memblock_alloc_nid() used to fallback to allocating anywhere by using
memblock_alloc() as a fallback.

However, some of my previous patches limit memblock_alloc() to the region
covered by MEMBLOCK_ALLOC_ACCESSIBLE which is not quite what we want
for memblock_alloc_try_nid().

So we fix it by explicitely using MEMBLOCK_ALLOC_ANYWHERE.

Not that so far only sparc uses memblock_alloc_nid() and it hasn't been updated
to clamp the accessible zone yet. Thus the temporary "breakage" should have
no effect.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

The former is now strict, it will fail if it cannot honor the allocation
within the node, while the later implements the previous semantic which
falls back to allocating anywhere.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

We now provide a default (weak) implementation of memblock_nid_range()
which uses the early_pfn_map[] if CONFIG_ARCH_POPULATES_NODE_MAP
is set. Sparc still needs to use its own method due to the way
the pages can be scattered between nodes.

This implementation is inefficient due to our main algorithm and
callback construct wanting to work on an ascending addresses bases
while early_pfn_map[] would rather work with nid's (it's unsorted
at that stage). But it should work and we can look into improving
it subsequently, possibly using arch compile options to chose a
different algorithm alltogether.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

To constraint the search of a region between two boundaries,
which will be used by the new NUMA aware allocator among others.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Some archs such as ARM want to avoid coalescing accross things such
as the lowmem/highmem boundary or similar. This provides the option
to control it via an arch callback for which a weak default is provided
which always allows coalescing.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

When one of the array gets full, we resize it. After much thinking and
a few iterations of that code, I went back to on-demand resizing using
the (new) internal memblock_find_base() function, which is pretty much what
Yinghai initially proposed, though there some differences in the details.

To work this relies on the default alloc limit being set sensibly by
the architecture.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Some shuffling is needed for doing array resize so we may as well
put some sense into the ordering of the functions in the whole memblock.c
file. No code change. Added some comments.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

This will be used by the array resize code and might prove useful
to some arch code as well at which point it can be made non-static.

Also add comment as to why aligning size is important
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---

v2. Fix loss of size alignment
v3. Fix result code

It's a real PITA to have to search for it in the middle
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

This function will be used to locate a free area to put the new memblock
arrays when attempting to resize them. memblock_alloc_region() is gone,
the two callsites now call memblock_add_region().
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
v2. Fix membase_alloc_nid_region() conversion

Since we allocate one more than needed, why not do a bit of sanity checking
here to ensure we don't walk past the end of the array ?
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

This is in preparation for having resizable arrays.

Note that we still allocate one more than needed, this is unchanged from
the previous implementation.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Right now, both the "memory" and "reserved" memblock_type structures have
a "size" member. It represents the calculated memory size in the former
case and is unused in the latter.

This moves it out to the main memblock structure instead
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Let's not waste space and cycles on archs that don't support >32-bit
physical address space.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

The RMA (RMO is a misnomer) is a concept specific to ppc64 (in fact
server ppc64 though I hijack it on embedded ppc64 for similar purposes)
and represents the area of memory that can be accessed in real mode
(aka with MMU off), or on embedded, from the exception vectors (which
is bolted in the TLB) which pretty much boils down to the same thing.

We take that out of the generic MEMBLOCK data structure and move it into
arch/powerpc where it belongs, renaming it to "RMA" while at it.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

This introduce memblock.current_limit which is used to limit allocations
from memblock_alloc() or memblock_alloc_base(..., MEMBLOCK_ALLOC_ACCESSIBLE).

The old MEMBLOCK_ALLOC_ANYWHERE changes value from 0 to ~(u64)0 and can still
be used with memblock_alloc_base() to allocate really anywhere.

It is -no-longer- cropped to MEMBLOCK_REAL_LIMIT which disappears.

Note to archs: I'm leaving the default limit to MEMBLOCK_ALLOC_ANYWHERE. I
strongly recommend that you ensure that you set an appropriate limit
during boot in order to guarantee that an memblock_alloc() at any time
results in something that is accessible with a simple __va().

The reason is that a subsequent patch will introduce the ability for
the array to resize itself by reallocating itself. The MEMBLOCK core will
honor the current limit when performing those allocations.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Nobody uses it anymore. It's semantics were ... weird
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

CC: Russell King <linux@arm.linux.org.uk>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

CC: David S. Miller <davem@davemloft.net>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

CC: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

CC: Michal Simek <monstr@monstr.eu>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Walk memblock's using for_each_memblock() and use memblock_region_base/end_pfn() for
getting to PFNs.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Instead of the deprecated memblock_find()
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

The implementation is pretty much similar. There is a -small- added
overhead by having another function call and the address shift.

If that becomes a concern, I suppose we could actually have memblock
itself expose a memblock_pfn_valid() which then ARM can use directly
with an appropriate #define...
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

To make it fast, we steal ARM's binary search for memblock_is_memory()
and we use that to also the replace existing implementation of
memblock_is_reserved().
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

All callers expect a boolean result which is true if the region
overlaps a reserved region. However, the implementation actually
returns -1 if there is no overlap, and a region index (0 based)
if there is.

Make it behave as callers (and common sense) expect.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

* git://git.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-2.6-nmw:
  GFS2: Fix recovery stuck bug (try #2)
  GFS2: Fix typo in stuffed file data copy handling
  Revert "GFS2: recovery stuck on transaction lock"
  GFS2: Make "try" lock not try quite so hard
  GFS2: remove dependency on __GFP_NOFAIL
  GFS2: Simplify gfs2_write_alloc_required
  GFS2: Wait for journal id on mount if not specified on mount command line
  GFS2: Use nobh_writepage

* 'linux-next' of git://git.infradead.org/ubi-2.6:
  UBI: do not warn unnecessarily
  UBI: do not print message about corruptes PEBs if we have none of them
  UBI: improve delete-compatible volumes handling
  UBI: fix error message and compilation warnings
  UBI: generate random image_seq when formatting MTD devices
  UBI: improve ECC error message
  UBI: improve corrupted flash handling
  UBI: introduce eraseblock counter variables
  UBI: introduce a new IO return code
  UBI: simplify IO error codes

* 'linux-next' of git://git.infradead.org/ubifs-2.6:
  UBIFS: fix a memory leak on error path.
  UBIFS: fix GC LEB recovery
  UBIFS: use ERR_CAST
  UBIFS: check return code

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ericvh/v9fs: (22 commits)
  9p: fix sparse warnings in new xattr code
  fs/9p: remove sparse warning in vfs_inode
  fs/9p: destroy fid on failed remove
  fs/9p: Prevent parallel rename when doing fid_lookup
  fs/9p: Add support user. xattr
  net/9p: Implement TXATTRCREATE 9p call
  net/9p: Implement attrwalk 9p call
  9p: Implement LOPEN
  fs/9p: This patch implements TLCREATE for 9p2000.L protocol.
  9p: Implement TMKDIR
  9p: Implement TMKNOD
  9p: Define and implement TSYMLINK for 9P2000.L
  9p: Define and implement TLINK for 9P2000.L
  9p: Define and implement TLINK for 9P2000.L
  9p: Implement client side of setattr for 9P2000.L protocol.
  9p: getattr client implementation for 9P2000.L protocol.
  fs/9p: Pass the correct user credentials during attach
  net/9p: Handle the server returned error properly
  9p: readdir implementation for 9p2000.L
  9p: Make use of iounit for read/write
  ...

* 'for-linus' of git://oss.sgi.com/xfs/xfs: (49 commits)
  xfs simplify and speed up direct I/O completions
  xfs: move aio completion after unwritten extent conversion
  direct-io: move aio_complete into ->end_io
  xfs: fix big endian build
  xfs: clean up xfs_bmap_get_bp
  xfs: simplify xfs_truncate_file
  xfs: kill the b_strat callback in xfs_buf
  xfs: remove obsolete osyncisosync mount option
  xfs: clean up filestreams helpers
  xfs: fix gcc 4.6 set but not read and unused statement warnings
  xfs: Fix build when CONFIG_XFS_POSIX_ACL=n
  xfs: fix unsigned underflow in xfs_free_eofblocks
  xfs: use GFP_NOFS for page cache allocation
  xfs: fix memory reclaim recursion deadlock on locked inode buffer
  xfs: fix xfs_trans_add_item() lockdep warnings
  xfs: simplify and remove xfs_ireclaim
  xfs: don't block on buffer read errors
  xfs: move inode shrinker unregister even earlier
  xfs: remove a dmapi leftover
  xfs: writepage always has buffers
  ...