Change the signature of Split to have no count,
assuming a full split, and rename the existing
Split with a count to SplitN.
Do the same to package bytes.
Add a gofix module.
R=adg, dsymonds, alex.brainman, rsc
CC=golang-dev
https://golang.org/cl/4661051
grsec needs the FIXED flag to be provided to mmap, which
works now. That said, when the allocation fails to be made
in the specific address, we're still given back a writable
page. This change will unmap that page to avoid using
twice the amount of memory needed.
It'd also be pretty easy to avoid the extra system calls
once we detected that the flag is needed, but I'm not sure
if that edge case is worth the effort.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/4634086
All but two packages depend on net:
debug/proc
os/signal
With this change, we can produce
a working build with GOOS=plan9.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/4639053
Correct a few error messages (libcgo -> runtime/cgo)
and delete old nacl_386.c file too.
Fixes#1657.
R=iant
CC=golang-dev
https://golang.org/cl/4603057
5a: add SQRTF and SQRTD
5l: add ASQRTF and ASQRTD
Use ARMv7 VFP VSQRT instruction to speed up math.Sqrt
R=rsc, dave, m
CC=golang-dev
https://golang.org/cl/4551082
This change was adapted from gccgo's libgo/runtime/mem.c at
Ian Taylor's suggestion. It fixes all.bash failing with
"address space conflict: map() =" on amd64 Linux with kernel
version 2.6.32.8-grsec-2.1.14-modsign-xeon-64.
With this change, SysMap will use MAP_FIXED to allocate its desired
address space, after first calling mincore to check that there is
nothing else mapped there.
R=iant, dave, n13m3y3r, rsc
CC=golang-dev
https://golang.org/cl/4438091
breaks Mac build
««« original CL description
runtime: use HOST_CC to compile mkversion
HOST_CC is set in Make.inc, so use that rather
than hardcoding quietgcc
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/4515163
»»»
R=iant
CC=golang-dev
https://golang.org/cl/4515168
Works around bug in kernel implementation on old ARM5 kernels.
Bug was fixed on 26 Nov 2007 (between 2.6.23 and 2.6.24) but
old kernels persist.
Fixes#1750.
R=dfc, golang-dev
CC=golang-dev
https://golang.org/cl/4436072
The g->sched.sp saved stack pointer and the
g->stackbase and g->stackguard stack bounds
can change even while "the world is stopped",
because a goroutine has to call functions (and
therefore might split its stack) when exiting a
system call to check whether the world is stopped
(and if so, wait until the world continues).
That means the garbage collector cannot access
those values safely (without a race) for goroutines
executing system calls. Instead, save a consistent
triple in g->gcsp, g->gcstack, g->gcguard during
entersyscall and have the garbage collector refer
to those.
The old code was occasionally seeing (because of
the race) an sp and stk that did not correspond to
each other, so that stk - sp was not the number of
stack bytes following sp. In that case, if sp < stk
then the call scanblock(sp, stk - sp) scanned too
many bytes (anything between the two pointers,
which pointed into different allocation blocks).
If sp > stk then stk - sp wrapped around.
On 32-bit, stk - sp is a uintptr (uint32) converted
to int64 in the call to scanblock, so a large (~4G)
but positive number. Scanblock would try to scan
that many bytes and eventually fault accessing
unmapped memory. On 64-bit, stk - sp is a uintptr (uint64)
promoted to int64 in the call to scanblock, so a negative
number. Scanblock would not scan anything, possibly
causing in-use blocks to be freed.
In short, 32-bit platforms would have seen either
ineffective garbage collection or crashes during garbage
collection, while 64-bit platforms would have seen
either ineffective or incorrect garbage collection.
You can see the invalid arguments to scanblock in the
stack traces in issue 1620.
Fixes#1620.
Fixes#1746.
R=iant, r
CC=golang-dev
https://golang.org/cl/4437075
runtime: memory allocated by OS not in usable range
runtime: out of memory: cannot allocate 1114112-byte block (2138832896 in use)
throw: out of memory
runtime.throw+0x40 /Users/rsc/g/go/src/pkg/runtime/runtime.c:102
runtime.throw(0x1fffd, 0x101)
runtime.mallocgc+0x2af /Users/rsc/g/go/src/pkg/runtime/malloc.c:60
runtime.mallocgc(0x100004, 0x0, 0x1, 0x1, 0xc093, ...)
runtime.mal+0x40 /Users/rsc/g/go/src/pkg/runtime/malloc.c:289
runtime.mal(0x100004, 0x20bc4)
runtime.new+0x26 /Users/rsc/g/go/src/pkg/runtime/malloc.c:296
runtime.new(0x100004, 0x8fe84000, 0x20bc4)
main.main+0x29 /Users/rsc/x.go:11
main.main()
runtime.mainstart+0xf /Users/rsc/g/go/src/pkg/runtime/386/asm.s:93
runtime.mainstart()
runtime.goexit /Users/rsc/g/go/src/pkg/runtime/proc.c:178
runtime.goexit()
----- goroutine created by -----
_rt0_386+0xbf /Users/rsc/g/go/src/pkg/runtime/386/asm.s:80
R=iant, r
CC=golang-dev
https://golang.org/cl/4444073
In a GOROOT path a backslash is a path separator
not an escape character. For example, `C:\go`.
Fixes gotest error:
version.go:3: unknown escape sequence: g
R=rsc
CC=golang-dev
https://golang.org/cl/4437076
Used to fault trying to access l->list->next
when l->list == nil after MCentral_AllocList.
Now prints
runtime: out of memory: no room in arena for 65536-byte allocation (536870912 in use)
throw: out of memory
followed by stack trace.
Fixes#1650.
R=r, dfc
CC=golang-dev
https://golang.org/cl/4446062
Avoid getting out of synch when a function, such as main.init,
has no associated line number information. Without this the
function before main.init can skip the PC all the way to the
next function, which will cause the next function's line table
to be associated with main.init, and leave subsequent
functions with the wrong line numbers.
R=rsc
CC=golang-dev
https://golang.org/cl/4426055
go/types: update for export data format change
reflect: require package qualifiers to match during interface check
runtime: require package qualifiers to match during interface check
test: fixed bug324, adapt to be silent
Fixes#1550.
Issue 1536 remains open.
R=gri, ken2, r
CC=golang-dev
https://golang.org/cl/4442071
* Reduces malloc counts during gob encoder/decoder test from 6/6 to 3/5.
The current reflect uses Set to mean two subtly different things.
(1) If you have a reflect.Value v, it might just represent
itself (as in v = reflect.NewValue(42)), in which case calling
v.Set only changed v, not any other data in the program.
(2) If you have a reflect Value v derived from a pointer
or a slice (as in x := []int{42}; v = reflect.NewValue(x).Index(0)),
v represents the value held there. Changing x[0] affects the
value returned by v.Int(), and calling v.Set affects x[0].
This was not really by design; it just happened that way.
The motivation for the new reflect implementation was
to remove mallocs. The use case (1) has an implicit malloc
inside it. If you can do:
v := reflect.NewValue(0)
v.Set(42)
i := v.Int() // i = 42
then that implies that v is referring to some underlying
chunk of memory in order to remember the 42; that is,
NewValue must have allocated some memory.
Almost all the time you are using reflect the goal is to
inspect or to change other data, not to manipulate data
stored solely inside a reflect.Value.
This CL removes use case (1), so that an assignable
reflect.Value must always refer to some other piece of data
in the program. Put another way, removing this case would
make
v := reflect.NewValue(0)
v.Set(42)
as illegal as
0 = 42.
