I am seeing a slightly greater than 2M disparity between the amount of
memory reported free from nr_free_pages, and the number of pages with
page->count==0.
It looks roughly like the difference is the number of resered pages
(kernel memory?) + the 380k gap between 640k and 1M.
But the amount of difference is not constant :(, so I'm not sure.
Any clues?
Eric
>Mem-info:
>Free pages: 18204kB
^^^^^^^ 18204/4 = 4551 pages
> ( 1*4kB 15*8kB 18*16kB 10*32kB 3*64kB 135*128kB = 18204kB)
>Swap cache: add 0/0, delete 0/0, find 0/0
>Free swap: 913564kB
>8192 pages of RAM
>5101 free pages
^^^^ == 20404kB or 550 pages more.
>455 reserved pages
If these count it is only 95 pages == 380kB!
>2243 pages shared
>0 pages swap cached
>Buffer memory: 1220kB
>Buffer heads: 1256
>Buffer blocks: 1220
> CLEAN: 909 buffers, 46 used (last=414), 0 locked, 0 protected, 0 dirty
> LOCKED: 273 buffers, 23 used (last=245), 0 locked, 0 protected, 0 dirty
> DIRTY: 1 buffers, 0 used (last=0), 0 locked, 0 protected, 1 dirty
>Networking buffers in use : 0
>Total network buffer allocations : 100
>Total failed network buffer allocs : 0
>IP fragment buffer size : 0
>Mem-info:
>Free pages: 9428kB
^^^^^ = 2357 pages
> ( 3*4kB 1*8kB 2*16kB 3*32kB 1*64kB 72*128kB = 9428kB)
>Swap cache: add 0/0, delete 0/0, find 0/0
>Free swap: 913564kB
>8192 pages of RAM
>2909 free pages
^^^^ now it's 552 pages different
>455 reserved pages
^^^ less this it's 97 pages == 388Kb different
>2716 pages shared
>0 pages swap cached
>Buffer memory: 2628kB
>Buffer heads: 2652
>Buffer blocks: 2628
> CLEAN: 2218 buffers, 43 used (last=409), 0 locked, 0 protected, 0 dirty
> LOCKED: 362 buffers, 26 used (last=356), 0 locked, 0 protected, 0 dirty
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu