Summary: The idea behind wildcopy is that it can be cheaper to copy more bytes (say 8) than it is to copy less (say, 3). This change takes that further by exploiting some properties:
1. it's almost always OK to copy 16 bytes instead of 8, which means fewer copy instructions, and fewer branches
2. A 16 byte chunk size means that ~90% of wildcopy invocations will have a trip count of 1, so branch prediction will be improved.
Speedup on Xeon E5-2680v4 is in the range of 3-5%.
Measured wildcopy length distributions on silesia.tar:
level <=8 <=16 <=24 >24
1 78.05% 11.49% 3.52% 6.94%
3 82.14% 8.99% 2.44% 6.43%
6 85.81% 6.51% 2.92% 4.76%
8 83.02% 7.31% 3.64% 6.03%
10 84.13% 6.67% 3.29% 5.91%
15 77.58% 7.55% 5.21% 9.66%
16 80.07% 7.20% 3.98% 8.75%
Test Plan: benchmark silesia, make check
as suggested in #1441.
generally U32 and unsigned are the same thing,
except when they are not ...
case : 32-bit compilation for MIPS (uint32_t == unsigned long)
A vast majority of transformation consists in transforming U32 into unsigned.
In rare cases, it's the other way around (typically for internal code, such as seeds).
Among a few issues this patches solves :
- some parameters were declared with type `unsigned` in *.h,
but with type `U32` in their implementation *.c .
- some parameters have type unsigned*,
but the caller user a pointer to U32 instead.
These fixes are useful.
However, the bulk of changes is about %u formating,
which requires unsigned type,
but generally receives U32 values instead,
often just for brevity (U32 is shorter than unsigned).
These changes are generally minor, or even annoying.
As a consequence, the amount of code changed is larger than I would expect for such a patch.
Testing is also a pain :
it requires manually modifying `mem.h`,
in order to lie about `U32`
and force it to be an `unsigned long` typically.
On a 64-bit system, this will break the equivalence unsigned == U32.
Unfortunately, it will also break a few static_assert(), controlling structure sizes.
So it also requires modifying `debug.h` to make `static_assert()` a noop.
And then reverting these changes.
So it's inconvenient, and as a consequence,
this property is currently not checked during CI tests.
Therefore, these problems can emerge again in the future.
I wonder if it is worth ensuring proper distinction of U32 != unsigned in CI tests.
It's another restriction for coding, adding more frustration during merge tests,
since most platforms don't need this distinction (hence contributor will not see it),
and while this can matter in theory, the number of platforms impacted seems minimal.
Thoughts ?
answering #1407.
Also : removed obsolete function ZSTD_setDStreamParameter()
which could only be used with one parameter (DStream_p_maxWindowSize).
Now replaced by ZSTD_DCtx_setWindowSize() (which exists since a few revisions)
not sure why it only triggers now,
this code has been around for a while.
Introduced a new error code : dstBuffer_null,
I couldn't express anything even remotely similar with existing error codes set.
isolate all logic associated with block decompression
into its own module.
zstd_decompress is still in charge
of context creation/destruction,
frames, headers, streaming, special blocks, etc.
Compressed blocks themselves are now handled within zstd_decompress_block .
since corrupted bitstreams can generate too large values.
This slightly reduces the benefits from clang on my laptop.
gcc results and code generation are not affected.
while reviewing #1364,
I found a decompression speed improvement.
On my laptop, the new code decompresses +5-6% faster on clang
and +2-3% faster on gcc.
not bad for an accidental optimization...
We could undersize the literals buffer by up to 11 bytes,
due to a combination of 2 bugs:
* The literals buffer didn't have `WILDCOPY_OVERLENGTH` extra
space, like it is supposed to.
* We didn't check the literals buffer size in `ZSTD_sufficientBuff()`.
It's not necessary to ensure that no job is ongoing.
The pool is only expanded, existing threads are preserved.
In case of error, the only option is to return NULL and terminate the thread pool anyway.
There were 2 competing set of debug functions
within zstd_internal.h and bitstream.h.
They were mostly duplicate, and required care to avoid messing with each other.
There is now a single implementation, shared by both.
Significant change :
The macro variable ZSTD_DEBUG does no longer exist,
it has been replaced by DEBUGLEVEL,
which required modifying several source files.
removed "cached" structure.
prices are now saved in the optimal table.
Primarily done for simplification.
Might improve speed by a little.
But actually, and surprisingly, also improves ratio in some circumstances.
ensure that, when frequency[symbol]==0,
result is (tableLog + 1) bits
with both upper-bit and fractional-bit estimates.
Also : enable BIT_DEBUG in /tests
This edge case is only possible with the new optimal encoding selector,
since before zstd would always choose `set_basic` for small numbers of
sequences.