It would also make this illegal:
x := 0
v := reflect.NewValue(x)
v.Set(42)
for the same reason. (Note that right now, v.Set(42) "succeeds"
but does not change the value of x.)
If you really wanted to make v refer to x, you'd start with &x
and dereference it:
x := 0
v := reflect.NewValue(&x).Elem() // v = *&x
v.Set(42)
It's pretty rare, except in tests, to want to use NewValue and then
call Set to change the Value itself instead of some other piece of
data in the program. I haven't seen it happen once yet while
making the tree build with this change.
For the same reasons, reflect.Zero (formerly reflect.MakeZero)
would also return an unassignable, unaddressable value.
This invalidates the (awkward) idiom:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.PointTo(v)
which, when the API changed, turned into:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.Set(v.Addr())
In both, it is far from clear what the code is trying to do. Now that
it is possible, this CL adds reflect.New(Type) Value that does the
obvious thing (same as Go's new), so this code would be replaced by:
pv := ... some Ptr Value we have ...
pv.Set(reflect.New(pv.Type().Elem()))
The changes just described can be confusing to think about,
but I believe it is because the old API was confusing - it was
conflating two different kinds of Values - and that the new API
by itself is pretty simple: you can only Set (or call Addr on)
a Value if it actually addresses some real piece of data; that is,
only if it is the result of dereferencing a Ptr or indexing a Slice.
If you really want the old behavior, you'd get it by translating:
v := reflect.NewValue(x)
into
v := reflect.New(reflect.Typeof(x)).Elem()
v.Set(reflect.NewValue(x))
Gofix will not be able to help with this, because whether
and how to change the code depends on whether the original
code meant use (1) or use (2), so the developer has to read
and think about the code.
You can see the effect on packages in the tree in
https://golang.org/cl/4423043/.
R=r
CC=golang-dev
https://golang.org/cl/4435042
. Missing declaration of runtime.brk_();
. Argument v in runtime.SysReserve() is not used;
(I'd prefer a Plan 9-type solution...)
R=golang-dev, r, r2
CC=golang-dev
https://golang.org/cl/4368076
The list elements are already being allocated out of a
single memory buffer. We can drop the Link* pointer
following and the memory it requires, replacing it with
index operations.
The change also keeps a channel from containing a pointer
back into its own allocation block, which would create a
cycle. Blocks involved in cycles are not guaranteed to be
finalized properly, and channels depend on finalizers to
free OS-level locks on some systems. The self-reference
was keeping channels from being garbage collected.
runtime-gdb.py will need to be updated in order to dump
the content of buffered channels with the new data structure.
Fixes#1676.
R=ken2, r
CC=golang-dev
https://golang.org/cl/4411045
in gdb, 'info goroutines' and 'goroutine <n> <cmd> were crashing
because the 'g' and 'm' structures had changed a bit.
R=rsc
CC=golang-dev
https://golang.org/cl/4289077
On darwin amd64 it was impossible to create more that ~132 threads. While
investigating I noticed that go consumes almost 1TB of virtual memory per
OS thread and the reason for such a small limit of OS thread was because
process was running out of virtual memory. While looking at bsdthread_create
I noticed that on amd64 it wasn't using PTHREAD_START_CUSTOM.
If you look at http://fxr.watson.org/fxr/source/bsd/kern/pthread_synch.c?v=xnu-1228
you will see that in that case darwin will use stack pointer as stack size,
allocating huge amounts of memory for stack. This change fixes the issue
and allows for creation of up to 2560 OS threads (which appears to be some
Mac OS X limit) with relatively small virtual memory consumption.
R=rsc
CC=golang-dev
https://golang.org/cl/4289075
Fixes#1641.
Actually it side steps the real issue, which is that the
setitimer(2) implementation on OS X is not useful for
profiling of multi-threaded programs. I filed the below
using the Apple Bug Reporter.
/*
Filed as Apple Bug Report #9177434.
This program creates a new pthread that loops, wasting cpu time.
In the main pthread, it sleeps on a condition that will never come true.
Before doing so it sets up an interval timer using ITIMER_PROF.
The handler prints a message saying which thread it is running on.
POSIX does not specify which thread should receive the signal, but
in order to be useful in a user-mode self-profiler like pprof or gprof
http://code.google.com/p/google-perftoolshttp://www.delorie.com/gnu/docs/binutils/gprof_25.html
it is important that the thread that receives the signal is the one
whose execution caused the timer to expire.
Linux and FreeBSD handle this by sending the signal to the process's
queue but delivering it to the current thread if possible:
http://lxr.linux.no/linux+v2.6.38/kernel/signal.c#L802
807 /*
808 * Now find a thread we can wake up to take the signal off the queue.
809 *
810 * If the main thread wants the signal, it gets first crack.
811 * Probably the least surprising to the average bear.
812 * /
http://fxr.watson.org/fxr/source/kern/kern_sig.c?v=FREEBSD8;im=bigexcerpts#L1907
1914 /*
1915 * Check if current thread can handle the signal without
1916 * switching context to another thread.
1917 * /
On those operating systems, this program prints:
$ ./a.out
signal on cpu-chewing looper thread
signal on cpu-chewing looper thread
signal on cpu-chewing looper thread
signal on cpu-chewing looper thread
signal on cpu-chewing looper thread
signal on cpu-chewing looper thread
signal on cpu-chewing looper thread
signal on cpu-chewing looper thread
signal on cpu-chewing looper thread
signal on cpu-chewing looper thread
$
The OS X kernel does not have any such preference. Its get_signalthread
does not prefer current_thread(), in contrast to the other two systems,
so the signal gets delivered to the first thread in the list that is able to
handle it, which ends up being the main thread in this experiment.
http://fxr.watson.org/fxr/source/bsd/kern/kern_sig.c?v=xnu-1456.1.26;im=excerpts#L1666
$ ./a.out
signal on sleeping main thread
signal on sleeping main thread
signal on sleeping main thread
signal on sleeping main thread
signal on sleeping main thread
signal on sleeping main thread
signal on sleeping main thread
signal on sleeping main thread
signal on sleeping main thread
signal on sleeping main thread
$
The fix is to make get_signalthread use the same heuristic as
Linux and FreeBSD, namely to use current_thread() if possible
before scanning the process thread list.
*/
#include <sys/time.h>
#include <sys/signal.h>
#include <pthread.h>
#include <unistd.h>
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
static void handler(int);
static void* looper(void*);
static pthread_t pmain, ploop;
int
main(void)
{
struct itimerval it;
struct sigaction sa;
pthread_cond_t cond;
pthread_mutex_t mu;
memset(&sa, 0, sizeof sa);
sa.sa_handler = handler;
sa.sa_flags = SA_RESTART;
memset(&sa.sa_mask, 0xff, sizeof sa.sa_mask);
sigaction(SIGPROF, &sa, 0);
pmain = pthread_self();
pthread_create(&ploop, 0, looper, 0);
memset(&it, 0, sizeof it);
it.it_interval.tv_usec = 10000;
it.it_value = it.it_interval;
setitimer(ITIMER_PROF, &it, 0);
pthread_mutex_init(&mu, 0);
pthread_mutex_lock(&mu);
pthread_cond_init(&cond, 0);
for(;;)
pthread_cond_wait(&cond, &mu);
return 0;
}
static void
handler(int sig)
{
static int nsig;
pthread_t p;
p = pthread_self();
if(p == pmain)
printf("signal on sleeping main thread\n");
else if(p == ploop)
printf("signal on cpu-chewing looper thread\n");
else
printf("signal on %p\n", (void*)p);
if(++nsig >= 10)
exit(0);
}
static void*
looper(void *v)
{
for(;;);
}
R=r
CC=golang-dev
https://golang.org/cl/4273113
Also fix comment.
