Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
userbinator
on July 25, 2020
|
parent
|
context
|
favorite
| on:
Avoid character-by-character processing when perfo...
Things like this are why I wish x86 would've exploited its CISC-ness to do more autovectorisation in hardware, something like this:
mov rsi, ptr mov rcx, size repz andb 80h ; while(rcx-- && !(al = *rsi++ & 0x80)) ; setz al
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: