Averaged times (in microseconds) for drawing an 800x600 rectangle
are listed below. The summary is: around a 100x improvement.
draw.Draw call times were typically linear in the number of pixels
touched (i.e. drawing an 800x600 rectangle took 100x as much time as
drawing an 80x60 rectangle).
Before this change, there was only the general-but-slow code path.
When drawing any src with a 50%-opaque mask: 237300 us
When drawing any src with a nil mask: 50100 us
After this change, the 50%-opaque mask case is unchanged.
For an *image.RGBA dst and nil mask and...
...a uniform color (i.e. an image.ColorImage) src: 282 us
...another *image.RGBA src: 615 us.
For the curious, an intermediate implementation detected the special
cases but used simple nested for loops instead of the built-in copy
function. The respective times (compared to 282 and 615 for the
final implementation, or 50100 for the original) were 3110 and 3573.
Times were measured with 8g/8l on my laptop. I haven't tried gccgo
or other architectures.
R=r, rsc
CC=golang-dev
https://golang.org/cl/201048
Marshalling:
* Fixes several silly bugs.
* Support the RawContents type.
* Support the RawValue type.
* Recurse into nested interface{}.
Both directions:
* Better handling of SETs. You can now tag an element in a
structure with "set" to get the correct tag for serialisation.
* For types that aren't in a structure, you can now name them
with "SET" on the end.
* SETs are no longer implicitly treated as SEQUENCEs.
R=rsc
CC=golang-dev
https://golang.org/cl/201049
Make sure to print a time zone when formatting even if none is defined.
Add a comment introducing lookupTimezone (not lookupTimeZone).
Fixes isse 577.
R=rsc
CC=golang-dev
https://golang.org/cl/196090
Added special cases to comments for asin.go and fabs.go.
Added Trunc() to floor.go and floor_386.s. Fixed formatting
error in hypot_386.s Added new functions Acosh, Asinh,
Atanh, Copysign, Erf, Erfc, Expm1, and Log1p. Added
386 FPU version of Fmod. Added tests, benchmarks, and
precision to expected results in all_test.go. Edited
makefile so it all compiles.
R=rsc
CC=golang-dev
https://golang.org/cl/195052
* renamed channels to say what gets sent
* use channel closed status instead of racy check of boolean
R=nigeltao_golang
CC=golang-dev
https://golang.org/cl/196065
In particular, add field-value tokenizer which respects quoting rules.
The code is intended for use in tokenizing the Transfer-Encoding and Trailer fields.
The lexing function is not connected to the main parsing code yet (in the next CL).
R=rsc
CC=golang-dev
https://golang.org/cl/190085
(1) http.Response must close resp.Body after writing.
(2) Case when resp.Body != nil and resp.ContentLength = 0 should not be
treated as an error in Response.Write, because this is what
ReadResponse often returns.
(3) Changed body.th to body.hdr for readability.
R=rsc
CC=golang-dev
https://golang.org/cl/194084
Seems to be enabled spuriously during godoc (can't see why),
producing errors like:
parser.parseDir: src/pkg/http/server.go:159:16: 'Write' declared already at src/pkg/http/request.go:140:21 (and 4 more errors)
R=r
CC=golang-dev
https://golang.org/cl/194119
- provide scope to parse functions; if non-nil, parser uses the scope
to declare and lookup identifiers
- resolve forward references where possible
R=rsc
CC=golang-dev
https://golang.org/cl/194098
Hostname reads the file /proc/sys/kernel/hostname to determine
the value it returns. Some people set this to a Fully Qualified
Doamin Name. At least one implementation of /bin/hostname
truncates the name it gets (often from the "uname" system call)
at the first dot unless it is given a "-f" flag. This change makes
the unit test also truncate at the first dot and checks if the strings
then match. This seems more portable than adding an extra flag
to the called /bin/hostname program.
R=rsc
CC=golang-dev
https://golang.org/cl/181097
Added 386 FPU version of Hypot; modified all_test.go to test
Hypot with large arguments. Also edited sqrt.go to remove
Sqrt(0) as a special case.
R=rsc
CC=golang-dev
https://golang.org/cl/186180
bonus: type switch now detects multiple uses of identical interface types.
bonus: interface types are now order-independent, following the spec.
R=ken2
CC=golang-dev
https://golang.org/cl/194053
5g/6g/8g: add import statements to export metadata, mapping package path to package name.
recognize "" as the path of the package in export metadata.
use "" as the path of the package in object symbol names.
5c/6c/8c, 5a/6a/8a: rewrite leading . to "". so that ·Sin means Sin in this package.
5l/6l/8l: rewrite "" in symbol names as object files are read.
gotest: handle new symbol names.
gopack: handle new import lines in export metadata.
Collectively, these changes eliminate the assumption of a global
name space in the object file formats. Higher level pieces such as
reflect and the computation of type hashes still depend on the
assumption; we're not done yet.
R=ken2, r, ken3
CC=golang-dev
https://golang.org/cl/186263
Other code fixing:
- Fixed bugs in get32.
- Fix code for parsing display string (as a new function).
- Fix code for connecting to X server. The old code only work
if the server is listening to TCP port, otherwise it doesn't
work (at least in my PC).
R=nigeltao_golang, rsc, jhh
CC=golang-dev
https://golang.org/cl/183111
need better management of mach semaphores eventually
but this avoids allocating them for uncontended Locks.
R=r
CC=agl1, golang-dev
https://golang.org/cl/190079
limitations:
poor error handling
teardown not done
exporter must send, importer must receive
testing is rudimentary at best
R=rsc
CC=golang-dev
https://golang.org/cl/186234
Was trying to interpret raw body as chunked body.
Add test for ReadResponse.
Fixes#544.
R=r, petar-m
CC=golang-dev, shadowice
https://golang.org/cl/190068
Wrong position information for identifier
could lead to destructive reformatting of
source via hg gofmt.
R=rsc
CC=golang-dev
https://golang.org/cl/189100
- Identifiers refer now to the language entity (Object)
that they denote. At the moment this is at best an
approximation.
- Initial data structures for language entities (Objects)
and expression types (Type) independent of the actual
type notations.
- Initial support for declaring and looking up identifiers.
- Updated various dependent files and added support functions.
- Extensively tested to avoid breakage. This is an AST change.
R=rsc
CC=golang-dev, rog
https://golang.org/cl/189080
* move memory code into $GOOS-specific directory.
* allow printing of static strings < 256 bytes.
(dynamic strings will bump maxstring as they are allocated.)
* use cgo2c for runtime.mal.
R=r, dho
CC=golang-dev
https://golang.org/cl/186143
Emphasis on minimal interface and fast scanning.
Recognizes all Go literals by default. Easily
configurable to recognize different whitespace
characters and tokens.
Provides detailed position information for each
token.
R=rsc, r
CC=golang-dev
https://golang.org/cl/181160
Added special case tests to all_test.go for Fmod. Fixed Fmod [hung
for Fmod(+/-Inf, <finite>)]. Also added test for Ceil in all_test.go.
R=rsc
CC=golang-dev
https://golang.org/cl/186076
now printed using the semicolon-free style.
Removed NoSemis flag and mechanism dealing with
optional semicolons from go/printer.
Updated all go/printer output files using the
semi-colon free style.
Changes have no formatting impact on checked in
go code under src and misc.
R=rsc, r
CC=golang-dev
https://golang.org/cl/184068
Addresses some cases and respects position of
closing parenthesis; but doesn't work yet in
all cases. Heuristic; will need to be replaced
by a real solution eventually. Does not cause
existing code to be reformatted.
R=r
CC=golang-dev, rsc
https://golang.org/cl/186043
The model is that formatString is a a representation of a standard time,
and that Format converts the time to that representation.
Standard representaitons are defined for ANSIC, RFC850, RFC1123, and ISO8601.
There's also a humane Kitchen fomat: 3:04PM.
R=rsc, benolive, cw
CC=golang-dev
https://golang.org/cl/181130
used only for debugging, debug.go is not normally part of the package source.
also add a dump program to call it.
R=rsc
CC=golang-dev
https://golang.org/cl/183075
(I was looking at this code accidentally because of some gofmt
issues and thought that one could write this more effectively.
You may have deliberately chosen not to use ranges here to make
the index range clearer. Just let me know.)
R=agl, agl1
CC=golang-dev
https://golang.org/cl/181084
Listener contains private members and 6g now enforces that private
members cannot be assigned outside of their package.
R=rsc
CC=golang-dev
https://golang.org/cl/183073
This was convenient for me to have without being forced
to parse the regexp myself. I'd understand if it's not
really wanted, but I also think that some meta information
about compiled regexps would be fine.
R=r, rsc
CC=golang-dev
https://golang.org/cl/183044
as it is not needed anymore (only one impl.
of vector package).
Makefile, vector_test.go, and nogen_test.go
were modified manually (find/replace), the
other files (intvector_test.go, strinvector_test.go
are generated).
Runs all tests.
R=r
https://golang.org/cl/182041
Manual changes to the following files:
src/pkg/Makefile
src/pkg/exp/vector/Makefile (now: src/pkg/container/vector/Makefile)
R=rsc, r
CC=golang-dev
https://golang.org/cl/181041
- use an interface {Get()}
- implement Get for maps, slices
- for slices, retrieves the address of the end of the array, which will give the
same value for every slice of the same array.
R=rsc
CC=golang-dev
https://golang.org/cl/179129
(Thanks to ken and rsc for pointing this out)
rsc:
ken pointed out that there's a race in the new
one-lock-per-channel code. the issue is that
if one goroutine has gone to sleep doing
select {
case <-c1:
case <-c2:
}
and then two more goroutines try to send
on c1 and c2 simultaneously, the way that
the code makes sure only one wins is the
selgen field manipulation in dequeue:
// if sgp is stale, ignore it
if(sgp->selgen != sgp->g->selgen) {
//prints("INVALID PSEUDOG POINTER\n");
freesg(c, sgp);
goto loop;
}
// invalidate any others
sgp->g->selgen++;
but because the global lock is gone both
goroutines will be fiddling with sgp->g->selgen
at the same time.
This results in a 7% slowdown in the single threaded case for a
ping-pong microbenchmark.
Since the cas predominantly succeeds, adding a simple check first
didn't make any difference.
R=rsc
CC=golang-dev
https://golang.org/cl/180068
This is not a complete JPEG implementation (e.g. it does not handle
progressive JPEGs or restart markers), but I was able to take a photo
with my phone, and view the resultant JPEG in pure Go.
The decoder is simple, but slow. The Huffman decoder in particular
should be easily improvable, but optimization is left to future
changelists. Being able to inline functions in the inner loop should
also help performance.
The output is not pixel-for-pixel identical to libjpeg, although
identical behavior isn't necessarily a goal, since JPEG is a lossy
codec. There are at least two reasons for the discrepancy.
First, the inverse DCT algorithm used is the same as Plan9's
src/cmd/jpg, which has different rounding errors from libjpeg's
default IDCT implementation. Note that libjpeg actually has three
different IDCT implementations: one floating point, and two fixed
point. Out of those four, Plan9's seemed the simplest to understand,
partly because it has no #ifdef's or C macros.
Second, for 4:2:2 or 4:2:0 chroma sampling, this implementation does
nearest neighbor upsampling, compared to libjpeg's triangle filter
(e.g. see h2v1_fancy_upsample in jdsample.c).
The difference from the first reason is typically zero, but sometimes
1 (out of 256) in YCbCr space, or double that in RGB space. The
difference from the second reason can be as large as 8/256 in YCbCr
space, in regions of steep chroma gradients. Informal eyeballing
suggests that the net difference is typically imperceptible, though.
R=r
CC=golang-dev, rsc
https://golang.org/cl/164056
parsing and printing to new syntax.
Use -oldparser to parse the old syntax,
use -oldprinter to print the old syntax.
2) Change default gofmt formatting settings
to use tabs for indentation only and to use
spaces for alignment. This will make the code
alignment insensitive to an editor's tabwidth.
Use -spaces=false to use tabs for alignment.
3) Manually changed src/exp/parser/parser_test.go
so that it doesn't try to parse the parser's
source files using the old syntax (they have
new syntax now).
4) gofmt -w src misc test/bench
5th and last set of files.
R=rsc
CC=golang-dev
https://golang.org/cl/180050
parsing and printing to new syntax.
Use -oldparser to parse the old syntax,
use -oldprinter to print the old syntax.
2) Change default gofmt formatting settings
to use tabs for indentation only and to use
spaces for alignment. This will make the code
alignment insensitive to an editor's tabwidth.
Use -spaces=false to use tabs for alignment.
3) Manually changed src/exp/parser/parser_test.go
so that it doesn't try to parse the parser's
source files using the old syntax (they have
new syntax now).
4) gofmt -w src misc test/bench
4th set of files.
R=rsc
CC=golang-dev
https://golang.org/cl/180049
parsing and printing to new syntax.
Use -oldparser to parse the old syntax,
use -oldprinter to print the old syntax.
2) Change default gofmt formatting settings
to use tabs for indentation only and to use
spaces for alignment. This will make the code
alignment insensitive to an editor's tabwidth.
Use -spaces=false to use tabs for alignment.
3) Manually changed src/exp/parser/parser_test.go
so that it doesn't try to parse the parser's
source files using the old syntax (they have
new syntax now).
4) gofmt -w src misc test/bench
3rd set of files.
R=rsc
CC=golang-dev
https://golang.org/cl/180048
parsing and printing to new syntax.
Use -oldparser to parse the old syntax,
use -oldprinter to print the old syntax.
2) Change default gofmt formatting settings
to use tabs for indentation only and to use
spaces for alignment. This will make the code
alignment insensitive to an editor's tabwidth.
Use -spaces=false to use tabs for alignment.
3) Manually changed src/exp/parser/parser_test.go
so that it doesn't try to parse the parser's
source files using the old syntax (they have
new syntax now).
4) gofmt -w src misc test/bench
1st set of files.
R=rsc
CC=agl, golang-dev, iant, ken2, r
https://golang.org/cl/180047
parsing and printing to new syntax.
Use -oldparser to parse the old syntax,
use -oldprinter to print the old syntax.
2) Change default gofmt formatting settings
to use tabs for indentation only and to use
spaces for alignment. This will make the code
alignment insensitive to an editor's tabwidth.
Use -spaces=false to use tabs for alignment.
3) Manually changed src/exp/parser/parser_test.go
so that it doesn't try to parse the parser's
source files using the old syntax (they have
new syntax now).
4) gofmt -w src misc test/bench
2nd set of files.
R=rsc
CC=golang-dev
https://golang.org/cl/179067
mustgetc reports unexpected EOF as SyntaxError. using
mustgetc seems to be a better approach than letting the
caller handle unexpected EOF every time.
name: the second if statement should explicitly return
ok==false.
R=rsc
https://golang.org/cl/174083