The only caller of chanrecv initializes the value to false, so
this patch makes no difference at present. But it seems like
the right thing to do.
R=rsc
CC=golang-dev
https://golang.org/cl/4312053
* Change use of m->g0 stack (aka scheduler stack).
* Provide runtime.mcall(f) to invoke f() on m->g0 stack.
* Replace scheduler loop entry with runtime.mcall(schedule).
Runtime.mcall eliminates the need for fake scheduler states that
exist just to run a bit of code on the m->g0 stack
(Grecovery, Gstackalloc).
The elimination of the scheduler as a loop that stops and
starts using gosave and gogo fixes a bad interaction with the
way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled)
C functions on that stack, and then when calling back into Go,
it sets m->g0->sched.sp below the added call frames, so that
other uses of m->g0's stack will not interfere with those frames.
Unfortunately, gogo (longjmp) back to the scheduler loop at
this point would end up running scheduler with the lower
sp, which no longer points at a valid stack frame for
a call to scheduler. If scheduler then wrote any function call
arguments or local variables to where it expected the stack
frame to be, it would overwrite other data on the stack.
I realized this possibility while debugging a problem with
calling complex Go code in a Go -> C -> Go cgo callback.
This wasn't the bug I was looking for, it turns out, but I believe
it is a real bug nonetheless. Switching to runtime.mcall, which
only adds new frames to the stack and never jumps into
functions running in existing ones, fixes this bug.
* Move cgo-related code out of proc.c into cgocall.c.
* Add very large comment describing cgo call sequences.
* Simpilify, regularize cgo function implementations and names.
* Add test suite as misc/cgo/test.
Now the Go -> C path calls cgocall, which calls asmcgocall,
and the C -> Go path calls cgocallback, which calls cgocallbackg.
The shuffling, which affects mainly the callback case, moves
most of the callback implementation to cgocallback running
on the m->curg stack (not the m->g0 scheduler stack) and
only while accounted for with $GOMAXPROCS (between calls
to exitsyscall and entersyscall).
The previous callback code did not block in startcgocallback's
approximation to exitsyscall, so if, say, the garbage collector
were running, it would still barge in and start doing things
like call malloc. Similarly endcgocallback's approximation of
entersyscall did not call matchmg to kick off new OS threads
when necessary, which caused the bug in issue 1560.
Fixes#1560.
R=iant
CC=golang-dev
https://golang.org/cl/4253054
This functionality might be used in environments
where programs are limited to a single thread,
to simulate a select-driven network server. It is
not exposed via the standard runtime API.
R=r, r2
CC=golang-dev
https://golang.org/cl/4254041
Using the kernel-supplied compare-and-swap code
on linux/arm means that runtime doesn't have to care
whether this is GOARM=5 or GOARM=6 anymore.
Fixes#1494.
R=r, r2
CC=golang-dev
https://golang.org/cl/4245043
The pointer will eventually let us find *T given T.
This CL just makes room for it, always storing a zero.
R=r, r2
CC=golang-dev
https://golang.org/cl/4221046
In CL 4188061 I changed malg to allocate the requested
number of bytes n, not n+StackGuard, so that the
allocations would use rounder numbers.
The allocation of the signal stack asks for 32k and
then used g->stackguard as the base, but g->stackguard
is StackGuard bytes above the base. Previously, asking
for 32k meant getting 32k+StackGuard bytes, so using
g->stackguard as the base was safe. Now, the actual base
must be computed, so that the signal handler does not
run StackGuard bytes past the top of the stack.
Was causing flakiness mainly in programs that use the
network, because they sometimes write to closed network
connections, causing SIGPIPEs. Was also causing problems
in the doc/progs test.
Also fix Makefile so that changes to stack.h trigger rebuild.
R=bradfitzgo, r, r2
CC=golang-dev
https://golang.org/cl/4230044
Avoids deadlocks like the one below, in which a stack split happened
in order to call lock(&stacks), but then the stack unsplit cannot run
because stacks is now locked.
The only code calling stackalloc that wasn't on a scheduler
stack already was malg, which creates a new goroutine.
runtime.futex+0x23 /home/rsc/g/go/src/pkg/runtime/linux/amd64/sys.s:139
runtime.futex()
futexsleep+0x50 /home/rsc/g/go/src/pkg/runtime/linux/thread.c:51
futexsleep(0x5b0188, 0x300000003, 0x100020000, 0x4159e2)
futexlock+0x85 /home/rsc/g/go/src/pkg/runtime/linux/thread.c:119
futexlock(0x5b0188, 0x5b0188)
runtime.lock+0x56 /home/rsc/g/go/src/pkg/runtime/linux/thread.c:158
runtime.lock(0x5b0188, 0x7f0d27b4a000)
runtime.stackfree+0x4d /home/rsc/g/go/src/pkg/runtime/malloc.goc:336
runtime.stackfree(0x7f0d27b4a000, 0x1000, 0x8, 0x7fff37e1e218)
runtime.oldstack+0xa6 /home/rsc/g/go/src/pkg/runtime/proc.c:705
runtime.oldstack()
runtime.lessstack+0x22 /home/rsc/g/go/src/pkg/runtime/amd64/asm.s:224
runtime.lessstack()
----- lessstack called from goroutine 2 -----
runtime.lock+0x56 /home/rsc/g/go/src/pkg/runtime/linux/thread.c:158
runtime.lock(0x5b0188, 0x40a5e2)
runtime.stackalloc+0x55 /home/rsc/g/go/src/pkg/runtime/malloc.c:316
runtime.stackalloc(0x1000, 0x4055b0)
runtime.malg+0x3d /home/rsc/g/go/src/pkg/runtime/proc.c:803
runtime.malg(0x1000, 0x40add9)
runtime.newproc1+0x12b /home/rsc/g/go/src/pkg/runtime/proc.c:854
runtime.newproc1(0xf840027440, 0x7f0d27b49230, 0x0, 0x49f238, 0x40, ...)
runtime.newproc+0x2f /home/rsc/g/go/src/pkg/runtime/proc.c:831
runtime.newproc(0x0, 0xf840027440, 0xf800000010, 0x44b059)
...
R=r, r2
CC=golang-dev
https://golang.org/cl/4216045
A terminal panic (one that prints a stack trace and exits)
has been calling runtime.breakpoint before calling exit,
so that if running under a debugger, the debugger can
take control. When not running under a debugger, though,
this causes an additional SIGTRAP on Unix and pop-up
dialogs on Windows.
Support for debugging Go programs has gotten good
enough that we can rely on the debugger to set its own
breakpoint on runtime.exit if it wants to look around.
R=r, r2
CC=golang-dev
https://golang.org/cl/4222043
The existing code assumed that signals only arrived
while executing on the goroutine stack (g == m->curg),
not while executing on the scheduler stack (g == m->g0).
Most of the signal handling trampolines correctly saved
and restored g already, but the sighandler C code did not
have access to it.
Some rewriting of assembly to make the various
implementations as similar as possible.
Will need to change Windows too but I don't
understand how sigtramp gets called there.
R=r
CC=golang-dev
https://golang.org/cl/4203042
With this change, a panic trace due to a signal arriving while
running on the scheduler stack during a lessstack
(a stack unsplit) will trace through the lessstack to show
the state of the goroutine that was unsplitting its stack.
