maybe with some
closures inside) will be able to be more precise about inputs
that escape only back to outputs, but that is not implemented yet.
R=ken2
CC=golang-dev, lvd
https://golang.org/cl/6304050
CL 4313064 fixed its test case but did not address a
general enough problem:
type T1 struct { F *T2 }
type T2 T1
type T3 T2
could still end up copying the definition of T1 for T2
before T1 was done being evaluated, or T3 before T2
was done.
In order to propagate the updates correctly,
record a copy of an incomplete type for re-execution
once the type is completed. Roll back CL 4313064.
Fixes#3709.
R=ken2
CC=golang-dev, lstoakes
https://golang.org/cl/6301059
This is part 1 of a 2 part changelist. Part 2 contains the mechanical
change to parse.go to compare atoms (ints) instead of strings.
The overall effect of the two changes are:
benchmark old ns/op new ns/op delta
BenchmarkParser 4462274 4058254 -9.05%
BenchmarkRawLevelTokenizer 913202 912917 -0.03%
BenchmarkLowLevelTokenizer 1268626 1267836 -0.06%
BenchmarkHighLevelTokenizer 1947305 1968944 +1.11%
R=rsc
CC=andybalholm, golang-dev, r
https://golang.org/cl/6305053
The reordering speedup in CL 6245068 changed the semantics
of %#v by delaying the clearing of some flags. Restore the old
semantics and add a test.
Fixes#3706.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6302048
CL 6250075 removed AI_MASK mask on all BSD variants,
however FreeBSD's AI_MASK does not include AI_V4MAPPED
and AI_ALL, and its libc is strict about the ai_flags.
This will fix the FreeBSD builder.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6305054
On netbsd/386, tv_sec is a 64-bit integer for both timeval and timespec.
Fix the time handling code so that it works correctly.
R=golang-dev, rsc, m4dh4tt3r
CC=golang-dev
https://golang.org/cl/6256056
Thanks to Dave Cheney for the magic words "comm page".
benchmark old ns/op new ns/op delta
BenchmarkNow 197 33 -83.05%
This should make profiling a little better on OS X.
The raw time saved is unlikely to matter: what likely matters
more is that it seems like OS X sends profiling signals on the
way out of system calls more often than it should; avoiding
the system call should increase the accuracy of cpu profiles.
The 386 version would be similar but needs to do different
math for CPU speeds less than 1 GHz. (Apparently Apple has
never shipped a 64-bit CPU with such a slow clock.)
R=golang-dev, bradfitz, dave, minux.ma, r
CC=golang-dev
https://golang.org/cl/6275056
amd64 was done in CL 6275056.
We don't attempt to handle machines with clock speeds
less than 1 GHz. Those will fall back to the system call.
benchmark old ns/op new ns/op delta
BenchmarkNow 364 38 -89.53%
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6307045
Using an int64 for a block size doesn't make
sense on 32bit platforms but extracts a performance
penalty dealing with double word quantities on Arm.
linux/arm
benchmark old ns/op new ns/op delta
BenchmarkGobDecode 155401600 144589300 -6.96%
BenchmarkGobEncode 72772220 62460940 -14.17%
BenchmarkGzip 5822632 2604797 -55.26%
BenchmarkGunzip 326321 151721 -53.51%
benchmark old MB/s new MB/s speedup
BenchmarkGobDecode 4.94 5.31 1.07x
BenchmarkGobEncode 10.55 12.29 1.16x
R=golang-dev, rsc, bradfitz
CC=golang-dev
https://golang.org/cl/6272047
The original implementation of closures created the
underlying top-level function during walk, which is fairly
late in the compilation process and caused ordering-based
complications due to earlier stages that had to be repeated
any number of times.
Create the underlying function during typecheck, much
earlier, so that later stages can be run just once.
The result is a simpler compilation sequence.
R=ken2
CC=golang-dev
https://golang.org/cl/6279049
although the comment says it uses libc's getenv, without NOPLAN9DEFINES
it actually uses p9getenv which strdups.
R=golang-dev, dave, rsc
CC=golang-dev
https://golang.org/cl/6285046
The recent shuffle in parsing formats exposed probably unintentional
behavior in time.Parse, namely that it was mostly ignoring ".99999"
in the format, producing the following behavior:
fmt.Println(time.Parse("03:04:05.999 MST", "12:00:00.888 PDT")) // error (.888 unexpected)
fmt.Println(time.Parse("03:04:05.999", "12:00:00")) // error (input too short)
fmt.Println(time.Parse("03:04:05.999 MST", "12:00:00 PDT")) // ok (extra bytes on input make it ok)
http://play.golang.org/p/ESJ1UYXzq2
API CHANGE:
This CL makes all three examples valid: ".999" can match an
empty string or else a fractional second with at most nine digits.
Fixes#3701.
R=r, r
CC=golang-dev
https://golang.org/cl/6267045
An attempt to profit from CL 6176043 (fix to superpolinomial
runtime of karatsuba multiplication) and determine a better
karatsuba threshold. The result indicates that 32 is still
a reasonable value. Left the threshold as is (== 32), but
made some minor changes to the calibrate code which are
worthwhile saving (use of existing benchmarking code for
better results, better use of package time).
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6260062
pipe2 is equivalent to pipe with flags set to 0.
However, pipe2 was only added recently. Using pipe
instead improves compatibility with NetBSD 5.
R=jsing
CC=golang-dev
https://golang.org/cl/6268045
The cleanup also makes it ~5% faster, but that's
not the point of this CL.
Optimizations can come in future CLs.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6286043
It's very unfortunate that the type of Data field of struct
RawSockaddr is [14]uint8 on Linux/ARM instead of [14]int8
on all the others.
btw, it should be [14]int8 according to my header files.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6275050
Move address info flags to per-platform files. This is needed to
enable cgo on NetBSD (and later OpenBSD), as some of the currently
used AI_* defines do not exist on these platforms.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6250075
Use perfect cuckoo hash, to avoid binary search.
Define Atom bits as offset+len in long string instead
of enumeration, to avoid string headers.
Before: 1909 string bytes + 6060 tables = 7969 total data
After: 1406 string bytes + 2048 tables = 3454 total data
benchmark old ns/op new ns/op delta
BenchmarkLookup 83878 64681 -22.89%
R=nigeltao, r
CC=golang-dev
https://golang.org/cl/6262051
Ceil to 4.81 from 20.6 ns/op
Floor to 4.37 from 13.5 ns/op
Trunc to 3.97 from 14.3 ns/op
Also changed three MOVSDs to MOVAPDs in log_amd64.s
R=rsc, golang-dev
CC=golang-dev
https://golang.org/cl/6262048
Currently walk() doesn't check for err == SkipDir when iterating
a directory list, but such promise is made in the docs for WalkFunc.
Fixes#3486.
R=rsc, r
CC=golang-dev
https://golang.org/cl/6257059
Now that gri has made go/parser 15% faster, I offer this
change to slow back down cmd/api ~proportionately, adding
FreeBSD to the go1-checked set of platforms.
Really we should have done this earlier. This will prevent us
from breaking FreeBSD compatibility accidentally in the
future.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6279044
To avoid goroutines during init, the nextItem function was a
clever workaround. Now that init goroutines are permitted,
restore the original, simpler design.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6282043
- only compute current line position if needed
(i.e., if a comment is present)
- added benchmark
benchmark old ns/op new ns/op delta
BenchmarkParse 10902990 9313330 -14.58%
benchmark old MB/s new MB/s speedup
BenchmarkParse 5.31 6.22 1.17x
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6270043
Saving the code in case we improve things enough that
it matters later, but at least right now it is not worth doing.
R=ken2
CC=golang-dev
https://golang.org/cl/6248071
The previous code was preparing arrays of entries that would be
filled if there was one entry every 128 bytes. Moving to a 4096
byte interval reduces the overhead per megabyte of address space
to 2kB from 64kB (on 64-bit systems).
The performance impact will be negative for very small MemProfileRate.
test/bench/garbage/tree2 -heapsize 800000000 (default memprofilerate)
Before: mprof 65993056 bytes (1664 bucketmem + 65991392 addrmem)
After: mprof 1989984 bytes (1680 bucketmem + 1988304 addrmem)
R=golang-dev, rsc
CC=golang-dev, remy
https://golang.org/cl/6257069
The previous heap profile format did not include buckets with
zero used bytes. Also add several missing MemStats fields in
debug mode.
R=golang-dev, rsc
CC=golang-dev, remy
https://golang.org/cl/6249068
Drop expecttaken function in favor of extra argument
to gbranch and bgen. Mark loop condition as likely to
be true, so that loops are generated inline.
The main benefit here is contiguous code when trying
to read the generated assembly. It has only minor effects
on the timing, and they mostly cancel the minor effects
that aligning function entry points had. One exception:
both changes made Fannkuch faster.
Compared to before CL 6244066 (before aligned functions)
benchmark old ns/op new ns/op delta
BenchmarkBinaryTree17 4222117400 4201958800 -0.48%
BenchmarkFannkuch11 3462631800 3215908600 -7.13%
BenchmarkGobDecode 20887622 20899164 +0.06%
BenchmarkGobEncode 9548772 9439083 -1.15%
BenchmarkGzip 151687 152060 +0.25%
BenchmarkGunzip 8742 8711 -0.35%
BenchmarkJSONEncode 62730560 62686700 -0.07%
BenchmarkJSONDecode 252569180 252368960 -0.08%
BenchmarkMandelbrot200 5267599 5252531 -0.29%
BenchmarkRevcomp25M 980813500 985248400 +0.45%
BenchmarkTemplate 361259100 357414680 -1.06%
Compared to tip (aligned functions):
benchmark old ns/op new ns/op delta
BenchmarkBinaryTree17 4140739800 4201958800 +1.48%
BenchmarkFannkuch11 3259914400 3215908600 -1.35%
BenchmarkGobDecode 20620222 20899164 +1.35%
BenchmarkGobEncode 9384886 9439083 +0.58%
BenchmarkGzip 150333 152060 +1.15%
BenchmarkGunzip 8741 8711 -0.34%
BenchmarkJSONEncode 65210990 62686700 -3.87%
BenchmarkJSONDecode 249394860 252368960 +1.19%
BenchmarkMandelbrot200 5273394 5252531 -0.40%
BenchmarkRevcomp25M 996013800 985248400 -1.08%
BenchmarkTemplate 360620840 357414680 -0.89%
R=ken2
CC=golang-dev
https://golang.org/cl/6245069
On 6l and 8l, this is a real instruction, guaranteed to
cause an 'undefined instruction' exception.
On 5l, we simulate it as BL to address 0.
The plan is to use it as a signal to the linker that this
point in the instruction stream cannot be reached
(hence the changes to nofollow). This will help the
compiler explain that panicindex and friends do not
return without having to put a list of these functions
in the linker.
R=ken2
CC=golang-dev
https://golang.org/cl/6255064
16 seems pretty standard on x86 for function entry.
I don't know if ARM would benefit, so I used just 4
(single instruction alignment).
This has a minor absolute effect on the current timings.
