This CL adjusts code referring to src/pkg to refer to src.
Immediately after submitting this CL, I will submit
a change doing 'hg mv src/pkg/* src'.
That change will be too large to review with Rietveld
but will contain only the 'hg mv'.
This CL will break the build.
The followup 'hg mv' will fix it.
For more about the move, see golang.org/s/go14nopkg.
runtime: implement time.now in assembly on plan9, solaris, windows
These all used a C implementation that contained 64-bit divide by 1000000000.
On 32-bit systems that ends up in the 64-bit C divide support, which makes
other calls and ends up using a fair amount of stack. We could convert them
to Go but then they'd still end up in software 64-bit divide code. That would
be okay, because Go code can split the stack, but it's still unnecessary.
Write time·now in assembly, just like on all the other systems, and use the
actual hardware support for 64/32 -> 64/32 division. This cuts the software
routines out entirely.
The actual code to do the division is copied and pasted from the sys_darwin_*.s files.
LGTM=alex.brainman
R=golang-codereviews, alex.brainman
CC=aram, golang-codereviews, iant, khr, r
https://golang.org/cl/136300043
panic: httptest: failed to listen on a port: listen tcp 127.0.0.1:0:
listen: An operation on a socket could not be performed because the
system lacked sufficient buffer space or because a queue was full.
Since we can't seem to understand what the test is trying to test,
and because it is causing problems on multiple systems,
delete it.
This is one of those "how did this ever work?" bugs.
The current build failures are happening because
a fault comes up while executing on m->curg on a
system-created thread using an m obtained from needm,
but TLS is set to m->g0, not m->curg. On fault,
sigtramp starts executing, assumes r10 (g) might be
incorrect, reloads it from TLS, and gets m->g0, not
m->curg. Then sighandler dutifully pushes a call to
sigpanic onto the stack and returns to it.
We're now executing on the m->curg stack but with
g=m->g0. Sigpanic does a stack split check, sees that
the SP is not in range (50% chance depending on relative
ordering of m->g0's and m->curg's stacks), and then
calls morestack. Morestack sees that g=m->g0 and
crashes the program.
The fix is to replace every change of g in asm_arm.s
with a call to a function that both updates g and
saves the updated g to TLS.
Why did it start happening? That's unclear.
Unfortunately there were other bugs in the initial
checkin that mask exactly which of a sequence of
CLs started the behavior where sigpanic would end
up tripping the stack split.
Fixes arm build.
Fixes #8675.
LGTM=iant
R=golang-codereviews, iant
CC=dave, golang-codereviews, khr, minux, r
https://golang.org/cl/135570043
After the three pending CLs listed below, there will be no more .goc files.
134580043 runtime: move stubs.goc code into runtime.c 133670043 runtime: fix windows syscalls for copying stacks 141180043 runtime: eliminate Go -> C -> block paths for Solaris
LGTM=bradfitz
R=golang-codereviews, bradfitz, dave
CC=golang-codereviews, iant, r
https://golang.org/cl/132680043
Syscall and everything it calls must be nosplit:
we cannot split a stack once Syscall has been invoked,
because we don't know which of its arguments are
pointers.
Increase NOSPLIT reservation from 192 to 384 bytes.
The problem is that the non-Unix systems (Solaris and Windows)
just can't make system calls in a small amount of space,
and then worse they do things that are complex enough
to warrant calling runtime.throw on failure.
We don't have time to rewrite the code to use less stack.
I'm not happy about this, but it's still a small amount.
The good news is that we're doing this to get to only
using copying stacks for stack growth. Once that is true,
we can drop the default stack size from 8k to 4k, which
should more than make up for the bytes we're losing here.
The gp->panicwrap adjustment is just fatally flawed.
Now that there is a Panic.argp field, update that instead.
That can be done on entry only, so that unwinding doesn't
need to worry about undoing anything. The wrappers
emit a few more instructions in the prologue but everything
else in the system gets much simpler.
It also fixes (without trying) a broken test I never checked in.
Fixes #7491.
LGTM=khr
R=khr
CC=dvyukov, golang-codereviews, iant, r
https://golang.org/cl/135490044
If there is doubt about passing arguments correctly
(as there is in this test), there should be doubt about
getting the results back intact too. Using 0 and 1
(especially 0 for success) makes it easy to get a PASS
accidentally when the return value is not actually
being propagated. Use less common values.
LGTM=iant
R=golang-codereviews, iant
CC=golang-codereviews, r
https://golang.org/cl/141110043
runtime: handle nil ptr load/store in arm software floating point
We cannot let a real panic start there, because there is C code
on the stack, and worse, there is an assembly frame with a
saved copy of the registers and we have no idea which ones
are pointers.
Instead, detect the nil ptr load/store and return out of the C
and assembly into a stub that will start the call to sigpanic.
Fixes GOARM=5 build.
LGTM=iant
R=golang-codereviews, iant
CC=dave, golang-codereviews, minux, r
https://golang.org/cl/138130043
sigprof and setcpuprofilerate coordinate the enabling/disabling
of the handler using a Mutex. This has always been a bit dodgy:
setcpuprofilerate must be careful to turn off signals before acquiring
the lock to avoid a deadlock.
Now the lock implementations use onM, and onM isn't okay on the
signal stack. We know how to make it okay, but it's more work than
is probably worth doing.
Since this is super-dodgy anyway, replace the lock with a simple
cas loop. It is only contended if setcpuprofilerate is being called,
and that doesn't happen frequently enough to care about the
raw speed or about using futexes/semaphores.
TBR to fix freebsd/amd64 and dragonfly/amd64 builds.
Happy to make changes in a follow-up CL.
The general kernel system call interface
takes 6 arguments: R0, R1, R2, R3, R4, R5.
Syscall is for calls that only need 3.
The amd64 and 386 versions zero the extra arg registers,
but the arm version does not.
func utimensat calls Syscall with 3 arguments.
The kernel expects a 4th argument.
That turns out to be whatever is in R3 at the time of the call.
CL 137160043 changed various pieces of code and apparently
changed the value left in R3 at the time of utimensat's Syscall.
This causes the kernel to return EINVAL.
Change linux/arm Syscall to zero R3, R4, R5, so that calls will
behave deterministically, even if they pass too few arguments.
Arguably, utimensat could be fixed too, but the predictable
zeroing is certainly worth doing, and once done utimensat's
use of Syscall is fine.
1. If onM is called on a g0 stack, it just calls the given function.
2. If onM is called on a gsignal stack, it calls badonm.
3. If onM is called on a curg stack, it switches to the g0 stack
and then calls the function.
In cases 1 and 2, if the program then crashes (and badonm always does),
we want to see what called onM, but the traceback stops at onM.
In case 3, the traceback must stop at onM, because the g0
stack we are renting really does stop at onM.
The current code stops the traceback at onM to handle 3,
at the cost of making 1 and 2 crash with incomplete traces.
Change traceback to scan past onM but in case 3 make it look
like on the rented g0 stack, onM was called from mstart.
The traceback already knows that mstart is a top-of-stack function.
Alternate fix at CL 132610043 but I think this one is cleaner.
This CL makes 3 the exception, while that CL makes 1 and 2 the exception.
Submitting TBR to try to get better stack traces out of the
freebsd/amd64 builder, but happy to make changes in a
followup CL.
cmd/dist: make textflag.h available in runtime, avoid android/linux conflicts
1) cmd/dist was copying textflag.h to the build include directory,
but only after compiling package runtime. So other packages could
use it, just not runtime. Copy earlier, so that runtime can use it too.
2) We decided for android that anything marked linux is also included
in the build. The generated linux-specific files in cmd/dist must therefore
have explicit +build !android tags, or else you can't have simultaneous
linux/arm and android/arm builds in a single client. The tag was already
there for at least one file, but it was missing from many others.
cmd/api: don't depend on os/user or USER to check api
The -nocgo builder failed because it has cgo disabled
as well as no USER environment variable:
http://build.golang.org/log/2250abb82f5022b72a12997b8ff89fcdeff094c9
# Checking API compatibility.
Error getting current user: user: Current not implemented on linux/amd64
exit status 1
The original conversion in CL 132090043 cut up
the function in an attempt to avoid converting most
of the code to Go. This contorts the control flow.
