Rosetta 2 is remarkably fast when compared to other x86-on-ARM emulators. I ve spent a little time looking at how it works, out of idle curiosity, and found it to be quite unusual, so I figured I d put together my notes. Continue reading
In my recent post on optimising zlib decompression for the Apple M1, I used a loop that refilled a bit-buffer and decoded a huffman code each iteration, based on variant 4 from Fabian Giesen s Reading bits in far too many Continue reading
DEFLATE is a relatively slow compression algorithm from 1991, which (along with its wrapper format, zlib) is incredibly widely used, for example in the PNG, Zip and Gzip file formats and the HTTP, SSL, and SSH protocols. As such, I Continue reading
Variable length non-self-synchronising prefix codes (like x86 instructions and Huffman codes) are hard to decode in parallel, as each word must be decoded to figure out its length, before the next can be found and decoded. I came up with Continue reading
CRC32 is a checksum first proposed in 1961, and now used in a wide variety of performance sensitive contexts, from file formats (zip, png, gzip) to filesystems (ext4, btrfs) and protocols (like ethernet and SATA). So, naturally, a lot of Continue reading
I was inspired by Daniel Lemire s blog post, Converting integers to fix-digit representations quickly (and the follow up Converting integers to decimal strings faster with AVX-512) to try solving the same problem using the Neon SIMD instructions on the Apple Continue reading
I came up with a (seemingly) new method to encode bitmask immediate values on ARM64. This really isn t worth optimising clarity and verifiability are more important but it s a fun bit-twiddling problem, and the solution I came up Continue reading
Out-of-order processors have to keep track of multiple in-flight operations at once, and they use a variety of different buffers and queues to do so. I ve been trying to characterise and measure some of these buffers in the Apple M1 Continue reading
Many people, myself included, have held the belief that Spectre exploits need to know, understand, and manipulate microarchitectural details that are specific to a given processor design. Published Spectre PoCs generally use techniques such as cache analysis, and flushing lines Continue reading
In the words of Tom Lehrer, this is completely pointless, but may prove useful to some of you some day, perhaps in a somewhat bizarre set of circumstances. The problem is as follows: suppose you re working in a programming environment Continue reading