The main hope is that it will make them more consistent from
run to run.
benchmark old ns/op new ns/op delta
BenchmarkBinaryTree17 4222117400 4140739800 -1.93%
BenchmarkFannkuch11 3462631800 3259914400 -5.85%
BenchmarkGobDecode 20887622 20620222 -1.28%
BenchmarkGobEncode 9548772 9384886 -1.72%
BenchmarkGzip 151687 150333 -0.89%
BenchmarkGunzip 8742 8741 -0.01%
BenchmarkJSONEncode 62730560 65210990 +3.95%
BenchmarkJSONDecode 252569180 249394860 -1.26%
BenchmarkMandelbrot200 5267599 5273394 +0.11%
BenchmarkRevcomp25M 980813500 996013800 +1.55%
BenchmarkTemplate 361259100 360620840 -0.18%
R=ken2
CC=golang-dev
https://golang.org/cl/6244066
The code was inconsistent about when it used
brchain(x) and when it used x directly, with the result
that you could end up emitting code for brchain(x) but
leave the jump pointing at an unemitted x.
R=ken2
CC=golang-dev
https://golang.org/cl/6250077
This bug has been introduced in the following revision:
changeset: 11404:26dceba5c610
user: Ivan Krasin <krasin@golang.org>
date: Mon Jan 23 09:19:39 2012 -0500
summary: compress/flate: reduce memory pressure at cost of additional arithmetic operation.
This is the review page for that CL: https://golang.org/cl/5555070/
R=rsc, imkrasin
CC=golang-dev
https://golang.org/cl/6249067
The correct procid is needed for unparking LWPs on NetBSD - always
initialise procid in minit() so that cgo works correctly. The non-cgo
case already works correctly since procid is initialised via
lwp_create().
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6257071
On NetBSD a cgo enabled binary has more than 32 sections - bump NSECTS
so that we can actually link them successfully.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6261052
filterPaeth takes []byte arguments instead of byte arguments,
which avoids some redudant computation of the previous pixel
in the inner loop.
Also eliminate a bounds check in decoding the up filter.
benchmark old ns/op new ns/op delta
BenchmarkDecodeGray 3139636 2812531 -10.42%
BenchmarkDecodeNRGBAGradient 12341520 10971680 -11.10%
BenchmarkDecodeNRGBAOpaque 10740780 9612455 -10.51%
BenchmarkDecodePaletted 1819535 1818913 -0.03%
BenchmarkDecodeRGB 8974695 8178070 -8.88%
R=rsc
CC=golang-dev
https://golang.org/cl/6243061
A block with finalizer might also be profiled. The special bit
is needed to unregister the block from the profile. It will be
unset only when the block is freed.
Fixes#3668.
R=golang-dev, rsc
CC=golang-dev, remy
https://golang.org/cl/6249066
The check for Stringer etc. can only fire if the test is not a builtin, so avoid
the expensive check if we know there's no chance.
Also put in a fast path for pad, which saves a more modest amount.
benchmark old ns/op new ns/op delta
BenchmarkSprintfEmpty 148 152 +2.70%
BenchmarkSprintfString 585 497 -15.04%
BenchmarkSprintfInt 441 396 -10.20%
BenchmarkSprintfIntInt 718 603 -16.02%
BenchmarkSprintfPrefixedInt 676 621 -8.14%
BenchmarkSprintfFloat 1003 953 -4.99%
BenchmarkManyArgs 2945 2312 -21.49%
BenchmarkScanInts 1704152 1734441 +1.78%
BenchmarkScanRecursiveInt 1837397 1828920 -0.46%
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6245068
This prevents clients from seeing RSTs and missing the response
body.
TCP stacks vary. The included test failed on Darwin before but
passed on Linux.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6256066
It was only being used for (*Stmt).Exec, not Query, and not for
the same two methods on *DB.
This unifies (*Stmt).Exec's old inline code into the old
subsetArgs function, renaming it in the process (changing the
old word "subset" to "driver", mostly converted earlier)
Fixes#3640
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6258045
It's sad to introduce a new macro, but rnd shows up consistently
in profiles, and the function call overwhelms the two arithmetic
instructions it performs.
R=r
CC=golang-dev
https://golang.org/cl/6260051
Plan 9 versions for amd64 have 2 megabyte pages.
This also fixes the logic for 32-bit vs 64-bit Plan 9,
making 64-bit the default, and adds logic to generate
a symbols table.
R=golang-dev, rsc, rminnich, ality, 0intro
CC=golang-dev, john
https://golang.org/cl/6218046
The old code generated for a bounds check was
CMP
JLT ok
CALL panicindex
ok:
...
The new code is (once the linker finishes with it):
CMP
JGE panic
...
panic:
CALL panicindex
which moves the calls out of line, putting more useful
code in each cache line. This matters especially in tight
loops, such as in Fannkuch. The benefit is more modest
elsewhere, but real.
From test/bench/go1, amd64:
benchmark old ns/op new ns/op delta
BenchmarkBinaryTree17 6096092000 6088808000 -0.12%
BenchmarkFannkuch11 6151404000 4020463000 -34.64%
BenchmarkGobDecode 28990050 28894630 -0.33%
BenchmarkGobEncode 12406310 12136730 -2.17%
BenchmarkGzip 179923 179903 -0.01%
BenchmarkGunzip 11219 11130 -0.79%
BenchmarkJSONEncode 86429350 86515900 +0.10%
BenchmarkJSONDecode 334593800 315728400 -5.64%
BenchmarkRevcomp25M 1219763000 1180767000 -3.20%
BenchmarkTemplate 492947600 483646800 -1.89%
And 386:
benchmark old ns/op new ns/op delta
BenchmarkBinaryTree17 6354902000 6243000000 -1.76%
BenchmarkFannkuch11 8043769000 7326965000 -8.91%
BenchmarkGobDecode 19010800 18941230 -0.37%
BenchmarkGobEncode 14077500 13792460 -2.02%
BenchmarkGzip 194087 193619 -0.24%
BenchmarkGunzip 12495 12457 -0.30%
BenchmarkJSONEncode 125636400 125451400 -0.15%
BenchmarkJSONDecode 696648600 685032800 -1.67%
BenchmarkRevcomp25M 2058088000 2052545000 -0.27%
BenchmarkTemplate 602140000 589876800 -2.04%
To implement this, two new instruction forms:
JLT target // same as always
JLT $0, target // branch expected not taken
JLT $1, target // branch expected taken
The linker could also emit the prediction prefixes, but it
does not: expected taken branches are reversed so that the
expected case is not taken (as in example above), and
the default expectaton for such a jump is not taken
already.
