Motivation
This is a "simpler" alternative to #9416 which fixes the same
CompositeByteBuf bugs described there, originally reported by @jingene
in #9398.
Modifications
- Add fields to Component class for the original buffer along with its
adjustment, which may be different to the already-stored unwrapped
buffer. Use it in appropriate places to ensure correctness and
equivalent behaviour to that prior to the earlier optimizations
- Add comments explaining purpose of each of the Component fields
- Unwrap more kinds of buffers in newComponent method to extend scope of
the existing indirection-reduction optimization
- De-duplicate common buffer consolidation logic
- Unit test for the original bug provided by @jingene
Result
- Correct behaviour / fixed bugs
- Some code deduplication / simplification
- Unwrapping optimization applied to more types of buffers
The downside is increased mem footprint from the two new fields, and
additional allocations in some specific cases, though those should be
rare.
Co-authored-by: jingene <jingene0206@gmail.com>
Motivation:
The Netty classes are initialized at build time by default for GraalVM Native Image compilation. This is configured via the `--initialize-at-build-time=io.netty` option. While this reduces start-up time it can lead to some problems:
- The class initializer of `io.netty.buffer.PooledByteBufAllocator` looks at the maximum memory size to compute the size of internal buffers. If the class initializer runs during image generation, then the buffers are sized according to the very large heap size that the image generator uses, and Netty allocates several arrays that are 16 MByte. The fix is to initialize the following 3 classes at run time: `io.netty.buffer.PooledByteBufAllocator,io.netty.buffer.ByteBufAllocator,io.netty.buffer.ByteBufUtil`. This fix was dependent on a GraalVM Native Image fix that was included in 19.2.0.
- The class initializer of `io.netty.handler.ssl.util.ThreadLocalInsecureRandom` needs to be initialized at runtime to ensure that the generated values are trully random and not fixed for each generated image.
- The class initializers of `io.netty.buffer.AbstractReferenceCountedByteBuf` and `io.netty.util.AbstractReferenceCounted` compute field offsets. While the field offset recomputation is necessary for correct execution as a native image these initializers also have logic that depends on the presence/absence of `sun.misc.Unsafe`, e.g., via the `-Dio.netty.noUnsafe=true` flag. The fix is to push these initializers to runtime so that the field offset lookups (and the logic depending on them) run at run time. This way no manual substitutions are necessary either.
Modifications:
Add `META-INF/native-image` configuration files that correctly trigger the inialization of the above classes at run time via `--initialize-at-run-time=...` flags.
Result:
Fixes the initialisation issues described above for Netty executables built with GraalVM.
Motivation:
AbstractByteBuf.indexOf(...) currently delegates to ByteBufUtils.indexOf(...) which will create a new ByteBufProcessor on each call. This is done to reduce overhead of bounds-checks. Unfortunally while this reduces bounds checks it produces a lot of GC. We can just implement our own version in AbstractByteBuf which makes use of _getByte(...) and so does no bound checks as well but also not need to create any garbage.
Modifications:
Write optimized implementation of indexOf(...) for AbstractByteBuf
Result:
Fixes https://github.com/netty/netty/issues/9499.
Motivation:
If all we need is the FileChannel we should better use RandomAccessFile as FileInputStream and FileOutputStream use a finalizer.
Modifications:
Replace FileInputStream and FileOutputStream with RandomAccessFile when possible.
Result:
Fixes https://github.com/netty/netty/issues/8078.
Motivation
#1802 fixed ByteBuf implementations to ensure that the whole buffer
region is preserved when capacity is increased, not just the readable
part. The behaviour is still different however when the capacity is
_decreased_ - data outside the currently-readable region is zeroed.
Modifications
Update ByteBuf capacity(int) implementations to also copy the whole
buffer region when the new capacity is less than the current capacity.
Result
Consistent behaviour of ByteBuf#capacity(int) regardless of whether the
new capacity is greater than or less than the current capacity.
Motivation
Underlying array allocations in UnpooledHeapByteBuf are intended be done
via the protected allocateArray(int) method, so that they can be tracked
and/or overridden by subclasses, for example
UnpooledByteBufAllocator$InstrumentedUnpooledHeapByteBuf or #8015. But
it looks like an explicit allocation was missed in the copy(int,int)
method.
Modification
Just use alloc().heapBuffer(...) for the allocation
Result
No possibility of "missing" array allocations when ByteBuf#copy is used.
Motivation:
#9224 introduced overrides of ByteBufUtil#writeUtf8(...) and related
methods to operate on a sub-CharSequence directly to save having to
allocate substrings, but it missed an edge case where the subsequence
does not extend to the end of the CharSequence and the last char in the
sequence is a high surrogate.
Due to the catch-IndexOutOfBoundsException optimization that avoids an
additional bounds check, it would be possible to read past the specified
end char index and successfully decode a surrogate pair which would
otherwise result in a '?' byte being written.
Modifications:
- Check for end-of-subsequence before reading next char after a high
surrogate is encountered in the
writeUtf8(AbstractByteBuf,int,CharSequence,int,int) and
utf8BytesNonAscii methods
- Add unit test for this edge case
Result:
Bug is fixed.
This removes the bounds-check-avoidance optimization but it does not
appear to have a measurable impact on benchmark results, including when
the char sequence contains many surrogate pairs (which should be rare in
any case).
* Correctly take length of ByteBufInputStream into account for readLine() / readByte()
Motivation:
ByteBufInputStream did not correctly take the length into account when validate bounds for readLine() / readByte() which could lead to read more then allowed.
Modifications:
- Correctly take length into account
- Add unit tests
- Fix existing unit test
Result:
Correctly take length of ByteBufInputStream into account.
Related to https://github.com/netty/netty/pull/9306.
Motivation:
buffer.isReadable() should not be used to limit the amount of data that can be read as the amount may be less then was is readable.
Modification:
- Use available() which takes the length into account
- Add unit test
Result:
Fixes https://github.com/netty/netty/issues/9305
Motivation:
Fix the issue of incorrectly calculating the number of dump rows when using prettyHexDumpmethod in ByteBufUtil. The way to find the remainder is either length % 16 or length & 15
Modification:
Fixed the way to calculate the remainder
Result:
Fixed#9301
Motivation:
Some methods that either override others or are implemented as part of implementation an interface did miss the `@Override` annotation
Modifications:
Add missing `@Override`s
Result:
Code cleanup
Motivation:
At the moment EmptyByteBuf.getCharSequence(0,...) will return null while it must return a "".
Modifications:
- Let EmptyByteBuf.getCharSequence(0,...) return ""
- Add unit test
Result:
Fixes https://github.com/netty/netty/issues/9271.
Motivation
It would be useful to be able to write UTF-8 encoded subsequence of
CharSequence characters to a ByteBuf without needing to create a
temporary object via CharSequence#subSequence().
Modification
Add overloads of ByteBufUtil writeUtf8, reserveAndWriteUtf8 and
utf8Bytes methods which take explicit subsequence bounds.
Result
More efficient writing of substrings to byte buffers possible
Motivation
There's quite a lot of duplicate/equivalent logic across the various
concrete ByteBuf implementations. We could take this even further but
for now I've focused on the PooledByteBuf sub-hierarchy.
Modifications
- Move common logic/methods into existing PooledByteBuf abstract
superclass
- Shorten PooledByteBuf.capacity(int) method implementation
Result
Less code to maintain
Motivation
While digging around looking at something else I noticed that these
share a lot of logic and it would be nice to reduce that duplication.
Modifications
Have UnpooledUnsafeDirectByteBuf extend UnpooledDirectByteBuf and make
adjustments to ensure existing behaviour remains unchanged.
The most significant addition needed to UnpooledUnsafeDirectByteBuf was
re-overriding the getPrimitive/setPrimitive methods to revert back to
the AbstractByteBuf versions which include bounds checks
(UnpooledDirectByteBuf excludes these as an optimization, relying on
those done by underlying ByteBuffer).
Result
~200 fewer lines, less duplicate logic.
Motivation:
1. Users will be able to use an optimized version of
`UnpooledHeapByteBuf` and override behavior of methods if required.
2. Consistency with `UnpooledDirectByteBuf`, `UnpooledHeapByteBuf`, and
`UnpooledUnsafeDirectByteBuf`.
Modifications:
- Add `public` access modifier to `UnpooledUnsafeHeapByteBuf` class and
ctor;
Result:
Public access for optimized version of `UnpooledHeapByteBuf`.
Motivation
There are a few minor inconsistencies / redundant operations in the
ByteBuf implementations which would be good to fix.
Modifications
- Unnecessary ByteBuffer.duplicate() performed in
CompositeByteBuf.nioBuffer(int,int)
- Add missing checkIndex(...) check to
ReadOnlyByteBufferBuf.nioBuffer(int,int)
- Remove duplicate bounds check in
ReadOnlyByteBufferBuf.getBytes(int,byte[],int,int)
- Omit redundant bounds check in
UnpooledHeapByteBuf.getBytes(int,ByteBuffer)
Result
More consistency and slightly less overhead
Motivation
ByteBuf capacity is automatically increased as needed up to maxCapacity
when writing beyond the buffer's current capacity. However there's no
way to tell in general whether such an increase will result in a
relatively costly internal buffer re-allocation.
For unpooled buffers it always does, in pooled cases it depends on the
size of the associated chunk of allocated memory, which I don't think is
currently exposed in any way.
It would sometimes be useful to know where this limit is when making
external decisions about whether to reuse or preemptively reallocate.
It would also be advantageous to take this limit into account when
auto-increasing the capacity during writes, to defer such reallocation
until really necessary.
Modifications
Introduce new AbstractByteBuf.maxFastWritableBytes() method which will
return a value >= writableBytes() and <= maxWritableBytes().
Make use of the new method in the sizing decision made by the
AbstractByteBuf.ensureWritable(...) methods.
Result
Less reallocation/copying.
