Renaming the C SysAlloc will let Go define a prototype without exporting it.
For use in cpuprof.goc's translation to Go.
LGTM=mdempsky
R=golang-codereviews, mdempsky
CC=golang-codereviews, iant
https://golang.org/cl/140060043
Remove C version of GC.
Convert freeOSMemory to Go.
Restore g0 check in GC.
Remove unknownGCPercent check in GC,
it's initialized explicitly now.
LGTM=rsc
R=golang-codereviews, rsc
CC=golang-codereviews, khr
https://golang.org/cl/139910043
Mutex is consistent with package sync, and when in the
unexported Go form it avoids having a conflcit between
the type (now mutex) and the function (lock).
LGTM=iant
R=golang-codereviews, iant
CC=dvyukov, golang-codereviews, r
https://golang.org/cl/133140043
uintptr or uint64 in the runtime C were turning into uint in the Go,
bool was turning into uint8, and so on. Fix that.
Also delete Go wrappers for C functions.
The C functions can be called directly now
(but still eventually need to be converted to Go).
LGTM=bradfitz, minux, iant
R=golang-codereviews, bradfitz, iant, minux
CC=golang-codereviews, khr, r
https://golang.org/cl/138740043
Before, a slice with cap=0 or a string with len=0 might have its
base pointer pointing beyond the actual slice/string data into
the next block. The collector had to ignore slices and strings with
cap=0 in order to avoid misinterpreting the base pointer.
Now, a slice with cap=0 or a string with len=0 still has a base
pointer pointing into the actual slice/string data, no matter what.
The collector can now always scan the pointer, which means
strings and slices are no longer special.
Fixes#8404.
LGTM=khr, josharian
R=josharian, khr, dvyukov
CC=golang-codereviews
https://golang.org/cl/112570044
A whole thread is too much for background scavenger that sleeps all the time anyway.
We already have sysmon thread that can do this work.
Also remove g->isbackground and simplify enter/exitsyscall.
LGTM=rsc
R=golang-codereviews, rsc
CC=golang-codereviews, khr, rlh
https://golang.org/cl/108640043
This is based on the crash dump provided by Alan
and on mental experiments:
sweep 0 74
fatal error: gc: unswept span
runtime stack:
runtime.throw(0x9df60d)
markroot(0xc208002000, 0x3)
runtime.parfordo(0xc208002000)
runtime.gchelper()
I think that when we moved all stacks into heap,
we introduced a bunch of bad data races. This was later
worsened by parallel stack shrinking.
Observation 1: exitsyscall can allocate a stack from heap at any time (including during STW).
Observation 2: parallel stack shrinking can (surprisingly) grow heap during marking.
Consider that we steadily grow stacks of a number of goroutines from 8K to 16K.
And during GC they all can be shrunk back to 8K. Shrinking will allocate lots of 8K
stacks, and we do not necessary have that many in heap at this moment. So shrinking
can grow heap as well.
Consequence: any access to mheap.allspans in GC (and otherwise) must take heap lock.
This is not true in several places.
Fix this by protecting accesses to mheap.allspans and introducing allspans cache for marking,
similar to what we use for sweeping.
LGTM=rsc
R=golang-codereviews, rsc
CC=adonovan, golang-codereviews, khr, rlh
https://golang.org/cl/126510043
These are required for chans, semaphores, timers, etc.
LGTM=khr
R=golang-codereviews, khr
CC=golang-codereviews, rlh, rsc
https://golang.org/cl/123640043
Newly allocated memory is subtracted from inuse, while it was never added to inuse.
Span leftovers are subtracted from both inuse and idle,
while they were never added.
Fixes#8544.
Fixes#8430.
LGTM=khr, cookieo9
R=golang-codereviews, khr, cookieo9
CC=golang-codereviews, rlh, rsc
https://golang.org/cl/130200044
Currently we do the following dance after sweeping a span:
1. lock mcentral
2. remove the span from a list
3. unlock mcentral
4. unmark span
5. lock mheap
6. insert the span into heap
7. unlock mheap
8. lock mcentral
9. observe empty list
10. unlock mcentral
11. lock mheap
12. grab the span
13. unlock mheap
14. mark span
15. lock mcentral
16. insert the span into empty list
17. unlock mcentral
This change short-circuits this sequence to nothing,
that is, we just cache and use the span after sweeping.
This gives us functionality similar (even better) to tcmalloc's transfer cache.
benchmark old ns/op new ns/op delta
BenchmarkMalloc8 22.2 19.5 -12.16%
BenchmarkMalloc16 31.0 26.6 -14.19%
LGTM=khr
R=golang-codereviews, khr
CC=golang-codereviews, rlh, rsc
https://golang.org/cl/119550043
Eliminating use of this extension makes it easier to port the Go runtime
to other compilers. This CL also disables the extension in cc to prevent
accidental use.
LGTM=rsc, khr
R=rsc, aram, khr, dvyukov
CC=axwalk, golang-codereviews
https://golang.org/cl/106790044
FlagNoGC is unused now.
FlagNoInvokeGC is unneeded as we don't invoke GC
on g0 and when holding locks anyway.
mal/malloc have very few uses and you never remember
the exact set of flags they use and the difference between them.
Moreover, eventually we need to give exact types to all allocations,
something what mal/malloc do not support.
LGTM=khr
R=golang-codereviews, khr
CC=golang-codereviews, rsc
https://golang.org/cl/117580043
Full spans can't be passed to UncacheSpan since we get rid of free.
LGTM=rsc
R=golang-codereviews
CC=golang-codereviews, khr, rsc
https://golang.org/cl/119490044
We call scanblock for lots of small root pieces
e.g. for every stack frame args and locals area.
Every scanblock invocation calls getempty/putempty,
which accesses lock-free stack shared among all worker threads.
One-element local cache allows most scanblock calls
to proceed without accessing the shared stack.
LGTM=rsc
R=golang-codereviews, rlh
CC=golang-codereviews, khr, rsc
https://golang.org/cl/121250043
Several reasons:
1. Significantly simplifies runtime.
2. This code proved to be buggy.
3. Free is incompatible with bump-the-pointer allocation.
4. We want to write runtime in Go, Go does not have free.
5. Too much code to free env strings on startup.
LGTM=khr
R=golang-codereviews, josharian, tracey.brendan, khr
CC=bradfitz, golang-codereviews, r, rlh, rsc
https://golang.org/cl/116390043
This change introduces gomallocgc, a Go clone of mallocgc.
Only a few uses have been moved over, so there are still
lots of uses from C. Many of these C uses will be moved
over to Go (e.g. in slice.goc), but probably not all.
What should remain of C's mallocgc is an open question.
LGTM=rsc, dvyukov
R=rsc, khr, dave, bradfitz, dvyukov
CC=golang-codereviews
https://golang.org/cl/108840046
Implement the design described in:
https://docs.google.com/document/d/1v4Oqa0WwHunqlb8C3ObL_uNQw3DfSY-ztoA-4wWbKcg/pub
Summary of the changes:
GC uses "2-bits per word" pointer type info embed directly into bitmap.
Scanning of stacks/data/heap is unified.
The old spans types go away.
Compiler generates "sparse" 4-bits type info for GC (directly for GC bitmap).
Linker generates "dense" 2-bits type info for data/bss (the same as stacks use).
Summary of results:
-1680 lines of code total (-1000+ in mgc0.c only)
-25% memory consumption
-3-7% binary size
-15% GC pause reduction
-7% run time reduction
LGTM=khr
R=golang-codereviews, rsc, christoph, khr
CC=golang-codereviews, rlh
https://golang.org/cl/106260045
redo stack allocation. This is mostly the same as
the original CL with a few bug fixes.
1. add racemalloc() for stack allocations
2. fix poolalloc/poolfree to terminate free lists correctly.
3. adjust span ref count correctly.
4. don't use cache for sizes >= StackCacheSize.
Should fix bugs and memory leaks in original changelist.
««« original CL description
undo CL 104200047 / 318b04f28372
Breaks windows and race detector.
TBR=rsc
««« original CL description
runtime: stack allocator, separate from mallocgc
In order to move malloc to Go, we need to have a
separate stack allocator. If we run out of stack
during malloc, malloc will not be available
to allocate a new stack.
Stacks are the last remaining FlagNoGC objects in the
GC heap. Once they are out, we can get rid of the
distinction between the allocated/blockboundary bits.
(This will be in a separate change.)
Fixes#7468Fixes#7424
LGTM=rsc, dvyukov
R=golang-codereviews, dvyukov, khr, dave, rsc
CC=golang-codereviews
https://golang.org/cl/104200047
»»»
TBR=rsc
CC=golang-codereviews
https://golang.org/cl/101570044
»»»
LGTM=dvyukov
R=dvyukov, dave, khr, alex.brainman
CC=golang-codereviews
https://golang.org/cl/112240044
Breaks windows and race detector.
TBR=rsc
««« original CL description
runtime: stack allocator, separate from mallocgc
In order to move malloc to Go, we need to have a
separate stack allocator. If we run out of stack
during malloc, malloc will not be available
to allocate a new stack.
Stacks are the last remaining FlagNoGC objects in the
GC heap. Once they are out, we can get rid of the
distinction between the allocated/blockboundary bits.
(This will be in a separate change.)
Fixes#7468Fixes#7424
LGTM=rsc, dvyukov
R=golang-codereviews, dvyukov, khr, dave, rsc
CC=golang-codereviews
https://golang.org/cl/104200047
»»»
TBR=rsc
CC=golang-codereviews
https://golang.org/cl/101570044
In order to move malloc to Go, we need to have a
separate stack allocator. If we run out of stack
during malloc, malloc will not be available
to allocate a new stack.
Stacks are the last remaining FlagNoGC objects in the
GC heap. Once they are out, we can get rid of the
distinction between the allocated/blockboundary bits.
(This will be in a separate change.)
Fixes#7468Fixes#7424
LGTM=rsc, dvyukov
R=golang-codereviews, dvyukov, khr, dave, rsc
CC=golang-codereviews
https://golang.org/cl/104200047
Use a real type for Gs instead of scanning them conservatively.
Zero the schedlink pointer when it is dead.
Update #7820
LGTM=rsc
R=rsc, dvyukov
CC=golang-codereviews
https://golang.org/cl/89360043
Trying to make GODEBUG=gcdead=1 work with liveness
and in particular ambiguously live variables.
1. In the liveness computation, mark all ambiguously live
variables as live for the entire function, except the entry.
They are zeroed directly after entry, and we need them not
to be poisoned thereafter.
2. In the liveness computation, compute liveness (and deadness)
for all parameters, not just pointer-containing parameters.
Otherwise gcdead poisons untracked scalar parameters and results.
3. Fix liveness debugging print for -live=2 to use correct bitmaps.
(Was not updated for compaction during compaction CL.)
4. Correct varkill during map literal initialization.
Was killing the map itself instead of the inserted value temp.
5. Disable aggressive varkill cleanup for call arguments if
the call appears in a defer or go statement.
6. In the garbage collector, avoid bug scanning empty
strings. An empty string is two zeros. The multiword
code only looked at the first zero and then interpreted
the next two bits in the bitmap as an ordinary word bitmap.
For a string the bits are 11 00, so if a live string was zero
length with a 0 base pointer, the poisoning code treated
the length as an ordinary word with code 00, meaning it
needed poisoning, turning the string into a poison-length
string with base pointer 0. By the same logic I believe that
a live nil slice (bits 11 01 00) will have its cap poisoned.
Always scan full multiword struct.
7. In the runtime, treat both poison words (PoisonGC and
PoisonStack) as invalid pointers that warrant crashes.
Manual testing as follows:
- Create a script called gcdead on your PATH containing:
#!/bin/bash
GODEBUG=gcdead=1 GOGC=10 GOTRACEBACK=2 exec "$@"
- Now you can build a test and then run 'gcdead ./foo.test'.
- More importantly, you can run 'go test -short -exec gcdead std'
to run all the tests.
Fixes#7676.
While here, enable the precise scanning of slices, since that was
disabled due to bugs like these. That now works, both with and
without gcdead.
Fixes#7549.
LGTM=khr
R=khr
CC=golang-codereviews
https://golang.org/cl/83410044
Reduce footprint of liveness bitmaps by about 5x.