Fix `FSE_readNCount()` to support buffers < 4 bytes.
Credit to OSS-Fuzz
Estimate the cost for using FSE modes `set_basic`, `set_compressed`, and
`set_repeat`, and select the one with the lowest cost.
* The cost of `set_basic` is computed using the cross-entropy cost
function `ZSTD_crossEntropyCost()`, using the normalized default count
and the count.
* The cost of `set_repeat` is computed using `FSE_bitCost()`. We check the
previous table to see if it is able to represent the distribution.
* The cost of `set_compressed` is computed with the entropy cost function
`ZSTD_entropyCost()`, together with the cost of writing the normalized
count `ZSTD_NCountCost()`.
for FSE symbols.
While it seems to work, the gains are negligible compared to rough maxNbBits evaluation.
There are even a few losses sometimes, that still need to be explained.
Furthermode, there are still cases where btlazy2 does a better job than btopt,
which seems rather strange too.
for proper estimation of symbol's weights
when using dictionary compression.
Note : using only huffman costs is not good enough,
presumably because sequence symbol costs are incorrect.
* Expose the reference external sequences API for zstdmt.
Allows external sequences of any length, which get split when necessary.
* Reset the LDM window when the context is reset.
* Store the maximum number of LDM sequences.
* Sequence generation now returns the number of last literals.
* Fix sequence generation to not throw out the last literals when blocks of
more than 1 MB are encountered.
* Replaced a non-breaking space and an en dash with a plain space and
a hyphen.
* This means the files are simple ASCII and less likely to run into
codepage issues.
clang only claims compatibility with gcc 4.2.
Consequently, recent patch which reserved DYNAMIC_BMI2 for gcc >= 4.8
also disabled it for clang.
fix : __clang__ is now enough to enable DYNAMIC_BMI2
(associated with other existing conditions : x64/x64, !bmi2)
Update code documentation, and properly names a few "magic constants".
Also, HUF_compress_internal() gets a cleaner way
to determine size of tables inside workspace.
* `ZSTD_ldm_generateSequences()` generates the LDM sequences and
stores them in a table. It should work with any chunk size, but
is currently only called one block at a time.
* `ZSTD_ldm_blockCompress()` emits the pre-defined sequences, and
instead of encoding the literals directly, it passes them to a
secondary block compressor. The code to handle chunk sizes greater
than the block size is currently commented out, since it is unused.
The next PR will uncomment exercise this code.
* During optimal parsing, ensure LDM `minMatchLength` is at least
`targetLength`. Also don't emit repcode matches in the LDM block
compressor. Enabling the LDM with the optimal parser now actually improves
the compression ratio.
* The compression ratio is very similar to before. It is very slightly
different, because the repcode handling is slightly different. If I remove
immediate repcode checking in both branches the compressed size is exactly
the same.
* The speed looks to be the same or better than before.
Up Next (in a separate PR)
--------------------------
Allow sequence generation to happen prior to compression, and produce more
than a block worth of sequences. Expose some API for zstdmt to consume.
This will test out some currently untested code in
`ZSTD_ldm_blockCompress()`.
Pathological samples may result in literal section being incompressible.
This case is now detected,
and literal distribution is replaced by one that can be written into the dictionary.
The deep fuzzer tests caught a subtle bug that was probably there for a long time.
The impact of the bug is not a crash, or any other clear error signal,
rather, it reduces performance, by cutting data into smaller blocks.
Eventually, the following test would fail because it produces too many 1-byte blocks,
requiring more space than buffer can provide :
`./zstreamtest_asan --mt -s3514 -t1678312 -i1678314`
The root scenario is as follows :
- Create context, initialize it using explicit parameters or a `cdict` to pin them down, set `pledgedSrcSize=1`
- The compression parameters will not be adapted, but `windowSize` and `blockSize` will be automatically set to `1`.
`windowSize` and `blockSize` are dynamic values, set within `ZSTD_resetCCtx_internal()`.
The automatic adaptation makes it possible to generate smaller contexts for smaller input sizes.
- Complete compression
- New compression with same context, using same parameters, but `pledgedSrcSize=ZSTD_CONTENTSIZE_UNKNOWN`
trigger "continue mode"
- Continue mode doesn't modify blockSize, because it used to depend on `windowLog` only,
but in fact, it also depends on `pledgedSrcSize`.
- The "old" blocksize (1) is still there,
next compression will use this value to cut input into blocks,
resulting in more blocks and worse performance than necessary performance.
Given the scenario, and its possible variants, I'm surprised it did not show up before.
But I suspect it did show up, it's just that it never triggered an error, because "worse performance" is not a trigger.
