> As far I have understood the buffer cache is for buffering block > devices data and the page cache for buffering filesystems' file data. Linux has no separate buffer cache, there is only a page cache. When the block device layer needs a buffer to read in data, it gets a page from page cache and attaches a buffer head to it (see grow_buffers()). Each time grow_buffers() gets a page, it increments a global buffermem_pages. What "free" reports as 'buffers' and 'cached' is buffermem_pages and (page_cache_size - buffermem_pages). > But when the command "cat /dev/hda > /dev/null" runs, the "free" > command reports that 'cached' is growing and 'buffers' are shrinking. > Can anyone figure out what have I misunderstood? My guess is that block layer is freeing the buffers (i.e freeing buffer heads and decrementing the page usage count), but the page remains in page cache - may be because it is still being referenced. Not a very convincing explanation, but that's all I can think of. Hope this helps, Ravi. __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