1. Mark all liveness bitmap symbols as 4-byte aligned
(they were aligned to a larger size by default).
2. The bitmap data is a bitmap count n followed by n bitmaps.
Each bitmap begins with its own count m giving the number
of bits. All the m's are the same for the n bitmaps.
Emit this bitmap length once instead of n times.
3. Many bitmaps within a function have the same bit values,
but each call site was given a distinct bitmap. Merge duplicate
bitmaps so that no bitmap is written more than once.
4. Many functions end up with the same aggregate bitmap data.
We used to name the bitmap data funcname.gcargs and funcname.gclocals.
Instead, name it gclocals.<md5 of data> and mark it dupok so
that the linker coalesces duplicate sets. This cut the bitmap
data remaining after step 3 by 40%; I was not expecting it to
be quite so dramatic.
Applied to "go build -ldflags -w code.google.com/p/go.tools/cmd/godoc":
bitmaps pclntab binary on disk
before this CL 1326600 1985854 12738268
4-byte align 1154288 (0.87x) 1985854 (1.00x) 12566236 (0.99x)
one bitmap len 782528 (0.54x) 1985854 (1.00x) 12193500 (0.96x)
dedup bitmap 414748 (0.31x) 1948478 (0.98x) 11787996 (0.93x)
dedup bitmap set 245580 (0.19x) 1948478 (0.98x) 11620060 (0.91x)
While here, remove various dead blocks of code from plive.c.
Fixes#6929.
Fixes#7568.
LGTM=khr
R=khr
CC=golang-codereviews
https://golang.org/cl/83630044
GODEBUG=allocfreetrace=1:
The allocfreetrace=1 mode prints a stack trace for each block
allocated and freed, and also a stack trace for each garbage collection.
It was implemented by reusing the heap profiling support: if allocfreetrace=1
then the heap profile was effectively running at 1 sample per 1 byte allocated
(always sample). The stack being shown at allocation was the stack gathered
for profiling, meaning it was derived only from the program counters and
did not include information about function arguments or frame pointers.
The stack being shown at free was the allocation stack, not the free stack.
If you are generating this log, you can find the allocation stack yourself, but
it can be useful to see exactly the sequence that led to freeing the block:
was it the garbage collector or an explicit free? Now that the garbage collector
runs on an m0 stack, the stack trace for the garbage collector was never interesting.
Fix all these problems:
1. Decouple allocfreetrace=1 from heap profiling.
2. Print the standard goroutine stack traces instead of a custom format.
3. Print the stack trace at time of allocation for an allocation,
and print the stack trace at time of free (not the allocation trace again)
for a free.
4. Print all goroutine stacks at garbage collection. Having all the stacks
means that you can see the exact point at which each goroutine was
preempted, which is often useful for identifying liveness-related errors.
GODEBUG=gcdead=1:
This mode overwrites dead pointers with a poison value.
Detect the poison value as an invalid pointer during collection,
the same way that small integers are invalid pointers.
LGTM=khr
R=khr
CC=golang-codereviews
https://golang.org/cl/81670043
Currently it's possible that bgsweep finishes before all spans
have been swept (we only know that sweeping of all spans has *started*).
In such case bgsweep may fail wake up runfinq goroutine when it needs to.
finq may still be nil at this point, but some finalizers may be queued later.
Make bgsweep to wait for sweeping to *complete*, then it can decide
whether it needs to wake up runfinq for sure.
Update #7533
LGTM=rsc
R=rsc
CC=golang-codereviews
https://golang.org/cl/75960043
Change two-bit stack map entries to encode:
0 = dead
1 = scalar
2 = pointer
3 = multiword
If multiword, the two-bit entry for the following word encodes:
0 = string
1 = slice
2 = iface
3 = eface
That way, during stack scanning we can check if a string
is zero length or a slice has zero capacity. We can avoid
following the contained pointer in those cases. It is safe
to do so because it can never be dereferenced, and it is
desirable to do so because it may cause false retention
of the following block in memory.
Slice feature turned off until issue 7564 is fixed.
