Patch series "lib/lzo: run-length encoding support", v5.
Following on from the previous lzo-rle patchset:
https://lkml.org/lkml/2018/11/30/972
This patchset contains only the RLE patches, and should be applied on
top of the non-RLE patches ( https://lkml.org/lkml/2019/2/5/366 ).
Previously, some questions were raised around the RLE patches. I've
done some additional benchmarking to answer these questions. In short:
- RLE offers significant additional performance (data-dependent)
- I didn't measure any regressions that were clearly outside the noise
One concern with this patchset was around performance - specifically,
measuring RLE impact separately from Matt Sealey's patches (CTZ & fast
copy). I have done some additional benchmarking which I hope clarifies
the benefits of each part of the patchset.
Firstly, I've captured some memory via /dev/fmem from a Chromebook with
many tabs open which is starting to swap, and then split this into 4178
4k pages. I've excluded the all-zero pages (as zram does), and also the
no-zero pages (which won't tell us anything about RLE performance).
This should give a realistic test dataset for zram. What I found was
that the data is VERY bimodal: 44% of pages in this dataset contain 5%
or fewer zeros, and 44% contain over 90% zeros (30% if you include the
no-zero pages). This supports the idea of special-casing zeros in zram.
Next, I've benchmarked four variants of lzo on these pages (on 64-bit
Arm at max frequency): baseline LZO; baseline + Matt Sealey's patches
(aka MS); baseline + RLE only; baseline + MS + RLE. Numbers are for
weighted roundtrip throughput (the weighting reflects that zram does
more compression than decompression).
https://drive.google.com/file/d/1VLtLjRVxgUNuWFOxaGPwJYhl_hMQXpHe/view?usp=sharing
Matt's patches help in all cases for Arm (and no effect on Intel), as
expected.
RLE also behaves as expected: with few zeros present, it makes no
difference; above ~75%, it gives a good improvement (50 - 300 MB/s on
top of the benefit from Matt's patches).
Best performance is seen with both MS and RLE patches.
Finally, I have benchmarked the same dataset on an x86-64 device. Here,
the MS patches make no difference (as expected); RLE helps, similarly as
on Arm. There were no definite regressions; allowing for observational
error, 0.1% (3/4178) of cases had a regression > 1 standard deviation,
of which the largest was 4.6% (1.2 standard deviations). I think this
is probably within the noise.
https://drive.google.com/file/d/1xCUVwmiGD0heEMx5gcVEmLBI4eLaageV/view?usp=sharing
One point to note is that the graphs show RLE appears to help very
slightly with no zeros present! This is because the extra code causes
the clang optimiser to change code layout in a way that happens to have
a significant benefit. Taking baseline LZO and adding a do-nothing line
like "__builtin_prefetch(out_len);" immediately before the "goto next"
has the same effect. So this is a real, but basically spurious effect -
it's small enough not to upset the overall findings.
This patch (of 3):
When using zram, we frequently encounter long runs of zero bytes. This
adds a special case which identifies runs of zeros and encodes them
using run-length encoding.
This is faster for both compression and decompresion. For high-entropy
data which doesn't hit this case, impact is minimal.
Compression ratio is within a few percent in all cases.
This modifies the bitstream in a way which is backwards compatible
(i.e., we can decompress old bitstreams, but old versions of lzo cannot
decompress new bitstreams).
Link: http://lkml.kernel.org/r/20190205155944.16007-2-dave.rodgman@arm.com
Signed-off-by: Dave Rodgman <dave.rodgman@arm.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Markus F.X.J. Oberhumer <markus@oberhumer.com>
Cc: Matt Sealey <matt.sealey@arm.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nitin Gupta <nitingupta910@gmail.com>
Cc: Richard Purdie <rpurdie@openedhand.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>
Cc: Sonny Rao <sonnyrao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This fix ensures that we never meet an integer overflow while adding
255 while parsing a variable length encoding. It works differently from
commit 206a81c ("lzo: properly check for overruns") because instead of
ensuring that we don't overrun the input, which is tricky to guarantee
due to many assumptions in the code, it simply checks that the cumulated
number of 255 read cannot overflow by bounding this number.
The MAX_255_COUNT is the maximum number of times we can add 255 to a base
count without overflowing an integer. The multiply will overflow when
multiplying 255 by more than MAXINT/255. The sum will overflow earlier
depending on the base count. Since the base count is taken from a u8
and a few bits, it is safe to assume that it will always be lower than
or equal to 2*255, thus we can always prevent any overflow by accepting
two less 255 steps.
This patch also reduces the CPU overhead and actually increases performance
by 1.1% compared to the initial code, while the previous fix costs 3.1%
(measured on x86_64).
The fix needs to be backported to all currently supported stable kernels.
Reported-by: Willem Pinckaers <willem@lekkertech.net>
Cc: "Don A. Bailey" <donb@securitymouse.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This reverts commit 206a81c ("lzo: properly check for overruns").
As analysed by Willem Pinckaers, this fix is still incomplete on
certain rare corner cases, and it is easier to restart from the
original code.
Reported-by: Willem Pinckaers <willem@lekkertech.net>
Cc: "Don A. Bailey" <donb@securitymouse.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The lzo decompressor can, if given some really crazy data, possibly
overrun some variable types. Modify the checking logic to properly
detect overruns before they happen.
Reported-by: "Don A. Bailey" <donb@securitymouse.com>
Tested-by: "Don A. Bailey" <donb@securitymouse.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This commit updates the kernel LZO code to the current upsteam version
which features a significant speed improvement - benchmarking the Calgary
and Silesia test corpora typically shows a doubled performance in
both compression and decompression on modern i386/x86_64/powerpc machines.
Signed-off-by: Markus F.X.J. Oberhumer <markus@oberhumer.com>
Rename the source file to match the function name and thereby
also make room for a possible future even slightly faster
"non-safe" decompressor version.
Signed-off-by: Markus F.X.J. Oberhumer <markus@oberhumer.com>
