Bill O'Farrell [Tue, 31 Oct 2017 23:18:48 +0000 (19:18 -0400)]
cmd/link: enable c-shared and c-archive mode on s390x
Adding s390x to the list of architectures that support c-shared and c-archive.
Required adding load-time initialization (via _rt0_s390x_linux_lib) and adding s390x
to the c-shared and c-archive tests.
Change-Id: I75883b2891c310fe8ce7f08c27b06895c074e123
Reviewed-on: https://go-review.googlesource.com/74910 Reviewed-by: Michael Munday <mike.munday@ibm.com>
Austin Clements [Wed, 15 Nov 2017 22:43:05 +0000 (14:43 -0800)]
runtime: remove another TODO
I experimented with having the compiler spill the two registers that
are clobbered by the write barrier fast path, but it slightly slows
down compilebench, which is a good write barrier benchmark:
Austin Clements [Wed, 15 Nov 2017 22:26:33 +0000 (14:26 -0800)]
runtime: remove TODO
I experimented with changing the write barrier to take the value in SI
rather than AX to improve register allocation. It had no effect on
performance and only made the "hello world" text 0.07% smaller, so
let's just remove the comment.
Daniel Theophanes [Tue, 14 Nov 2017 16:53:56 +0000 (08:53 -0800)]
database/sql: allow OpenConnector in a driver.Driver interface
While driver.Connector was previously added to allow non-string
connection arguments and access to the context, most users of
the sql package will continue to rely on a string DSN.
Allow drivers to implement a string DSN to Connector interface
that both allows a single parsing of the string DSN and uses
the Connector interface which passes available context to
the driver dialer.
Fixes #22713
Change-Id: Ia0b862262f4c4670effe2538d0d6d43733fea18d
Reviewed-on: https://go-review.googlesource.com/77550 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
I realized this simplification was possible when writing the vet loop
(just above the code being modified here) but never circled back
to make the compiler loop match.
Russ Cox [Tue, 14 Nov 2017 15:57:14 +0000 (10:57 -0500)]
cmd/go/internal/work: do not write trivial.c when testing gcc flags
CL 61111 disabled the writing of trivial.c in -n mode, which
made -n mode at least inconsistent with regular mode in
how it was testing for flags. We think that both were getting
the same answer, so avoid creating the file in both modes
to make sure.
If this CL turns out to be wrong, then when we revert it we
should make sure that the empty file is written even in -n mode,
because this check affects the command-line flags printed
by other commands in that mode.
Change-Id: I0a050bfc148fe5a9d430a153d7816b2821277f0d
Reviewed-on: https://go-review.googlesource.com/78115
Run-TryBot: Russ Cox <rsc@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Russ Cox [Thu, 16 Nov 2017 02:14:21 +0000 (21:14 -0500)]
cmd/go: fix swig support and run swig tests during run.bash
(The tests only run when swig is already installed on the local system.)
Change-Id: I172d106a68cfc746a1058f5a4bcf6761bab88912
Reviewed-on: https://go-review.googlesource.com/78175 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Joe Tsai [Wed, 15 Nov 2017 19:27:10 +0000 (11:27 -0800)]
archive/tar: partially revert sparse file support
This CL removes the following APIs:
type SparseEntry struct{ ... }
type Header struct{ SparseHoles []SparseEntry; ... }
func (*Header) DetectSparseHoles(f *os.File) error
func (*Header) PunchSparseHoles(f *os.File) error
func (*Reader) WriteTo(io.Writer) (int, error)
func (*Writer) ReadFrom(io.Reader) (int, error)
This API was added during the Go1.10 dev cycle, and are safe to remove.
The rationale for reverting is because Header.DetectSparseHoles and
Header.PunchSparseHoles are functionality that probably better belongs in
the os package itself.
The other API like Header.SparseHoles, Reader.WriteTo, and Writer.ReadFrom
perform no OS specific logic and only perform the actual business logic of
reading and writing sparse archives. Since we do know know what the API added to
package os may look like, we preemptively revert these non-OS specific changes
as well by simply commenting them out.
Russ Cox [Thu, 16 Nov 2017 16:49:33 +0000 (11:49 -0500)]
cmd/go/internal/cache: fix build failure introduced in CL 78176
Maybe a bad git merge - not sure.
In any event, I do miss the trybots.
Noticed while fixing: change print-to-stderr+panic
to pure panic, just so that the test (which catches the panic)
does not print any errors before passing.