R=r
CC=golang-dev
https://golang.org/cl/4206042
Fix problems found.
On amd64, various library routines had bigger
stack frames than expected, because large function
calls had been added.
runtime.assertI2T: nosplit stack overflow
120 assumed on entry to runtime.assertI2T
8 after runtime.assertI2T uses 112
0 on entry to runtime.newTypeAssertionError
-8 on entry to runtime.morestack01
runtime.assertE2E: nosplit stack overflow
120 assumed on entry to runtime.assertE2E
16 after runtime.assertE2E uses 104
8 on entry to runtime.panic
0 on entry to runtime.morestack16
-8 after runtime.morestack16 uses 8
runtime.assertE2T: nosplit stack overflow
120 assumed on entry to runtime.assertE2T
16 after runtime.assertE2T uses 104
8 on entry to runtime.panic
0 on entry to runtime.morestack16
-8 after runtime.morestack16 uses 8
runtime.newselect: nosplit stack overflow
120 assumed on entry to runtime.newselect
56 after runtime.newselect uses 64
48 on entry to runtime.printf
8 after runtime.printf uses 40
0 on entry to vprintf
-8 on entry to runtime.morestack16
runtime.selectdefault: nosplit stack overflow
120 assumed on entry to runtime.selectdefault
56 after runtime.selectdefault uses 64
48 on entry to runtime.printf
8 after runtime.printf uses 40
0 on entry to vprintf
-8 on entry to runtime.morestack16
runtime.selectgo: nosplit stack overflow
120 assumed on entry to runtime.selectgo
0 after runtime.selectgo uses 120
-8 on entry to runtime.gosched
On arm, 5c was tagging functions NOSPLIT that should
not have been, like the recursive function printpanics:
printpanics: nosplit stack overflow
124 assumed on entry to printpanics
112 after printpanics uses 12
108 on entry to printpanics
96 after printpanics uses 12
92 on entry to printpanics
80 after printpanics uses 12
76 on entry to printpanics
64 after printpanics uses 12
60 on entry to printpanics
48 after printpanics uses 12
44 on entry to printpanics
32 after printpanics uses 12
28 on entry to printpanics
16 after printpanics uses 12
12 on entry to printpanics
0 after printpanics uses 12
-4 on entry to printpanics
R=r, r2
CC=golang-dev
https://golang.org/cl/4188061
BSD and Darwin require an extra page between
end and the first mapping, and Windows has various
memory in the way too.
Fixes#1464.
R=r, r2
CC=golang-dev
https://golang.org/cl/4167041
GC is still single-threaded.
Multiple threads will happen in another CL.
Garbage collection pauses are typically
about half as long as they were before this CL.
R=brainman, iant, r
CC=golang-dev
https://golang.org/cl/3975046
Follow morestack, so that crashes during a stack split
give complete traces. Also mark stack segment boundaries
as an aid to debugging.
Correct various line number bugs with yet another attempt
at interpreting the pc/ln table. This one has a chance at
being correct, because I based it on reading src/cmd/ld/lib.c
instead of on reading the documentation.
Fixes#1138.
Fixes#1430.
Fixes#1461.
throw: runtime: split stack overflow
runtime.throw+0x3e /home/rsc/g/go2/src/pkg/runtime/runtime.c:78
runtime.throw(0x81880af, 0xf75c8b18)
runtime.newstack+0xad /home/rsc/g/go2/src/pkg/runtime/proc.c:728
runtime.newstack()
runtime.morestack+0x4f /home/rsc/g/go2/src/pkg/runtime/386/asm.s:184
runtime.morestack()
----- morestack called from stack: -----
runtime.new+0x1a /home/rsc/g/go2/src/pkg/runtime/malloc.c:288
runtime.new(0x1, 0x0, 0x0)
gongo.makeBoard+0x33 /tmp/Gongo/gongo_robot_test.go:344
gongo.makeBoard(0x809d238, 0x1, 0xf76092c8, 0x1)
----- stack segment boundary -----
gongo.checkEasyScore+0xcc /tmp/Gongo/gongo_robot_test.go:287
gongo.checkEasyScore(0xf764b710, 0x0, 0x809d238, 0x1)
gongo.TestEasyScore+0x8c /tmp/Gongo/gongo_robot_test.go:255
gongo.TestEasyScore(0xf764b710, 0x818a990)
testing.tRunner+0x2f /home/rsc/g/go2/src/pkg/testing/testing.go:132
testing.tRunner(0xf764b710, 0xf763b5dc, 0x0)
runtime.goexit /home/rsc/g/go2/src/pkg/runtime/proc.c:149
runtime.goexit()
R=ken2, r
CC=golang-dev
https://golang.org/cl/4000053
If the same directory was used for multiple builds,
it was possible for a stale version.go to contain the
wrong definitions for $GOOS and $GOARCH, because
they can change even if the hg version does not.
Split into multiple files to fix.
R=r, r2
CC=golang-dev
https://golang.org/cl/4124050
Shame on me: I fixed the same bug in 6l in 8691fcc6a66e
(https://golang.org/cl/2609041) and neglected
to look at 5l and 8l to see if they were affected.
On the positive side, the check I added in that CL is the
one that detected this bug.
Fixes#1457.
R=ken2
CC=golang-dev
https://golang.org/cl/3981052
The callback mechanism has been made more flexible.
Eliminated one round of argument copying in Syscall.
Faster Get/SetLastError implemented.
Added gettime for gc perf profiling.
R=rsc, brainman, mattn, rog
CC=golang-dev
https://golang.org/cl/4058046
The sanity checking in pass 2 is wrong
when a select is offering to communicate in
either direction on a channel and neither case
is immediately ready.
R=ken2
CC=golang-dev
https://golang.org/cl/3991047
The old heap maps used a multilevel table, but that
was overkill: there are only 1M entries on a 32-bit
machine and we can arrange to use a dense address
range on a 64-bit machine.
The heap map is in bss. The assumption is that if
we don't touch the pages they won't be mapped in.
Also moved some duplicated memory allocation
code out of the OS-specific files.
R=r
CC=golang-dev
https://golang.org/cl/4118042
It is unmaintained and untested, and I think it's broken too.
It was a toy to show that Go can run on real hardware,
and it served its purpose.
The source code will of course remain in the repository
history, so it could be brought back if needed later.
R=r, r2, uriel
CC=golang-dev
https://golang.org/cl/3996047
Close of closed channel panics.
Receive from closed channel never panics,
even if done repeatedly.
Fixes#1349.
Fixes#1419.
R=gri, iant, ken2, r, gri1, r2, iant2, rog, albert.strasheim, niemeyer, ejsherry
CC=golang-dev
https://golang.org/cl/3989042
The functionality we want (shared ppid) is implied
by CLONE_THREAD already, and CLONE_PARENT
causes problems if the Go program is pid 1 (init).
See issue 1406 for more details.
Fixes#1406.
R=adg, iant
CC=golang-dev
https://golang.org/cl/3971044
The o+i*p approach to visiting select cases in random
order stops being fair when there is some case that
is never ready. If that happens, then the case that follows
it in the order gets more chances than the others.
In general the only way to ensure fairness is to make
all permutations equally likely. I've done that by computing
one explicitly.
Makes the permutations correct for n >= 4 where
previously they were broken. For n > 12, there's not
enough randomness to do a perfect job but this should
still be much better than before.
Fixes#1425.
R=r, ken2, ejsherry
CC=golang-dev
https://golang.org/cl/4037043
The fault was lucky: when it wasn't faulting it was silently
copying a word from some other block and later putting
that same word back. If some other goroutine had changed
that word of memory in the interim, too bad.
The ARM code was inconsistent about whether the
"argument frame" included the saved LR. Including it made
some things more regular but mostly just caused confusion
in the places where the regularity broke. Now the rule
reflects reality: argp is always a pointer to arguments,
never a saved link register.
Renamed struct fields to make meaning clearer.
Running ARM in QEMU, package time's gotest:
* before: 27/58 failed
* after: 0/50
R=r, r2
CC=golang-dev
https://golang.org/cl/3993041
In this specific package crosscall2 is already defined in a .S
file anyhow. This avoids a warning about mismatched
alignment.
R=rsc
CC=golang-dev
https://golang.org/cl/4000043
I missed that environment is used during runtime setup,
well before go init() functions run. Implemented os-dependent
runtime.goenvs functions to allow for different unix, plan9 and
windows versions of environment discovery.
R=rsc, paulzhol
CC=golang-dev
https://golang.org/cl/3787046
If we don't do this, then when C code calls back to Go code
which panics, we lose space on the scheduler stack. If that
happens a lot, eventually there is no space left on the
scheduler stack.
R=rsc
CC=golang-dev
https://golang.org/cl/3898042
A cursory reading of the cgo code suggests this
should be necessary, though I don't have access
to a FreeBSD machine for testing.
R=rsc, adg
CC=golang-dev
https://golang.org/cl/3746047
#pragma dynexport is no longer needed for
this use of cgo, since the gcc and gc code are
now linked together into the same binary.
It may still be necessary later.
On the Mac, you cannot use the GOT to resolve
symbols that exist in the current binary, so 6l and 8l
translate the GOT-loading mov instructions into lea
instructions.
On ELF systems, we could use the GOT for those
symbols, but for consistency 6l and 8l apply the
same translation.
The translation is sketchy in the extreme
(depending on the relocation being in a mov
instruction) but it verifies that the instruction
is a mov before rewriting it to lea.
Also makes typedefs global across files.
Fixes#1335.
Fixes#1345.
R=iant, r
CC=golang-dev
https://golang.org/cl/3650042
The recent linker changes broke NaCl support
a month ago, and there are no known users of it.
The NaCl code can always be recovered from the
repository history.
R=adg, r
CC=golang-dev
https://golang.org/cl/3671042
Too many programs complain that we even try.
This was a bit of security paranoia and not worth
the bother.
Fixes#1340.
R=r, r2
CC=golang-dev
https://golang.org/cl/3579042
cc: same
runtime: test cc alignment (required moving #define of offsetof to runtime.h)
fix bug260
Fixes#482.
Fixes#609.
R=ken2, r
CC=golang-dev
https://golang.org/cl/3563042
Formerly known as libcgo.
Almost no code here is changing; the diffs
are shown relative to the originals in libcgo.
R=r
CC=golang-dev
https://golang.org/cl/3420043
8l was broken by commit 7ac0d2eed9, it caused .data to be page aligned in the file - which is not how Plan 9 expects things to be.
Also .rodata was layed out in a similar fashion.
Not sure when signame was introduced, but added a stub.
Removed the symo assignment in asm.c as it is not currently used.
Fix runtime breakage after commit 629c065d36 which prefixes all external symbols with runtime·.
R=rsc
CC=golang-dev
https://golang.org/cl/2674041
Prefix all external symbols in runtime by runtime·,
to avoid conflicts with possible symbols of the same
name in linked-in C libraries. The obvious conflicts
are printf, malloc, and free, but hide everything to
avoid future pain.
The symbols left alone are:
** known to cgo **
_cgo_free
_cgo_malloc
libcgo_thread_start
initcgo
ncgocall
** known to linker **
_rt0_$GOARCH
_rt0_$GOARCH_$GOOS
text
etext
data
end
pclntab
epclntab
symtab
esymtab
** known to C compiler **
_divv
_modv
_div64by32
etc (arch specific)
Tested on darwin/386, darwin/amd64, linux/386, linux/amd64.
Built (but not tested) for freebsd/386, freebsd/amd64, linux/arm, windows/386.
R=r, PeterGo
CC=golang-dev
https://golang.org/cl/2899041
Adds softfloat64 to generic runtime
(will be discarded by linker when unused)
and adds test for it. I used the test to check
the software code against amd64 hardware
and then check the software code against
the arm and its simulation of hardware.
The latter should have been a no-op (testing
against itself) but turned up a bug in 5c causing
the vlrt.c routines to miscompile.
These changes make the cmath, math,
and strconv tests pass without any special
accommodations for arm.
R=ken2
CC=golang-dev
https://golang.org/cl/2713042
Just enough to make mov instructions work,
which in turn is enough to make strconv work
when it avoids any floating point calculations.
That makes a bunch of other packages pass
their tests.
Should suffice until hardware floating point
is available.
Enable package tests that now pass
(some due to earlier fixes).
Looks like there is a new integer math bug
exposed in the fmt and json tests.
R=ken2
CC=golang-dev
https://golang.org/cl/2638041
The frame that gets allocated is for both
the args and the autos. If together they
exceed the default frame size, we need to
tell morestack about both so that it allocates
a large enough frame.
Sanity check stack pointer in morestack
to catch similar bugs.
R=ken2
CC=golang-dev
https://golang.org/cl/2609041
That is, move the pc/ln table and the symbol table
into the read-only data segment. This eliminates
the need for a special load command to map the
symbol table into memory, which makes the
information available on systems that couldn't handle
the magic load to 0x99000000, like NaCl and ARM QEMU
and Linux without config_highmem=y. It also
eliminates an #ifdef and some clumsy code to
find the symbol table on Windows.
The bad news is that the binary appears to be bigger
than it used to be. This is not actually the case, though:
the same amount of data is being mapped into memory
as before, and the tables are still read-only, so they're
still shared across multiple instances of the binary as
they were before. The difference is just that the tables
aren't squirreled away in some section that "size" doesn't
know to look at.
This is a checkpoint.
It probably breaks Windows and breaks NaCl more
than it used to be broken, but those will be fixed.
The logic involving -s needs to be revisited too.
Fixes#871.
R=ken2
CC=golang-dev
https://golang.org/cl/2587041
No multiple processes/locks, managed to compile
and run a hello.go (with print not fmt). Also test/sieve.go
seems to run until 439 and stops with a
'throw: all goroutines are asleep - deadlock!'
- just like runtime/tiny.
based on Russ's suggestions at:
http://groups.google.com/group/comp.os.plan9/browse_thread/thread/cfda8b82535d2d68/243777a597ec1612
Build instructions:
cd src/pkg/runtime
make clean && GOOS=plan9 make install
this will build and install the runtime.
When linking with 8l, you should pass -s to suppress symbol
generation in the a.out, otherwise the generated executable will not run.
This is runtime only, the porting of the toolchain has already
been done: http://code.google.com/p/go-plan9/source/browse
in the plan9-quanstro branch.
R=rsc
CC=golang-dev
https://golang.org/cl/2273041
Because the SB is only good for 8k and Go programs
tend to have much more data than that, SB doesn't
save very much. A fmt.Printf-based hello world program
has 360 kB text segment. Removing SB makes the text
500 bytes (0.14%) longer.
R=ken2, r2, ken3
CC=golang-dev
https://golang.org/cl/2487042
On systems where the mmap succeeds
(e.g., sysctl -w vm.mmap_min_addr=0)
it changes the signal code delivered for a
nil fault from ``page not mapped'' to
``invalid permissions for page.''
