This takes my "make fastcheck" from about 57 seconds to about 43 seconds.

The ccanlint patch is rather intrusive.  First, it adds a new field to
all the ccanlint tests, "key".  key is a shorter, still unique
description of the test (e.g. "valgrind").  The names I chose as keys
for all the tests are somewhat arbitrary and often don't reflect the
name of the .c source file (because some of those names are just too
darn long).  Second, it adds two new options to ccanlint:

   -l: list tests ccanlint performs
   -x: exclude tests (e.g. -x trailing_whitespace,valgrind)

It also adds a consistency check making sure all tests have unique
keys and names.

The primary goal of the ccanlint patch was so I could exclude the
valgrind test, which takes a really long time for some modules (I
think btree takes the longest, at around 2 minutes).  I'm not sure I
did it 100% correctly, so you'll want to review it first.

The btree patch gives the btree module an intuitive frontend
(btree_insert, btree_remove, btree_lookup) and a built-in ordering
function for strings.  Together, these make it easy to use the btree
module as a dynamic string map.

The charset patch makes utf8_validate reject the invalid codepoints
U+FFFE and U+FFFF .  Hopefully it's fully UTF-8 compliant now.

tdb transactions were designed to be robust against the machine
powering off, but interestingly were never designed to handle the case
where an administrator kill -9's a process during commit.  Because
recovery is only done on tdb_open, processes with the tdb already
mapped will simply use it despite it being corrupt and needing
recovery.

The solution to this is to check for recovery every time we grab a
data lock: we could have gained the lock because a process just died.
This has no measurable cost: here is the time for tdbtorture -s 0 -n 1
-l 10000:

Before:
	2.75 2.50 2.81 3.19 2.91 2.53 2.72 2.50 2.78 2.77 = Avg 2.75

After:
	2.81 2.57 3.42 2.49 3.02 2.49 2.84 2.48 2.80 2.43 = Avg 2.74
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

Reduce code duplication, and also gives us a central point for the next
patch which wants to cover all list locks.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

Now the transaction code uses the standard allrecord lock, that stops
us from trying to grab any per-record locks anyway.  We don't need to
have special noop lock ops for transactions.

This is a nice simplification: if you see brlock, you know it's really
going to grab a lock.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

Records themselves get (read) locked by the traversal code against delete.
Interestingly, this locking isn't done when the allrecord lock has been
taken, though the allrecord lock until recently didn't cover the actual
records (it now goes to end of file).

The write record lock, grabbed by the delete code, is not suppressed by
the allrecord lock, which causes us to punch a hole in that lock when we
release the write record lock.  Make this consistent: *no* record locks
of any kind when the allrecord lock is taken.

There's little point in ever shrinking the file, and it definitely breaks in the case where a process has died during a transaction commit and other processes have the tdb mapped.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

commit b37b452cb8c1f56b37b04abe7bffdede371ca361
Author: Rusty Russell <rusty@rustcorp.com.au>
Date:   Thu Feb 4 23:59:54 2010 +1030

    tdb: fix recovery reuse after crash
    
    If a process (or the machine) dies after just after writing the
    recovery head (pointing at the end of file), the recovery record will filled
    with 0x42.  This will not invoke a recovery on open, since rec.magic
    != TDB_RECOVERY_MAGIC.
    
    Unfortunately, the first transaction commit will happily reuse that
    area: tdb_recovery_allocate() doesn't check the magic.  The recovery
    record has length 0x42424242, and it writes that back into the
    now-valid-looking transaction header) for the next comer (which
    happens to be tdb_wipe_all in my tests).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

With killing children, CLEAR_IF_FIRST can happen quite a bit.

Now the transaction allrecord lock the standard one, and thus is cleaned
in tdb_release_extra_locks(), _tdb_transaction_cancel() doesn't need to
know what type it is.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

Centralize locking of all chains of the tdb; rename _tdb_lockall to
tdb_allrecord_lock and _tdb_unlockall to tdb_allrecord_unlock, and
tdb_brlock_upgrade to tdb_allrecord_upgrade.

Then we use this in the transaction code.  Unfortunately, if the transaction
code records that it has grabbed the allrecord lock read-only, write locks
will fail, so we treat this upgradable lock as a write lock, and mark it
as upgradable using the otherwise-unused offset field.

One subtlety: now the transaction code is using the allrecord_lock, the
tdb_release_extra_locks() function drops it for us, so we no longer need
to do it manually in _tdb_transaction_cancel.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

We were previously inconsistent with our "global" lock: the
transaction code grabbed it from FREELIST_TOP to end of file, and the
rest of the code grabbed it from FREELIST_TOP to end of the hash
chains.  Change it to always grab to end of file for simplicity.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

This was redundant before this patch series: it mirrored num_lockrecs
exactly.  It still does.

Also, skip useless branch when locks == 1: unconditional assignment is
cheaper anyway.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

This is pure overhead, but it centralizes the locking.  Realloc (esp. as
most implementations are lazy) is fast compared to the fnctl anyway.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

Rather than a boutique lock and a separate nest count, use our
newly-generic nested lock tracking for the active lock.

Note that the tdb_have_extra_locks() and tdb_release_extra_locks()
functions have to skip over this lock now it is tracked.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

This never nests, so it's overkill, but it centralizes the locking into
lock.c and removes the ugly flag in the transaction code to track whether
we have the lock or not.

Note that we have a temporary hack so this places a real lock, despite
the fact that we are in a transaction.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

Rather than a boutique lock and a separate nest count, use our
newly-generic nested lock tracking for the transaction lock.

Note that the tdb_have_extra_locks() and tdb_release_extra_locks()
functions have to skip over this lock now it is tracked.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

Factor out two loops which find locks; we are going to introduce a couple
more so a helper makes sense.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

Move locking intelligence back into lock.c.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>