Russ Cox [Sun, 12 Nov 2017 00:50:19 +0000 (19:50 -0500)]
cmd/go: implement go clean -testcache
Ian suggested that since test caching is not expected to be perfect
in all cases, we should allow users to clear the test cache separately
from clearing the entire build cache.
This CL adds 'go clean -testcache' to do that. The implementation
does not actually delete files (for that, use 'go clean -cache').
Instead, it writes down the current time, and future go tests will
ignore any cached test results written before that time.
Change-Id: I4f84065d7dfc2499fa3f203e9ab62e68d7f367c5
Reviewed-on: https://go-review.googlesource.com/78176 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Russ Cox [Fri, 10 Nov 2017 20:58:16 +0000 (15:58 -0500)]
cmd/go: run cover before cgo
If we're running coverage on a package using cgo, we need to
apply both cmd/cover and cmd/cgo as source transformers.
To date we've applied cgo, then cover.
Cover is very sensitive to the exact character position of
expressions in its input, though, and cgo is not, so swap
them, applying first cover and then cgo.
The only drawback here is that coverage formerly applied
to SWIG-generated cgo files, and now it does not.
I am not convinced anyone depended critically on that,
and probably the later analysis with go tool cover would
have tried to parse the original .swig file as a Go file and
gotten very confused.
Fixes #8726.
Fixes #9212.
Fixes #9479.
Change-Id: I777c8b64f7726cb117d59e03073954abc6dfa34d
Reviewed-on: https://go-review.googlesource.com/77155 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Russ Cox [Fri, 10 Nov 2017 20:46:56 +0000 (15:46 -0500)]
cmd/go: make file:line for cgo files look like non-cgo files
Passing the absolute path to cgo puts the absolute path in the
generated file's //line directives, which then shows that path
in the compiler output, which the go command can then
make relative to the current directory, same as it does for
other compiler output.
Change-Id: Ia2064fea40078c46fd97e3a3b8c9fa1488f913e3
Reviewed-on: https://go-review.googlesource.com/77154 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Russ Cox [Fri, 10 Nov 2017 16:45:16 +0000 (11:45 -0500)]
cmd/cgo: modify source as text, not as AST
Cgo has always operated by rewriting the AST and invoking go/printer.
This CL converts it to use the AST to make decisions but then apply
its edits directly to the underlying source text. This approach worked
better in rsc.io/grind (used during the C to Go conversion) and also
more recently in cmd/cover. It guarantees that all comments and
line numbers are preserved exactly.
This eliminates a lot of special concern about comments and
problems with cgo not preserving meaningful comments.
Combined with the CL changing cmd/cover to use the same
approach, it means that the combination of applying cgo and
applying cover still guarantees all comments and line numbers
are preserved exactly.
This sets us up to fix some cgo vs cover bugs by swapping
the order in which they run during the go command.
This also sets up #16623 a bit: the edit list being
accumulated here is nearly exactly what you'd want
to pass to the compiler for that issue.
Change-Id: I7611815be22e7c5c0d4fc3fa11832c42b32c4eb3
Reviewed-on: https://go-review.googlesource.com/77153 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Russ Cox [Thu, 16 Nov 2017 15:29:50 +0000 (10:29 -0500)]
encoding/csv: restore Go 1.9 quoted \r\n handling in Reader
CL 52810 changed Reader to interpret a quoted \r\n as a raw \r\n
when reading fields. This seems likely to break existing users, and
discussion on both #21201 (the original issue that triggered the change)
and #22746 (discussing whether to revert the change) failed to identify
a single motivating example for this change. To avoid breaking existing
users for no clear reason, revert the change.
The Reader has been rewritten in the interim so this is not a git revert
but instead and adjustment (and slight simplification) of the new Reader.
Fixes #22746.
Change-Id: Ie857b2f4b1359a207d085b6d3c3a6d440a997d12
Reviewed-on: https://go-review.googlesource.com/78295 Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Russ Cox [Tue, 14 Nov 2017 03:32:19 +0000 (22:32 -0500)]
cmd/internal/obj, cmd/trace: restore bounds checks dropped in CL 56950
CL 56950 correctly identified code with checks that were impossible.
But instead of correcting the checks it deleted them.
This CL corrects the code to check what was meant.