While debugging the onM signal stack bug,
I reconverted sigqueue.goc in its entirety.
This restores the original control flow, which is
much easier to understand.
The current conversion is correct, it's just complex
and will be hard to maintain. The new one is as
readable as the original code.
I uploaded sigqueue.goc as the initial copy of
sigqueue.go in the CL, so if you view the diffs
of sigqueue.go comparing against patch set 2 [sic]
it will show the actual starting point.
For example:
https://golang.org/cl/136160043/diff2/20001:60001/src/pkg/runtime/sigqueue.go
LGTM=dvyukov, iant
R=golang-codereviews, dvyukov, iant
CC=golang-codereviews, khr, r
https://golang.org/cl/136160043
The arm5 build breakage at CL 139110043 was caused by
calling funcPC on a lessstack defined as a struct{}.
That symbol ended up with a non-4-aligned address,
which caused the memory fault that broke the builders.
The definition of lessstack was fixed in CL 140880043.
Tracking that down suggested that it would be worth
looking for the same bug elsewhere in the directory.
This is the only one I found.
LGTM=bradfitz
R=golang-codereviews, dave, bradfitz
CC=dvyukov, golang-codereviews, iant, khr, r
https://golang.org/cl/134410043
The implementation and use patterns of onM assume
that they run on either the m->curg or m->g0 stack.
Calling onM from m->gsignal has two problems:
(1) When not on g0, onM switches to g0 and then "back" to curg.
If we didn't start at curg, bad things happen.
(2) The use of scalararg/ptrarg to pass C arguments and results
assumes that there is only one onM call at a time.
If a gsignal starts running, it may have interrupted the
setup/teardown of the args for an onM on the curg or g0 stack.
Using scalararg/ptrarg itself would smash those.
We can fix (1) by remembering what g was running before the switch.
We can fix (2) by requiring that uses of onM that might happen
on a signal handling stack must save the old scalararg/ptrarg
and restore them after the call, instead of zeroing them.
The only sane way to do this is to introduce a separate
onM_signalsafe that omits the signal check, and then if you
see a call to onM_signalsafe you know the surrounding code
must preserve the old scalararg/ptrarg values.
(The implementation would be that onM_signalsafe just calls
fn if on the signal stack or else jumps to onM. It's not necessary
to have two whole copies of the function.)
(2) is not a problem if the caller and callee are both Go and
a closure is used instead of the scalararg/ptrarg slots.
For now, I think we can avoid calling onM from code executing
on gsignal stacks, so just reject it.
In the long term, (2) goes away (as do the scalararg/ptrarg slots)
once everything is in Go, and at that point fixing (1) would be
trivial and maybe worth doing just for regularity.
LGTM=iant
R=golang-codereviews, iant
CC=dvyukov, golang-codereviews, khr, r
https://golang.org/cl/135400043
Instead of making asmcgocall call asmcgocall_errno,
make both load args into registers and call a shared
assembly function.
On amd64, this costs 1 word in the asmcgocall_errno path
but saves 3 words in the asmcgocall path, and the latter
is what happens on critical nosplit paths on Windows.
On arm, this fixes build failures: asmcgocall was writing
the arguments for asmcgocall_errno into the wrong
place on the stack. Passing them in registers avoids the
decision entirely.
On 386, this isn't really needed, since the nosplit paths
have twice as many words to work with, but do it for consistency.
net: fix parsing literal IP address in builtin dns stub resolver
This CL fixes a bug introduced by CL 128820043 which is that
builtin dns stub resolver doesn't work well with literal IPv6
address namesever entries in /etc/resolv.conf.
Also simplifies resolv.conf parser and adds more test cases.
Rob Pike [Wed, 3 Sep 2014 22:57:03 +0000 (15:57 -0700)]
text/template: 0xef is an integer, not a floating-point value.
