Fast reader/writer locks
Posted Feb 13, 2003 19:16 UTC (Thu) by metacircles
Parent article: Fast reader/writer locks
What happened to the vsyscall work that Andrea (I think) was doing? From memory, it used a similar pair-of-counters scheme for updates, but it also avoided the need for an actual syscall, by storing the data (and counters) on a page visible in both kernel and user-space
I thought it had been merged, at least in the x86-64 port. This was a while ago, though.
to post comments)