Change-Id: Ic89222184ee4fa5cacccae12d750601a9438ac8d
Reviewed-on: https://go-review.googlesource.com/78113
Run-TryBot: Russ Cox <rsc@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Russ Cox [Thu, 16 Nov 2017 02:29:04 +0000 (02:29 +0000)]
Revert "math/rand: make Perm match Shuffle"
This reverts CL 55972.
Reason for revert: this changes Perm's behavior unnecessarily.
I asked for this change originally but I now regret it.
Reverting so that I don't have to justify it in Go 1.10 release notes.
Edited to keep the change to rand_test.go, which seems to have
been mostly unrelated.
Fixes #22744.
Change-Id: If8bb1bcde3ced0db2fdcd0aa65ab128613686c66
Reviewed-on: https://go-review.googlesource.com/78195
Run-TryBot: Russ Cox <rsc@golang.org> Reviewed-by: Emmanuel Odeke <emm.odeke@gmail.com> Reviewed-by: Rob Pike <r@golang.org> Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reason for revert: the CL moves a parser for what appears to be an
Android-specific file format into the main code and makes it available
on all platforms. Android-specific file formats should be limited to
Android.
Russ Cox [Tue, 14 Nov 2017 19:35:39 +0000 (14:35 -0500)]
io: eliminate full copy of copy loop in CopyN
CL 60630 claimed to and did “improve performance of CopyN”
but in doing so introduced a second copy of the I/O copying loop.
This code is subtle and easy to get wrong and the last thing we
need is of two copies that can drift out of sync. Even the newly
introduced copy contains various subtle changes that are not
obviously semantically equivalent to the original. (They probably
are, but it's not obvious.)
Although the CL description does not explain further what the
important optimization was, it appears that the most critical
one was not allocating a 32kB buffer for CopyN(w, r, 512).
This CL deletes the forked copy of copy and instead applies
the buffer size restriction optimization directly to copy itself.
CL 60630 reported:
name old time/op new time/op delta
CopyNSmall-4 5.09µs ± 1% 2.25µs ±86% -55.91% (p=0.000 n=11+14)
CopyNLarge-4 114µs ±73% 121µs ±72% ~ (p=0.701 n=14+14)
Starting with that CL as the baseline, this CL does not change a ton:
name old time/op new time/op delta
CopyNSmall-8 370ns ± 1% 411ns ± 1% +11.18% (p=0.000 n=16+14)
CopyNLarge-8 18.2µs ± 1% 18.3µs ± 1% +0.63% (p=0.000 n=19+20)
It does give up a small amount of the win of 60630 but preserves
the bulk of it, with the benefit that we will not need to debug these
two copies drifting out of sync in the future.
Russ Cox [Fri, 10 Nov 2017 04:58:42 +0000 (23:58 -0500)]
cmd/cover: do not report coverage for assembly functions
cover -func mode was reporting a coverage for function
declarations without bodies - assembly functions.
Since we are not annotating their code, we have no data
for those functions and should not report them at all.
Fixes #6880.
Change-Id: I4b8cd90805accf61f54e3ee167f54f4dc10c7c59
Reviewed-on: https://go-review.googlesource.com/77152 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Russ Cox [Fri, 10 Nov 2017 04:55:15 +0000 (23:55 -0500)]
cmd/cover: add //line comment pointing to original file
Now that cover does not modify the formatting of the original file
or add any newline characters, we can make it print a //line comment
pointing back at the original, and compiler errors and panics will
report accurate line numbers.
Fixes #6329.
Fixes #15757.
Change-Id: I7b0e386112c69beafe69e0d47c5f9e9abc87c0f5
Reviewed-on: https://go-review.googlesource.com/77151
Run-TryBot: Russ Cox <rsc@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Russ Cox [Wed, 15 Nov 2017 20:19:31 +0000 (15:19 -0500)]
database/sql/driver: document that Execer, Queryer need not be implemented
CL 21663 allowed drivers to implement ExecerContext without
also implementing Execer, and similarly QueryerContext without
Queryer, but it did not make that clear in the documentation.
This CL updates the documentation.
Change-Id: I9a4accaac32edfe255fe7c0b0907d4c1014322b4
Reviewed-on: https://go-review.googlesource.com/78129 Reviewed-by: Daniel Theophanes <kardianos@gmail.com>
Russ Cox [Tue, 14 Nov 2017 18:43:02 +0000 (13:43 -0500)]
encoding/hex: make Decode, Decoder, DecodeString agree about partial results and errors
CL 70210 added Decoder for #21590, and in doing so it changed
the existing func Decode to return partial results for decoding errors.
That seems like a good change to make to Decode, but it was
untested (except as used by Decoder), inconsistent with DecodeString
in all error cases, and inconsistent with Decoder in not returning
partial results for odd-length input strings.
This CL makes Decode, DecodeString, and Decoder all agree about
the handling of partial results (they are returned) and error
precedence (the error earliest in the input is reported),
and it documents and tests this.
Kevin Burke [Fri, 14 Jul 2017 22:20:27 +0000 (16:20 -0600)]
crypto/cipher: use raw bytes for keys in docs
Using ASCII values for keys is a bad idea since it makes them vastly
easier to guess. Instead, use the same method as the examples in the
golang.org/x/crypto/nacl package to load keys from a hex value.
Changing the key required updating the ciphertext in many of the
examples.
I am still worried about the fact the examples ask the user to
authenticate messages; authentication isn't trivial, and to be honest
it may be better to steer people to a higher level primitive like
secretbox, unless people really need AES.
Troels Thomsen [Tue, 14 Nov 2017 22:22:19 +0000 (23:22 +0100)]
net: Forget lookups for canceled contexts
A sequential lookup using any non-canceled context has a risk of
returning the result of the previous lookup for a canceled context (i.e.
an error).
This is already prevented for timed out context by forgetting the host
immediately and extending this to also compare the error to
`context.Canceled` resolves this issue.
Russ Cox [Wed, 15 Nov 2017 16:47:31 +0000 (11:47 -0500)]
archive/zip: fix handling of replacement rune in UTF8 check
The replacement rune is a valid rune and can appear as itself in valid UTF8
(it encodes as three bytes). To check for invalid UTF8 it is necessary to
look for utf8.DecodeRune returning the replacement rune and size==1.
Russ Cox [Wed, 15 Nov 2017 20:07:44 +0000 (15:07 -0500)]
database/sql/driver: rename ResetSessioner to SessionResetter
Originally we tried the strict -er suffix as the rule in this case
but eventually we decided it was too awkward: io.WriteByter
became io.ByteWriter. By analogy, here the interface should be
named SessionResetter instead of the awkward ResetSessioner.
This change should not affect any drivers that have already
implemented the interface, because the method name is not changing.
(This was added during the Go 1.10 cycle and has not been
released yet, so we can change it.)
Russ Cox [Tue, 14 Nov 2017 19:01:11 +0000 (14:01 -0500)]
encoding/xml: move unexported const out of exported const block
CL 58210 introduced this constant for reasons I don't understand.
It should not be in the exported const block, which will pollute
godoc output with a "... unexported" notice.
Also since we already have a constant named xmlnsPrefix for "xmlns",
it is very confusing to also have xmlNamespacePrefix for "xml".
If we must have the constant at all, rename it to xmlPrefix.
Russ Cox [Tue, 14 Nov 2017 18:23:10 +0000 (13:23 -0500)]
encoding/csv: rename ParseError.RecordLine to .StartLine
A record can span multiple lines (the whole reason for the extra field),
so the important fact is that it's the _start_ of the record.
Make that clear in the name.
(This API was added during the Go 1.10 cycle so it can still be cleaned up.)
Russ Cox [Tue, 14 Nov 2017 17:38:11 +0000 (12:38 -0500)]
encoding/binary: make new example a bit more idiomatic
Mainly get rid of the weird zero-value struct literal,
but while we're here also group and order things a bit better:
first the reader, then the data, then the call (which takes reader then data).
Russ Cox [Tue, 14 Nov 2017 02:30:22 +0000 (21:30 -0500)]
bytes, strings: restore O(1) behavior of IndexAny(s, "") and LastIndexAny(s, "")
CL 65851 (bytes) and CL 65910 (strings) “improve[d] readability”
by removing the special case that bypassed the whole function body
when chars == "". In doing so, yes, the function was unindented a
level, which is nice, but the runtime of that case went from O(1) to O(n)
where n = len(s).
I don't know if anyone's code depends on the O(1) behavior in this case,
but quite possibly someone's does.
This CL adds the special case back, with a comment to prevent future
deletions, and without reindenting each function body in full.
Joe Tsai [Wed, 15 Nov 2017 18:49:35 +0000 (10:49 -0800)]
archive/tar: change error prefix
Change error message prefix from "tar:" to "archive/tar:" to maintain
backwards compatibility with Go1.9 and earlier in the unfortunate event
that someone is relying on string parsing of errors.