Motivation
A small thread-safety bug was introduced during the internal
optimizations of ComponentByteBuf made a while back in #8437. When there
is a single component which was added as a slice,
internalNioBuffer(int,int) will currently return the unwrapped slice's
un-duplicated internal NIO buffer. This is not safe since it could be
modified concurrently with other usage of that parent buffer.
Modifications
Delegate internalNioBuffer to nioBuffer in this case, which returns a
duplicate. This matches what's done in derived buffers in general
(for the same reason). Add unit test.
Result
Fixed possible thread-safety bug
Motivation
Direct buffers are normally preferred when interfacing with raw
sockets. Currently netty will only return direct io buffers (for reading
from a channel) when a platform has unsafe. However, this is
inconsistent with the write-side (filterOutboundMessage) where a direct
byte buffer will be returned if pooling is enabled. This means that
environments without unsafe (and no manual netty configurations) end up
with many pooled heap byte buffers for reading, many pooled direct byte
buffers for writing, and jdk pooled byte buffers (for reading).
Modifications
This commit modifies the AbstractByteBufAllocator to return a direct
byte buffer for io handling when the platform has unsafe or direct byte
buffers are pooled.
Result:
Use direct buffers when direct buffers are pooled for IO.
* Fix incorrect behavior of ReadOnlyByteBufferBuf.getBytes(int,ByteBuffer)
Motivation
It currently will succeed when the destination is larger than the source
range, but the ByteBuf javadoc states this should be a failure, as is
the case with all the other implementations.
Modifications
- Fix logic to fail the bounds check in this case
- Remove explicit null check which isn't done in any equivalent method
- Add unit test
Result
More correct/consistent behaviour
Motivation:
1f93bd3 introduced a regression that could lead to not have the lastAccessed field correctly null'ed out when the endOffset of the internal Component == CompositeByteBuf.readerIndex()
Modifications:
- Correctly null out the lastAccessed field in any case
- Add unit tests
Result:
Fixes regression in CompositeByteBuf.discard*ReadBytes()
Motivation
AbstractReferenceCounted and AbstractReferenceCountedByteBuf contain
duplicate logic for managing the volatile refcount in an optimized and
consistent manner, which increased in complexity in #8583. It's possible
to extract this into a common helper class now that all access is via an
AtomicIntegerFieldUpdater.
Modifications
- Move duplicate logic into a shared ReferenceCountUpdater class
- Incorporate some additional simplification for the most common single
increment/decrement cases (fewer checks/operations)
Result
Less code duplication, better encapsulation of the "non-trivial"
internal volatile refcount manipulation
Motivation:
The CompositeByteBuf discardReadBytes / discardReadComponents methods are currently quite inefficient, including when there are no read components to discard. We would like to call the latter more frequently in ByteToMessageDecoder#COMPOSITE_CUMULATOR.
In the same context it would be beneficial to perform a "shallow copy" of a composite buffer (for example when it has a refcount > 1) to avoid having to allocate and copy the contained bytes just to obtain an "independent" cumulation.
Modifications:
- Optimize discardReadBytes() and discardReadComponents() implementations (start at first comp rather than performing a binary search for the readerIndex).
- New addFlattenedComponents(boolean,ByteBuf) method which performs a shallow copy if the provided buffer is also composite and avoids adding any empty buffers, plus unit test.
- Other minor optimizations to avoid unnecessary checks.
Results:
discardReadXX methods are faster, composite buffers can be easily appended without deepening the buffer "tree" or retaining unused components.
Motivation:
We synchronize on the chunk.arena when produce the String returned by PoolSubpage.toString() which may raise a NPE when chunk == null. Chunk == null for the head of the linked-list and so a NPE may raised by a debugger. This NPE can never happen in real code tho as we never access toString() of the head.
Modifications:
Add null checks and so fix the possible NPE
Result:
No NPE when using a debugger and inspect the PooledByteBufAllocator.
Motivation:
In MemoryRegionCache.Entry we use the Recycler to reduce GC pressure and churn. The problem is that these will also be recycled when the PoolThreadCache is collected and finalize() is called. This then can have the effect that we try to load class but the WebApp is already stoped.
This will produce an stacktrace like this on Tomcat:
```
19-Mar-2019 15:53:21.351 INFO [Finalizer] org.apache.catalina.loader.WebappClassLoaderBase.checkStateForResourceLoading Illegal access: this web application instance has been stopped already. Could not load [java.util.WeakHashMap]. The following stack trace is thrown for debugging purposes as well as to attempt to terminate the thread which caused the illegal access.
java.lang.IllegalStateException: Illegal access: this web application instance has been stopped already. Could not load [java.util.WeakHashMap]. The following stack trace is thrown for debugging purposes as well as to attempt to terminate the thread which caused the illegal access.
