Checking whether the current allocation needs to be profiled is
currently branch-y and weirdly a lot of code. The branches are
generally predictable, but it's a surprising number of instructions.
Part of the problem is that MemProfileRate is just a global that can be
set at any time, so we need to load it and check certain settings
explicitly. In an ideal world, we would just always subtract from
nextSample and have a single branch to take the slow path if we
subtract below zero.
If MemProfileRate were a function, we could trash all the nextSample
values intentionally in each mcache. This would be slow, but
MemProfileRate changes rarely while the malloc hot path is well, hot.
Unfortunate...
Although this ideal world is, AFAICT, impossible, we can still get
close. If we cache the value of MemProfileRate in each mcache, then we
can force malloc to take the slow path whenever MemProfileRate changes.
This does require two additional loads, but crucially, these loads are
independent of everything else in mallocgc. Furthermore, the branch
dependent on those loads is incredibly predictable in practice.
This CL on its own has little-to-no impact on mallocgc. But this
codepath is going to be duplicated in several places in the next CL, so
it'll pay to simplify it. Also, we're very much trying to remedy a
death-by-a-thousand-cuts situation, and malloc is currently still kind
of a monster -- it will not help if mallocgc isn't really streamlined
itself.
Lastly, there's a nice property now that all nextSample values get
immediately re-sampled when MemProfileRate changes.
Change-Id: I6443d0cf9bd7861595584442b675ac1be8ea3455
Reviewed-on: https://go-review.googlesource.com/c/go/+/615815
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
This change brings back a minor optimization lost in the Go 1.22 cycle
wherein the 8-byte pointer-ful span class spans would have the pointer
bitmap written ahead of time in bulk, because there's only one possible
pattern.
│ before │ after │
│ sec/op │ sec/op vs base │
MallocTypeInfo8-4 25.13n ± 1% 23.59n ± 2% -6.15% (p=0.002 n=6)
Change-Id: I135b84bb1d5b7e678b841b56430930bc73c0a038
Reviewed-on: https://go-review.googlesource.com/c/go/+/614256
Reviewed-by: Keith Randall <khr@golang.org>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Keith Randall <khr@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Cleanup and friction reduction
For #65355.
Change-Id: Ia14c9dc584a529a35b97801dd3e95b9acc99a511
Reviewed-on: https://go-review.googlesource.com/c/go/+/600436
Reviewed-by: Keith Randall <khr@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Keith Randall <khr@golang.org>
mspan.freeindex and nelems can fit into uint16 for all possible
values. Use uint16 instead of uintptr.
Change-Id: Ifce20751e81d5022be1f6b5cbb5fbe4fd1728b1b
Reviewed-on: https://go-review.googlesource.com/c/go/+/451359
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Run 'unconvert -safe -apply' (https://github.com/mdempsky/unconvert)
Change-Id: I24b7cd7d286cddce86431d8470d15c5f3f0d1106
GitHub-Last-Rev: 022e75384c
GitHub-Pull-Request: golang/go#62662
Reviewed-on: https://go-review.googlesource.com/c/go/+/528696
Auto-Submit: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change adds an API to the runtime for arenas. A later CL can
potentially export it as an experimental API, but for now, just the
runtime implementation will suffice.
The purpose of arenas is to improve efficiency, primarily by allowing
for an application to manually free memory, thereby delaying garbage
collection. It comes with other potential performance benefits, such as
better locality, a better allocation strategy, and better handling of
interior pointers by the GC.
This implementation is based on one by danscales@google.com with a few
significant differences:
* The implementation lives entirely in the runtime (all layers).
* Arena chunks are the minimum of 8 MiB or the heap arena size. This
choice is made because in practice 64 MiB appears to be way too large
of an area for most real-world use-cases.
* Arena chunks are not unmapped, instead they're placed on an evacuation
list and when there are no pointers left pointing into them, they're
allowed to be reused.
* Reusing partially-used arena chunks no longer tries to find one used
by the same P first; it just takes the first one available.
* In order to ensure worst-case fragmentation is never worse than 25%,
only types and slice backing stores whose sizes are 1/4th the size of
a chunk or less may be used. Previously larger sizes, up to the size
of the chunk, were allowed.
* ASAN, MSAN, and the race detector are fully supported.
* Sets arena chunks to fault that were deferred at the end of mark
termination (a non-public patch once did this; I don't see a reason
not to continue that).
For #51317.
Change-Id: I83b1693a17302554cb36b6daa4e9249a81b1644f
Reviewed-on: https://go-review.googlesource.com/c/go/+/423359
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
[this is a retry of CL 407035 + its revert CL 422395. The content is unchanged]
Use just 1 bit per word to record the ptr/nonptr bitmap.
Use word-sized operations to manipulate the bitmap, so we can operate
on up to 64 ptr/nonptr bits at a time.
Use a separate bitmap, one bit per word of the ptr/nonptr bitmap,
to encode a no-more-pointers signal. Since we can check 64 ptr/nonptr
bits at once, knowing the exact last pointer location is not necessary.
As a followon CL, we should make the gcdata bitmap an array of
uintptr instead of an array of byte, so we can load 64 bits of it at once.
Similarly for the processing of gc programs.
Change-Id: Ica5eb622f5b87e647be64f471d67b02732ef8be6
Reviewed-on: https://go-review.googlesource.com/c/go/+/422634
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
This reverts commit b589208c8c.
Reason for revert: Bug somewhere in this code, causing wasm and maybe linux/386 to fail.
Change-Id: I5e1e501d839584e0219271bb937e94348f83c11f
Reviewed-on: https://go-review.googlesource.com/c/go/+/422395
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Keith Randall <khr@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Use just 1 bit per word to record the ptr/nonptr bitmap.
Use word-sized operations to manipulate the bitmap, so we can operate
on up to 64 ptr/nonptr bits at a time.
Use a separate bitmap, one bit per word of the ptr/nonptr bitmap,
to encode a no-more-pointers signal. Since we can check 64 ptr/nonptr
bits at once, knowing the exact last pointer location is not necessary.
This cleans up the bitmap implementation significantly, which will
hopefully make it faster. TODO: measure
As a followon CL, we should make the gcdata bitmap an array of
uintptr instead of an array of byte, so we can load 64 bits of it at once.
Similarly for the processing of gc programs.
Change-Id: I18151b1876d9543599800dec51e2a1b19df97d49
Reviewed-on: https://go-review.googlesource.com/c/go/+/407035
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Keith Randall <khr@google.com>
CL 377516 made it so that memory metrics are truly monotonic, but also
updated how heapLive tracked allocated memory to also be monotonic.
The result is that cached spans with allocated memory aren't fully
accounted for by the GC, causing it to make a worse assumption (the
exact mechanism is at this time unknown), resulting in a memory
regression, especially for smaller heaps.
This change is a partial revert of CL 377516 that makes heapLive a
non-monotonic overestimate again, which appears to resolve the
regression.
For #53738.
Change-Id: I5c51067abc0b8e0a6b89dd8dbd4a0be2e8c0c1b2
Reviewed-on: https://go-review.googlesource.com/c/go/+/416417
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently the consistent total allocation stats are managed as uintptrs,
which means they can easily overflow on 32-bit systems. Fix this by
storing these stats as uint64s. This will cause some minor performance
degradation on 32-bit systems, but there really isn't a way around this,
and it affects the correctness of the metrics we export.
Fixes#52680.
Change-Id: I7e6ca44047d46b4bd91c6f87c2d29f730e0d6191
Reviewed-on: https://go-review.googlesource.com/c/go/+/403758
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Austin Clements <austin@google.com>
Fundamentally, all of these memstats exist to serve the runtime in
managing memory. For the sake of simpler testing, couple these stats
more tightly with the GC.
This CL was mostly done automatically. The fields had to be moved
manually, but the references to the fields were updated via
gofmt -w -r 'memstats.<field> -> gcController.<field>' *.go
For #48409.
Change-Id: Ic036e875c98138d9a11e1c35f8c61b784c376134
Reviewed-on: https://go-review.googlesource.com/c/go/+/397678
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This will be used by the memory limit computation to determine
overheads.
For #48409.
Change-Id: Iaa4e26e1e6e46f88d10ba8ebb6b001be876dc5cd
Reviewed-on: https://go-review.googlesource.com/c/go/+/394220
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Right now we export alloc count metrics via the runtime/metrics package
and mark them as monotonic, but that's not actually true. As an
optimization, the runtime assumes a span is always fully allocated
before being uncached, and updates the accounting as such. In the rare
case that it's wrong, the span has enough information to back out what
did not get allocated.
This change uses 16 bits of padding in the mspan to house another field
that represents the amount of mspan slots filled just as the mspan is
cached. This is information is enough to get an exact count, allowing us
to make the metrics truly monotonic.
Change-Id: Iaff3ca43f8745dc1bbb0232372423e014b89b920
Reviewed-on: https://go-review.googlesource.com/c/go/+/377516
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This change moves heapLive and heapScan updates on gcController into a
method for better testability. It's also less error-prone because code
that updates these fields needs to remember to emit traces and/or call
gcController.revise; this method now handles those cases.
For #44167.
Change-Id: I3d6f2e7abb22def27c93feacff50162b0b074da2
Reviewed-on: https://go-review.googlesource.com/c/go/+/309275
Trust: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Currently, the runtime zeroes allocations in several ways. First, small
object spans are always zeroed if they come from mheap, and their slots
are zeroed later in mallocgc if needed. Second, large object spans
(objects that have their own spans) plumb the need for zeroing down into
mheap. Thirdly, large objects that have no pointers have their zeroing
delayed until after preemption is reenabled, but before returning in
mallocgc.