Fixes #22740
Change-Id: I59039c59818a0599e9d3b06bb5a531aa22a389b8
Reviewed-on: https://go-review.googlesource.com/77933 Reviewed-by: roger peppe <rogpeppe@gmail.com>
Daniel Martí [Sun, 29 Oct 2017 21:04:01 +0000 (21:04 +0000)]
go/printer: indent lone comments in composite lits
If a composite literal contains any comments on their own lines without
any elements, the printer would unindent the comments.
The comments in this edge case are written when the closing '}' is
written. Indent and outdent first so that the indentation is
interspersed before the comment is written.
Also note that the go/printer golden tests don't show the exact same
behaviour that gofmt does. Added a TODO to figure this out in a separate
CL.
While at it, ensure that the tree conforms to gofmt. The changes are
unrelated to this indentation fix, however.
Fixes #22355.
Change-Id: I5ac25ac6de95a236f1e123479127cc4dd71e93fe
Reviewed-on: https://go-review.googlesource.com/74232
Run-TryBot: Daniel Martí <mvdan@mvdan.cc> Reviewed-by: Robert Griesemer <gri@golang.org>
Austin Clements [Tue, 14 Nov 2017 23:08:32 +0000 (15:08 -0800)]
runtime: fix gctrace STW CPU time and CPU fraction
The CPU time reported in the gctrace for STW phases is simply
work.stwprocs times the wall-clock duration of these phases. However,
work.stwprocs is set to gcprocs(), which is wrong for multiple
reasons:
1. gcprocs is intended to limit the number of Ms used for mark
termination based on how well the garbage collector actually
scales, but the gctrace wants to report how much CPU time is being
stolen from the application. During STW, that's *all* of the CPU,
regardless of how many the garbage collector can actually use.
2. gcprocs assumes it's being called during STW, so it limits its
result to sched.nmidle+1. However, we're not calling it during STW,
so sched.nmidle is typically quite small, even if GOMAXPROCS is
quite large.
Fix this by setting work.stwprocs to min(ncpu, GOMAXPROCS). This also
fixes the overall GC CPU fraction, which is based on the computed CPU
times.
Marko Mudrinic [Wed, 15 Nov 2017 12:56:01 +0000 (12:56 +0000)]
crypto/sha1, crypto/sha256: fix typo fisrt -> first
Change-Id: I9bb7568ca16b420254796eb38dfb86c5ca3399d7
Reviewed-on: https://go-review.googlesource.com/77890 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Ian Lance Taylor [Sat, 14 Oct 2017 01:26:10 +0000 (18:26 -0700)]
cmd/cgo: permit passing string values directly between Go and C
Permit the C preamble to use the _GoString_ type. Permit Go code to
pass string values directly to those C types. Add accessors for C
code to retrieve sizes and pointers.
Fixes #6907
Change-Id: I190c88319ec88a3ef0ddb99f342a843ba69fcaa3
Reviewed-on: https://go-review.googlesource.com/70890
Run-TryBot: Ian Lance Taylor <iant@golang.org> Reviewed-by: Austin Clements <austin@google.com>
wei xiao [Mon, 28 Nov 2016 02:55:43 +0000 (10:55 +0800)]
runtime: IndexByte and memclr perf improvements on arm64
Update runtime asm_arm64.s and memclr_arm64.s to improve performance by using
SIMD instructions to do more in parallel. It shows improvement on bytes, html
and go1 benchmarks (particualrly regexp, which uses IndexByte frequently).
Tw [Sat, 24 Jun 2017 06:23:17 +0000 (14:23 +0800)]
net/http: don't set Content-Type with empty body automatically
We set Content-Type to "text/plain; charset=utf-8" even with blank body
before. Let's strip this unnecessary header though it's harmless in most
cases.
Unless you go back and read the hash package documentation, it's
not clear that all the hash packages implement marshaling and
unmarshaling. Document the behaviour specifically in each package
that implements it as it this is hidden behaviour and easy to miss.
Ian Lance Taylor [Fri, 10 Nov 2017 17:48:39 +0000 (09:48 -0800)]
runtime: call amd64 VDSO entry points on large stack
If the Linux kernel was built with CONFIG_OPTIMIZE_INLINING=n and was
built with hardening options turned on, GCC will insert a stack probe
in the VDSO function that requires a full page of stack space.
The stack probe can corrupt memory if another thread is using it.
Avoid sporadic crashes by calling the VDSO on the g0 or gsignal stack.
While we're at it, align the stack as C code expects. We've been
getting away with a misaligned stack, but it's possible that the VDSO
code will change in the future to break that assumption.
Benchmarks show a 11% hit on time.Now, but it's only 6ns.
Michael Stapelberg [Tue, 14 Nov 2017 12:46:03 +0000 (04:46 -0800)]
net/mail: treat comment in address as display name
I verified this change on a corpus of > 200 GB of emails since the mid-90s. With
this change, more addresses parse than before, and anything which parsed before
still parses.
In said corpus, I came across the edge case of comments preceding an
addr-spec (with angle brackets!), e.g. “(John Doe) <john@example.com>”, which
does not satisfy the conditions to be treated as a fallback, as per my reading
of RFC2822.
This change does not parse quoted-strings within comments (a corresponding TODO
is in the code), but I have not seen that in the wild.
Hana Kim [Tue, 14 Nov 2017 18:17:44 +0000 (13:17 -0500)]
doc: correct supported gdb version info
GDB 7.5 recognizes DWARF4 by default.
GDB 7.5 release note does not explicitly mention DWARF4 support
but according to GCC 4.8 release note
https://gcc.gnu.org/gcc-4.8/changes.html
"DWARF4 is now the default when generating DWARF debug
information. ...
GDB 7.5, Valgrind 3.8.0 and elfutils 0.154 debug information
consumers support DWARF4 by default."
Change-Id: I56b011c7c38fbc103bbd366ceaea3b709c66ab7f
Reviewed-on: https://go-review.googlesource.com/77570 Reviewed-by: Alberto Donizetti <alb.donizetti@gmail.com>
Brian Kessler [Tue, 14 Nov 2017 06:05:45 +0000 (22:05 -0800)]
math/big: clarify comment on lehmerGCD overflow
A clarifying comment was added to indicate that overflow of a
single Word is not possible in the single digit calculation.
Lehmer's paper includes a proof of the bounds on the size of the
cosequences (u0, u1, u2, v0, v1, v2).
Change-Id: I98127a07aa8f8fe44814b74b2bc6ff720805194b
Reviewed-on: https://go-review.googlesource.com/77451 Reviewed-by: Robert Griesemer <gri@golang.org>
Akhil Indurti [Tue, 14 Nov 2017 16:40:09 +0000 (11:40 -0500)]
net/http/httputil: allow ReverseProxy to call ModifyResponse on failed requests
Previously when RoundTrip returned a non-nil error, the proxy returned a
StatusBadGateway error, instead of first calling ModifyResponse. This
commit first calls ModifyResponse, whether or not the error returned
from RoundTrip is nil.
Also closes response body when ModifyResponse returns an error. See #22658.
Daniel Theophanes [Mon, 13 Nov 2017 22:25:19 +0000 (14:25 -0800)]
database/sql: do not leak the connectionResetter goroutine
Before terminating the connectionResetter goroutine the connection
pool processes all of the connections on the channel to unlock the
driverConn instances so everthing can shutdown cleanly. However
the channel was never closed so the goroutine hangs on the range.
Close the channel prior to ranging over it. Also prevent additional
connections from being sent to the resetter after the connection
pool has been closed.
Joe Tsai [Mon, 13 Nov 2017 21:48:49 +0000 (13:48 -0800)]
hash: document that the encoded state may contain input in plaintext
The cryptographic checksums operate in blocks of 64 or 128 bytes,
which means that the last 128 bytes or so of the input may be encoded
in its original (plaintext) form as part of the state.
Document this so users do not falsely assume that the encoded state
carries no reversible information about the input.
Austin Clements [Thu, 9 Nov 2017 22:55:45 +0000 (17:55 -0500)]
runtime: don't elide wrapper functions that call panic or at TOS
CL 45412 started hiding autogenerated wrapper functions from call
stacks so that call stack semantics better matched language semantics.
This is based on the theory that the wrapper function will call the
"real" function and all the programmer knows about is the real
function.
However, this theory breaks down in two cases:
1. If the wrapper is at the top of the stack, then it didn't call
anything. This can happen, for example, if the "stack" was actually
synthesized by the user.
2. If the wrapper panics, for example by calling panicwrap or by
dereferencing a nil pointer, then it didn't call the wrapped
function and the user needs to see what panicked, even if we can't
attribute it nicely.