Update #7549
LGTM=rsc
R=golang-codereviews, bradfitz, rsc
CC=golang-codereviews
https://golang.org/cl/76380043
The existing code did not have a clear notion of whether
memory has been actually reserved. It checked based on
whether in 32-bit mode or 64-bit mode and (on GNU/Linux) the
requested address, but it confused the requested address and
the returned address.
LGTM=rsc
R=rsc, dvyukov
CC=golang-codereviews, michael.hudson
https://golang.org/cl/79610043
Instead, split the underlying storage in half and
free just half of it.
Shrinking without copying lets us reclaim storage used
by a previously profligate Go routine that has now blocked
inside some C code.
To shrink in place, we need all stacks to be a power of 2 in size.
LGTM=rsc
R=golang-codereviews, rsc
CC=golang-codereviews
https://golang.org/cl/69580044
Two memory allocator bug fixes.
- efence is not maintaining the proper heap metadata
to make eventual memory reuse safe, so use SysFault.
- now that our heap PageSize is 8k but most hardware
uses 4k pages, SysAlloc and SysReserve results must be
explicitly aligned. Do that in a few more call sites and
document this fact in malloc.h.
Fixes#7448.
LGTM=iant
R=golang-codereviews, josharian, iant
CC=dvyukov, golang-codereviews
https://golang.org/cl/71750048
On stack overflow, if all frames on the stack are
copyable, we copy the frames to a new stack twice
as large as the old one. During GC, if a G is using
less than 1/4 of its stack, copy the stack to a stack
half its size.
TODO
- Do something about C frames. When a C frame is in the
stack segment, it isn't copyable. We allocate a new segment
in this case.
- For idempotent C code, we can abort it, copy the stack,
then retry. I'm working on a separate CL for this.
- For other C code, we can raise the stackguard
to the lowest Go frame so the next call that Go frame
makes triggers a copy, which will then succeed.
- Pick a starting stack size?
The plan is that eventually we reach a point where the
stack contains only copyable frames.
LGTM=rsc
R=dvyukov, rsc
CC=golang-codereviews
https://golang.org/cl/54650044
MCaches now hold a MSpan for each sizeclass which they have
exclusive access to allocate from, so no lock is needed.
Modifying the heap bitmaps also no longer requires a cas.
runtime.free gets more expensive. But we don't use it
much any more.
It's not much faster on 1 processor, but it's a lot
faster on multiple processors.
benchmark old ns/op new ns/op delta
BenchmarkSetTypeNoPtr1 24 23 -0.42%
BenchmarkSetTypeNoPtr2 33 34 +0.89%
BenchmarkSetTypePtr1 51 49 -3.72%
BenchmarkSetTypePtr2 55 54 -1.98%
benchmark old ns/op new ns/op delta
BenchmarkAllocation 52739 50770 -3.73%
BenchmarkAllocation-2 33957 34141 +0.54%
BenchmarkAllocation-3 33326 29015 -12.94%
BenchmarkAllocation-4 38105 25795 -32.31%
BenchmarkAllocation-5 68055 24409 -64.13%
BenchmarkAllocation-6 71544 23488 -67.17%
BenchmarkAllocation-7 68374 23041 -66.30%
BenchmarkAllocation-8 70117 20758 -70.40%
LGTM=rsc, dvyukov
R=dvyukov, bradfitz, khr, rsc
CC=golang-codereviews
https://golang.org/cl/46810043
Package runtime's C functions written to be called from Go
started out written in C using carefully constructed argument
lists and the FLUSH macro to write a result back to memory.
For some functions, the appropriate parameter list ended up
being architecture-dependent due to differences in alignment,
so we added 'goc2c', which takes a .goc file containing Go func
declarations but C bodies, rewrites the Go func declaration to
equivalent C declarations for the target architecture, adds the
needed FLUSH statements, and writes out an equivalent C file.
That C file is compiled as part of package runtime.
Native Client's x86-64 support introduces the most complex
alignment rules yet, breaking many functions that could until
now be portably written in C. Using goc2c for those avoids the
breakage.
Separately, Keith's work on emitting stack information from
the C compiler would require the hand-written functions
to add #pragmas specifying how many arguments are result
parameters. Using goc2c for those avoids maintaining #pragmas.