All of this has two consequences:
1. Spans for small objects that come from mheap are sometimes
unnecessarily zeroed, even if the mallocgc call that created them
doesn't need the object slot to be zeroed.
2. This is all messy and difficult to reason about.
This CL simplifies this code, resolving both (1) and (2). First, it
recognizes that zeroing in mheap is unnecessary for small object spans;
mallocgc and its callees in mcache and mcentral, by design, are *always*
able to deal with non-zeroed spans. They must, for they deal with
recycled spans all the time. Once this fact is made clear, the only
remaining use of zeroing in mheap is for large objects.
As a result, this CL lifts mheap zeroing for large objects into
mallocgc, to parallel all the other codepaths in mallocgc. This is makes
the large object allocation code less surprising.
Next, this CL sets the flag for the delayed zeroing explicitly in the one
case where it matters, and inverts and renames the flag from isZeroed to
delayZeroing.
Finally, it adds a check to make sure that only pointer-free allocations
take the delayed zeroing codepath, as an extra safety measure.
Benchmark results: https://perf.golang.org/search?q=upload:20211028.8
Inspired by tapir.liu@gmail.com's CL 343470.
Change-Id: I7e1296adc19ce8a02c8d93a0a5082aefb2673e8f
Reviewed-on: https://go-review.googlesource.com/c/go/+/359477
Trust: Michael Knyszek <mknyszek@google.com>
Reviewed-by: David Chase <drchase@google.com>
Since all callers of getMCache appear to have mp available,
we pass the mp to getMCache, and reduce one call to getg.
And after modification, getMCache is also inlined.
Change-Id: Ib7880c118336acc026ecd7c60c5a88722c3ddfc7
Reviewed-on: https://go-review.googlesource.com/c/go/+/349329
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Trust: Michael Knyszek <mknyszek@google.com>
Trust: Carlos Amedee <carlos@golang.org>
If something "huge" is allocated, and the zeroing is trivial (no pointers
involved) then zero it by chunks in a loop so that preemption can occur,
not all in a single non-preemptible call.
Benchmarking suggests that 256K is the best chunk size.
Updates #42642.
Change-Id: I94015e467eaa098c59870e479d6d83bc88efbfb4
Reviewed-on: https://go-review.googlesource.com/c/go/+/270943
Trust: David Chase <drchase@google.com>
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Currently tiny allocations are not represented in either MemStats or
runtime/metrics, but they're represented in MemStats (indirectly) via
Mallocs. Add them to runtime/metrics by first merging
memstats.tinyallocs into consistentHeapStats (just for simplicity; it's
monotonic so metrics would still be self-consistent if we just read it
atomically) and then adding /gc/heap/tiny/allocs:objects to the list of
supported metrics.
Change-Id: Ie478006ab942a3e877b4a79065ffa43569722f3d
Reviewed-on: https://go-review.googlesource.com/c/go/+/312909
Trust: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change moves certain important but internal-only GC statistics from
memstats into gcController. These statistics are mainly used in pacing
the GC, so it makes sense to keep them in the pacer's state.
This CL was mostly generated via
rf '
ex . {
memstats.gc_trigger -> gcController.trigger
memstats.triggerRatio -> gcController.triggerRatio
memstats.heap_marked -> gcController.heapMarked
memstats.heap_live -> gcController.heapLive
memstats.heap_scan -> gcController.heapScan
}
'
except for a few special cases, like updating names in comments and when
these fields are used within gcControllerState methods (at which point
they're accessed through the reciever).
For #44167.
Change-Id: I6bd1602585aeeb80818ded24c07d8e6fec992b93
Reviewed-on: https://go-review.googlesource.com/c/go/+/306598
Trust: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change modifies the consistent stats implementation to keep the
per-P sequence counter on each P instead of each mcache. A valid mcache
is not available everywhere that we want to call e.g. allocSpan, as per
issue #42339. By decoupling these two, we can add a mechanism to allow
contexts without a P to update stats consistently.
In this CL, we achieve that with a mutex. In practice, it will be very
rare for an M to update these stats without a P. Furthermore, the stats
reader also only needs to hold the mutex across the update to "gen"
since once that changes, writers are free to continue updating the new
stats generation. Contention could thus only arise between writers
without a P, and as mentioned earlier, those should be rare.
A nice side-effect of this change is that the consistent stats acquire
and release API becomes simpler.
Fixes#42339.
Change-Id: Ied74ab256f69abd54b550394c8ad7c4c40a5fe34
Reviewed-on: https://go-review.googlesource.com/c/go/+/267158
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Trust: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change moves the responsibility of throwing if an mcache is not
available to the caller, because the inlining cost of throw is set very
high in the compiler. Even if it was reduced down to the cost of a usual
function call, it would still be too expensive, so just move it out.
This choice also makes sense in the context of #42339 since we're going
to have to handle the case where we don't have an mcache to update stats
in a few contexts anyhow.
