shawnps [Sat, 7 Jan 2017 16:23:11 +0000 (08:23 -0800)]
all: fix misspellings
Change-Id: I429637ca91f7db4144f17621de851a548dc1ce76
Reviewed-on: https://go-review.googlesource.com/34923 Reviewed-by: Ian Lance Taylor <iant@golang.org> Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Matthew Dempsky [Fri, 6 Jan 2017 20:25:32 +0000 (12:25 -0800)]
net: disable RFC 6724 Rule 9 for IPv4 addresses
Rule 9 arguably doesn't make sense for IPv4 addresses, and so far it
has only caused problems (#13283, #18518). Disable it until we hear
from users that actually want/need it.
Russ Cox [Fri, 6 Jan 2017 05:54:24 +0000 (00:54 -0500)]
runtime: fix corruption crash/race between select and stack growth
To implement the blocking of a select, a goroutine builds a list of
offers to communicate (pseudo-g's, aka sudog), one for each case,
queues them on the corresponding channels, and waits for another
goroutine to complete one of those cases and wake it up. Obviously it
is not OK for two other goroutines to complete multiple cases and both
wake the goroutine blocked in select. To make sure that only one
branch of the select is chosen, all the sudogs contain a pointer to a
shared (single) 'done uint32', which is atomically cas'ed by any
interested goroutines. The goroutine that wins the cas race gets to
wake up the select. A complication is that 'done uint32' is stored on
the stack of the goroutine running the select, and that stack can move
during the select due to stack growth or stack shrinking.
The relevant ordering to block and unblock in select is:
1. Lock all channels.
2. Create list of sudogs and queue sudogs on all channels.
3. Switch to system stack, mark goroutine as asleep,
unlock all channels.
4. Sleep until woken.
5. Wake up on goroutine stack.
6. Lock all channels.
7. Dequeue sudogs from all channels.
8. Free list of sudogs.
9. Unlock all channels.
There are two kinds of stack moves: stack growth and stack shrinking.
Stack growth happens while the original goroutine is running.
Stack shrinking happens asynchronously, during garbage collection.
While a channel listing a sudog is locked by select in this process,
no other goroutine can attempt to complete communication on that
channel, because that other goroutine doesn't hold the lock and can't
find the sudog. If the stack moves while all the channel locks are
held or when the sudogs are not yet or no longer queued in the
channels, no problem, because no goroutine can get to the sudogs and
therefore to selectdone. We only need to worry about the stack (and
'done uint32') moving with the sudogs queued in unlocked channels.
Stack shrinking can happen any time the goroutine is stopped.
That code already acquires all the channel locks before doing the
stack move, so it avoids this problem.
Stack growth can happen essentially any time the original goroutine is
running on its own stack (not the system stack). In the first half of
the select, all the channels are locked before any sudogs are queued,
and the channels are not unlocked until the goroutine has stopped
executing on its own stack and is asleep, so that part is OK. In the
second half of the select, the goroutine wakes up on its own goroutine
stack and immediately locks all channels. But the actual call to lock
might grow the stack, before acquiring any locks. In that case, the
stack is moving with the sudogs queued in unlocked channels. Not good.
One goroutine has already won a cas on the old stack (that goroutine
woke up the selecting goroutine, moving it out of step 4), and the
fact that done = 1 now should prevent any other goroutines from
completing any other select cases. During the stack move, however,
sudog.selectdone is moved from pointing to the old done variable on
the old stack to a new memory location on the new stack. Another
goroutine might observe the moved pointer before the new memory
location has been initialized. If the new memory word happens to be
zero, that goroutine might win a cas on the new location, thinking it
can now complete the select (again). It will then complete a second
communication (reading from or writing to the goroutine stack
incorrectly) and then attempt to wake up the selecting goroutine,
which is already awake.
The scribbling over the goroutine stack unexpectedly is already bad,
but likely to go unnoticed, at least immediately. As for the second
wakeup, there are a variety of ways it might play out.
* The goroutine might not be asleep.
That will produce a runtime crash (throw) like in #17007:
runtime: gp: gp=0xc0422dcb60, goid=2299, gp->atomicstatus=8
runtime: g: g=0xa5cfe0, goid=0, g->atomicstatus=0
fatal error: bad g->status in ready
Here, atomicstatus=8 is copystack; the second, incorrect wakeup is
observing that the selecting goroutine is in state "Gcopystack"
instead of "Gwaiting".
* The goroutine might be sleeping in a send on a nil chan.
If it wakes up, it will crash with 'fatal error: unreachable'.
* The goroutine might be sleeping in a send on a non-nil chan.
If it wakes up, it will crash with 'fatal error: chansend:
spurious wakeup'.
* The goroutine might be sleeping in a receive on a nil chan.
If it wakes up, it will crash with 'fatal error: unreachable'.
* The goroutine might be sleeping in a receive on a non-nil chan.
If it wakes up, it will silently (incorrectly!) continue as if it
received a zero value from a closed channel, leaving a sudog queued on
the channel pointing at that zero vaue on the goroutine's stack; that
space will be reused as the goroutine executes, and when some other
goroutine finally completes the receive, it will do a stray write into
the goroutine's stack memory, which may cause problems. Then it will
attempt the real wakeup of the goroutine, leading recursively to any
of the cases in this list.
* The goroutine might have been running a select in a finalizer
(I hope not!) and might now be sleeping waiting for more things to
finalize. If it wakes up, as long as it goes back to sleep quickly
(before the real GC code tries to wake it), the spurious wakeup does
no harm (but the stack was still scribbled on).
* The goroutine might be sleeping in gcParkAssist.
If it wakes up, that will let the goroutine continue executing a bit
earlier than we would have liked. Eventually the GC will attempt the
real wakeup of the goroutine, leading recursively to any of the cases
in this list.
* The goroutine cannot be sleeping in bgsweep, because the background
sweepers never use select.
* The goroutine might be sleeping in netpollblock.
If it wakes up, it will crash with 'fatal error: netpollblock:
corrupted state'.
* The goroutine might be sleeping in main as another thread crashes.
If it wakes up, it will exit(0) instead of letting the other thread
crash with a non-zero exit status.
* The goroutine cannot be sleeping in forcegchelper,
because forcegchelper never uses select.
* The goroutine might be sleeping in an empty select - select {}.
If it wakes up, it will return to the next line in the program!
* The goroutine might be sleeping in a non-empty select (again).
In this case, it will wake up spuriously, with gp.param == nil (no
reason for wakeup), but that was fortuitously overloaded for handling
wakeup due to a closing channel and the way it is handled is to rerun
the select, which (accidentally) handles the spurious wakeup
correctly:
if cas == nil {
// This can happen if we were woken up by a close().
// TODO: figure that out explicitly so we don't need this loop.
goto loop
}
Before looping, it will dequeue all the sudogs on all the channels
involved, so that no other goroutine will attempt to wake it.
Since the goroutine was blocked in select before, being blocked in
select again when the spurious wakeup arrives may be quite likely.
In this case, the spurious wakeup does no harm (but the stack was
still scribbled on).
* The goroutine might be sleeping in semacquire (mutex slow path).
If it wakes up, that is taken as a signal to try for the semaphore
again, not a signal that the semaphore is now held, but the next
iteration around the loop will queue the sudog a second time, causing
a cycle in the wakeup list for the given address. If that sudog is the
only one in the list, when it is eventually dequeued, it will
(due to the precise way the code is written) leave the sudog on the
queue inactive with the sudog broken. But the sudog will also be in
the free list, and that will eventually cause confusion.
* The goroutine might be sleeping in notifyListWait, for sync.Cond.
If it wakes up, (*Cond).Wait returns. The docs say "Unlike in other
systems, Wait cannot return unless awoken by Broadcast or Signal,"
so the spurious wakeup is incorrect behavior, but most callers do not
depend on that fact. Eventually the condition will happen, attempting
the real wakeup of the goroutine and leading recursively to any of the
cases in this list.
* The goroutine might be sleeping in timeSleep aka time.Sleep.
If it wakes up, it will continue running, leaving a timer ticking.
When that time bomb goes off, it will try to ready the goroutine
again, leading to any one of the cases in this list.
* The goroutine cannot be sleeping in timerproc,
because timerproc never uses select.
* The goroutine might be sleeping in ReadTrace.
If it wakes up, it will print 'runtime: spurious wakeup of trace
reader' and return nil. All future calls to ReadTrace will print
'runtime: ReadTrace called from multiple goroutines simultaneously'.
Eventually, when trace data is available, a true wakeup will be
attempted, leading to any one of the cases in this list.
None of these fatal errors appear in any of the trybot or dashboard
logs. The 'bad g->status in ready' that happens if the goroutine is
running (the most likely scenario anyway) has happened once on the
dashboard and eight times in trybot logs. Of the eight, five were
atomicstatus=8 during net/http tests, so almost certainly this bug.
The other three were atomicstatus=2, all near code in select,
but in a draft CL by Dmitry that was rewriting select and may or may
not have had its own bugs.
This bug has existed since Go 1.4. Until then the select code was
implemented in C, 'done uint32' was a C stack variable 'uint32 done',
and C stacks never moved. I believe it has become more common recently
because of Brad's work to run more and more tests in net/http in
parallel, which lengthens race windows.
The fix is to run step 6 on the system stack,
avoiding possibility of stack growth.
Fixes #17007 and possibly other mysterious failures.
Austin Clements [Fri, 23 Dec 2016 00:30:23 +0000 (17:30 -0700)]
runtime: update big mgc.go comment
The comment describing the overall GC algorithm at the top of mgc.go
has gotten woefully out-of-date (and was possibly never
correct/complete). Update it to reflect the current workings of the
GC and the set of phases that we now divide it into.
Change-Id: I02143c0ebefe9d4cd7753349dab8045f0973bf95
Reviewed-on: https://go-review.googlesource.com/34711 Reviewed-by: Rick Hudson <rlh@golang.org>
Russ Cox [Fri, 6 Jan 2017 06:17:24 +0000 (01:17 -0500)]
net/http: better failure in TestTransportPersistConnLeak
If one of the c.Get(ts.URL) results in an error, the child goroutine
calls t.Errorf, but the test goroutine gets stuck waiting for <-gotReqCh,
so the test hangs and the program is eventually killed (after 10 minutes!).
Whatever might have been printed to t.Errorf is never seen.
Adjust test so that the test fails cleanly in this case.
Still trying to debug why c.Get might fail.
It seems to have something to do with occasional connection
failures on macOS Sierra.
Austin Clements [Fri, 6 Jan 2017 14:44:41 +0000 (09:44 -0500)]
runtime: use 4K as the boundary of legal pointers
Currently, the check for legal pointers in stack copying uses
_PageSize (8K) as the minimum legal pointer. By default, Linux won't
let you map under 64K, but
1) it's less clear what other OSes allow or will allow in the future;
2) while mapping the first page is a terrible idea, mapping anywhere
above that is arguably more justifiable;
3) the compiler only assumes the first physical page (4K) is never
mapped.
Make the runtime consistent with the compiler and more robust by
changing the bad pointer check to use 4K as the minimum legal pointer.
This came out of discussions on CLs 34663 and 34719.
Lion Yang [Wed, 4 Jan 2017 22:25:59 +0000 (06:25 +0800)]
x/crypto/chacha20poly1305: fix detection of BMI on AMD64
This change uses runtime.support_bmi2 as an additional condition
to examine the usability of AVX2 version algorithm, fixes
the crash on the platfrom which supports AVX2 but not support BMI2.
Fixes #18512
Change-Id: I408c0844ae2eb242dacf70cb9e8cec1b8f3bd941
Reviewed-on: https://go-review.googlesource.com/34851 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Lion Yang [Wed, 4 Jan 2017 21:13:53 +0000 (05:13 +0800)]
crypto: detect BMI usability on AMD64 for sha1 and sha256
The existing implementations on AMD64 only detects AVX2 usability,
when they also contains BMI (bit-manipulation instructions).
These instructions crash the running program as 'unknown instructions'
on the architecture, e.g. i3-4000M, which supports AVX2 but not
support BMI.
This change added the detections for BMI1 and BMI2 to AMD64 runtime with
two flags as the result, `support_bmi1` and `support_bmi2`,
in runtime/runtime2.go. It also completed the condition to run AVX2 version
in packages crypto/sha1 and crypto/sha256.
Fixes #18512
Change-Id: I917bf0de365237740999de3e049d2e8f2a4385ad
Reviewed-on: https://go-review.googlesource.com/34850 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Russ Cox [Wed, 4 Jan 2017 19:12:56 +0000 (14:12 -0500)]
.gitignore: fix attempt at rooted paths
When I wrote the lines
bin/
pkg/
I was trying to match just the top-level bin and pkg directories, and I put the
final slash in because 'git help gitignore' says:
o If the pattern does not contain a slash /, Git treats it as a shell
glob pattern and checks for a match against the pathname relative
to the location of the .gitignore file (relative to the toplevel of
the work tree if not from a .gitignore file).
o Otherwise, Git treats the pattern as a shell glob suitable for
consumption by fnmatch(3) with the FNM_PATHNAME flag: wildcards in
the pattern will not match a / in the pathname. For example,
"Documentation/*.html" matches "Documentation/git.html" but not
"Documentation/ppc/ppc.html" or
"tools/perf/Documentation/perf.html".
Putting a trailing slash was my way of opting in to the "rooted path" semantics
without looking different from the surrounding rooted paths like "src/go/build/zcgo.go".
But HA HA GIT FOOLED YOU! above those two bullets the docs say:
o If the pattern ends with a slash, it is removed for the purpose of
the following description, ...
Change all the patterns to use a leading slash for "rooted" behavior.
This bit me earlier today because I had a perfectly reasonable source
code directory go/src/cmd/go/testdata/src/empty/pkg that was
not added by 'git add empty'.
Change-Id: I6f8685b3c5be22029c33de9ccd735487089a1c03
Reviewed-on: https://go-review.googlesource.com/34832 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Brad Fitzpatrick [Thu, 5 Jan 2017 00:52:04 +0000 (00:52 +0000)]
lib/time: update tzdata to 2016j
Fixes #18500
Change-Id: I4dddd1b99aecf86b9431b0c14f452152dff9b95a
Reviewed-on: https://go-review.googlesource.com/34816 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Brad Fitzpatrick [Wed, 4 Jan 2017 21:23:00 +0000 (21:23 +0000)]
net/http/httputil: make DumpRequest and DumpRequestOut recognize http.NoBody
Fixes #18506
Change-Id: I6b0b107296311178938609e878e1ef47a30a463f
Reviewed-on: https://go-review.googlesource.com/34814 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Brad Fitzpatrick [Wed, 4 Jan 2017 21:03:24 +0000 (21:03 +0000)]
net/http: make Server cancel its ReadTimeout between requests
Fixes #18447
Change-Id: I5d60c3632a5ce625d3bac9d85533ce689e301707
Reviewed-on: https://go-review.googlesource.com/34813 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Michael Marineau [Tue, 3 Jan 2017 08:15:05 +0000 (00:15 -0800)]
runtime: check sched_getaffinity return value
Android on ChromeOS uses a restrictive seccomp filter that blocks
sched_getaffinity, leading this code to index a slice by -errno.
Change-Id: Iec09a4f79dfbc17884e24f39bcfdad305de75b37
Reviewed-on: https://go-review.googlesource.com/34794 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
This will make diagnosis easier when there are multiple function calls per line.
Change-Id: I9881713c5671b847b84a0df0115f57e7cba17d72
Reviewed-on: https://go-review.googlesource.com/34730 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Vladimir Stefanovic [Tue, 27 Dec 2016 15:26:41 +0000 (16:26 +0100)]
runtime: fix SP alignment in mips{,le} sigfwd
Fixes misc/cgo/testsigfwd, enabled for mips{,le} with the next commit
(https://golang.org/cl/34646).
