Messages in this thread | | | From | David Laight <> | Subject | RE: [PATCH 0/2] Optimize memchr() | Date | Mon, 15 Aug 2022 10:59:26 +0000 |
| |
From: Pavel Machek > Sent: 12 August 2022 20:07 > > Hi! > > > This patche series optimized "memchr()" and add a macro for > > "memchr_inv()" so that both funtions can use it to generate bit mask. > > > > The original implementaion of "memchr()" is based on byte-wise comparison, > > which do not fully use 64-bit or 32-bit register in CPU. We implement a > > word-wise comparison so that at least 4 bytes can be compared at the same > > time. The optimized "memchr()" is nearly 4x faster than the original one > > for long strings. In Linux Kernel, we find that the length of the string > > Well... how much slower is it for short strings?
And cold cache??
David
> > searched by "memchr()" is up to 512 bytes in drivers/misc/lkdtm/heap.c. > > In our test, the optimized version is about 20% faster if the target > > character is at the end of the string when going through a 512-byte > > string. > > "What is the average length passed to memchr" would be more useful question. > > Best regards, > Pavel
- Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK Registration No: 1397386 (Wales)
| |