Also, add getMCache to the list of functions that should be inlined to
prevent future regressions.
getMCache is called on the allocation fast path and because its not
inlined actually causes a significant regression (~10%) in some
microbenchmarks.
Fixes#42305.
Change-Id: I64ac5e4f26b730bd4435ea1069a4a50f55411ced
Reviewed-on: https://go-review.googlesource.com/c/go/+/267157
Trust: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
This change moves the mcache-local malloc stats into the
consistentHeapStats structure so the malloc stats can be managed
consistently with the memory stats. The one exception here is
tinyAllocs for which moving that into the global stats would incur
several atomic writes on the fast path. Microbenchmarks for just one CPU
core have shown a 50% loss in throughput. Since tiny allocation counnt
isn't exposed anyway and is always blindly added to both allocs and
frees, let that stay inconsistent and flush the tiny allocation count
every so often.
Change-Id: I2a4b75f209c0e659b9c0db081a3287bf227c10ca
Reviewed-on: https://go-review.googlesource.com/c/go/+/247039
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Trust: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change adds a global set of heap statistics which are similar
to existing memory statistics. The purpose of these new statistics
is to be able to read them and get a consistent result without stopping
the world. The goal is to eventually replace as many of the existing
memstats statistics with the sharded ones as possible.
The consistent memory statistics use a tailor-made synchronization
mechanism to allow writers (allocators) to proceed with minimal
synchronization by using a sequence counter and a global generation
counter to determine which set of statistics to update. Readers
increment the global generation counter to effectively grab a snapshot
of the statistics, and then iterate over all Ps using the sequence
counter to ensure that they may safely read the snapshotted statistics.
To keep statistics fresh, the reader also has a responsibility to merge
sets of statistics.
These consistent statistics are computed, but otherwise unused for now.
Upcoming changes will integrate them with the rest of the codebase and
will begin to phase out existing statistics.
Change-Id: I637a11f2439e2049d7dccb8650c5d82500733ca5
Reviewed-on: https://go-review.googlesource.com/c/go/+/247037
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Trust: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change adds a function getMCache which returns the current P's
mcache if it's available, and otherwise tries to get mcache0 if we're
bootstrapping. This function will come in handy as we need to replicate
this behavior in multiple places in future changes.
Change-Id: I536073d6f6dc6c6390269e613ead9f8bcb6e7f98
Reviewed-on: https://go-review.googlesource.com/c/go/+/246976
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Trust: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change renames a bunch of malloc statistics stored in the mcache
that are all named with the "local_" prefix. It also renames largeAlloc
to allocLarge to prevent a naming conflict, and next_sample because it
would be the last mcache field with the old C naming style.
Change-Id: I29695cb83b397a435ede7e9ad5c3c9be72767ea3
Reviewed-on: https://go-review.googlesource.com/c/go/+/246969
Trust: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Now that local_scan is the last mcache-based statistic that is flushed
by purgecachedstats, and heap_scan and gcController.revise may be
interacted with concurrently, we don't need to flush heap_scan at
arbitrary locations where the heap is locked, and we don't need
purgecachedstats and cachestats anymore. Instead, we can flush
local_scan at the same time we update heap_live in refill, so the two
updates may share the same revise call.
Clean up unused functions, remove code that would cause the heap to get
locked in the allocSpan when it didn't need to (other than to flush
local_scan), and flush local_scan explicitly in a few important places.
Notably we need to flush local_scan whenever we flush the other stats,
but it doesn't need to be donated anywhere, so have releaseAll do the
flushing. Also, we need to flush local_scan before we set heap_scan at
the end of a GC, which was previously handled by cachestats. Just do so
explicitly -- it's not much code and it becomes a lot more clear why we
need to do so.
Change-Id: I35ac081784df7744d515479896a41d530653692d
Reviewed-on: https://go-review.googlesource.com/c/go/+/246968
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Trust: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change makes local_tinyallocs work like the rest of the malloc
stats and doesn't flush local_tinyallocs, instead making that the
source-of-truth.
Change-Id: I3e6cb5f1b3d086e432ce7d456895511a48e3617a
Reviewed-on: https://go-review.googlesource.com/c/go/+/246967
Trust: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change removes mcentral.nmalloc and adds mcache.local_nsmallalloc
which fulfills the same role but may be accessed non-atomically. It also
moves responsibility for updating heap_live and local_nsmallalloc into
mcache functions.
As a result of this change, mcache is now the sole source-of-truth for
malloc stats. It is also solely responsible for updating heap_live and
performing the various operations required as a result of updating
heap_live. The overall improvement here is in code organization:
previously malloc stats were fairly scattered, and now they have one
single home, and nearly all the required manipulations exist in a single
file.