Change-Id: I2bec894b0492fd4d84dd73a4faa19eafca760107
Reviewed-on: https://go-review.googlesource.com/34645 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Daniel Theophanes [Mon, 26 Dec 2016 19:33:46 +0000 (11:33 -0800)]
database/sql: prevent Tx.rollback from racing Tx.close
Previously Tx.done was being set in close, but in a Tx
rollback and Commit are the real closing methods,
and Tx.close is just a helper common to both. Prior to this
change a multiple rollback statements could be called, one
would enter close and begin closing it while the other was
still in rollback breaking it. Fix that by setting done
in rollback and Commit, not in Tx.close.
Austin Clements [Fri, 23 Dec 2016 01:17:32 +0000 (18:17 -0700)]
runtime: document MemStats.BySize fields
Change-Id: Iae8cdcd84e9b5f5d7c698abc6da3fc2af0ef839a
Reviewed-on: https://go-review.googlesource.com/34710 Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>
Emmanuel Odeke [Fri, 23 Dec 2016 11:23:10 +0000 (03:23 -0800)]
test: lock in test for _ assignment evaluation/zerodivide panic
Fixes #5790.
Fixes #18421.
* Lock in _ = x1/x2 divide by zero runtime panics since
it is actually evaluated and not discarded as in previous
versions before Go1.8.
* Update a test that was skipping over zerodivide tests
that expected runtime panics, enabling us to check for
the expected panics.
Change-Id: I0af0a6ecc19345fa9763ab2e35b275fb2d9d0194
Reviewed-on: https://go-review.googlesource.com/34712 Reviewed-by: Keith Randall <khr@golang.org>
Raul Silvera [Wed, 21 Dec 2016 18:16:15 +0000 (10:16 -0800)]
cmd/pprof: Re-enable weblist and disasm
Previous changes started using the full filename for object files
on graph nodes, instead of just the file basename. The basename
was still being used when selecting mappings to disassemble for
weblist and disasm commands, causing a mismatch.
This fixes #18385. It was already fixed on the upstream pprof.
Brad Fitzpatrick [Thu, 1 Dec 2016 21:45:59 +0000 (21:45 +0000)]
net/http: restore Transport's Request.Body byte sniff in limited cases
In Go 1.8, we'd removed the Transport's Request.Body
one-byte-Read-sniffing to disambiguate between non-nil Request.Body
with a ContentLength of 0 or -1. Previously, we tried to see whether a
ContentLength of 0 meant actually zero, or just an unset by reading a
single byte of the Request.Body and then stitching any read byte back
together with the original Request.Body.
That historically has caused many problems due to either data races,
blocking forever (#17480), or losing bytes (#17071). Thus, we removed
it in both HTTP/1 and HTTP/2 in Go 1.8. Unfortunately, during the Go
1.8 beta, we've found that a few people have gotten bitten by the
behavior change on requests with methods typically not containing
request bodies (e.g. GET, HEAD, DELETE). The most popular example is
the aws-go SDK, which always set http.Request.Body to a non-nil value,
even on such request methods. That was causing Go 1.8 to send such
requests with Transfer-Encoding chunked bodies, with zero bytes,
confusing popular servers (including but limited to AWS).
This CL partially reverts the no-byte-sniffing behavior and restores
it only for GET/HEAD/DELETE/etc requests, and only when there's no
Transfer-Encoding set, and the Content-Length is 0 or -1.
Updates #18257 (aws-go) bug
And also private bug reports about non-AWS issues.
Updates #18407 also, but haven't yet audited things enough to declare
it fixed.
Change-Id: Ie5284d3e067c181839b31faf637eee56e5738a6a
Reviewed-on: https://go-review.googlesource.com/34668
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Elias Naur [Thu, 22 Dec 2016 17:31:17 +0000 (18:31 +0100)]
runtime: skip floating point hardware check on Android
CL 33652 removed the fake auxv for Android, and replaced it with
a /proc/self/auxv fallback. When /proc/self/auxv is unreadable,
however, hardware capabilities detection won't work and the runtime
will mistakenly think that floating point hardware is unavailable.
Fix this by always assuming floating point hardware on Android.
Manually tested on a Nexus 5 running Android 6.0.1. I suspect the
android/arm builder has a readable /proc/self/auxv and therefore
does not trigger the failure mode.