R=golang-dev, gri, r, dave
CC=golang-dev
https://golang.org/cl/6248049
Implement the (3-per-family) Noah's Ark clause (i.e. don't put
more than three identical elements on the list of active formatting
elements.
Also, when running tests, sort attributes by name before dumping
them.
Pass 4 additional tests with Noah's Ark clause (including one
that needs attributes to be sorted).
Pass 5 additional, unrelated tests because of sorting attributes.
R=nigeltao, rsc
CC=golang-dev
https://golang.org/cl/6247056
CanonicalHeaderKey didn't allocate, but it did use unnecessary
CPU in the hot path, deciding it didn't need to allocate.
I considered using constants for all these common header keys
but I didn't think it would be prettier. "Content-Length" looks
better than contentLength or hdrContentLength, etc.
R=golang-dev, dave
CC=golang-dev
https://golang.org/cl/6255053
Comment on cache keys above connectMethod says "http to proxy, http
anywhere after that", however in reality target address was always
included, which prevented http requests to different target
addresses to reuse the same http proxy connection.
R=golang-dev, r, rsc, bradfitz
CC=golang-dev
https://golang.org/cl/5901064
CL 5956051 introduced too many call != nil checks, so
attempt to improve this by splitting logic into three
distinct parts.
R=r
CC=golang-dev
https://golang.org/cl/6248048
for expr1, expr2 = range slice
was assigning to expr1 and expr2 in sequence
instead of in parallel. Now it assigns in parallel,
as it should. This matters for things like
for i, x[i] = range slice.
Fixes#3464.
R=ken2
CC=golang-dev
https://golang.org/cl/6252048
This is from CL 5451105 but was dropped from that CL.
See also CL 6137051.
The only change compared to 5451105 is to check for
h != nil in reflect·mapiterinit; allowing use of nil maps
must have happened after that original CL.
Fixes#3573.
R=golang-dev, dave, r
CC=golang-dev
https://golang.org/cl/6215078
Remove redundant checks for integration points.
Ignore null bytes in text.
Don't break out of foreign content for a <font> tag unless it
has a color, face, or size attribute.
Check for MathML text integration points when breaking out of
foreign content.
Pass two new tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6256045
The bulk of the gains come from hoisting the modulo ops outside of
the inner loop.
Reducing the digest type from 8 bytes to 4 bytes gains another 1% on
the hash/adler32 micro-benchmark.
Benchmarks for $GOOS,$GOARCH = linux,amd64 below.
hash/adler32 benchmark:
benchmark old ns/op new ns/op delta
BenchmarkAdler32KB 1660 1364 -17.83%
image/png benchmark:
benchmark old ns/op new ns/op delta
BenchmarkDecodeGray 2466909 2425539 -1.68%
BenchmarkDecodeNRGBAGradient 9884500 9751705 -1.34%
BenchmarkDecodeNRGBAOpaque 8511615 8379800 -1.55%
BenchmarkDecodePaletted 1366683 1330677 -2.63%
BenchmarkDecodeRGB 6987496 6884974 -1.47%
BenchmarkEncodePaletted 6292408 6040052 -4.01%
BenchmarkEncodeRGBOpaque 19780680 19178440 -3.04%
BenchmarkEncodeRGBA 80738600 79076800 -2.06%
Wall time for Denis Cheremisov's PNG-decoding program given in
https://groups.google.com/group/golang-nuts/browse_thread/thread/22aa8a05040fdd49
Before: 2.44s
After: 2.26s
Delta: -7%
R=rsc
CC=golang-dev
https://golang.org/cl/6251044
When client fails to write a request is sends caller that error,
however server might have failed to read that request in the mean
time and replied with that error. When client then reads the
response the call would no longer be pending, so call will be nil
Handle this gracefully by discarding such server responses
R=golang-dev, r
CC=golang-dev, rsc
https://golang.org/cl/5956051
* Eliminate bounds check on known small shifts.
* Rewrite x<<s | x>>(32-s) as a rotate (constant s).
* More aggressive (but still minimal) range analysis.
R=ken, dave, iant
CC=golang-dev
https://golang.org/cl/6209077
The previous attempt to explain this got it backwards (all the more reason to be
sad we couldn't make the two functions behave the same).
Fixes#3669.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6249051
There's no need for the 16-bit arithmetic here,
and it tickles a long-standing compiler bug.
Fix the exp code not to use 16-bit math and
create an explicit test for the compiler bug.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6256048
- interface methods appeared under VarDecl in search results
(long-standing TODO)
- don't walk parts of AST which contain no indexable material
(minor performance tuning)
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6228047
The documentation says so, but in the case of a normalized
integral Rat, the denominator was a new value. Changed the
internal representation to use an Int to represent the
denominator (with the sign ignored), so a reference to it
can always be returned.
Clarified documentation and added test cases.
Fixes#3521.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6237045
* Shift/rotate by constant doesn't have to stop subprop. (also in 8g)
* Remove redundant MOVLQZX instructions.
* An attempt at issuing loads early.
Good for 0.5% on a good day, might not be worth keeping.
Need to understand more about whether the x86
looks ahead to what loads might be coming up.
R=ken2, ken
CC=golang-dev
https://golang.org/cl/6203091
Detect HTML integration points and MathML text integration points.
At these points, process tokens as HTML, not as foreign content.
Pass 33 more tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6249044
Import updated test data from the WebKit Subversion repository (SVN revision 118111).
Some of the old tests were failing because we were HTML5 compliant, but the tests weren't.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6228049
I needed this to explore per-GOOS/GOARCH differences in pkg
syscall for a recent CL. Others may find it useful too.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6236046
- there is no label scope at package level
- open/close all scopes symmetrically now
that there is only one parse entry point
(parseFile)
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6230047
CL 5823055 removed a line introduced in Linux/ARM cgo support.
Because readsym() now returns nil for "$a", "$d" mapping symbols,
no matter the settings of `needSym', we still have to guard against
them in ldelf().
R=golang-dev, dave, rsc
CC=golang-dev
https://golang.org/cl/6220073
This quiets all.bash noise for upcoming features we know about.
The all.bash warnings will now only print for things not in next.txt
(or in next.txt but not in the API).
Once an API is frozen, we rename next.txt to a new frozen file
(like go1.txt)
Fixes#3651
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6218069
Handle text, comment, and doctype tokens in afterBodyIM, afterAfterBodyIM,
and afterAfterFramesetIM.
Pass three more tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6231043
Also: simplified some existing tests.
No support for Rats for now because the precision-preserving
default notation (fractions of the form a/b) is not a valid
JSON value.
Fixes#3657.
R=golang-dev, bradfitz, rsc
CC=golang-dev
https://golang.org/cl/6211079
Currently, if you pass some data to a template as an interface (e.g. interface{})
and extract that value that value as a parameter for a function, it fails, saying
wrong type.
This is because it is only looking at the interface type, not the interface content.
This CL uses the underlying content as the parameter to the func.
Fixes#3642.
R=golang-dev, r, r
CC=golang-dev
https://golang.org/cl/6218052
Introduce a newsym() to cmd/lib.c to add a symbol but don't add
them to hash table.
Introduce a new bit flag SHIDDEN and bit mask SMASK to handle hidden
and/or local symbols in ELF symbol tables. Though we still need to order
the symbol table entries correctly.
Fix for issue 3261 comment #9.
For CL 5822049.
R=iant, rsc
CC=golang-dev
https://golang.org/cl/5823055
ld -r could generate multiple section symbols for the same section,
but with different values, we have to take that into account.
Fixes#3322.
Part of issue 3261.
For CL 5822049.
R=golang-dev, iant, rsc, iant
CC=golang-dev
https://golang.org/cl/5823059
Now that we've fixed the Expect: test, this CL should be okay.
««« original CL description
net/http: revert 97d027b3aa68
Revert the following change set:
changeset: 13018:97d027b3aa68
user: Gustavo Niemeyer <gustavo@niemeyer.net>
date: Mon Apr 23 22:00:16 2012 -0300
summary: net/http: allow clients to disable keep-alive
This broke a test on Windows 64 and somebody else
will have to check.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6112054
»»»
Fixes#3540.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6228046
In both the web and command line tool,
the comment is shown after the declaration.
But in the code the comment is obviously before.
Make the text not refer to a specific order.
R=r, dsymonds
CC=golang-dev
https://golang.org/cl/6206094
The speedup is a combination of unrolling/specializing
the actual code and also making the compiler generate better code.
Go 1.0.1 (size: 1239 code + 320 data = 1559 total)
md5.BenchmarkHash1K 1000000 7178 ns/op 142.64 MB/s
md5.BenchmarkHash8K 200000 56834 ns/op 144.14 MB/s
Partial unroll (size: 1115 code + 256 data = 1371 total)
md5.BenchmarkHash1K 5000000 2513 ns/op 407.37 MB/s
md5.BenchmarkHash8K 500000 19406 ns/op 422.13 MB/s
Complete unroll (size: 1900 code + 0 data = 1900 code)
md5.BenchmarkHash1K 5000000 2442 ns/op 419.18 MB/s
md5.BenchmarkHash8K 500000 18957 ns/op 432.13 MB/s
Comparing Go 1.0.1 and the complete unroll (this CL):
benchmark old MB/s new MB/s speedup
md5.BenchmarkHash1K 142.64 419.18 2.94x
md5.BenchmarkHash8K 144.14 432.13 3.00x
On the same machine, 'openssl speed md5' reports 441 MB/s
and 531 MB/s for our two cases, so this CL is at 90% and 80% of
those speeds, which is at least in the right ballpark.
OpenSSL is using carefully engineered assembly, so we are
unlikely to catch up completely.
Measurements on a Mid-2010 MacPro5,1.
R=golang-dev, bradfitz, agl
CC=golang-dev
https://golang.org/cl/6220046
This just eliminates some duplication.
Also add a pointer to RFC 1122, in case
this comes up again.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6229044
Fail more usefully, and Logf in one place instead of Errorf where
an error is acceptable.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6221059
Comment groups must end at the end of a line (or the
next non-comment token) if the group started on a line
with non-comment tokens.
This is important for correct computation of "lead"
and "line" comments (Doc and Comment fields in AST nodes).
Without this fix, the "line" comment for F1 in the
following example:
type T struct {
F1 int // comment1
// comment2
F2 int
}
is "// comment1// comment2" rather than just "// comment1".
This bug was present from Day 1 but only visible when
looking at export-filtered ASTs where only comments
associated with AST nodes are printed, and only in rare
cases (e.g, in the case above, if F2 where not exported,
godoc would show "// comment2" anyway because it was
considered part of the "line" comment for F1).