Change-Id: I7e93fa297c1debf17e3f2a0d68aeed28a9c6af00
Reviewed-on: https://go-review.googlesource.com/c/go/+/246966
Trust: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change makes nlargealloc and largealloc into mcache fields just
like nlargefree and largefree. These local fields become the new
source-of-truth. This change also moves the accounting for these fields
out of allocSpan (which is an inappropriate place for it -- this
accounting generally happens much closer to the point of allocation) and
into largeAlloc. This move is partially possible now that we can call
gcController.revise at that point.
Furthermore, this change moves largeAlloc into mcache.go and makes it a
method of mcache. While there's a little bit of a mismatch here because
largeAlloc barely interacts with the mcache, it helps solidify the
mcache as the first allocation layer and provides a clear place to
aggregate and manage statistics.
Change-Id: I37b5e648710733bb4c04430b71e96700e438587a
Reviewed-on: https://go-review.googlesource.com/c/go/+/246965
Trust: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change makes it so that various local malloc stats (excluding
heap_scan and local_tinyallocs) are no longer written first to mheap
fields but are instead accessed directly from each mcache.
This change is part of a move toward having stats be distributed, and
cleaning up some old code related to the stats.
Note that because there's no central source-of-truth, when an mcache
dies, it must donate its stats to another mcache. It's always safe to
donate to the mcache for the 0th P, so do that.
Change-Id: I2556093dbc27357cb9621c9b97671f3c00aa1173
Reviewed-on: https://go-review.googlesource.com/c/go/+/246964
Trust: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change deletes the old mcentral implementation from the code base
and the newMCentralImpl feature flag along with it.
Updates #37487.
Change-Id: Ibca8f722665f0865051f649ffe699cbdbfdcfcf2
Reviewed-on: https://go-review.googlesource.com/c/go/+/221184
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Currently mcentral is implemented as a couple of linked lists of spans
protected by a lock. Unfortunately this design leads to significant lock
contention.
The span ownership model is also confusing and complicated. In-use spans
jump between being owned by multiple sources, generally some combination
of a gcSweepBuf, a concurrent sweeper, an mcentral or an mcache.
So first to address contention, this change replaces those linked lists
with gcSweepBufs which have an atomic fast path. Then, we change up the
ownership model: a span may be simultaneously owned only by an mcentral
and the page reclaimer. Otherwise, an mcentral (which now consists of
sweep bufs), a sweeper, or an mcache are the sole owners of a span at
any given time. This dramatically simplifies reasoning about span
ownership in the runtime.
As a result of this new ownership model, sweeping is now driven by
walking over the mcentrals rather than having its own global list of
spans. Because we no longer have a global list and we traditionally
haven't used the mcentrals for large object spans, we no longer have
anywhere to put large objects. So, this change also makes it so that we
keep large object spans in the appropriate mcentral lists.
In terms of the static lock ranking, we add the spanSet spine locks in
pretty much the same place as the mcentral locks, since they have the
potential to be manipulated both on the allocation and sweep paths, like
the mcentral locks.
This new implementation is turned on by default via a feature flag
called go115NewMCentralImpl.
Benchmark results for 1 KiB allocation throughput (5 runs each):
name \ MiB/s go113 go114 gotip gotip+this-patch
AllocKiB-1 1.71k ± 1% 1.68k ± 1% 1.59k ± 2% 1.71k ± 1%
AllocKiB-2 2.46k ± 1% 2.51k ± 1% 2.54k ± 1% 2.93k ± 1%
AllocKiB-4 4.27k ± 1% 4.41k ± 2% 4.33k ± 1% 5.01k ± 2%
AllocKiB-8 4.38k ± 3% 5.24k ± 1% 5.46k ± 1% 8.23k ± 1%
AllocKiB-12 4.38k ± 3% 4.49k ± 1% 5.10k ± 1% 10.04k ± 0%
AllocKiB-16 4.31k ± 1% 4.14k ± 3% 4.22k ± 0% 10.42k ± 0%
AllocKiB-20 4.26k ± 1% 3.98k ± 1% 4.09k ± 1% 10.46k ± 3%
AllocKiB-24 4.20k ± 1% 3.97k ± 1% 4.06k ± 1% 10.74k ± 1%
AllocKiB-28 4.15k ± 0% 4.00k ± 0% 4.20k ± 0% 10.76k ± 1%
Fixes#37487.
Change-Id: I92d47355acacf9af2c41bf080c08a8c1638ba210
Reviewed-on: https://go-review.googlesource.com/c/go/+/221182
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Overflow of the comparison caused very large (>=1<<32) allocations to
sometimes not get sampled at all. Use uintptr so the comparison will
never overflow.
Fixes#33342
Tested on the example in 33342. I don't want to check a test in that
needs that much memory, however.
Change-Id: I51fe77a9117affed8094da93c0bc5f445ac2d3d3
Reviewed-on: https://go-review.googlesource.com/c/go/+/188017
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Currently there's an invariant in the runtime wherein the heap lock
can only be acquired on the system stack, otherwise a self-deadlock
could occur if the stack grows while the lock is held.