at org.apache.catalina.loader.WebappClassLoaderBase.checkStateForResourceLoading(WebappClassLoaderBase.java:1383)
at org.apache.catalina.loader.WebappClassLoaderBase.checkStateForClassLoading(WebappClassLoaderBase.java:1371)
at org.apache.catalina.loader.WebappClassLoaderBase.loadClass(WebappClassLoaderBase.java:1224)
at org.apache.catalina.loader.WebappClassLoaderBase.loadClass(WebappClassLoaderBase.java:1186)
at io.netty.util.Recycler$3.initialValue(Recycler.java:233)
at io.netty.util.Recycler$3.initialValue(Recycler.java:230)
at io.netty.util.concurrent.FastThreadLocal.initialize(FastThreadLocal.java:188)
at io.netty.util.concurrent.FastThreadLocal.get(FastThreadLocal.java:142)
at io.netty.util.Recycler$Stack.pushLater(Recycler.java:624)
at io.netty.util.Recycler$Stack.push(Recycler.java:597)
at io.netty.util.Recycler$DefaultHandle.recycle(Recycler.java:225)
at io.netty.buffer.PoolThreadCache$MemoryRegionCache$Entry.recycle(PoolThreadCache.java:478)
at io.netty.buffer.PoolThreadCache$MemoryRegionCache.freeEntry(PoolThreadCache.java:459)
at io.netty.buffer.PoolThreadCache$MemoryRegionCache.free(PoolThreadCache.java:430)
at io.netty.buffer.PoolThreadCache$MemoryRegionCache.free(PoolThreadCache.java:422)
at io.netty.buffer.PoolThreadCache.free(PoolThreadCache.java:279)
at io.netty.buffer.PoolThreadCache.free(PoolThreadCache.java:270)
at io.netty.buffer.PoolThreadCache.free(PoolThreadCache.java:241)
at io.netty.buffer.PoolThreadCache.finalize(PoolThreadCache.java:230)
at java.lang.System$2.invokeFinalize(System.java:1270)
at java.lang.ref.Finalizer.runFinalizer(Finalizer.java:102)
at java.lang.ref.Finalizer.access$100(Finalizer.java:34)
at java.lang.ref.Finalizer$FinalizerThread.run(Finalizer.java:217)
```
Beside this we also need to ensure we not try to lazy load SizeClass when the finalizer is used as it may not be present anymore if the ClassLoader is already destroyed.
This would produce an error like:
```
20-Mar-2019 11:26:35.254 INFO [Finalizer] org.apache.catalina.loader.WebappClassLoaderBase.checkStateForResourceLoading Illegal access: this web application instance has been stopped already. Could not load [io.netty.buffer.PoolArena$1]. The following stack trace is thrown for debugging purposes as well as to attempt to terminate the thread which caused the illegal access.
java.lang.IllegalStateException: Illegal access: this web application instance has been stopped already. Could not load [io.netty.buffer.PoolArena$1]. The following stack trace is thrown for debugging purposes as well as to attempt to terminate the thread which caused the illegal access.
at org.apache.catalina.loader.WebappClassLoaderBase.checkStateForResourceLoading(WebappClassLoaderBase.java:1383)
at org.apache.catalina.loader.WebappClassLoaderBase.checkStateForClassLoading(WebappClassLoaderBase.java:1371)
at org.apache.catalina.loader.WebappClassLoaderBase.loadClass(WebappClassLoaderBase.java:1224)
at org.apache.catalina.loader.WebappClassLoaderBase.loadClass(WebappClassLoaderBase.java:1186)
at io.netty.buffer.PoolArena.freeChunk(PoolArena.java:287)
at io.netty.buffer.PoolThreadCache$MemoryRegionCache.freeEntry(PoolThreadCache.java:464)
at io.netty.buffer.PoolThreadCache$MemoryRegionCache.free(PoolThreadCache.java:429)
at io.netty.buffer.PoolThreadCache$MemoryRegionCache.free(PoolThreadCache.java:421)
at io.netty.buffer.PoolThreadCache.free(PoolThreadCache.java:278)
at io.netty.buffer.PoolThreadCache.free(PoolThreadCache.java:269)
at io.netty.buffer.PoolThreadCache.free(PoolThreadCache.java:240)
at io.netty.buffer.PoolThreadCache.finalize(PoolThreadCache.java:229)
at java.lang.System$2.invokeFinalize(System.java:1270)
at java.lang.ref.Finalizer.runFinalizer(Finalizer.java:102)
at java.lang.ref.Finalizer.access$100(Finalizer.java:34)
at java.lang.ref.Finalizer$FinalizerThread.run(Finalizer.java:217)
```
Modifications:
- Only try to put the Entry back into the Recycler if the PoolThredCache is not destroyed because of the finalizer.
- Only try to access SizeClass if not triggered by finalizer.
Result:
No IllegalStateException anymoe when a webapp is reloaded in Tomcat that uses netty and uses the PooledByteBufAllocator.
Motivation:
The special case fixed in #8497 also requires that we keep a derived slice when trimming components in place, as done by the capacity(int) and discardReadBytes() methods.
Modifications:
Ensure that we keep a ref to trimmed components' original retained slice in capacity(int) and discardReadBytes() methods, so that it is released properly when the they are later freed. Add unit test which fails prior to the fix.
Result:
Edge case leak is eliminated.
Motivation:
PooledByteBufAllocator uses a PoolThreadCache per Thread that allocates / deallocates to minimize the performance overhead. This PoolThreadCache is trimmed after X allocations to free up buffers that are not allocated for a long time. This works out quite well when the app continues to allocate but fails if the app stops to allocate frequently (for whatever reason) and so a lot of memory is wasted and not given back to the arena / freed.
Modifications:
- Add a ThreadExecutorMap that offers multiple methods that wrap Runnable / ThreadFactory / Executor and allow to call ThreadExecutorMap.currentEventExecutor() to get the current executing EventExecutor for the calling Thread.
- Use these methods in the constructors of our EventExecutor implementations (which also covers the EventLoop implementations)
- Add io.netty.allocator.cacheTrimIntervalMillis system property which can be used to specify a fixed rate / interval on which we should try to trim the PoolThreadCache for a EventExecutor that allocates.
- Add PooledByteBufAllocator.trimCurrentThreadCache() to allow the user to trim the cache of the calling thread manually.
- Add testcases
- Introduce FastThreadLocal.getIfExists()
Result:
Allow to better / more frequently trim PoolThreadCache and so give back memory to the area / system.
Motivation:
We should run a CI job using J9 to ensure netty also works when using different JVMs.
Modifications:
- Adjust PooledByteBufAllocatorTest to be able to complete faster when using a JVM which takes longer when joining Threads (this seems to be the case with J9).
- Skip UDT tests on J9 as UDT is not supported there.
Result:
Be able to run CI against J9.
Motivation
There's some miscellaneous cleanup/simplification of CompositeByteBuf
which would help make the code a bit clearer.
Modifications
- Simplify web of constructors and addComponents methods, reducing
duplication of logic
- Rename `Component.freeIfNecessary()` method to just `free()`, which is
less confusing (see #8641)
- Make loop in addComponents0(...) method more verbose/readable (see
https://github.com/netty/netty/pull/8437#discussion_r232124414)
- Simplify addition/subtraction in setBytes(...) methods
Result
Smaller/clearer code
Motivation:
We have a utility method to check for > 0 and >0 arguments. We should use it.
Modification:
use checkPositive/checkPositiveOrZero instead of if statement.
Result:
Re-use utility method.
Motivation:
We need to update to a new checkstyle plugin to allow the usage of lambdas.
Modifications:
- Update to new plugin version.
- Fix checkstyle problems.
Result:
Be able to use checkstyle plugin which supports new Java syntax.
Motivation
In #8758, @doom369 reported an infinite loop bug in CompositeByteBuf
which was introduced in #8437.
This is the same small fix for that, along with fixes for two other bugs
found while re-inspecting the changes and adding unit tests.
Modification
- Replace recursive call to toComponentIndex with toComponentIndex0 as
intended
- Add missed "lastAccessed" racy cache invalidation in capacity(int)
method
- Fix incorrect determination of initial offset in non-zero cIndex case
of updateComponentOffsets method
- New unit tests for previously uncovered methods
Results
Fewer bugs.
Motivation:
The javadocs stating `IndexOutOfBoundsException` is thrown were
different from what `ByteBuf` actually did. We want to ensure the
Javadocs represent reality.
Modifications:
Updated javadocs on `write*`, `ensureWriteable`, `capacity`, and
`maxCapacity` methods.
Results:
Javadocs more closely match actual behaviour.
Motivation:
In versions of Netty prior to 4.1.31.Final, a CompositeByteBuf could be
created with any size (including potentially nonsensical negative
values). This behavior changed in e7737b993, which introduced a bounds
check to only allow for a component size greater than one. This broke
some existing use cases that attempted to create a byte buf with a
single component.
Modifications:
Lower the bounds check on numComponents to include the single component
case, but still throw an exception for anything less than one.
Add unit tests for the case of numComponents being less than, equal to,
and greater than this lower bound.
Result:
Return to the behavior of 4.1.30.Final, allowing one component, but
still include an explicit check against a lower bound.
Note that while creating a CompositeByteBuf with a single component is
in some ways a contradiction of the term "composite", this patch caters
for existing uses while excluding the clearly nonsensical case of asking
for a CompositeByteBuf with zero or fewer components.
Fixes#8613.
Motivation:
Often a temporary ByteBuffer is used which can be cached to reduce the GC pressure.
Modifications:
Cache the ByteBuffer in the PoolThreadCache as well.
Result:
Less GC.
Motivation
#8563 highlighted race conditions introduced by the prior optimistic
update optimization in 83a19d5650. These
were known at the time but considered acceptable given the perf
benefit in high contention scenarios.
This PR proposes a modified approach which provides roughly half the
gains but stronger concurrency semantics. Race conditions still exist
but their scope is narrowed to much less likely cases (releases
coinciding with retain overflow), and even in those
cases certain guarantees are still assured. Once release() returns true,
all subsequent release/retains are guaranteed to throw, and in
particular deallocate will be called at most once.