The above test is a special corner case, where performance is so impacted that it reaches an error case.
The fix works, but I'm not completely pleased.
I think the current code relies too much on implied relations between variables.
This will likely break again in the future when some related part of the code change.
Unfortunately, no time to make larger changes if we want to keep the release target for zstd v1.3.3.
So a longer term fix will have to be considered after the release.
To do : create a reliable test case which triggers this scenario for CI tests.
There was a flaw in the formula
which compared literal cost with match cost :
at a given position,
a non-null literal suite is going to be part of next sequence,
while if position ends a previous match, to immediately start another match,
next sequence will have a litlength of zero.
A litlength of zero has a non-null cost.
It follows that literals cost should be compared to match cost + litlength==0.
Not doing so gave a structural advantage to matches, which would be selected more often.
I believe that's what led to the creation of the strange heuristic which added a complex cost to matches.
The heuristic was actually compensating.
It was probably created through multiple trials, settling for best outcome on a given scenario (I suspect silesia.tar).
The problem with this heuristic is that it's hard to understand,
and unfortunately, any future change in the parser would impact the way it should be calculated and its effects.
The "proper" formula makes it possible to remove this heuristic.
Now, the problem is : in a head to head comparison, it's sometimes better, sometimes worse.
Note that all differences are small (< 0.01 ratio).
In general, the newer formula is better for smaller files (for example, calgary.tar and enwik7).
I suspect that's because starting statistics are pretty poor (another area of improvement).
However, for silesia.tar specifically, it's worse at level 22 (while being better at level 17, so even compression level has an impact ...).
It's a pity that zstd -22 gets worse on silesia.tar.
That being said, I like that the new code gets rid of strange variables,
which were introducing complexity for any future evolution (faster variants being in mind).
Therefore, in spite of this detrimental side effect, I tend to be in favor of it.
ZSTD_updateTree() expected to be followed by a Bt match finder, which would update zc->nextToUpdate.
With the new optimal match finder, it's not necessarily the case : a match might be found during repcode or hash3, and stops there because it reaches sufficient_len, without even entering the binary tree.
Previous policy was to nonetheless update zc->nextToUpdate, but the current position would not be inserted, creating "holes" in the btree, aka positions that will no longer be searched.
Now, when current position is not inserted, zc->nextToUpdate is not update, expecting ZSTD_updateTree() to fill the tree later on.
Solution selected is that ZSTD_updateTree() takes care of properly setting zc->nextToUpdate,
so that it no longer depends on a future function to do this job.
It took time to get there, as the issue started with a memory sanitizer error.
The pb would have been easier to spot with a proper `assert()`.
So this patch add a few of them.
Additionnally, I discovered that `make test` does not enable `assert()` during CLI tests.
This patch enables them.
Unfortunately, these `assert()` triggered other (unrelated) bugs during CLI tests, mostly within zstdmt.
So this patch also fixes them.
- Changed packed structure for gcc memory access : memory sanitizer would complain that a read "might" reach out-of-bound position on the ground that the `union` is larger than the type accessed.
Now, to avoid this issue, each type is independent.
- ZSTD_CCtxParams_setParameter() : @return provides the value of parameter, clamped/fixed appropriately.
- ZSTDMT : changed constant name to ZSTDMT_JOBSIZE_MIN
- ZSTDMT : multithreading is automatically disabled when srcSize <= ZSTDMT_JOBSIZE_MIN, since only one thread will be used in this case (saves memory and runtime).
- ZSTDMT : nbThreads is automatically clamped on setting the value.
this version has same speed as branch `opt`
which is itself 5-10% slower than branch `dev`
(no identified reason)
It does not compress exactly the same as `opt` or `dev`,
maybe because it doesn't stop search after repcodes,
leading to sometimes better compression, sometimes worse
(by a small margin).
warning : _extDict path does not work for the time being
This means that benchmark module works,
but file module will fail with large files (and high compression level).
Objective is to fuse _extDict path into current one,
in order to have a single parser to maintain.
* Maximum window size in 32-bit mode is 1GB, since allocations for 2GB fail
on my Mac.
* Maximum window size in 64-bit mode is 2GB, since that is the largest
power of 2 that works with the overflow prevention.
* Allow `--long=windowLog` to set the window log, along with
`--zstd=wlog=#`. These options also set the window size during
decompression, but don't override `--memory=#` if it is set.
* Present a helpful error message when the window size is too large during
decompression.
* The long range matcher defaults to a hash log 7 less than the window log,
which keeps it at 20 for window log 27.
* Keep the default long range matcher window size and the default maximum
window size at 27 for the API and CLI.
* Add tests that use the maximum window size and hash size for compression
and decompression.