This invariant is upheld and documented in a number of situations (e.g.
allocManual, freeManual) but there are other places where the invariant
is either not maintained at all which risks self-deadlock (e.g.
setGCPercent, gcResetMarkState, allocmcache) or is maintained but
undocumented (e.g. gcSweep, readGCStats_m).
This change adds go:systemstack to any function that acquires the heap
lock or adds a systemstack(func() { ... }) around the critical section,
where appropriate. It also documents the invariant on (*mheap).lock
directly and updates repetitive documentation to refer to that comment.
Fixes#32105.
Change-Id: I702b1290709c118b837389c78efde25c51a2cafb
Reviewed-on: https://go-review.googlesource.com/c/go/+/177857
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Austin Clements <austin@google.com>
This change cleans up references to MSpan, MCache, and MCentral in the
docs via a bunch of sed invocations to better reflect the Go names for
the equivalent structures (i.e. mspan, mcache, mcentral) and their
methods (i.e. MSpan_Sweep -> mspan.sweep).
Change-Id: Ie911ac975a24bd25200a273086dd835ab78b1711
Reviewed-on: https://go-review.googlesource.com/c/147557
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Currently, all mcaches are flushed during STW mark termination as a
root marking job. This is currently necessary because all spans must
be out of these caches before sweeping begins to avoid races with
allocation and to ensure the spans are in the state expected by
sweeping. We do it as a root marking job because mcache flushing is
somewhat expensive and O(GOMAXPROCS) and this parallelizes the work
across the Ps. However, it's also the last remaining root marking job
performed during mark termination.
This CL moves mcache flushing out of mark termination and performs it
lazily. We keep track of the last sweepgen at which each mcache was
flushed and as each P is woken from STW, it observes that its mcache
is out-of-date and flushes it.
The introduces a complication for spans cached in stale mcaches. These
may now be observed by background or proportional sweeping or when
attempting to add a finalizer, but aren't in a stable state. For
example, they are likely to be on the wrong mcentral list. To fix
this, this CL extends the sweepgen protocol to also capture whether a
span is cached and, if so, whether or not its cache is stale. This
protocol blocks asynchronous sweeping from touching cached spans and
makes it the responsibility of mcache flushing to sweep the flushed
spans.
This eliminates the last mark termination root marking job, which
means we can now eliminate that entire infrastructure.
Updates #26903. This implements lazy mcache flushing.
Change-Id: Iadda7aabe540b2026cffc5195da7be37d5b4125e
Reviewed-on: https://go-review.googlesource.com/c/134783
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Rick Hudson <rlh@golang.org>
mcache.refill acquires g.m.locks, which is pointless because the
caller itself absolutely must have done so already to prevent
ownership of mcache from shifting.
Also, mcache.refill's documentation is generally a bit out-of-date, so
this cleans this up.
Change-Id: Idc8de666fcaf3c3d96006bd23a8f307539587d6c
Reviewed-on: https://go-review.googlesource.com/138195
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Rick Hudson <rlh@golang.org>
These functions all serve essentially the same purpose. mlookup is
used in only one place and findObject in only three. Use
heapBitsForObject instead, which is the most optimized implementation.
(This may seem slightly silly because none of these uses care about
the heap bits, but we're about to split up the functionality of
heapBitsForObject anyway. At that point, findObject will rise from the
ashes.)
Change-Id: I906468c972be095dd23cf2404a7d4434e802f250
Reviewed-on: https://go-review.googlesource.com/85877
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Rick Hudson <rlh@golang.org>
These have never had a use - not even going back to when they were added
in C.
Change-Id: I143b6902b3bacb1fa83c56c9070a8adb9f61a844
Reviewed-on: https://go-review.googlesource.com/69119
Reviewed-by: Dave Cheney <dave@cheney.net>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Dave Cheney <dave@cheney.net>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Currently, we mix objects with pointers and objects without pointers
("noscan" objects) together in memory. As a result, for every object
we grey, we have to check that object's heap bits to find out if it's
noscan, which adds to the per-object cost of GC. This also hurts the
TLB footprint of the garbage collector because it decreases the
density of scannable objects at the page level.
This commit improves the situation by using separate spans for noscan
objects. This will allow a much simpler noscan check (in a follow up
CL), eliminate the need to clear the bitmap of noscan objects (in a
follow up CL), and improves TLB footprint by increasing the density of
scannable objects.
This is also a step toward eliminating dead bits, since the current
noscan check depends on checking the dead bit of the first word.
This has no effect on the heap size of the garbage benchmark.
We'll measure the performance change of this after the follow-up
optimizations.
This is a cherry-pick from dev.garbage commit d491e550c3. The only
non-trivial merge conflict was in updatememstats in mstats.go, where
we now have to separate the per-spanclass stats from the per-sizeclass
stats.
Change-Id: I13bdc4869538ece5649a8d2a41c6605371618e40
Reviewed-on: https://go-review.googlesource.com/41251
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Rick Hudson <rlh@golang.org>
Currently fixalloc does not zero memory it reuses. This is dangerous
with the hybrid barrier if the type may contain heap pointers, since
it may cause us to observe a dead heap pointer on reuse. It's also
error-prone since it's the only allocator that doesn't zero on
allocation (mallocgc of course zeroes, but so do persistentalloc and
sysAlloc). It's also largely pointless: for mcache, the caller
immediately memclrs the allocation; and the two specials types are
tiny so there's no real cost to zeroing them.
Change fixalloc to zero allocations by default.
The only type we don't zero by default is mspan. This actually
requires that the spsn's sweepgen survive across freeing and
reallocating a span. If we were to zero it, the following race would
be possible:
1. The current sweepgen is 2. Span s is on the unswept list.
2. Direct sweeping sweeps span s, finds it's all free, and releases s
to the fixalloc.
3. Thread 1 allocates s from fixalloc. Suppose this zeros s, including
s.sweepgen.
4. Thread 1 calls s.init, which sets s.state to _MSpanDead.
5. On thread 2, background sweeping comes across span s in allspans
and cas's s.sweepgen from 0 (sg-2) to 1 (sg-1). Now it thinks it
owns it for sweeping. 6. Thread 1 continues initializing s.
Everything breaks.
I would like to fix this because it's obviously confusing, but it's a
subtle enough problem that I'm leaving it alone for now. The solution
may be to skip sweepgen 0, but then we have to think about wrap-around
much more carefully.
Updates #17503.
Change-Id: Ie08691feed3abbb06a31381b94beb0a2e36a0613
Reviewed-on: https://go-review.googlesource.com/31368
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Rick Hudson <rlh@golang.org>
This covers basically all sysAlloc'd, persistentalloc'd, and
fixalloc'd types.
Change-Id: I0487c887c2a0ade5e33d4c4c12d837e97468e66b
Reviewed-on: https://go-review.googlesource.com/30941
Reviewed-by: Rick Hudson <rlh@golang.org>
This is a renaming of the field ref to the
more appropriate allocCount. The field
holds the number of objects in the span
that are currently allocated. Some throws
strings were adjusted to more accurately
convey the meaning of allocCount.
Change-Id: I10daf44e3e9cc24a10912638c7de3c1984ef8efe
Reviewed-on: https://go-review.googlesource.com/19518
Reviewed-by: Austin Clements <austin@google.com>
Instead of building a freelist from the mark bits generated
by the GC this CL allocates directly from the mark bits.
The approach moves the mark bits from the pointer/no pointer
heap structures into their own per span data structures. The
mark/allocation vectors consist of a single mark bit per
object. Two vectors are maintained, one for allocation and
one for the GC's mark phase. During the GC cycle's sweep
phase the interpretation of the vectors is swapped. The
mark vector becomes the allocation vector and the old
allocation vector is cleared and becomes the mark vector that
the next GC cycle will use.
Marked entries in the allocation vector indicate that the
object is not free. Each allocation vector maintains a boundary
between areas of the span already allocated from and areas
not yet allocated from. As objects are allocated this boundary
is moved until it reaches the end of the span. At this point
further allocations will be done from another span.
Since we no longer sweep a span inspecting each freed object
the responsibility for maintaining pointer/scalar bits in
the heapBitMap containing is now the responsibility of the
the routines doing the actual allocation.
This CL is functionally complete and ready for performance
tuning.
Change-Id: I336e0fc21eef1066e0b68c7067cc71b9f3d50e04
Reviewed-on: https://go-review.googlesource.com/19470
Reviewed-by: Austin Clements <austin@google.com>
The tree's pretty inconsistent about single space vs double space
after a period in documentation. Make it consistently a single space,
per earlier decisions. This means contributors won't be confused by
misleading precedence.
This CL doesn't use go/doc to parse. It only addresses // comments.
It was generated with:
$ perl -i -npe 's,^(\s*// .+[a-z]\.) +([A-Z]),$1 $2,' $(git grep -l -E '^\s*//(.+\.) +([A-Z])')
$ go test go/doc -update
Change-Id: Iccdb99c37c797ef1f804a94b22ba5ee4b500c4f7
Reviewed-on: https://go-review.googlesource.com/20022
Reviewed-by: Rob Pike <r@golang.org>
Reviewed-by: Dave Day <djd@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Currently, we update memstats.heap_live from mcache.local_cachealloc
whenever we lock the heap (e.g., to obtain a fresh span or to release
an unused span). However, under the right circumstances,
local_cachealloc can accumulate allocations up to the size of
the *entire heap* without flushing them to heap_live. Specifically,
since span allocations from an mcentral don't lock the heap, if a
large number of pages are held in an mcentral and the application
continues to use and free objects of that size class (e.g., the
BinaryTree17 benchmark), local_cachealloc won't be flushed until the
mcentral runs out of spans.