TBR=r
CC=golang-dev
https://golang.org/cl/2294041
* Add documentation about array arguments. Fixes issue 1125.
* Do not interpret x, y := z, w as special errno form. Fixes issue 952.
* Fix nested Go calls (brainman). Fixes issue 907.
R=r
CC=golang-dev
https://golang.org/cl/2214044
g is not in r15 anymore.
now it's in a per-thread memory segment,
which is valid even inside a signal handler,
so we can just refer to g directly.
Fixes#1082.
R=r
CC=golang-dev
https://golang.org/cl/2149045
Need to correct for deferproc's extra 2 words of stack or in some
cases (such as memory profiling) traceback can cause a crash.
Also bulletproof the closure test.
R=rsc
CC=golang-dev
https://golang.org/cl/2138047
Old code was using recursion to traverse object graph.
New code uses an explicit stack, cutting the per-pointer
footprint to two words during the recursion and avoiding
the standard allocator and stack splitting code.
in test/garbage:
Reduces parser runtime by 2-3%
Reduces Peano runtime by 40%
Increases tree runtime by 4-5%
R=r
CC=golang-dev
https://golang.org/cl/2150042
Based on the observation that a great number of the types that
are copied or compared in interfaces, maps, and channels are
word-sized, this uses specialized copy and equality functions
for them that use a word instead of 4 or 8 bytes. Seems to yield
0-6% improvements in performance in the benchmarks I've run.
For example, with the regexp benchmarks:
Before:
regexp.BenchmarkLiteral 500000 3.26 µs/op
regexp.BenchmarkNotLiteral 100000 13.67 µs/op
regexp.BenchmarkMatchClass 100000 18.72 µs/op
regexp.BenchmarkMatchClass_InRange 100000 20.04 µs/op
regexp.BenchmarkReplaceAll 100000 27.85 µs/op
After:
regexp.BenchmarkLiteral 500000 3.11 µs/op
regexp.BenchmarkNotLiteral 200000 13.29 µs/op
regexp.BenchmarkMatchClass 100000 17.65 µs/op
regexp.BenchmarkMatchClass_InRange 100000 18.49 µs/op
regexp.BenchmarkReplaceAll 100000 26.34 µs/op
R=rsc
CC=golang-dev
https://golang.org/cl/1967047
The runtime only passes 32 bits of file offset,
but the kernel wants 64 bits, so have to add
zeros explicitly in a copy of the arguments.
R=adg, Martin Neubauer
CC=golang-dev
https://golang.org/cl/1933044
Function to create a GoString with a known length so it can contain NUL
bytes anywhere in the string. Some C libraries have strings like this.
R=rsc
CC=golang-dev
https://golang.org/cl/2007042
Returns R14 and R15 to the available register pool.
Plays more nicely with ELF ABI C code.
In particular, our signal handlers will no longer crash
when a signal arrives during execution of a cgo C call.
Fixes#720.
R=ken2, r
CC=golang-dev
https://golang.org/cl/1847051
but with less precision than hardware counterparts.
fixed a number of tests to output BUG when they failed.
changed the runner to distinghuish between output
and output containing ^BUG
R=rsc
CC=dho, golang-dev
https://golang.org/cl/1778041
updated thread.c to provide destroylock, which seems to be
required to link.
updated README with different virtualization programs.
R=golang-dev, adg
CC=golang-dev
https://golang.org/cl/1746047
Cannot assume that g == m->curg at time of signal.
Must save actual g and restore.
Fixes flaky crashes with messages like
throw: malloc mlookup
throw: malloc/free - deadlock
throw: unwindstack on self
throw: free mlookup
(and probably others) when running cgo.
R=iant
CC=golang-dev
https://golang.org/cl/1648043
* Code for assignment, conversions now mirrors spec.
* Changed some snprint -> smprint.
* Renamed runtime functions to separate
interface conversions from type assertions:
convT2I, assertI2T, etc.
* Correct checking of \U sequences.
Fixes#840.
Fixes#830.
Fixes#778.
R=ken2
CC=golang-dev
https://golang.org/cl/1303042
This permits cgo callbacks to work when run in init code.
Otherwise cgocallback switches to the wrong stack address.
R=rsc
CC=golang-dev
https://golang.org/cl/1123043
Import _mulv from Inferno again, change R9 to R2.
Not sure what the other differences were for, but
they weren't working.
TBR=kaib
CC=golang-dev
https://golang.org/cl/1079041
When trying to regenerate src/pkg/runtime/darwin/386/defs.h
on a 64 bit capable Snow Leopard (OS X 10.6.3) system I
needed to add -f -m32 to godefs, as this OS and hardware
combination defaults to 64 bit compilation.
For safety, make the same change to the 32 bit FreeBSD
instructions in .../freebsd/defs.c. (Tested OK and no
problems introduced.)
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/1052042
This is required to make cgo export work on Darwin. Note that
this corrects the stack alignment when calling initcgo to that
required by gcc on amd64.
R=rsc
CC=golang-dev
https://golang.org/cl/907041
These functions are used to call from a C function back to a
Go function. This only includes 386 support.
R=rsc
CC=golang-dev
https://golang.org/cl/834045
tested on linux/amd64, linux/386, linux/arm, darwin/amd64, darwin/386.
freebsd untested; will finish in a separate CL.
for now all the panics are errorStrings.
richer structures can be added as necessary
once the mechanism is shaked out.
R=r
CC=golang-dev
https://golang.org/cl/906041
Could not take a signal on threads other than the main thread.
If you look at the spinning binary with dtrace, you can see a
fault happening over and over:
$ dtrace -n '
fbt::user_trap:entry /execname=="boot32" && self->count < 10/
{
self->count++;
printf("%s %x %x %x %x", probefunc, arg1, arg2, arg3, arg4);
stack();
tracemem(arg4, 256);
}'
dtrace: description 'fbt::user_trap:entry ' matched 1 probe
CPU ID FUNCTION:NAME
1 17015 user_trap:entry user_trap 0 10 79af0a0 79af0a0
mach_kernel`lo_alltraps+0x12a
0 1 2 3 4 5 6 7 8 9 a b c d e f 0123456789abcdef
0: 0e 00 00 00 37 00 00 00 00 00 00 00 1f 00 00 00 ....7...........
10: 1f 00 00 00 a8 33 00 00 00 00 00 01 00 00 00 00 .....3..........
20: 98 ba dc fe 07 09 00 00 00 00 00 00 98 ba dc fe ................
30: 06 00 00 00 0d 00 00 00 34 00 00 00 9e 1c 00 00 ........4.......
40: 17 00 00 00 00 02 00 00 ac 30 00 00 1f 00 00 00 .........0......
50: 00 00 00 00 00 00 00 00 0d 00 00 00 e0 e6 29 00 ..............).
60: 34 00 00 00 00 00 00 00 9e 1c 00 00 00 00 00 00 4...............
70: 17 00 00 00 00 00 00 00 00 02 00 00 00 00 00 00 ................
80: ac 30 00 00 00 00 00 00 1f 00 00 00 00 00 00 00 .0..............
90: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
a0: 48 00 00 00 10 00 00 00 85 00 00 00 a0 f2 29 00 H.............).
b0: 69 01 00 02 00 00 00 00 e6 93 04 82 ff 7f 00 00 i...............
c0: 2f 00 00 00 00 00 00 00 06 02 00 00 00 00 00 00 /...............
d0: 78 ee 42 01 01 00 00 00 1f 00 00 00 00 00 00 00 x.B.............
e0: 00 ed 9a 07 00 00 00 00 00 00 00 00 00 00 00 00 ................
f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
...