Change-Id: I95c3873803f9e17333c6cb8b9ff2016723104085
Reviewed-on: https://go-review.googlesource.com/34641 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Minux Ma <minux@golang.org>
Run-TryBot: Elias Naur <elias.naur@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Austin Clements [Tue, 20 Dec 2016 20:37:56 +0000 (15:37 -0500)]
runtime: avoid CreateThread panic when exiting process
On Windows, CreateThread occasionally fails with ERROR_ACCESS_DENIED.
We're not sure why this is, but the Wine source code suggests that
this can happen when there's a concurrent CreateThread and ExitProcess
in the same process.
Fix this by setting a flag right before calling ExitProcess and
halting if CreateThread fails and this flag is set.
Updates #18253 (might fix it, but we're not sure this is the issue and
can't reproduce it on demand).
Change-Id: I1945b989e73a16cf28a35bf2613ffab07577ed4e
Reviewed-on: https://go-review.googlesource.com/34616
TryBot-Result: Gobot Gobot <gobot@golang.org>
Run-TryBot: Austin Clements <austin@google.com> Reviewed-by: Alex Brainman <alex.brainman@gmail.com> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Kirill Smelkov [Thu, 1 Dec 2016 18:40:57 +0000 (21:40 +0300)]
cmd/compile/internal/ssa: add tests for BSWAP on stores on AMD64
Commit 10f75748 (CL 32222) taught AMD64 backend to rewrite series of
byte loads or stores with corresponding shifts into a single long or
quad load or store + appropriate BSWAP. However it did not added test
for stores - only loads were tested.
Fix it.
NOTE Tests for indexed stores are not added because 10f75748 did not add
support for indexed stores - only indexed loads were handled then.
Kirill Smelkov [Tue, 20 Dec 2016 09:02:39 +0000 (12:02 +0300)]
io: fix PipeWriter.Close to wake up Writes
Since commit cc62bed0 (CL 994043) the pipe deadlock when doing
Read+Close or Write+Close on same end was fixed, alas with test for
Read+Close case only.
Then commit 6d6f3381 (CL 4252057) made a thinko: in the writer path
p.werr is checked for != nil and then err is set but there is no break
from waiting loop unlike break is there in similar condition for reader.
Together with having only Read+Close case tested that made it to leave
reintroduced Write+Close deadlock unnoticed.
Fix it.
Implicitly this also fixes net.Pipe to conform to semantic of net.Conn
interface where Close is documented to unblock any blocked Read or Write
operations.
No test added to net/ since net.Pipe tests are "Assuming that the
underlying io.Pipe implementation is solid and we're just testing the
net wrapping". The test added in this patch should be enough to cover
the breakage.
Brad Fitzpatrick [Tue, 20 Dec 2016 19:32:47 +0000 (19:32 +0000)]
A+C: automated update
Now with a few more repos included.
Add Albert Yu (individual CLA)
Add Alessandro Baffa (individual CLA)
Add Alexandre Fiori (individual CLA)
Add Andrew Austin (individual CLA)
Add Andy Finkenstadt (individual CLA)
Add Antonio Bibiano (individual CLA)
Add Baiju Muthukadan (individual CLA)
Add Ben Lubar (individual CLA)
Add Euan Kemp (individual CLA)
Add Harry Moreno (individual CLA)
Add Jason Buberel (corporate CLA for Google Inc.)
Add Joop Kiefte (individual CLA)
Add Maksym Trykur (individual CLA)
Add Mathieu Olivier (individual CLA)
Add Nick Leli (individual CLA)
Add Nik Nyby (individual CLA)
Add Quinn Slack (corporate CLA for Sourcegraph Inc)
Add Rafal Jeczalik (individual CLA)
Add Raphael Geronimi (individual CLA)
Add Ryan Bagwell (individual CLA)
Add Steve Francia (corporate CLA for Google Inc.)
Add Tristan Colgate (individual CLA)
Add Фахриддин Балтаев (individual CLA)
Updates #12042
Change-Id: Iab98da8a7a9fd3ee54f716ea358b2d515e1e32c4
Reviewed-on: https://go-review.googlesource.com/34658 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Austin Clements [Tue, 20 Dec 2016 03:43:38 +0000 (22:43 -0500)]
runtime: avoid incorrect panic when a signal arrives during STW
Stop-the-world and freeze-the-world (used for unhandled panics) are
currently not safe to do at the same time. While a regular unhandled
panic can't happen concurrently with STW (if the P hasn't been
stopped, then the panic blocks the STW), a panic from a _SigThrow
signal can happen on an already-stopped P, racing with STW. When this
happens, freezetheworld sets sched.stopwait to 0x7fffffff and
stopTheWorldWithSema panics because sched.stopwait != 0.
Fix this by detecting when freeze-the-world happens before
stop-the-world has completely stopped the world and freeze the STW
operation rather than panicking.
Shenghou Ma [Sun, 18 Dec 2016 03:53:38 +0000 (22:53 -0500)]
doc/go1.8: document that CGO_ENABLED is sticky
Fixes #18363.
Change-Id: Ifc98506d33a6753cd7db8e505cf86d5626fbbad0
Reviewed-on: https://go-review.googlesource.com/34596 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Brad Fitzpatrick [Thu, 15 Dec 2016 05:53:01 +0000 (05:53 +0000)]
crypto/x509: speed up and deflake non-cgo Darwin root cert discovery
Piping into security verify-cert only worked on macOS Sierra, and was
flaky for unknown reasons. Users reported that the number of trusted
root certs stopped randomly jumping around once they switched to using
verify-cert against files on disk instead of /dev/stdin.
But even using "security verify-cert" on 150-200 certs took too
long. It took 3.5 seconds on my machine. More than 4 goroutines
hitting verify-cert didn't help much, and soon started to hurt
instead.
New strategy, from comments in the code:
// 1. Run "security trust-settings-export" and "security
// trust-settings-export -d" to discover the set of certs with some
// user-tweaked trusy policy. We're too lazy to parse the XML (at
// least at this stage of Go 1.8) to understand what the trust
// policy actually is. We just learn that there is _some_ policy.
//
// 2. Run "security find-certificate" to dump the list of system root
// CAs in PEM format.
//
// 3. For each dumped cert, conditionally verify it with "security
// verify-cert" if that cert was in the set discovered in Step 1.
// Without the Step 1 optimization, running "security verify-cert"
// 150-200 times takes 3.5 seconds. With the optimization, the
// whole process takes about 180 milliseconds with 1 untrusted root
// CA. (Compared to 110ms in the cgo path)
Fixes #18203
Change-Id: I4e9c11fa50d0273c615382e0d8f9fd03498d4cb4
Reviewed-on: https://go-review.googlesource.com/34389
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org> Reviewed-by: Quentin Smith <quentin@golang.org>
Joe Tsai [Tue, 20 Dec 2016 01:18:45 +0000 (17:18 -0800)]
Revert: "archive/zip: handle mtime in NTFS/UNIX/ExtendedTS extra fields"
This change reverts the following CLs:
CL/18274: handle mtime in NTFS/UNIX/ExtendedTS extra fields
CL/30811: only use Extended Timestamp on non-zero MS-DOS timestamps
We are reverting support for extended timestamps since the support was not
not complete. CL/18274 added full support for reading extended timestamp fields
and minimal support for writing them. CL/18274 is incomplete because it made
no changes to the FileHeader struct, so timezone information was lost when
reading and/or writing.
While CL/18274 was a step in the right direction, we should provide full
support for high precision timestamps in both the reader and writer.
This will probably require that we add a new field of type time.Time.
The complete fix is too involved to add in the time remaining for Go 1.8
and will be completed in Go 1.9.
Dhananjay Nakrani [Mon, 19 Dec 2016 03:25:37 +0000 (19:25 -0800)]
cmd/cover: retain un-attached compiler directives
Parser doesn't attach some compiler directives to anything in the tree.
We have to explicitely retain them in the generated code. This change,
makes cover explicitely print out any compiler directive that wasn't
handled in the ast.Visitor.
Fixes #18285.
Change-Id: Ib60f253815e92d7fc85051a7f663a61116e40a91
Reviewed-on: https://go-review.googlesource.com/34563
Run-TryBot: Rob Pike <r@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Rob Pike <r@golang.org> Reviewed-by: Russ Cox <rsc@golang.org>
Austin Clements [Fri, 16 Dec 2016 16:57:25 +0000 (11:57 -0500)]
runtime: clean up and improve reflect.methodValue comments
The runtime no longer hard-codes the offset of
reflect.methodValue.stack, so remove these obsolete comments. Also,
reflect.methodValue and runtime.reflectMethodValue must also agree
with reflect.makeFuncImpl, so update the comments on all three to
mention this.
This was pointed out by Minux on CL 31138.
Change-Id: Ic5ed1beffb65db76aca2977958da35de902e8e58
Reviewed-on: https://go-review.googlesource.com/34590 Reviewed-by: Keith Randall <khr@golang.org>
Matthew Dempsky [Mon, 19 Dec 2016 16:19:50 +0000 (08:19 -0800)]
cmd/compile: restore zero assignment optimization for non-pointer types
golang.org/cl/31572 disabled some write barrier optimizations, but
inadvertantly disabled optimizations for some non-pointer composite
literal assignments too.
Fixes #18370.
Change-Id: Ia25019bd3016b6ab58173298c7d16202676bce6b
Reviewed-on: https://go-review.googlesource.com/34564
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
Michael Hudson-Doyle [Mon, 12 Dec 2016 00:31:56 +0000 (13:31 +1300)]
cmd/compile, runtime: a different approach to duplicate itabs
golang.org/issue/17594 was caused by additab being called more than once for
an itab. golang.org/cl/32131 fixed that by making the itabs local symbols,
but that in turn causes golang.org/issue/18252 because now there are now
multiple itab symbols in a process for a given (type,interface) pair and
different code paths can end up referring to different itabs which breaks
lots of reflection stuff. So this makes itabs global again and just takes
care to only call additab once for each itab.
Fixes #18252
Change-Id: I781a193e2f8dd80af145a3a971f6a25537f633ea
Reviewed-on: https://go-review.googlesource.com/34173
Run-TryBot: Michael Hudson-Doyle <michael.hudson@canonical.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Crawshaw <crawshaw@golang.org>
Austin Clements [Wed, 14 Dec 2016 18:24:21 +0000 (13:24 -0500)]
runtime: cross-reference _func type better
It takes me several minutes every time I want to find where the linker
writes out the _func structures. Add some comments to make this
easier.
Change-Id: Ic75ce2786ca4b25726babe3c4fe9cd30c85c34e2
Reviewed-on: https://go-review.googlesource.com/34390 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Kevin Burke [Fri, 16 Dec 2016 15:50:08 +0000 (07:50 -0800)]
runtime/pprof: fix spelling in test
Change-Id: Id10e41fe396156106f63a4b29d673b31bea5358f
Reviewed-on: https://go-review.googlesource.com/34551 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Bryan C. Mills [Thu, 15 Dec 2016 22:21:13 +0000 (14:21 -0800)]
runtime: preserve callee-saved C registers in sigtramp
This fixes Linux and the *BSD platforms on 386/amd64.
A few OS/arch combinations were already saving registers and/or doing
something that doesn't clearly resemble the SysV C ABI; those have
been left alone.
Fixes #18328.
Change-Id: I6398f6c71020de108fc8b26ca5946f0ba0258667
Reviewed-on: https://go-review.googlesource.com/34501
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Brad Fitzpatrick [Thu, 15 Dec 2016 16:18:45 +0000 (16:18 +0000)]
net/http: deflake TestServerTimeouts maybe
I haven't been able to reproduce this one, but change a few suspect
things in this test. Notably, using the global "Get" function and thus
using the DefaultTransport was buggy in a parallel test. Then add some error
checks and close a TCP connection.
Hopefully the failure wasn't timing-related.
Fixes #18036 (I hope)
Change-Id: I4904e42e40b26d488cf82111424a1d4d46f42dae
Reviewed-on: https://go-review.googlesource.com/34490
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>