Second step in pushing down the page_table_lock.  Remove the temporary
bridging hack from __pud_alloc, __pmd_alloc, __pte_alloc: expect callers not
to hold page_table_lock, whether it's on init_mm or a user mm; take
page_table_lock internally to check if a racing task already allocated.

Convert their callers from common code.  But avoid coming back to change them
again later: instead of moving the spin_lock(&mm->page_table_lock) down,
switch over to new macros pte_alloc_map_lock and pte_unmap_unlock, which
encapsulate the mapping+locking and unlocking+unmapping together, and in the
end may use alternatives to the mm page_table_lock itself.

These callers all hold mmap_sem (some exclusively, some not), so at no level
can a page table be whipped away from beneath them; and pte_alloc uses the
"atomic" pmd_present to test whether it needs to allocate.  It appears that on
all arches we can safely descend without page_table_lock.
Signed-off-by: NHugh Dickins <hugh@veritas.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

One anomaly remains from when Andrea rationalized the responsibilities of
mmap_sem and page_table_lock: in dup_mmap we add vmas to the child holding its
page_table_lock, but not the mmap_sem which normally guards the vma list and
rbtree.  Which could be an issue for unuse_mm: though since it just walks down
the list (today with page_table_lock, tomorrow not), it's probably okay.  Will
need a memory barrier?  Oh, keep it simple, Nick and I agreed, no harm in
taking child's mmap_sem here.
Signed-off-by: NHugh Dickins <hugh@veritas.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Use the parent's oldmm throughout dup_mmap, instead of perversely going back
to current->mm.  (Can you hear the sigh of relief from those mpnts?  Usually I
squash them, but not today.)
Signed-off-by: NHugh Dickins <hugh@veritas.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

I was lazy when we added anon_rss, and chose to change as few places as
possible.  So currently each anonymous page has to be counted twice, in rss
and in anon_rss.  Which won't be so good if those are atomic counts in some
configurations.

Change that around: keep file_rss and anon_rss separately, and add them
together (with get_mm_rss macro) when the total is needed - reading two
atomics is much cheaper than updating two atomics.  And update anon_rss
upfront, typically in memory.c, not tucked away in page_add_anon_rmap.
Signed-off-by: NHugh Dickins <hugh@veritas.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

How is anon_rss initialized?  In dup_mmap, and by mm_alloc's memset; but
that's not so good if an mm_counter_t is a special type.  And how is rss
initialized?  By set_mm_counter, all over the place.  Come on, we just need to
initialize them both at once by set_mm_counter in mm_init (which follows the
memcpy when forking).
Signed-off-by: NHugh Dickins <hugh@veritas.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

The original vm_stat_account has fallen into disuse, with only one user, and
only one user of vm_stat_unaccount.  It's easier to keep track if we convert
them all to __vm_stat_account, then free it from its __shackles.
Signed-off-by: NHugh Dickins <hugh@veritas.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

The PF_NOFREEZE process flag should not be inherited when a thread is
forked.  This patch (as585) removes the flag from the child.

This problem is starting to show up more and more as drivers turn to the
kthread API instead of using kernel_thread().  As a result, their kernel
threads are now children of the kthread worker instead of modprobe, and
they inherit the PF_NOFREEZE flag.  This can cause problems during system
suspend; the kernel threads are not getting frozen as they ought to be.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Acked-by: NPavel Machek <pavel@ucw.cz>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Fix a problem wherein a new-born task is added to a dead CPU.
Signed-off-by: NSrivatsa Vaddagiri <vatsa@in.ibm.com>
Acked-by: NNick Piggin <nickpiggin@yahoo.com.au>
Acked-by: NShaohua Li <shaohua.li@intel.com>
Acked-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Patch to eliminate struct files_struct.file_lock spinlock on the reader side
and use rcu refcounting rcuref_xxx api for the f_count refcounter.  The
updates to the fdtable are done by allocating a new fdtable structure and
setting files->fdt to point to the new structure.  The fdtable structure is
protected by RCU thereby allowing lock-free lookup.  For fd arrays/sets that
are vmalloced, we use keventd to free them since RCU callbacks can't sleep.  A
global list of fdtable to be freed is not scalable, so we use a per-cpu list.
If keventd is already handling the current cpu's work, we use a timer to defer
queueing of that work.

Since the last publication, this patch has been re-written to avoid using
explicit memory barriers and use rcu_assign_pointer(), rcu_dereference()
premitives instead.  This required that the fd information is kept in a
separate structure (fdtable) and updated atomically.
Signed-off-by: NDipankar Sarma <dipankar@in.ibm.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

In order for the RCU to work, the file table array, sets and their sizes must
be updated atomically.  Instead of ensuring this through too many memory
barriers, we put the arrays and their sizes in a separate structure.  This
patch takes the first step of putting the file table elements in a separate
structure fdtable that is embedded withing files_struct.  It also changes all
the users to refer to the file table using files_fdtable() macro.  Subsequent
applciation of RCU becomes easier after this.
Signed-off-by: NDipankar Sarma <dipankar@in.ibm.com>
Signed-Off-By: NDavid Howells <dhowells@redhat.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Race is as follows. Process A forks process B, both being part of the same
session. Then, A calls disassociate_ctty while B forks C:

A				B
====				====
				fork()
				  copy_signal()
dissasociate_ctty()		....
				  attach_pid(p, PIDTYPE_SID, p->signal->session);

Now, C can have current->signal->tty pointing to a freed tty structure, as
it hasn't yet been added to the session group (to have its controlling tty
cleared on the diassociate_ctty() call).

This has shown up as an oops but could be even more serious.  I haven't
tried to create a test case, but a customer has verified that the patch
below resolves the issue, which was occuring quite frequently.  I'll try
and post the test case if i can.

The patch simply checks for a NULL tty *after* it has been attached to the
proper session group and clears it as necessary.  Alternatively, we could
simply do the tty assignment after the the process is added to the proper
session group.
Signed-off-by: NJason Baron <jbaron@redhat.com>
Cc: Roland McGrath <roland@redhat.com>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

An oversight.  We don't want to carry the IO scheduler's "we hold exclusive fs
resources" hint over to the child across fork().
Acked-by: NJens Axboe <axboe@suse.de>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

      Jeff Dike <jdike@addtoit.com>,
      Paolo 'Blaisorblade' Giarrusso <blaisorblade_spam@yahoo.it>,
      Bodo Stroesser <bstroesser@fujitsu-siemens.com>

Adds a new ptrace(2) mode, called PTRACE_SYSEMU, resembling PTRACE_SYSCALL
except that the kernel does not execute the requested syscall; this is useful
to improve performance for virtual environments, like UML, which want to run
the syscall on their own.

In fact, using PTRACE_SYSCALL means stopping child execution twice, on entry
and on exit, and each time you also have two context switches; with SYSEMU you
avoid the 2nd stop and so save two context switches per syscall.

Also, some architectures don't have support in the host for changing the
syscall number via ptrace(), which is currently needed to skip syscall
execution (UML turns any syscall into getpid() to avoid it being executed on
the host).  Fixing that is hard, while SYSEMU is easier to implement.

* This version of the patch includes some suggestions of Jeff Dike to avoid
  adding any instructions to the syscall fast path, plus some other little
  changes, by myself, to make it work even when the syscall is executed with
  SYSENTER (but I'm unsure about them). It has been widely tested for quite a
  lot of time.

* Various fixed were included to handle the various switches between
  various states, i.e. when for instance a syscall entry is traced with one of
  PT_SYSCALL / _SYSEMU / _SINGLESTEP and another one is used on exit.
  Basically, this is done by remembering which one of them was used even after
  the call to ptrace_notify().

* We're combining TIF_SYSCALL_EMU with TIF_SYSCALL_TRACE or TIF_SINGLESTEP
  to make do_syscall_trace() notice that the current syscall was started with
  SYSEMU on entry, so that no notification ought to be done in the exit path;
  this is a bit of a hack, so this problem is solved in another way in next
  patches.

* Also, the effects of the patch:
"Ptrace - i386: fix Syscall Audit interaction with singlestep"
are cancelled; they are restored back in the last patch of this series.

Detailed descriptions of the patches doing this kind of processing follow (but
I've already summed everything up).

* Fix behaviour when changing interception kind #1.

  In do_syscall_trace(), we check the status of the TIF_SYSCALL_EMU flag
  only after doing the debugger notification; but the debugger might have
  changed the status of this flag because he continued execution with
  PTRACE_SYSCALL, so this is wrong.  This patch fixes it by saving the flag
  status before calling ptrace_notify().

* Fix behaviour when changing interception kind #2:
  avoid intercepting syscall on return when using SYSCALL again.

  A guest process switching from using PTRACE_SYSEMU to PTRACE_SYSCALL
  crashes.

  The problem is in arch/i386/kernel/entry.S.  The current SYSEMU patch
  inhibits the syscall-handler to be called, but does not prevent
  do_syscall_trace() to be called after this for syscall completion
  interception.

  The appended patch fixes this.  It reuses the flag TIF_SYSCALL_EMU to
  remember "we come from PTRACE_SYSEMU and now are in PTRACE_SYSCALL", since
  the flag is unused in the depicted situation.

* Fix behaviour when changing interception kind #3:
  avoid intercepting syscall on return when using SINGLESTEP.

  When testing 2.6.9 and the skas3.v6 patch, with my latest patch and had
  problems with singlestepping on UML in SKAS with SYSEMU.  It looped
  receiving SIGTRAPs without moving forward.  EIP of the traced process was
  the same for all SIGTRAPs.