The memory dump shows a 32-bit exception frame:
x86_saved_state32
gs = 0x37
fs = 0
es = 0x1f
ds = 0x1f
edi = 0x33a8
esi = 0x01000000
ebp = 0
cr2 = 0xfedcba98
ebx = 0x0907
edx = 0
ecx = 0xfedcba98
eax = 0x06
trapno = 0x0d
err = 0x34
eip = 0x1c9e
cs = 0x17
efl = 0x0200
uesp = 0x30ac
ss = 0x1f
The cr2 of 0xfedcba98 is the address that the new thread read
to cause the fault, but note that the trap is now a GP fault with
error code 0x34, meaning it's moved past the cr2 problem and on
to an invaild segment selector. The 0x34 is suspiciously similar
to the 0x37 in gs, and sure enough, OS X forces gs to have
that value in the signal handler, and if your thread hasn't set
up that segment (known as USER_CTHREAD), you'll fault on the IRET
into the signal handler and never be able to handle a signal.
The kernel bug is that it forces segment 0x37 without making sure
it is a valid segment. Leopard also forced 0x37 but had the courtesy
to set it up first.
Since OS X requires us to set up that segment (using the
thread_fast_set_cthread_self system call), we might as well
use it instead of the more complicated i386_set_ldt call to
set up our per-OS thread storage.
Also add some more zeros to bsdthread_register for new arguments
in Snow Leopard (apparently unnecessary, but being careful).
Fixes#510.
R=r
CC=golang-dev
https://golang.org/cl/824046
data just read from the channel.
this will make it easier to
recognize when to garbage
collect and finalize.
R=rsc
CC=golang-dev
https://golang.org/cl/882043
1. Fix bug in GOMAXPROCS when trying to cut number of procs
Race could happen on any system but was
manifesting only on Xen hosted Linux.
2. Fix recover on ARM, where FP != caller SP.
R=r
CC=golang-dev
https://golang.org/cl/880043
log: add test of Caller
New regexp in log test is picky and will require some
maintenance, but it catches off-by-one mistakes too.
Fixes#710.
R=gri
CC=esko.luontola, golang-dev
https://golang.org/cl/887043
* correct symbol table size
* do not reorder functions in output
* traceback
* signal handling
* use same code for go + defer
* handle leaf functions in symbol table
R=kaib, dpx
CC=golang-dev
https://golang.org/cl/884041
* adds pass 3 to dequeue from channels eagerly
various other cleanup/churn:
* use switch on cas->send in each pass to
factor out common code.
* longer goto labels, commented at target
* be more agressive about can't happen:
throw instead of print + cope.
* use "select" instead of "selectgo" in errors
* use printf for debug prints when possible
R=ken2, ken3
CC=golang-dev, r
https://golang.org/cl/875041
main semantic change is to enforce single argument to panic.
runtime: change to 1-argument panic.
use String method on argument if it has one.
R=ken2, r
CC=golang-dev
https://golang.org/cl/812043
- mkall now generates syscall stabs automatically
- we can call unicode versions of winapi now
- introduce GetErrstr function to fetch error text given errno
- general cleanup
R=rsc
CC=golang-dev
https://golang.org/cl/562041
This keeps fragmentation from delaying
garbage collections (and causing more fragmentation).
Cuts fresh godoc (with indexes) from 261M to 166M (120M live).
Cuts toy wc program from 50M to 8M.
Fixes#647.
R=r, cw
CC=golang-dev
https://golang.org/cl/257041
because free needs to mark the block as freed to
coordinate with the garbage collector.
(in C++ free can blindly put the block on the free list,
no questions asked, so the cache saves some work.)
R=iant
CC=golang-dev
https://golang.org/cl/206069
* specialize sweepspan as sweepspan0 and sweepspan1.
* in sweepspan1, inline "free" to avoid expensive mlookup.
R=iant
CC=golang-dev
https://golang.org/cl/206060
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
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
* 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
(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
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
This change removes the necessity to have GOBIN in $PATH,
and also doesn't assume that the build is being run from
$GOROOT/src. This is a minimal set of necessary changes
to get Go to build happily from the FreeBSD ports
collection.
R=rsc
CC=golang-dev
https://golang.org/cl/171044
FreeBSD was passing stk as the new thread's stack base, while
stk is the top of the stack in go. The added check should cause
a trap if this ever comes up in any new ports, or regresses
in current ones.
R=rsc
CC=golang-dev
https://golang.org/cl/167055
On a microbenchmark that ping-pongs on lots of channels, this makes
the multithreaded case about 20% faster and the uniprocessor case
about 1% slower. (Due to cache effects, I expect.)
R=rsc, agl
CC=golang-dev
https://golang.org/cl/166043
the bash scripts and makefiles for building go didn't take into account
the fact $GOROOT / $GOBIN could both be directories containing whitespaces,
and was not possible to build it in such a situation.
this commit adjusts the various makefiles/scripts to make it aware of that
possibility, and now it builds successfully when using a path with whitespaces
as well.
Fixes#115.
R=rsc, dsymonds1
https://golang.org/cl/157067
* add runtime sliceslice1 for x[lo:]
* remove runtime arraytoslice, rewriting &arr into arr[0:len(arr)].
* port cgen_inline into 8g, 5g.
* use native memmove in maps
R=ken2
https://golang.org/cl/157106
In thread.c, we need to cast to whatever the native
size of intptr is on the system, but we only have
uintptr available. They're the same size, but can't
do signed casts without this one :).
R=rsc
CC=golang-dev
https://golang.org/cl/156073
* move memmove to arch-specific subdirectories
* add memmove for arm
* add copyright notices marking them as copied from Inferno
R=ken2
https://golang.org/cl/156061
the signal handling stack is a different size than
the normal stack, so it cannot be allocated using
the backup stack allocator.
Fixes#250.
R=agl1
CC=golang-dev
https://golang.org/cl/157044
cgo/libmach remain unimplemented. However, compilers, runtime,
and packages are 100%. I still need to go through and implement
missing syscalls (at least make sure they're all listed), but
for all shipped functionality, this is done. Ship! ;)
R=rsc, VenkateshSrinivas
https://golang.org/cl/152142
SELinux will cause mmap to fail when we request w+x memory unless the
user has configured their policies. We have a warning in make.bash,
but it's quite likely that the policy will be reset at some point and
then all their binaries start failing.
This patch prints a warning on Linux when mmap fails with EACCES.
R=rsc
CC=golang-dev
https://golang.org/cl/152086
selfree maintains a cache of Select structures for several sizes. In
newselect, we'll use an entry from the cache if one is found. However,
the Scase structures corresponding to a send may have been allocated
for the wrong size. In this case we'll write off the end of the Scase
into random memory and, generally, read some amount of junk in the
receive.
This patch fixes the issue by removing the cache, on the advice of
rsc.
R=rsc
CC=go-dev
http://go/go-review/1016002
when switching architectures.
fix bug twice: make sure clean removes the file,
and generate into architecture-specific file names.
R=r
http://go/go-review/1013018
call. This uses the header files from a 32-bit Ubuntu Hardy
system.
The use of _LOOSE_KERNEL_NAMES seems right. The
-D__ARCH_SI_UID_T works around a bug which appears to be fixed
in later Linux versions.
R=rsc
http://go/go-review/1013015
now that all arguments are passed on the stack.
go/test: passes 89% (310/345)
R=rsc
APPROVED=rsc
DELTA=33 (13 added, 14 deleted, 6 changed)
OCL=36009
CL=36022
because they are in package runtime.
another step to enforcing package boundaries.
