This is the simplest operation. Whenever a pages is requested, the system searches for it in the cache. If the page is found in the cache, the system decompresses it and places the result on the user address space. Otherwise, if the page is in the disk, the disk block is read and the page is decompressed as in the previous case. As read disk blocks are not placed in the cache, if another page from the same buffer is requested, a new disk read will be needed (remember that hits_do_read are not very frequent).