> where [tagK] is the cache line tag (physical address[1]) and [byteK] > is actual data. > well, i new what a cache _is_, i just wasn't sure i understood the meaning of a "cache line" ;) anyway, im still confused: if the cpu's caches are not shared amongst them, so why is there need to align data structure to different cache lines?? on second thought, i think i understand. (bare with me, and correct me if im wrong ): lets say we're talking about UMA machines. since each cpu's cache maps the same memory, when one of them dirties its cache, all caches need to be replenished (when containing the same address...) . so, to avoid this, when the data structures are aligned to different cache lines, no thrashing takes place. am i correct? -- ======================================================================== nir. -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