R=r
DELTA=732 (114 added, 93 deleted, 525 changed)
OCL=35811
CL=35824
5g. fixes to 64 bit code gen. added (finally) function to do
shifts properly.
go/test: passes 83% (287/342)
R=rsc
APPROVED=rsc
DELTA=156 (50 added, 53 deleted, 53 changed)
OCL=35589
CL=35616
* in 6l, -K already meant check for stack underflow.
add -KK to mean double-check stack overflows
even in nosplit functions.
* comment out print locks; they deadlock too easily
but are still useful to put back for special occasions.
* let runcgo assembly switch to scheduler stack
without involving scheduler directly. because runcgo
gets called from matchmg, it is too hard to keep it
from being called on other stacks.
R=r
DELTA=94 (65 added, 18 deleted, 11 changed)
OCL=35591
CL=35604
better mach binaries.
cgo working on darwin+linux amd64+386.
eliminated context switches - pi is 30x faster.
add libcgo to build.
on snow leopard:
- non-cgo binaries work; all tests pass.
- cgo binaries work on amd64 but not 386.
R=r
DELTA=2031 (1316 added, 626 deleted, 89 changed)
OCL=35264
CL=35304
* change ldt0setup to set GS itself; nacl won't let us do it.
* change breakpoint to INT $3 so 8l can translate to HLT for nacl.
* panic if closure is needed on nacl.
* do not try to access symbol table on nacl.
* mmap in 64kB chunks.
nacl support:
* system calls, threading, locks.
R=r
DELTA=365 (357 added, 5 deleted, 3 changed)
OCL=34880
CL=34906
params. a number of fixes to assembly routines that assumed R0
had the first arg. one stack offset fix, arm pushes the link
register on stack top.
go/test: passes 65% (235/364) tests
R=rsc
APPROVED=rsc
DELTA=20 (11 added, 0 deleted, 9 changed)
OCL=34809
CL=34812
* use //ffi comments in package import data
to generate relocation entries and library loads.
* call initffi in rt0.s if present
R=r
DELTA=117 (91 added, 3 deleted, 23 changed)
OCL=33739
CL=33750
introduced explicit "data" symbol instead of etext
to mark beginning of data, so that using larger
alignment (i.e. 4MB like GNU loader) doesn't
confuse garbage collector.
split dodata into dodata and dobss in preparation
for putting the dynamic data + headers in the data
segment instead of stuffed at the beginning of the binary.
R=r
DELTA=52 (37 added, 3 deleted, 12 changed)
OCL=33610
CL=33618
Since pres != nil was already checked and the routine sets *pres to false
and returns, the subsequent tests are unneeded.
R=rsc
APPROVED=rsc
DELTA=4 (0 added, 4 deleted, 0 changed)
OCL=33439
CL=33441
at PC, to deal with functions that end with a CALL
instruction. Special case known call-by-return functions.
R=rsc
APPROVED=rsc
DELTA=12 (6 added, 6 deleted, 0 changed)
OCL=33255
CL=33279
so that == on func means that the
functions originated in the same
execution of a func literal or definition.
before, there was an inconsistency:
func() {x++} != func() {x++}
but
func() {} == func() {}
this CL makes the second case != too,
just like
make(map[int]int) != make(map[int]int)
R=r
DELTA=202 (71 added, 62 deleted, 69 changed)
OCL=32393
CL=32398
saving of sp was too far away from use in scanstack;
the stack had changed since the sp was saved.
R=r
DELTA=9 (4 added, 2 deleted, 3 changed)
OCL=32232
CL=32237
if there is a goroutine waiting to run
and the init goroutine enters a system call,
entersyscall was trying to kick off a new
scheduler for the other goroutine, causing
a panic (new goroutines can't run until main.main).
R=r
DELTA=32 (32 added, 0 deleted, 0 changed)
OCL=31982
CL=31982
introduce NodeList* type in compiler to replace OLIST.
this clarifies where lists can and cannot occur.
list append and concatenation are now cheap.
the _r rules are gone from yacc.
rev and unrev are gone.
no more lists of lists.
the representation of assignments is a bit clunkier.
split into OAS (1=1) and OAS2 (2 or more on one side).
delete dead chanrecv3 code.
delay construction of func types.
R=ken
OCL=31745
CL=31762
runtime package for use by debugger,
which needs to make sure that all ptrace calls
about a given pid come from the same thread.
R=r
DELTA=175 (90 added, 63 deleted, 22 changed)
OCL=31546
CL=31558
* use structs instead of interfaces
* compiler lays out data structures ahead of time,
so no more parsing of strings.
* unified reflect data structures with interface
runtime data structures.
* richer data structures should enable reflection
on chans and maps, but not implemented here.
R=r,iant
DELTA=1179 (1179 added, 0 deleted, 0 changed)
OCL=31107
CL=31213
now that cgo2c can handle it,
merge x.c and x_go.cgo into
a single x.cgo, for x=float,malloc,sema.
R=r
DELTA=1950 (954 added, 996 deleted, 0 changed)
OCL=30951
CL=30964
align the output args separately from the input args,
change cgo2c to insert the necessary padding
when the two arg lists are concatenated in the c
translation.
for example, there is a runtime
func indexstring(s string, i int32) (b byte)
right now in 6g those arguments are aligned in one
struct with s at offset 0, i at 16, and b at 20.
soon the b byte will be in its own struct and structs
are 8 aligned, so it will be b at 24.
right now cgo2c generates:
void indexstring(string s, int32 i, byte b)
this CL makes it generate, in --6g mode:
void indexstring(string s, int32 i, uint32, byte b)
this is valid 6c input, although not valid gcc input.
(the code is being generated for 6c only anyway.)
also, allow C code to be mixed in among the Go funcs.
every instance of the token `func' is expected to start
a new go func.
R=iant
DELTA=145 (118 added, 0 deleted, 27 changed)
OCL=30949
CL=30963
echo back context of call in error if likely to be useful.
For example, if os.Open("/etc/passwd", os.O_RDONLY)
fails with syscall.EPERM, it returns as the os.Error
&PathError{
Op: "open",
Path: "/etc/passwd"
Error: os.EPERM
}
which formats as
open /etc/passwd: permission denied
Not converted:
datafmt
go/...
google/...
regexp
tabwriter
template
R=r
DELTA=1153 (561 added, 156 deleted, 436 changed)
OCL=30738
CL=30781
* keep coherent SP/PC in gobuf
(i.e., SP that would be in use at that PC)
* gogocall replaces setspgoto,
should work better in presence of link registers
* delete unused system calls
only amd64; 386 is now broken
R=r
DELTA=548 (183 added, 183 deleted, 182 changed)
OCL=30381
CL=30442
- morestack support for 5l and arm runtime
- argsize support in 5c, 5l, ar and nm. assembly code from 5a
will break in interesting ways unless NOSPLIT is specified
- explicit cond execution constants
- fix 5l output to use %d instead of %ld so that negative
values show.
- added a lot of code to arm/asm.s. runtime entry code almost
working currently aborts at gogo not implemented
R=rsc
APPROVED=rsc
DELTA=305 (125 added, 29 deleted, 151 changed)
OCL=30246
CL=30347
not observed: do not use malloc to allocate stacks
during garbage collection, because it would make the
malloc data structures change underfoot.
R=r
DELTA=6 (3 added, 0 deleted, 3 changed)
OCL=30323
CL=30326