What's missing is to handle switching from PTRACE_SYSCALL_EMU to
PTRACE_SINGLESTEP in a way very similar to what is done for the change from
PTRACE_SYSCALL_EMU to PTRACE_SYSCALL_TRACE.

I.e., after calling ptrace(PTRACE_SYSEMU), on the return path, the debugger is
notified and then wake ups the process; the syscall is executed (or skipped,
when do_syscall_trace() returns 0, i.e.  when using PTRACE_SYSEMU), and
do_syscall_trace() is called again.  Since we are on the return path of a
SYSEMU'd syscall, if the wake up is performed through ptrace(PTRACE_SYSCALL),
we must still avoid notifying the parent of the syscall exit.  Now, this
behaviour is extended even to resuming with PTRACE_SINGLESTEP.
Signed-off-by: NPaolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>
Cc: Jeff Dike <jdike@addtoit.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

dup_mmap of a VM_DONTCOPY vma forgot to lower the child's total_vm.  (But
no way does this account for the recent report of total_vm seen too low.)
Signed-off-by: NHugh Dickins <hugh@veritas.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

This updates the CFQ io scheduler to the new time sliced design (cfq
v3).  It provides full process fairness, while giving excellent
aggregate system throughput even for many competing processes.  It
supports io priorities, either inherited from the cpu nice value or set
directly with the ioprio_get/set syscalls.  The latter closely mimic
set/getpriority.

This import is based on my latest from -mm.
Signed-off-by: NJens Axboe <axboe@suse.de>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Consolidate balance-on-exec with balance-on-fork.  This is made easy by the
sched-domains RCU patches.

As well as the general goodness of code reduction, this allows the runqueues
to be unlocked during balance-on-fork.

schedstats is a problem.  Maybe just have balance-on-event instead of
distinguishing fork and exec?
Signed-off-by: NNick Piggin <nickpiggin@yahoo.com.au>
Acked-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Remove part of comment on linking new vma in dup_mmap: since anon_vma rmap
came in, try_to_unmap_one knows the vma without needing find_vma.  But add
a comment to note that here vma is inserted without mmap_sem.
Signed-off-by: NHugh Dickins <hugh@veritas.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Ingo recently introduced a great speedup for allocating new mmaps using the
free_area_cache pointer which boosts the specweb SSL benchmark by 4-5% and
causes huge performance increases in thread creation.

The downside of this patch is that it does lead to fragmentation in the
mmap-ed areas (visible via /proc/self/maps), such that some applications
that work fine under 2.4 kernels quickly run out of memory on any 2.6
kernel.

The problem is twofold:

  1) the free_area_cache is used to continue a search for memory where
     the last search ended.  Before the change new areas were always
     searched from the base address on.

     So now new small areas are cluttering holes of all sizes
     throughout the whole mmap-able region whereas before small holes
     tended to close holes near the base leaving holes far from the base
     large and available for larger requests.

  2) the free_area_cache also is set to the location of the last
     munmap-ed area so in scenarios where we allocate e.g.  five regions of
     1K each, then free regions 4 2 3 in this order the next request for 1K
     will be placed in the position of the old region 3, whereas before we
     appended it to the still active region 1, placing it at the location
     of the old region 2.  Before we had 1 free region of 2K, now we only
     get two free regions of 1K -> fragmentation.

The patch addresses thes issues by introducing yet another cache descriptor
cached_hole_size that contains the largest known hole size below the
current free_area_cache.  If a new request comes in the size is compared
against the cached_hole_size and if the request can be filled with a hole
below free_area_cache the search is started from the base instead.

The results look promising: Whereas 2.6.12-rc4 fragments quickly and my
(earlier posted) leakme.c test program terminates after 50000+ iterations
with 96 distinct and fragmented maps in /proc/self/maps it performs nicely
(as expected) with thread creation, Ingo's test_str02 with 20000 threads
requires 0.7s system time.

Taking out Ingo's patch (un-patch available per request) by basically
deleting all mentions of free_area_cache from the kernel and starting the
search for new memory always at the respective bases we observe: leakme
terminates successfully with 11 distinctive hardly fragmented areas in
/proc/self/maps but thread creating is gringdingly slow: 30+s(!) system
time for Ingo's test_str02 with 20000 threads.

Now - drumroll ;-) the appended patch works fine with leakme: it ends with
only 7 distinct areas in /proc/self/maps and also thread creation seems
sufficiently fast with 0.71s for 20000 threads.
Signed-off-by: NWolfgang Wander <wwc@rentec.com>
Credit-to: "Richard Purdie" <rpurdie@rpsys.net>
Signed-off-by: NKen Chen <kenneth.w.chen@intel.com>
Acked-by: Ingo Molnar <mingo@elte.hu> (partly)
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.

Let it rip!