Modifications
- Use even numbers internally (including -ve) for live refcounts
- "Final" release changes to odd number (equivalent to refcount 0)
- Retain still uses faster getAndAdd, release uses CAS loop
- First CAS attempt uses non-volatile read
- Thread.yield() after a failed CAS provides a net gain
Result
More (though not completely) robust concurrency semantics for ref
counting; increased latency under high contention, but still roughly
twice as fast as the original logic. Bench results to follow
Motivation:
ByteBuf is used everywhere so we should try hard to be able to make things inlinable. During benchmarks it showed that writeCharSequence(...) fails to inline writeUtf8 because it is too big even if its hots.
Modifications:
Move less common code-path to extra method to allow inlining.
Result:
Be able to inline writeUtf8 in most cases.
Motivation:
Often a temporary ByteBuffer is used which can be cached to reduce the GC pressure.
Modifications:
Add a Deque per PoolChunk which will be used for caching.
Result:
Less GC.
Motivation:
When we create new chunk with memory aligned, the offset of direct memory should be
'alignment - address & (alignment - 1)', not just 'address & (alignment - 1)'.
Modification:
Change offset calculating formula to offset = alignment - address & (alignment - 1) in PoolArena.DirectArena#offsetCacheLine and add a unit test to assert that.
Result:
Correctly calculate offset.
Motivation:
ByteBuf.retainedSlice() and similar methods produce sliced buffers with
an independent refcount to the buffer that they wrap.
One of the optimizations in 10539f4dc7 was
to use the ref to the unwrapped buffer object for added slices, but this
did not take into account the above special case when later releasing.
Thanks to @rkapsi for discovering this via #8495.
Modifications:
Since a reference to the slice is still kept in the Component class,
just changed Component.freeIfNecessary() to release the slice in
preference to the unwrapped buf.
Also added a unit test which reproduces the bug.
Result:
Fixes#8495
Motivation:
Two similar bugs were introduced by myself in separate recent PRs #8393
and #8464, while optimizing the assignment/handling of temporary arrays
in ByteBufUtil and UnsafeByteBufUtil.
The temp arrays allocated for buffering data written to an OutputStream
are incorrectly sized to the full length of the data to copy rather than
being capped at WRITE_CHUNK_SIZE.
Unfortunately one of these is in the 4.1.31.Final release, I'm really
sorry and will be more careful in future.
This kind of thing is tricky to cover in unit tests.
Modifications:
Revert the temp array allocations back to their original sizes.
Avoid making duplicate calls to ByteBuf.capacity() in a couple of places
in ByteBufUtil (unrelated thing I noticed, can remove it from this PR if
desired!)
Result:
Temporary byte arrays will be reverted to their originally intended
sizes.
Motivation:
#8388 introduced a reusable ThreadLocal<byte[]> for use in
decodeString(...). It can be used in more places in the buffer package
to avoid temporary allocations of small arrays.
Modifications:
Encapsulate use of the ThreadLocal in a static package-private
ByteBufUtil.threadLocalTempArray(int) method, and make use of it from a
handful of new places including ByteBufUtil.readBytes(...).
Result:
Fewer short-lived small byte array allocations.
Motivation:
CompositeByteBuf is a powerful and versatile abstraction, allowing for
manipulation of large data without copying bytes. There is still a
non-negligible cost to reading/writing however relative to "singular"
ByteBufs, and this can be mostly eliminated with some rework of the
internals.
My use case is message modification/transformation while zero-copy
proxying. For example replacing a string within a large message with one
of a different length
Modifications:
- No longer slice added buffers and unwrap added slices
- Components store target buf offset relative to position in
composite buf
- Less allocations, object footprint, pointer indirection, offset
arithmetic
- Use Component[] rather than ArrayList<Component>
- Avoid pointer indirection and duplicate bounds check, more
efficient backing array growth
- Facilitates optimization when doing bulk-inserts - inserting n
ByteBufs behind m is now O(m + n) instead of O(mn)
- Avoid unnecessary casting and method call indirection via superclass
- Eliminate some duplicate range/ref checks via non-checking versions of
toComponentIndex and findComponent
- Add simple fast-path for toComponentIndex(0); add racy cache of
last-accessed Component to findComponent(int)
- Override forEachByte0(...) and forEachByteDesc0(...) methods
- Make use of RecyclableArrayList in nioBuffers(int, int) (in line with
FasterCompositeByteBuf impl)
- Modify addComponents0(boolean,int,Iterable) to use the Iterable
directly rather than copy to an array first (and possibly to an
ArrayList before that)
- Optimize addComponents0(boolean,int,ByteBuf[],int) to not perform
repeated array insertions and avoid second loop for offset updates
- Simplify other logic in various places, in particular the general
pattern used where a sub-range is iterated over
- Add benchmarks to demonstrate some improvements
While refactoring I also came across a couple of clear bugs. They are
fixed in these changes but I will open another PR with unit tests and
fixes to the current version.
Result:
Much faster creation, manipulation, and access; many fewer allocations
and smaller footprint. Benchmark results to follow.
Motivation:
Unpooled.wrap(byte[]...) and Unpooled.wrap(ByteBuffer...) currently
allocate/copy an intermediate ByteBuf ArrayList and array, which can be
avoided.
Modifications:
- Define new internal ByteWrapper interface and add a CompositeByteBuf
constructor which takes a ByteWrapper with an array of the type that it
wraps, and modify the appropriate Unpooled.wrap(...) methods to take
advantage of it
- Tidy up other constructors in CompositeByteBuf to remove duplication
and misleading len arg (which is really an end offset into provided
array)
Result:
Less allocation/copying when wrapping byte[] and ByteBuffer arrays,
tidier code.
Motivation:
I came across two bugs:
- Components removed due to capacity reduction aren't released
- Offsets aren't set correctly on empty components that are added
between existing components
Modifications:
Add unit tests which expose these bugs, fix them.
Result:
Bugs are fixed
Motivation:
There are currently many more places where this could be used which were
possibly not considered when the method was added.
If https://github.com/netty/netty/pull/8388 is included in its current
form, a number of these places could additionally make use of the same
BYTE_ARRAYS threadlocal.
There's also a couple of adjacent places where an optimistically-pooled
heap buffer is used for temp byte storage which could use the
threadlocal too in preference to allocating a temp heap bytebuf wrapper.
For example
https://github.com/netty/netty/blob/4.1/buffer/src/main/java/io/netty/buffer/ByteBufUtil.java#L1417.
Modifications:
Replace new byte[] with PlatformDependent.allocateUninitializedArray()
where appropriate; make use of ByteBufUtil.getBytes() in some places
which currently perform the equivalent logic, including avoiding copy of
backing array if possible (although would be rare).
Result:
Further potential speed-up with java9+ and appropriate compile flags.
Many of these places could be on latency-sensitive code paths.
* Optimize AbstractByteBuf.getCharSequence() in US_ASCII case
Motivation:
Inspired by https://github.com/netty/netty/pull/8388, I noticed this
simple optimization to avoid char[] allocation (also suggested in a TODO
here).
Modifications:
Return an AsciiString from AbstractByteBuf.getCharSequence() if
requested charset is US_ASCII or ISO_8859_1 (latter thanks to
@Scottmitch's suggestion). Also tweak unit tests not to require Strings
and include a new benchmark to demonstrate the speedup.
Result:
Speed-up of AbstractByteBuf.getCharSequence() in ascii and iso 8859/1
cases
Motivation:
CompositeByteBuf.decompose(...) did not correctly slice the content and so produced an incorrect representation of the data.
Modifications:
- Rewrote implementation to fix bug and also improved it to reduce GC
- Add unit tests.
Result:
Fixes https://github.com/netty/netty/issues/8400.
Motivation:
While looking at the nice optimization done in
https://github.com/netty/netty/pull/8347 I couldn't help noticing the
logic could be simplified further. Apologies if this is just my OCD and
inappropriate!
Modifications:
Reduce amount of code used for ByteBufInputStream.readLine()
Result:
Slightly smaller and simpler code
Motivation:
Avoid creating any StringBuilder instance if
ByteBufInputStream::readLine isn't used
Modifications:
The StringBuilder instance is lazy allocated on demand and
are added new test case branches to address the increased
complexity of ByteBufInputStream::readLine
Result:
Reduced GC activity if ByteBufInputStream::readLine isn't used
Motivation:
We should just directly init the refCnt to 1 and not use the AtomicIntegerFieldUpdater.
Modifications:
Just assing directly to 1.
Result:
Cleaner code and possible a bit faster as the JVM / JIT may be able to optimize the first store easily.
Motiviation:
At the moment whenever ensureAccessible() is called in our ByteBuf implementations (which is basically on each operation) we will do a volatile read. That per-se is not such a bad thing but the problem here is that it will also reduce the the optimizations that the compiler / jit can do. For example as these are volatile it can not eliminate multiple loads of it when inline the methods of ByteBuf which happens quite frequently because most of them a quite small and very hot. That is especially true for all the methods that act on primitives.
It gets even worse as people often call a lot of these after each other in the same method or even use method chaining here.
The idea of the change is basically just ue a non-volatile read for the ensureAccessible() check as its a best-effort implementation to detect acting on already released buffers anyway as even with a volatile read it could happen that the user will release it in another thread before we actual access the buffer after the reference check.
Modifications:
- Try to do a non-volatile read using sun.misc.Unsafe if we can use it.
- Add a benchmark
Result:
Big performance win when multiple ByteBuf methods are called from a method.
With the change:
UnsafeByteBufBenchmark.setGetLongUnsafeByteBuf thrpt 20 281395842,128 ± 5050792,296 ops/s
Before the change:
UnsafeByteBufBenchmark.setGetLongUnsafeByteBuf thrpt 20 217419832,801 ± 5080579,030 ops/s
Motivation:
The JVM isn't always able to hoist out/reduce bounds checking (due to ref counting operations etc etc) hence making it configurable could improve performances for most CPU intensive use cases.
Modifications:
Each AbstractByteBuf bounds check has been tested against a new static final configuration property similar to checkAccessible ie io.netty.buffer.bytebuf.checkBounds.