The bug fix is very small (parser.go). The bulk of the
changes are additional test cases (parser_test.go).
The fix exposed a caching bug in go/printer via one of the
existing tests, hence the changes to printer.go.
As an aside, the fix removes the the need for empty lines
before an "// Output" comment for some special cases of
code examples (e.g.: src/pkg/strings/example_test.go, Count
example).
No impact on gofmt formatting of src, misc.
Fixes#3139.
R=rsc
CC=golang-dev
https://golang.org/cl/6209080
This includes the NIST test suite for ECDSA and alters the test to
parse and evaluate it.
R=golang-dev, bradfitz, rsc, b
CC=golang-dev
https://golang.org/cl/6219058
This fixes occasional 64-bit failures.
Maybe it will fix the 32-bit failures too,
so re-enable on 32-bit for now.
R=golang-dev, bradfitz, r, dvyukov
CC=golang-dev
https://golang.org/cl/6218050
Clean up flow of control.
Ignore </table>, </tbody>, </tfoot>, </thead>, </tr> if there is not
an appropriate element in table scope.
Pass 3 more tests.
R=golang-dev, nigeltao
CC=golang-dev
https://golang.org/cl/6206093
http://www.rsa.com/rsalabs/node.asp?id=2125:
NOTE: A new OID has been defined for the combination
of the v1.5 signature scheme and the SHA-224 hash function:
sha224WithRSAEncryption OBJECT IDENTIFIER ::=
Like the other sha*WithRSAEncryption OIDs in PKCS #1 v2.1,
this OID has NULL parameters.
The DigestInfo encoding for SHA-224 (see Section 9.2, Note 1) is:
(0x)30 2d 30 0d 06 09 60 86 48 01 65 03 04 02 04 05 00 04 1c || H
R=golang-dev, agl
CC=golang-dev
https://golang.org/cl/6208076
The main content of this CL is a test case checking the reported
issue 3511 and a tiny fix for it. A subsequent CL will refactor
the fix as proposed issue 3511.
Fixes#3511.
R=golang-dev, steven.hartland, bradfitz
CC=golang-dev
https://golang.org/cl/6013049
Two tests added in 820ffde8c are expected to fail until the fix
for Issue 3540 goes back in (pending Windows net fixes), so
make those tests just Logf for now, with a TODO to re-enable.
Add a new client test.
Rearrange the transport code to be more readable, and fix the
bug from 820ffde8c where the persistConn was being closed before
the body was fully ready.
Fixes#3644
Updates #1967 (not yet fixed, but should be after Issue 3540)
R=golang-dev, adg
CC=golang-dev
https://golang.org/cl/6211069
RawSockaddrDatalink and SockaddrDatalink need to match - make Data
have length 12 for both.
R=golang-dev, mikioh.mikioh
CC=golang-dev
https://golang.org/cl/6223051
Delete cases that just fall down to "anything else" action.
Handle </tbody>, </tfoot>, and </thead>.
R=golang-dev, nigeltao
CC=golang-dev
https://golang.org/cl/6203061
Connections did not close if Request.Close or Response.Close was true. This meant that if the user wanted the connection to close, or if the server requested it via "Connection: close", the connection would not be closed.
Fixes#1967.
R=golang-dev, rsc, bradfitz
CC=golang-dev
https://golang.org/cl/6201044
key and simple comparisson. Search is not yet implemented in this CL.
Changed some of the types of table_test.go to allow reuse in the new test.
Also reduced number of primary values for illegal runes to 1 (both map to
the same).
R=r
CC=golang-dev
https://golang.org/cl/6202062
This CL makes
type T struct { *U }
behave in a similar way to:
type T struct { U }
Fixes#3108.
R=golang-dev, rsc, gustavo
CC=golang-dev
https://golang.org/cl/5694044
With the timed semacquire patch
(kernel-tsemacquire) for Plan 9,
we can now properly do a timed
wait for the semaphore, in
semasleep.
R=golang-dev, rsc, rminnich, ality, r
CC=0intro, golang-dev, john, mirtchovski
https://golang.org/cl/6197046
Implement getcontext and sigprocmask for NetBSD - these will soon be
used by the thread handling code.
Also fix netbsd/386 signal handling - there is no sigreturn, just
return so that we hit the trampoline.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6215049
Before:
./x.go:6: first argument to append must be slice; have nil
After:
./x.go:6: first argument to append must be typed slice; have untyped nil
Fixes#3616.
R=ken2
CC=golang-dev
https://golang.org/cl/6209067
Update/correct NetBSD signal handling - most of this is needed due to
the correctly generated runtime definitions.
R=golang-dev, m4dh4tt3r, rsc
CC=golang-dev
https://golang.org/cl/6195079
Fix and regenerate runtime defs for NetBSD.
Whilst the mcontext struct can be handled across architectures,
the registers are provided as defines that index an array, rather
than as members of the struct. Since these are architecture
dependent, include them via a defs_netbsd_<arch>.go file.
R=golang-dev, m4dh4tt3r, rsc
CC=golang-dev
https://golang.org/cl/6190070
Parallel GC needs to know in advance how many helper threads will be there.
Hopefully it's the last patch before I can tackle parallel sweep phase.
The benchmarks are unaffected.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6200064
Set the TLS base using the _lwp_setprivate() syscall, instead of via
sysarch(). NetBSD tracks the pointer passed to _lwp_setprivate() and
restores this value when restoring mcontext. If sysarch() is used
directly, restoring an mcontext trashes the FS/GS value, resulting
in a segfault when we next try to access the TLS.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6206062
Empty parts can be either of the form:
a) "--separator\r\n", header (w/ trailing 2xCRLF), \r\n "--separator"...
or
b) "--separator\r\n", header (w/ trailing 2xCRLF), "--separator"...