The discriminator in the execution engine was stupid.
Add a test to the parse package too. The problem wasn't there
but the particular case ('e' in a hex integer) was not covered.
runtime: make entersyscall/exitsyscall safe for stack splits
It is fundamentally unsafe to grow the stack once someone
has made a call to syscall.Syscall. That function takes 6 uintptr
arguments, but depending on the call some are pointers.
In fact, some might be pointers to stack values, and we don't know which.
That makes it impossible to copy the stack somewhere else.
Since we want to delete all the stack splitting code, relying only
on stack copying, make sure that Syscall never needs to split the stack.
The only thing Syscall does is:
call entersyscall
make the system call
call exitsyscall
As long as we make sure that entersyscall and exitsyscall
can live in the nosplit region, they won't ask for more stack.
Do this by making entersyscall and exitsyscall set up the
stack guard so that any call to a function with a split check
will cause a crash. Then move non-essential slow-path
work onto the m stack using onM and mark the rest of the
work nosplit. The linker will verify that the chain of nosplits
fits in the total nosplit budget.
LGTM=iant
R=golang-codereviews, iant
CC=dvyukov, golang-codereviews, khr, r
https://golang.org/cl/140950043
Because symtab.c was partially converted before,
the diffs are not terribly useful.
The earlier conversion was trying to refactor or
clean up the code in addition to doing the translation.
It also made a mistake by redefining Func to be something
users could overwrite.
I undid those changes, making symtab.go a more
literal line-for-line translation of symtab.c instead.
LGTM=josharian
R=golang-codereviews, dave, bradfitz, josharian
CC=golang-codereviews, iant, khr, r
https://golang.org/cl/140880043
Rick Hudson [Wed, 3 Sep 2014 16:06:36 +0000 (12:06 -0400)]
runtime: Start and stop individual goroutines at gc safepoints
Code to bring goroutines to a gc safepoint one at a time,
do some work such as scanning, and restart the
goroutine, and then move on to the next goroutine.
Currently this code does not do much useful work
but this infrastructure will be critical to future
concurrent GC work.
Alex Brainman [Wed, 3 Sep 2014 04:17:04 +0000 (14:17 +1000)]
runtime: increase windows prof thread priority sooner
If system is busy burning cpu, it takes long time (about 300ms on
windows builders) to adjust prof thread priority. Once adjusted, prof
thread runs ahead of everyone else, but due to initial slowness, it
does not capture prof snapshots until start-up period is completed.
Change prof thread priority sooner, so it can start captures straight
away.
The code I wrote originally works for trivial functions
that are inlined at a call site in another package,
because that was how I wrote my local test.
Make hex(x) work for non-inlinable functions too.
LGTM=iant
R=golang-codereviews, iant
CC=golang-codereviews, r
https://golang.org/cl/140830043
Marko Tiikkaja [Tue, 2 Sep 2014 16:08:41 +0000 (09:08 -0700)]
database/sql: Avoid re-preparing statements when all connections are busy
Previously, if all connections were busy, we would always
re-prepare the statement on the connection we were assigned from
the pool. That meant that if all connections were busy most of the
time, the number of prepared statements for each connection would
keep increasing over time.
Instead, after getting a free connection, check to see if the
statement has already been prepared on it, and reuse the statement
handle if so.
runtime: fix race detector running Go code on g0 of non-main thread
It looks like this has just always been broken:
the race detector handles running Go code on g0 of the main thread
and on g0 of any extra threads created by non-Go code, but it does
not handle running Go code on g0 of non-main threads created by Go.
Handle that.
Should fix the race build failures on the dashboard.
We're running into this now because we are running more
and more Go code on g0.
Alex Brainman [Tue, 2 Sep 2014 01:54:24 +0000 (11:54 +1000)]
time: update generated zoneinfo_abbrs_windows.go
The file in repo has been updated recently, but all these changes
are gone off the web site now. It seems web site gets updated once
in a while, so we'll update our file occasionally.
LGTM=r
R=golang-codereviews, r
CC=golang-codereviews
https://golang.org/cl/140780043