For both reasons, use goc2c for as many Go-called C functions
as possible.
This CL is a replay of the bulk of CL 15400047 and CL 15790043,
both of which were reviewed as part of the NaCl port and are
checked in to the NaCl branch. This CL is part of bringing the
NaCl code into the main tree.
No new code here, just reformatting and occasional movement
into .h files.
LGTM=r
R=dave, alex.brainman, r
CC=golang-codereviews
https://golang.org/cl/65220044
This cleans up the code significantly, and it avoids any
possible problems with madvise zeroing out some but
not all of the data.
Fixes#6400.
LGTM=dave
R=dvyukov, dave
CC=golang-codereviews
https://golang.org/cl/57680046
Moves sweep phase out of stoptheworld by adding
background sweeper goroutine and lazy on-demand sweeping.
It turned out to be somewhat trickier than I expected,
because there is no point in time when we know size of live heap
nor consistent number of mallocs and frees.
So everything related to next_gc, mprof, memstats, etc becomes trickier.
At the end of GC next_gc is conservatively set to heap_alloc*GOGC,
which is much larger than real value. But after every sweep
next_gc is decremented by freed*GOGC. So when everything is swept
next_gc becomes what it should be.
For mprof I had to introduce 3-generation scheme (allocs, revent_allocs, prev_allocs),
because by the end of GC we know number of frees for the *previous* GC.
Significant caution is required to not cross yet-unknown real value of next_gc.
This is achieved by 2 means:
1. Whenever I allocate a span from MCentral, I sweep a span in that MCentral.
2. Whenever I allocate N pages from MHeap, I sweep until at least N pages are
returned to heap.
This provides quite strong guarantees that heap does not grow when it should now.
http-1
allocated 7036 7033 -0.04%
allocs 60 60 +0.00%
cputime 51050 46700 -8.52%
gc-pause-one 34060569 1777993 -94.78%
gc-pause-total 2554 133 -94.79%
latency-50 178448 170926 -4.22%
latency-95 284350 198294 -30.26%
latency-99 345191 220652 -36.08%
rss 101564416 101007360 -0.55%
sys-gc 6606832 6541296 -0.99%
sys-heap 88801280 87752704 -1.18%
sys-other 7334208 7405928 +0.98%
sys-stack 524288 524288 +0.00%
sys-total 103266608 102224216 -1.01%
time 50339 46533 -7.56%
virtual-mem 292990976 293728256 +0.25%
garbage-1
allocated 2983818 2990889 +0.24%
allocs 62880 62902 +0.03%
cputime 16480000 16190000 -1.76%
gc-pause-one 828462467 487875135 -41.11%
gc-pause-total 4142312 2439375 -41.11%
rss 1151709184 1153712128 +0.17%
sys-gc 66068352 66068352 +0.00%
sys-heap 1039728640 1039728640 +0.00%
sys-other 37776064 40770176 +7.93%
sys-stack 8781824 8781824 +0.00%
sys-total 1152354880 1155348992 +0.26%
time 16496998 16199876 -1.80%
virtual-mem 1409564672 1402281984 -0.52%
LGTM=rsc
R=golang-codereviews, sameer, rsc, iant, jeremyjackins, gobot
CC=golang-codereviews, khr
https://golang.org/cl/46430043
Currently windows crashes because early allocs in schedinit
try to allocate tiny memory blocks, but m->p is not yet setup.
I've considered calling procresize(1) earlier in schedinit,
but this refactoring is better and must fix the issue as well.
Fixes#7218.
R=golang-codereviews, r
CC=golang-codereviews
https://golang.org/cl/54570045
When growing slice take into account size of the allocated memory block.
Also apply the same optimization to string->[]byte conversion.
Fixes#6307.
benchmark old ns/op new ns/op delta
BenchmarkAppendGrowByte 4541036 4434108 -2.35%
BenchmarkAppendGrowString 59885673 44813604 -25.17%
LGTM=khr
R=khr
CC=golang-codereviews, iant, rsc
https://golang.org/cl/53340044
Russ Cox
Dmitriy Vyukov
Keith Randall
Sanjay Menakuru
Peter Collingbourne
Ian Lance Taylor