We never handled case b). In fact the RFC seems kinda vague about
it, but browsers seem to do a), and App Engine's synthetic POST
bodies after blob uploads is of form b).
So handle them both, and add a bunch of tests.
(I can't promise these are the last fixes to multipart, especially
considering its history, but I'm growing increasingly confident at
least, and I've never submitted a multipart CL with known bugs
outstanding, including this time.)
R=golang-dev, adg
CC=golang-dev
https://golang.org/cl/6212046
Unbalanced extra right parenthesis produced an internal error instead of
a more descriptive one.
Fixes#3406.
R=r, rsc
CC=golang-dev
https://golang.org/cl/6201063
Fix mkerrors.sh so that it works on NetBSD.
Remove directory mode bits from types - this already appears in errors.
Regenerate the z* files now that cgo is working.
R=golang-dev, m4dh4tt3r, r
CC=golang-dev
https://golang.org/cl/6201077
No perf/semantic changes, merely improves code health.
There were several questions as to why Once.Do uses
atomic.CompareAndSwap to do a store.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6208057
we can't add the division result to n during iteration, because it might
turn n into NaN or Inf.
R=golang-dev, rsc, iant, iant
CC=golang-dev
https://golang.org/cl/6197045
Use correct syscall numbers and arguments for NetBSD.
Provide a trampoline for signal returns (using signal API 3).
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6209048
Manually repair/update z* files for netbsd - this allows Go to
compile again on 386.
R=golang-dev, rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6194064
Also set maxContractLen automatically.
Note that the table size is much bigger than it needs to be.
Optimization is best done, though, when the language specific
tables are added.
R=r
CC=golang-dev
https://golang.org/cl/6167044
go help remote used to reference "go help importpath", which has
changed to "go help packages".
Fixes#3598.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6201065
Tested manually.
Fixes#3554.
Before:
$ cd $GOROOT/src/pkg
$ go list io
io
$ go list io/...
io
io/ioutil
$ cd $GOROOT/src/pkg/io
$ go list .
io
$ go list ./...
io/ioutil
After:
$ cd $GOROOT/src/pkg
$ go list io
io
$ go list io/...
io
io/ioutil
$ cd $GOROOT/src/pkg/io
$ go list .
io
$ go list ./...
io
io/ioutil
$ go list ././...
io
io/ioutil
$ go list ././.././io/...
io
io/ioutil
$ go list ../image
image
$ go list ../image/...
image
image/color
image/draw
image/gif
image/jpeg
image/png
$ go list ../.../template
html/template
text/template
$ cd $GOROOT/src/pkg
$ go list ./io
io
$ go list ./io/...
io
io/ioutil
$ go list ./.../pprof
net/http/pprof
runtime/pprof
$ go list ./compress
can't load package: package compress: no Go source files in /home/nigeltao/go/src/pkg/compress
$ go list ./compress/...
compress/bzip2
compress/flate
compress/gzip
compress/lzw
compress/zlib
$ cd $GOROOT/src/pkg/code.google.com
$ go list ./p/leveldb-go/...
code.google.com/p/leveldb-go/leveldb
code.google.com/p/leveldb-go/leveldb/crc
code.google.com/p/leveldb-go/leveldb/db
code.google.com/p/leveldb-go/leveldb/memdb
code.google.com/p/leveldb-go/leveldb/memfs
code.google.com/p/leveldb-go/leveldb/record
code.google.com/p/leveldb-go/leveldb/table
code.google.com/p/leveldb-go/manualtest/filelock
$ go list ./p/.../truetype
code.google.com/p/freetype-go/example/truetype
code.google.com/p/freetype-go/freetype/truetype
$ go list ./p/.../example
warning: "./p/.../example" matched no packages
$ go list ./p/.../example/...
code.google.com/p/freetype-go/example/freetype
code.google.com/p/freetype-go/example/gamma
code.google.com/p/freetype-go/example/raster
code.google.com/p/freetype-go/example/round
code.google.com/p/freetype-go/example/truetype
code.google.com/p/x-go-binding/example/imgview
code.google.com/p/x-go-binding/example/xgb
R=rsc
CC=golang-dev
https://golang.org/cl/6194056
These files change from exactly 10003 bytes long to 100003: a digit,
a '.', 100k digits, and a '\n'.
The magic constants in compress/flate/deflate_test.go change since
deflateInflateStringTests checks that the compressed form of e.txt
is not 'too large'. I'm not exactly sure how these numbers were
originally calculated (they were introduced in codereview 5554066
"make lazy matching work"); perhaps krasin@golang.org can comment.
My change was to increase the first one (no compression) to a tight
bound, and multiply all the others by 10.
Benchcmp numbers for compress/flate and compress/lzw below. LZW's
window size of 4096 is less than 10k, so shows no significant change.
Flate's window size is 32768, between 10k and 100k, and so the .*1e5
and .*1e6 benchmarks show a dramatic drop, since the compressed forms
are no longer a trivial forward copy of 10k digits repeated over and
over, but should now be more representative of real world usage.
