I compress with ZSTD_compress2. On ARM64 machines I have no issues. But on ARM32 I very rarely get a corrupt output. There are interesting patterns in the corruption, so I share one case. Most of the decompresse
Only thing needed then is to replace all .gz occurences with .zst in the code. I am now testing it with a 1.9TB database and zstd compression level 11, to see if it works properly. However, I am in no way a really experienced C dev or anything related to mysql. So I will ...
But indeed, if we reach this point, data is certainly corrupt, so it would be useful to report it. However, an additional issue is that the bitstream is almost always over-read. This is part of a technique designed to reduce the amount of overhead from FSE bitstreams. Essentially, the...
Filesystem the data is stored on Compiler versions OS version:uname -aoutput CPU:lscpuoutput My best guess is that your filesystem is somehow different, and maybe exposing some bug in the zstd CLI's I/O code. $ zstd-dev-22k14-gcc-6.5.0_sl_lx32 -19 --zstd=wlog=25 -c silesia.tar...
The persistent compression state does not need to be publicly documented, nor stable across versions or platforms. If the persisted compression state is invalid/corrupt, it should be ignored. This could take the form of a--state STATE_FILEswitch that could be used as follows: ...
Zstandard - Fast real-time compression algorithm. Contribute to mtola/zstd development by creating an account on GitHub.
miss a bit of compression, but other solutions can corrupt the tree */ if (match[matchLength] < ip[matchLength]) { /* necessarily within correct buffer */ /* match is smaller than current */ *smallerPtr = matchIndex; /* update smaller idx */ commonLengthSmaller = matchLengt...
Zstandard - Fast real-time compression algorithm. Contribute to mtola/zstd development by creating an account on GitHub.
Zstandard - Fast real-time compression algorithm. Contribute to mtola/zstd development by creating an account on GitHub.