This commit modifies the traceback logic to include the wrapper
function in both of these cases.
Michael Munday [Tue, 26 Sep 2017 11:32:26 +0000 (07:32 -0400)]
crypto/elliptic: reduce allocations in s390x P256 code
The Go compiler assumes that pointers escape when passed into assembly
functions. To override this behavior we can annotate assembly functions
with go:noescape, telling the compiler that we know pointers do not
escape from it.
By annotating the assembly functions in the s390x P256 code in this way
we enable more variables to be allocated on the stack rather than
the heap, reducing the number of heap allocations required to execute
this code:
name old alloc/op new alloc/op delta
SignP256 3.66kB ± 0% 2.64kB ± 0% -27.95% (p=0.008 n=5+5)
VerifyP256 4.46kB ± 0% 1.23kB ± 0% -72.40% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
SignP256 40.0 ± 0% 31.0 ± 0% -22.50% (p=0.008 n=5+5)
VerifyP256 41.0 ± 0% 24.0 ± 0% -41.46% (p=0.008 n=5+5)
Change-Id: Id526c30c9b04b2ad79a55d76cab0e30cc8d60402
Reviewed-on: https://go-review.googlesource.com/66230
Run-TryBot: Michael Munday <mike.munday@ibm.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Daniel Martí [Sun, 22 Oct 2017 16:18:51 +0000 (17:18 +0100)]
cmd/compile: remove some more gotos in gc
Split typecheckrange into two, separating the bigger chunk of code that
takes care of the range expression. It had to sometimes exit early,
which was done via a goto in the larger func. This lets us simplify many
declarations and the flow of the code. While at it, also replace the
toomany int with a bool.
In the case of walkselect, split it into two funcs too since using a
defer for all the trailing work would be a bit much. It also lets us
simplify the declarations and the flow of the code, since now
walkselectcases has a narrower scope and straightforward signature.
Also replace the gotos in typecheckaste with a lineno defer.
Passes toolstash -cmp on std cmd.
Change-Id: Iacfaa0a34c987c44f180a792c473558785cf6823
Reviewed-on: https://go-review.googlesource.com/72374
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Kevin Burke [Wed, 19 Jul 2017 14:21:30 +0000 (07:21 -0700)]
doc: mention the scratch repo
If people are interested in contributing to Go, but not sure what
change they'd like to make just yet, we can point them to the scratch
repo, so they can go through the process of submitting and merging
something now, and make more useful contributions later.
My evidence that sending people to the scratch repo would encourage
future contributions is that a number of people who went through the
workshop at Gophercon have continued to send CL's after submitting to
the scratch repo, even though I doubt they planned to before going
through the workshop.
Keith Randall [Mon, 13 Nov 2017 18:52:35 +0000 (10:52 -0800)]
cmd/compile: add test for array decomposition
This test fails on 1.9.2, but is ok on tip.
CL 77331 has both the 1.9.2 fix and this test, and is on the 1.9 release branch.
This CL is just the test, and is on HEAD. The buggy code doesn't exist on tip.
Update #22683
Change-Id: I04a24bd6c2d3068e18ca81da3347e2c1366f4447
Reviewed-on: https://go-review.googlesource.com/77332
Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Joe Tsai [Fri, 10 Nov 2017 04:14:47 +0000 (20:14 -0800)]
encoding/json: always ignore embedded pointers to unexported struct types
CL 60410 fixes a bug in reflect that allows assignments to an embedded
field of a pointer to an unexported struct type.
This breaks the json package because unmarshal is now unable to assign
a newly allocated struct to such fields.
In order to be consistent in the behavior for marshal and unmarshal,
this CL changes both marshal and unmarshal to always ignore
embedded pointers to unexported structs.
Fixes #21357
Change-Id: If62ea11155555e61115ebb9cfa5305caf101bde5
Reviewed-on: https://go-review.googlesource.com/76851
Run-TryBot: Joe Tsai <thebrokentoaster@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Leigh McCulloch [Mon, 13 Nov 2017 06:27:06 +0000 (06:27 +0000)]
doc: add a link to the code of conduct
GitHub has defined a set of standard meta files to include with projects
hosted on GitHub. According to the GitHub Insights Community page for
go project the only one missing is the code of conduct.
Go has a code of conduct on it's website and we should link to it in the
prescribed `.github/CODE_OF_CONDUCT.md` so that people can find it in
the standard location on GitHub. This would be consistent with the
contribution guidelines that are linked to in `.github/CONTRIBUTING.md`.
griesemer [Fri, 10 Nov 2017 19:38:51 +0000 (11:38 -0800)]
cmd/compile: record original and absolute file names for line directives
Also, with this change, error locations don't print absolute positions
in [] brackets following positions relative to line directives. To get
the absolute positions as well, specify the -L flag.
Russ Cox [Fri, 10 Nov 2017 04:39:12 +0000 (23:39 -0500)]
cmd/cover: modify source as text, not as AST
cmd/cover rewrites Go source code to add coverage annotations.
The approach to date has been to parse the code to AST, analyze it,
rewrite the AST, and print it back out. This approach fails to preserve
line numbers in the original code and has a very difficult time with
comments, because go/printer does as well.
This CL changes cmd/cover to decide what to modify based on the
AST but to apply the modifications as purely textual substitutions.
In this way, cmd/cover can be sure it never adds or removes a newline
character, nor a comment, so all line numbers and comments are
preserved.
This also allows us to emit a single //line comment at the beginning
of the translated file and have the compiler report errors with
correct line numbers in the original file.
Fixes #6329.
Fixes #15757.
Change-Id: Ia95f6f894bb498e80d1f91fde56cd4a8009d7f9b
Reviewed-on: https://go-review.googlesource.com/77150 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Russ Cox [Fri, 10 Nov 2017 19:50:03 +0000 (14:50 -0500)]
cmd/go: more refined handling of cmd/compile magic environment variables
Per discussion with David Chase, need to check GOSSAHASH$n
for increasing n until one is missing. Also if GSHS_LOGFILE is set,
the compiler writes to that file, so arrange never to cache in that case.
Change-Id: I3931b4e296251b99abab9bbbbbdcf94ae8c1e2a6
Reviewed-on: https://go-review.googlesource.com/77111 Reviewed-by: David Chase <drchase@google.com>
Russ Cox [Fri, 10 Nov 2017 19:44:31 +0000 (14:44 -0500)]
cmd/go: cache and replay command output during build
It's nice that
go build -gcflags=-m errors
go build -gcflags=-m errors
uses the cache for the second command.
Even nicer is to make the second command
print the same output as the first command.
Fixes #22587.
Change-Id: I64350839f01c86c9a095d9d22f6924cd7a0b9105
Reviewed-on: https://go-review.googlesource.com/77110 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Daniel Martí [Sun, 12 Nov 2017 16:55:11 +0000 (16:55 +0000)]
cmd/compile: use stringer on types.EType
Another one that is possible thanks to the new -trimprefix stringer
flag.
The only subtle difference is that, in the previous version, some values
such as TUNSAFEPTR were stringified as "TUNSAFEPTR" instead of
"UNSAFEPTR". The new String method is always consistent in removing the
"T" prefix.
Passes toolstash -cmp on std cmd.
Change-Id: I68407f391795403dfcbbfa68c813018c0235bbb5
Reviewed-on: https://go-review.googlesource.com/77250
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com> Reviewed-by: Robert Griesemer <gri@golang.org>
Adam Langley [Mon, 30 Oct 2017 02:20:33 +0000 (19:20 -0700)]
crypto/x509: handle name constraints with cryptobyte
This allows better precision and (the motivation) empty strings to
be handled correctly. With that in place tests for the behaviour of
empty name constraints can be added.
Also fixes a compatibility issue with NSS. See #22616.
David du Colombier [Fri, 10 Nov 2017 20:46:41 +0000 (21:46 +0100)]
cmd/go: skip TestGoTestJSON on uniprocessor machines
CL 76873 added TestGoTestJSON. However, this test
is only succeeding on SMP machines.
This change skips TestGoTestJSON on uniprocessor machines.
Fixes #22665.
Change-Id: I3989d3331fb71193a25a3f0bbb84ff3e1b730890
Reviewed-on: https://go-review.googlesource.com/77130 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Alex Brainman [Mon, 28 Aug 2017 02:21:44 +0000 (12:21 +1000)]
net: ignore duplicate interfaces in TestInterfaceHardwareAddrWithGetmac
Sometimes getmac lists many interfaces for the same MAC address,
while Interfaces returns only single name for that address. Adjust
the test to ignore the names that are not returned by the Interfaces.