Result:
Any user could disable ByteBuf bounds checking in order to get extra performances.
* Allow to use native transports when sun.misc.Unsafe is not present on the system
Motivation:
We should be able to use the native transports (epoll / kqueue) even when sun.misc.Unsafe is not present on the system. This is especially important as Java11 will be released soon and does not allow access to it by default.
Modifications:
- Correctly disable usage of sun.misc.Unsafe when -PnoUnsafe is used while running the build
- Correctly increment metric when UnpooledDirectByteBuf is allocated. This was uncovered once -PnoUnsafe usage was fixed.
- Implement fallbacks in all our native transport code for when sun.misc.Unsafe is not present.
Result:
Fixes https://github.com/netty/netty/issues/8229.
In nioBuffer(int,int) in CompositeByteBuf , we create a sub-array of nioBuffers for the components that are in range, then concatenate all the components in range into a single bigger buffer.
However, if the call to nioBuffers() returned only one sub-buffer, then we are copying it to a newly-allocated buffer "merged" for no reason.
Motivation:
Profiler for Spark shows a lot of time spent in put() method inside nioBuffer(), while usually no copy of data is required.
Modification:
This change skips this last step and just returns a duplicate of the single buffer returned by the call to nioBuffers(), which will in most implementation not copy the data
Result:
No copy when the source is only 1 buffer
Motivation:
We need to add special handling for WrappedCompositeByteBuf as these also extend AbstractByteBuf, otherwise we will not correctly adjust / read the writerIndex during processing.
Modifications:
- Add instanceof checks for WrappedCompositeByteBuf as well.
- Add testcases
Result:
Fixes https://github.com/netty/netty/issues/8152.
Motivation:
5b1fe611a6 introduced the usage of a finalizer as last resort for PoolThreadCache. As we may call free() from the FastThreadLocal.onRemoval(...) and finalize() we need to guard against multiple calls as otherwise we will corrupt internal state (that is used for metrics).
Modifications:
Use AtomicBoolean to guard against multiple calls of PoolThreadCache.free().
Result:
No more corruption of internal state caused by calling PoolThreadCache.free() multuple times.
Motivation:
Recent PR https://github.com/netty/netty/pull/8040 introduced
Unpooled.wrappedUnmodifiableBuffer(ByteBuf...) which has the same
behaviour but wraps the provided array directly. This is preferred for
most uses (including varargs-based use) and if there are any unusual
cases of an explicit array which is re-used before the ByteBuf is
finished with, it can just be copied first.
Modifications:
Added @Deprecated annotation and javadoc to
Unpooled.unmodifiableBuffer(ByteBuf...).
Result:
Unpooled.unmodifiableBuffer(ByteBuf...) will be deprecated.
Motivation:
ObjectCleaner does start a Thread to handle the cleaning of resources which leaks into the users application. We should not use it in netty itself to make things more predictable.
Modifications:
- Remove usage of ObjectCleaner and use finalize as a replacement when possible.
- Clarify javadocs for FastThreadLocal.onRemoval(...) to ensure its clear that remove() is not guaranteed to be called when the Thread completees and so this method is not enough to guarantee cleanup for this case.
Result:
Fixes https://github.com/netty/netty/issues/8017.
Motivation:
Unpooled.unmodifiableBuffer() is currently used to efficiently write
arrays of ByteBufs via FixedCompositeByteBuf, but involves an allocation
and content-copy of the provided ByteBuf array which in many (most?)
cases shouldn't be necessary.
Modifications:
Modify the internal FixedCompositeByteBuf class to support wrapping the
provided ByteBuf array directly. Control this behaviour with a
constructor flag and expose the "unsafe" version via a new
Unpooled.wrappedUnmodifiableBuffer(ByteBuf...) method.
Result:
Less garbage on IO paths. I would guess pretty much all existing usage
of unmodifiableBuffer() could use the copy-free version but assume it's
not safe to change its default behaviour.
Motivation:
Eliminate avoidable backing array reallocations when constructing
composite ByteBufs from existing buffer arrays/Iterables. This also
applies to the Unpooled.wrappedBuffer(...) methods.
Modifications:
Ensure the initial components ComponentList is sized at least as large
as the provided buffer array/Iterable in the CompositeByteBuffer
constructors.
In single-arg Unpooled.wrappedBuffer(...) methods, set maxNumComponents
to the count of provided buffers, rather than a fixed default of 16. It
seems likely that most usage of these involves wrapping a list without
subsequent modification, particularly since they return a ByteBuf rather
than CompositeByteBuf. If a different/larger max is required there are
already the wrappedBuffer(int, ...) variants.
In fact the current behaviour could be considered inconsistent - if you
call Unpooled.wrappedBuffer(int, ByteBuf) with a single buffer, you
might expect to subsequently be able to add buffers to it (since you
specified a max related to consolidation), but it will in fact return
just a slice of the provided ByteBuf.
Result:
Fewer and smaller allocations in some cases when using CompositeByteBufs
or Unpooled.wrappedBuffer(...).