compress/flate:
benchmark old MB/s new MB/s speedup
BenchmarkDecodeDigitsSpeed1e4 16.58 16.52 1.00x
BenchmarkDecodeDigitsSpeed1e5 68.09 18.10 0.27x
BenchmarkDecodeDigitsSpeed1e6 124.63 18.35 0.15x
BenchmarkDecodeDigitsDefault1e4 17.21 17.12 0.99x
BenchmarkDecodeDigitsDefault1e5 118.28 19.19 0.16x
BenchmarkDecodeDigitsDefault1e6 295.62 20.52 0.07x
BenchmarkDecodeDigitsCompress1e4 17.22 17.17 1.00x
BenchmarkDecodeDigitsCompress1e5 118.19 19.21 0.16x
BenchmarkDecodeDigitsCompress1e6 295.59 20.55 0.07x
BenchmarkEncodeDigitsSpeed1e4 8.18 8.19 1.00x
BenchmarkEncodeDigitsSpeed1e5 43.22 12.84 0.30x
BenchmarkEncodeDigitsSpeed1e6 80.76 13.48 0.17x
BenchmarkEncodeDigitsDefault1e4 6.29 6.19 0.98x
BenchmarkEncodeDigitsDefault1e5 31.63 3.60 0.11x
BenchmarkEncodeDigitsDefault1e6 52.97 3.24 0.06x
BenchmarkEncodeDigitsCompress1e4 6.20 6.19 1.00x
BenchmarkEncodeDigitsCompress1e5 31.59 3.59 0.11x
BenchmarkEncodeDigitsCompress1e6 53.18 3.25 0.06x
compress/lzw:
benchmark old MB/s new MB/s speedup
BenchmarkDecoder1e4 21.99 22.09 1.00x
BenchmarkDecoder1e5 22.77 22.71 1.00x
BenchmarkDecoder1e6 22.90 22.90 1.00x
BenchmarkEncoder1e4 21.04 21.19 1.01x
BenchmarkEncoder1e5 22.06 22.06 1.00x
BenchmarkEncoder1e6 22.16 22.28 1.01x
R=rsc
CC=golang-dev, krasin
https://golang.org/cl/6207043
I'm not sure where the BOM came from, originally.
http://www.gutenberg.org/files/74/74.txt doesn't have it, although
a fresh download of that URL gives me "\r\n"s instead of plain "\n"s,
and the extra line "Character set encoding: ASCII". Maybe Project
Gutenberg has changed their server configuration since we added that
file to the Go repo.
Anyway, this change is just manually excising the BOM from the start
of the file, leaving pure ASCII.
R=r, bradfitz
CC=golang-dev, krasin, rsc
https://golang.org/cl/6197061
dictates a CJK rune is only part of a certain specified range if it
is explicitly defined in the Unicode Codepoint Database.
Fixed the code and some of the tests accordingly.
R=r
CC=golang-dev
https://golang.org/cl/6160044
This is a follow-up to CL 5978051.
Use kernel cas64 helper if we can, fallback to LDREXD/STREXD if
we are on ARMv6 or higher, and to lock-emulated cas64 if on ARMv5.
A future CL will fix {Add,Load,Store}{Int,Uint}64 and issue 3331.
R=bradfitz, rsc
CC=golang-dev
https://golang.org/cl/6034048
Decode AT_RANDOM, AT_HWCAP, and AT_PLATFORM.
This CL only make use of AT_RANDOM, but future CLs will make use of the others.
R=dave, rsc
CC=golang-dev
https://golang.org/cl/5978051
Some newer Linux distributions (Ubuntu ARM at least) use a new multiarch
directory organization, where dynamic linker is no longer in the hardcoded
path in our linker.
For example, Ubuntu 12.04 ARM hardfloat places its dynamic linker at
/lib/arm-linux-gnueabihf/ld-linux.so.3
Ref: http://lackof.org/taggart/hacking/multiarch/
Also, to support Debian GNU/kFreeBSD as a FreeBSD variant, we need this capability, so it's part of issue 3533.
This CL add a new pragma (#pragma dynlinker "path") to cc.
R=iant, rsc
CC=golang-dev
https://golang.org/cl/6086043
1. In CL 5989057, I made a mistake in the last minute change.
"MOVW.W R4, -4(SP)" should really be "MOVW.W R4, -4(R13)",
as 5l will rewrite offset for SP.
2. misc/cgo/test/issue1560.go tests for parallel sleep of 1s,
but on ARM, the deadline is frequently missed, so change sleep
time to 2s on ARM.
R=golang-dev, dave, rsc
CC=golang-dev
https://golang.org/cl/6202043
This adds proper note handling for Plan 9,
and fixes the issue of properly killing go procs.
Without this change, the first go proc that dies
(using runtime·exit()) would kill all the running
go procs. Proper signal handling is needed.
R=golang-dev, ality, rminnich, rsc
CC=golang-dev, john, mirtchovski
https://golang.org/cl/5617048
Instead use a new type, "Note", whose underlying
type is just a string. This change allows us to
remove the exported os.Plan9Note type.
R=bradfitz, seed, rsc
CC=golang-dev
https://golang.org/cl/6015046
Part 1 of CL 5601044 (cgo: Linux/ARM support)
Limitation: doesn't support thumb library yet.
R=golang-dev, dave, rsc
CC=golang-dev
https://golang.org/cl/5991065
crypto.Hash.New() changed to panicking when the hash function isn't
linked in, but crypto/x509 still expects it to return nil.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6175047
This makes it possible to inline the prefetch of upcoming
memory addresses during garbage collection, instead of
needing to flush registers, make a function call, and
reload registers. On garbage collection-heavy workloads,
this results in a 5% speedup.
Fixes#3493.
R=dvyukov, ken, r, dave
CC=golang-dev
https://golang.org/cl/5990066
The first bug was that tertiary ignorables had the same colElem as
implicit colElems, yielding unexpected results. The current encoding
ensures that a non-implicit colElem is never 0. This fix uncovered
another bug of the trie that indexed incorrectly into the null block.
This was caused by an unfinished optimization that would avoid the
need to max out the most-significant bits of continuation bytes.
This bug was also present in the trie used in exp/norm and has been
fixed there as well. The appearence of the bug was rare, as the lower
blocks happened to be nearly nil.
R=r
CC=golang-dev
https://golang.org/cl/6127070