This is a problem because, unlike many of the memory statistics that
are purely informative, heap_live is used to determine when the
garbage collector should start and how hard it should work.
This commit eliminates local_cachealloc, instead atomically updating
heap_live directly. To control contention, we do this only when
obtaining a span from an mcentral. Furthermore, we make heap_live
conservative: allocating a span assumes that all free slots in that
span will be used and accounts for these when the span is
allocated, *before* the objects themselves are. This is important
because 1) this triggers the GC earlier than necessary rather than
potentially too late and 2) this leads to a conservative GC rate
rather than a GC rate that is potentially too low.
Alternatively, we could have flushed local_cachealloc when it passed
some threshold, but this would require determining a threshold and
would cause heap_live to underestimate the true value rather than
overestimate.
Fixes#12199.
name old time/op new time/op delta
BinaryTree17-12 2.88s ± 4% 2.88s ± 1% ~ (p=0.470 n=19+19)
Fannkuch11-12 2.48s ± 1% 2.48s ± 1% ~ (p=0.243 n=16+19)
FmtFprintfEmpty-12 50.9ns ± 2% 50.7ns ± 1% ~ (p=0.238 n=15+14)
FmtFprintfString-12 175ns ± 1% 171ns ± 1% -2.48% (p=0.000 n=18+18)
FmtFprintfInt-12 159ns ± 1% 158ns ± 1% -0.78% (p=0.000 n=19+18)
FmtFprintfIntInt-12 270ns ± 1% 265ns ± 2% -1.67% (p=0.000 n=18+18)
FmtFprintfPrefixedInt-12 235ns ± 1% 234ns ± 0% ~ (p=0.362 n=18+19)
FmtFprintfFloat-12 309ns ± 1% 308ns ± 1% -0.41% (p=0.001 n=18+19)
FmtManyArgs-12 1.10µs ± 1% 1.08µs ± 0% -1.96% (p=0.000 n=19+18)
GobDecode-12 7.81ms ± 1% 7.80ms ± 1% ~ (p=0.425 n=18+19)
GobEncode-12 6.53ms ± 1% 6.53ms ± 1% ~ (p=0.817 n=19+19)
Gzip-12 312ms ± 1% 312ms ± 2% ~ (p=0.967 n=19+20)
Gunzip-12 42.0ms ± 1% 41.9ms ± 1% ~ (p=0.172 n=19+19)
HTTPClientServer-12 63.7µs ± 1% 63.8µs ± 1% ~ (p=0.639 n=19+19)
JSONEncode-12 16.4ms ± 1% 16.4ms ± 1% ~ (p=0.954 n=19+19)
JSONDecode-12 58.5ms ± 1% 57.8ms ± 1% -1.27% (p=0.000 n=18+19)
Mandelbrot200-12 3.86ms ± 1% 3.88ms ± 0% +0.44% (p=0.000 n=18+18)
GoParse-12 3.67ms ± 2% 3.66ms ± 1% -0.52% (p=0.001 n=18+19)
RegexpMatchEasy0_32-12 100ns ± 1% 100ns ± 0% ~ (p=0.257 n=19+18)
RegexpMatchEasy0_1K-12 347ns ± 1% 347ns ± 1% ~ (p=0.527 n=18+18)
RegexpMatchEasy1_32-12 83.7ns ± 2% 83.1ns ± 2% ~ (p=0.096 n=18+19)
RegexpMatchEasy1_1K-12 509ns ± 1% 505ns ± 1% -0.75% (p=0.000 n=18+19)
RegexpMatchMedium_32-12 130ns ± 2% 129ns ± 1% ~ (p=0.962 n=20+20)
RegexpMatchMedium_1K-12 39.5µs ± 2% 39.4µs ± 1% ~ (p=0.376 n=20+19)
RegexpMatchHard_32-12 2.04µs ± 0% 2.04µs ± 1% ~ (p=0.195 n=18+17)
RegexpMatchHard_1K-12 61.4µs ± 1% 61.4µs ± 1% ~ (p=0.885 n=19+19)
Revcomp-12 540ms ± 2% 542ms ± 4% ~ (p=0.552 n=19+17)
Template-12 69.6ms ± 1% 71.2ms ± 1% +2.39% (p=0.000 n=20+20)
TimeParse-12 357ns ± 1% 357ns ± 1% ~ (p=0.883 n=18+20)
TimeFormat-12 379ns ± 1% 362ns ± 1% -4.53% (p=0.000 n=18+19)
[Geo mean] 62.0µs 61.8µs -0.44%
name old time/op new time/op delta
XBenchGarbage-12 5.89ms ± 2% 5.81ms ± 2% -1.41% (p=0.000 n=19+18)
Change-Id: I96b31cca6ae77c30693a891cff3fe663fa2447a0
Reviewed-on: https://go-review.googlesource.com/17748
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
