[PATCH *] 2.4.17 rmap based VM #10

From: Rik van Riel (riel@conectiva.com.br)
Date: Tue Jan 01 2002 - 23:41:25 EST


Hi,

It seems the cause for the livelock wasn't what I thought it was
for rmap-9, I finally ran into a piece of old, broken code, right
in the middle of page_launder(). Now the machine boots and runs
decently with only 1 MB of swappable memory left (the rest consumed
by the swap bitmap, the kernel itself, profiling buffer, ...).
This version seems to work nicely even under some pretty strange
corner cases.

The 10th version of the reverse mapping based VM is now available.
This is an attempt at making a more robust and flexible VM
subsystem, while cleaning up a lot of code at the same time. The patch
is available from:

           http://surriel.com/patches/2.4/2.4.17-rmap-10
and http://linuxvm.bkbits.net/

My big TODO items for a next release are:
  - fix page_launder() so it doesn't submit the whole
    inactive_dirty list for writeout in one go

rmap 10:
  - fix the livelock for real (yeah right), turned out
    to be a stupid bug in page_launder_zone() (me)
  - to make sure the VM subsystem doesn't monopolise
    the CPU, let kswapd and some apps sleep a bit under
    heavy stress situations (me)
  - let __GFP_HIGH allocations dig a little bit deeper
    into the free page pool, the SCSI layer seems fragile (me)
rmap 9:
  - improve comments all over the place (Michael Cohen)
  - don't panic if page_remove_rmap() cannot find the
    rmap in question, it's possible that the memory was
    PG_reserved and belonging to a driver, but the driver
    exited and cleared the PG_reserved bit (me)
  - fix the VM livelock by replacing > by >= in a few
    critical places in the pageout code (me)
  - treat the reclaiming of an inactive_clean page like
    allocating a new page, calling try_to_free_pages()
    and/or fixup_freespace() if required (me)
  - when low on memory, don't make things worse by
    doing swapin_readahead (me)
rmap 8:
  - add ANY_ZONE to the balancing functions to improve
    kswapd's balancing a bit (me)
  - regularize some of the maximum loop bounds in
    vmscan.c for cosmetic purposes (William Lee Irwin)
  - move page_address() to architecture-independent
    code, now the removal of page->virtual is portable (William Lee Irwin)
  - speed up free_area_init_core() by doing a single
    pass over the pages and not using atomic ops (William Lee Irwin)
  - documented the buddy allocator in page_alloc.c (William Lee Irwin)
rmap 7:
  - clean up and document vmscan.c (me)
  - reduce size of page struct, part one (William Lee Irwin)
  - add rmap.h for other archs (untested, not for ARM) (me)
rmap 6:
  - make the active and inactive_dirty list per zone,
    this is finally possible because we can free pages
    based on their physical address (William Lee Irwin)
  - cleaned up William's code a bit (me)
  - turn some defines into inlines and move those to
    mm_inline.h (the includes are a mess ...) (me)
  - improve the VM balancing a bit (me)
  - add back inactive_target to /proc/meminfo (me)
rmap 5:
  - fixed recursive buglet, introduced by directly
    editing the patch for making rmap 4 ;))) (me)
rmap 4:
  - look at the referenced bits in page tables (me)
rmap 3:
  - forgot one FASTCALL definition (me)
rmap 2:
  - teach try_to_unmap_one() about mremap() (me)
  - don't assign swap space to pages with buffers (me)
  - make the rmap.c functions FASTCALL / inline (me)
rmap 1:
  - fix the swap leak in rmap 0 (Dave McCracken)
rmap 0:
  - port of reverse mapping VM to 2.4.16 (me)

Rik

-- 
Shortwave goes a long way:  irc.starchat.net  #swl

http://www.surriel.com/ http://distro.conectiva.com/

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Mon Jan 07 2002 - 21:00:16 EST