Rusty Russell [Tue, 18 Jan 2011 00:38:34 +0000 (11:08 +1030)]
failtest: failtest_restore.h as an antidote to function overload.
This makes some cases simpler, where you no longer want malloc etc.
to fail.
Brad Hards [Sun, 16 Jan 2011 06:50:51 +0000 (17:50 +1100)]
wwviaudio: spelling fix
Brad Hards [Sun, 16 Jan 2011 06:50:24 +0000 (17:50 +1100)]
tdb2: spelling fix
Brad Hards [Sun, 16 Jan 2011 06:50:03 +0000 (17:50 +1100)]
tdb: spelling fixes
Brad Hards [Sun, 16 Jan 2011 06:49:20 +0000 (17:49 +1100)]
tap: spelling fix
Brad Hards [Sun, 16 Jan 2011 06:48:53 +0000 (17:48 +1100)]
talloc: spelling fix.
Brad Hards [Sun, 16 Jan 2011 06:48:20 +0000 (17:48 +1100)]
rbtree: spelling fixes
Brad Hards [Sun, 16 Jan 2011 06:22:39 +0000 (17:22 +1100)]
opt: spelling fixes.
Brad Hards [Sun, 16 Jan 2011 06:22:14 +0000 (17:22 +1100)]
nfs: spelling fixes.
Brad Hards [Sun, 16 Jan 2011 06:21:39 +0000 (17:21 +1100)]
list: spelling fixes
Brad Hards [Sun, 16 Jan 2011 06:21:12 +0000 (17:21 +1100)]
iscsi: spelling fixes.
Brad Hards [Sun, 16 Jan 2011 06:19:51 +0000 (17:19 +1100)]
hash: spelling fix
Brad Hards [Sun, 16 Jan 2011 06:19:12 +0000 (17:19 +1100)]
btree: spelling fix
Brad Hards [Sun, 16 Jan 2011 06:17:36 +0000 (17:17 +1100)]
ciniparser: spelling fixes
Brad Hards [Sun, 16 Jan 2011 06:16:46 +0000 (17:16 +1100)]
junkcode: spelling fixes.
Brad Hards [Sun, 16 Jan 2011 06:15:32 +0000 (17:15 +1100)]
array: spelling fixes.
Brad Hards [Sun, 16 Jan 2011 05:22:46 +0000 (16:22 +1100)]
str: change example to match function being explained.
Looks like the example was copy-n-pasted from the previous entry.
Rusty Russell [Mon, 17 Jan 2011 05:50:15 +0000 (16:20 +1030)]
ccanlint: check examples actually mention thing they are demonstrating.
Brad suggested this after finding one such cut & paste in str:
rusty@vivaldi:~/devel/cvs/ccan/ccan/str$ ../../tools/ccanlint/ccanlint
Example: sections demonstrate appropriate function (examples_relevant): FAIL
/home/rusty/devel/cvs/ccan/ccan/str/str.h:64:Example for strcount doesn't mention it
Total score: 37/38
Rusty Russell [Mon, 17 Jan 2011 05:49:13 +0000 (16:19 +1030)]
ccanlint: fix trailing / in -d.
Before:
$ tools/ccanlint/ccanlint -d ccan/talloc/
Module's source code has no trailing whitespace (no_trailing_whitespace): FAIL
: Total score: 0/1
After:
$ tools/ccanlint/ccanlint -d ccan/talloc/
Module's source code has no trailing whitespace (no_trailing_whitespace): FAIL
talloc: Total score: 0/1
Rusty Russell [Mon, 17 Jan 2011 05:47:49 +0000 (16:17 +1030)]
ccanlint: score_file_error() takes printf-format
We simply build up the error string in score_file_error; a bit different
but simpler than current behaviour. We keep around struct file_error
because some tests need it.
Rusty Russell [Mon, 17 Jan 2011 05:34:42 +0000 (16:04 +1030)]
ccanlint: fix score for info_summary_single_line
My refactoring introduced a bug: we need to set score=1 for the success case.
Brad Hards [Mon, 17 Jan 2011 03:02:36 +0000 (13:32 +1030)]
ciniparser: Separate summary line and description text for metadata
Brad Hards [Mon, 17 Jan 2011 03:02:36 +0000 (13:32 +1030)]
block_pool: Change metadata layout so initial description is all on one line.
With the original layout, we got:
Summary:
An efficient allocator for blocks that don't need to be
....
Description:
resized or freed.
Rusty Russell [Mon, 17 Jan 2011 03:52:29 +0000 (14:22 +1030)]
ccanlint: neaten info_summary_single_line with new doc_section info.
Rusty Russell [Mon, 17 Jan 2011 03:44:43 +0000 (14:14 +1030)]
ccanlint: keep mapping back to original source line for doc_extract
Makes reporting of documentation easier.
Brad Hards [Mon, 17 Jan 2011 03:02:36 +0000 (13:32 +1030)]
ccanlint: add test case for metadata summary line on a single separate line.
Rusty Russell [Thu, 13 Jan 2011 08:55:01 +0000 (19:25 +1030)]
htable: fix type of cmpfn in htable_type
It in fact takes an object and a key to compare, not two keys.
The test case had the key as first element of the object, so it worked,
but ccanlint lost track of module dependencies due to this bug, and thus
would build submodules multiple times.
Rusty Russell [Mon, 10 Jan 2011 05:55:22 +0000 (16:25 +1030)]
rbtree: fix allocation failre paths.
Rusty Russell [Mon, 10 Jan 2011 05:17:24 +0000 (15:47 +1030)]
rbtree: don't use temporary context to destroy rbtree
It leads to a memory leak if the allocation fails (as we reparent onto NULL).
Also, the extra allocation in the failure path increases test time for
failtest (under valgrind) by a factor of 2.
Rusty Russell [Mon, 10 Jan 2011 05:53:46 +0000 (16:23 +1030)]
rbtree: use failtest to check handling of allocation failures.
Unfortunately this means we have to reduce run-many from 1000 nodes to 100
(as forking under valgrind is really slow: test takes about 2 minutes with
100 nodes).
Rusty Russell [Mon, 10 Jan 2011 05:51:24 +0000 (16:21 +1030)]
rbtree: fix memory leak in tests
The data is not made a child of the tree: the nodes are made children of the
data. So we must explicitly free the data objects.
Rusty Russell [Mon, 10 Jan 2011 05:55:48 +0000 (16:25 +1030)]
rbtree: vary insert and delete orders in test
This covers more code than simply doing an ordered delete/insert.
Rusty Russell [Mon, 10 Jan 2011 03:36:40 +0000 (14:06 +1030)]
talloc: allow replacement allocator
This allows us to both allocators which handle failure themselves, and
allocators which insert failures.
Rusty Russell [Mon, 10 Jan 2011 05:32:03 +0000 (16:02 +1030)]
talloc: fixed a use after free error
(Import from SAMBA commit
6f51a1f45bf4de062cce7a562477e8140630a53d):
this is the minimal fix for the problem Rusty found. I previously
thought that the best fix would be to change tc->parent to be valid
for all pointers, but that is expensive for realloc with large numbers
of child pointers, which is much more commmon than I expected it to
be.
Rusty Russell [Mon, 10 Jan 2011 03:31:47 +0000 (14:01 +1030)]
talloc: use failtest to test failure paths.
Rusty Russell [Mon, 10 Jan 2011 04:12:38 +0000 (14:42 +1030)]
failtest: new module.
A module designed to help test "never fails" functions like malloc.
Rusty Russell [Mon, 10 Jan 2011 05:30:48 +0000 (16:00 +1030)]
tap: add fail callback
This is useful for failtest, so we can abort on first failure.
Rusty Russell [Sun, 9 Jan 2011 01:26:41 +0000 (11:56 +1030)]
ccanlint: fix and simplify depends-accurate (with strreg)
Rusty Russell [Sat, 8 Jan 2011 09:39:25 +0000 (20:09 +1030)]
ccanlint: fix total score for running examples
Don't count examples which didn't compile (expected, as we mangle them in
various different ways).
Rusty Russell [Sat, 8 Jan 2011 10:13:59 +0000 (20:43 +1030)]
ccanlint: fix uninitialized variable
Rusty Russell [Sat, 8 Jan 2011 08:07:19 +0000 (18:37 +1030)]
ccanlint: use strreg for section extraction.
Makes it simpler and clearer.
Rusty Russell [Sat, 8 Jan 2011 08:06:56 +0000 (18:36 +1030)]
Makefile: append git revision to "make scores"
Good for testing ccanlint changes.
Rusty Russell [Sat, 8 Jan 2011 02:45:35 +0000 (13:15 +1030)]
str_talloc: strreg
Useful wrapper for extended POSIX regular expressions.
Rusty Russell [Fri, 7 Jan 2011 23:47:37 +0000 (10:17 +1030)]
str: strcount
Useful routine to count number of matches in a string.
Rusty Russell [Thu, 6 Jan 2011 03:50:26 +0000 (14:20 +1030)]
str: clean up tests so ccanlint doesn't complain about memory leaking.
Rusty Russell [Fri, 7 Jan 2011 11:51:29 +0000 (22:21 +1030)]
ccanlint: use positive description for test_pass_valgrind_noleaks
A bit more awkward, but more consistent with everything else.
Rusty Russell [Fri, 7 Jan 2011 11:50:44 +0000 (22:20 +1030)]
ccanlint: print keys in output
Since test keys are used for --target=, this is useful.
Rusty Russell [Fri, 7 Jan 2011 11:50:13 +0000 (22:20 +1030)]
ccanlint: rename files to match keys
Rusty Russell [Fri, 7 Jan 2011 11:50:04 +0000 (22:20 +1030)]
ccanlint: rename structures to match keys
Rusty Russell [Fri, 7 Jan 2011 11:49:49 +0000 (22:19 +1030)]
ccanlint: list dependencies by key
Joey Adams also pointed out that we should use strings for the dependency
lists. Moving them into the structure also somewhat simplifies it.
Rusty Russell [Fri, 7 Jan 2011 11:48:41 +0000 (22:18 +1030)]
ccanlint: rename test keys
Joey Adams rightly points out that the current keys are a mess: ideally the
filenames, test keys and structure names in ccanlint should be the same.
First step is to make the test names all regular, of basic form <noun>_<verb>
(eg "tests_exist" rather than "has-tests").
Rusty Russell [Fri, 7 Jan 2011 02:52:43 +0000 (13:22 +1030)]
Add scores/ directory to .gitignore.
Joey Adams [Thu, 6 Jan 2011 20:35:51 +0000 (15:35 -0500)]
ccanlint: added --test-dep-graph option
This option prints the dependency graph of ccanlint's tests
in Graphviz .dot format.
Sample usage:
ccanlint --test-dep-graph | dot -Tpng > out.png && eog out.png
Joey Adams [Thu, 6 Jan 2011 20:12:18 +0000 (15:12 -0500)]
ccanlint: fix segfault caused by tests not depending on the "info" test.
These tests:
"depends-exist" (compulsory_tests/check_depends_exist.c)
"info-documentation" (tests/has_info_documentation.c)
used m->info_file without checking if it was NULL,
leading to a segfault when no _info file was present.
Some other tests also used m->info_file without depending on "info",
but are taken care of indirectly by this patch.
Joey Adams [Thu, 6 Jan 2011 15:50:04 +0000 (10:50 -0500)]
opt: Fix warnings with gcc-4.5 (same approach as commit
6535bde)
&*ptr is used in some other macros, but at a glance, they look like
cases where the pointer shouldn't be NULL . Didn't change those,
and if we get more warnings, we'll cross that bridge when we get to it.
For now, I suppose they are just free NULL checks.
Rusty Russell [Thu, 6 Jan 2011 01:54:05 +0000 (12:24 +1030)]
daemonize: set stderr to /dev/null.
Rusty Russell [Thu, 6 Jan 2011 01:45:03 +0000 (12:15 +1030)]
daemonize: make valgrind happy.
Rusty Russell [Thu, 6 Jan 2011 01:37:37 +0000 (12:07 +1030)]
daemonize: use BSD-MIT as License: string in _info
The parenthesized thing is confusing and ccanlint dislikes it.
Rusty Russell [Thu, 6 Jan 2011 01:37:05 +0000 (12:07 +1030)]
ccanlint: allow BSD-MIT for MIT license.
There are a large number of BSD variants out there, be explicit.
Rusty Russell [Thu, 6 Jan 2011 00:56:47 +0000 (11:26 +1030)]
ccanlint: fix parsing bug which believes lines starting with - are a section header.
Rusty Russell [Tue, 4 Jan 2011 10:41:47 +0000 (21:11 +1030)]
ccanlint: have valgrind fail with an error, always
The upcoming failtest module can only tell that a child failed when it
exits with a non-zero error. So we need this, although it means for ccanlint
it still needs to look at output to distinguish a memory leak from a real
error.
Rusty Russell [Tue, 4 Jan 2011 10:41:47 +0000 (21:11 +1030)]
ccanlint: compile modules required by examples.
If an example #includes <ccan/foo/...> we assume it needs module foo,
but we would fail instead of building it if it isn't built.
Rusty Russell [Tue, 4 Jan 2011 10:41:47 +0000 (21:11 +1030)]
ccanlint: make get_manifest cache manifests
As we start doing more building of dependencies, this saves us effort.
Rusty Russell [Tue, 4 Jan 2011 10:42:41 +0000 (21:12 +1030)]
compiler: NORETURN
Rusty Russell [Sat, 1 Jan 2011 06:44:05 +0000 (17:14 +1030)]
tlist: typesafe variant of list module.
I chose not to do the "macro creates set of routines" approach, as
we can be almost as safe with a struct containing a pointer to the member
type.
Rusty Russell [Sat, 1 Jan 2011 06:43:04 +0000 (17:13 +1030)]
list: LIST_HEAD_INIT
I find hiding the declaration in LIST_HEAD() a bit weird.
Rusty Russell [Sat, 1 Jan 2011 06:42:40 +0000 (17:12 +1030)]
list: improve test coverage to 100%
In particular, we test the corruption cases.
Rusty Russell [Sat, 1 Jan 2011 06:41:15 +0000 (17:11 +1030)]
list: list_del_from()
Deletion from a specific list.
Rusty Russell [Sat, 1 Jan 2011 06:45:10 +0000 (17:15 +1030)]
list: test CCAN_LIST_DEBUG=1 case
Rusty Russell [Sat, 1 Jan 2011 06:40:54 +0000 (17:10 +1030)]
list: rename debug_list to list_debug, use list_check_node in list_del
When CCAN_LIST_DEBUG was defined, we were previously calling
list_check() in list_del instead of list_check_node(), which caused a
warning.
We also should stick within the "list_" prefix namespace, so rename
"debug_list" to "list_debug".
Rusty Russell [Sat, 1 Jan 2011 06:38:58 +0000 (17:08 +1030)]
list: implement list_check_node to check list from node rather than head
Rusty Russell [Sat, 1 Jan 2011 06:45:41 +0000 (17:15 +1030)]
config.h: HAVE_FLEXIBLE_ARRAY_MEMBER
Rusty Russell [Sat, 1 Jan 2011 06:42:00 +0000 (17:12 +1030)]
ccanlint: give a bonus for 100% code coverage
Rusty Russell [Thu, 23 Dec 2010 02:35:55 +0000 (13:05 +1030)]
tdb: add test for tdb_summary
Rusty Russell [Thu, 23 Dec 2010 02:14:48 +0000 (12:44 +1030)]
tdb: remove tally dependency
We lose the histograms, but this prepares it for upstream merge.
Rusty Russell [Thu, 23 Dec 2010 02:14:25 +0000 (12:44 +1030)]
tdb: fix uncoalesced record count in tdb_summary.
One free record is not "uncoalesced".
Rusty Russell [Thu, 23 Dec 2010 02:13:54 +0000 (12:43 +1030)]
tdb: combine dead_space helper from summary.c and check.c
Remove code duplication, and allow unit tests (which #include both).
Rusty Russell [Thu, 16 Dec 2010 05:33:41 +0000 (16:03 +1030)]
typesafe_cb: Fix warnings with gcc-4.5:
Test compiled with warnings:
/home/rusty/devel/cvs/ccan/ccan/typesafe_cb/test/compile_ok-typesafe_cb-NULL.c:/home/rusty/devel/cvs/ccan/ccan/typesafe_cb/test/compile_ok-typesafe_cb-NULL.c: In function ‘main’:
/home/rusty/devel/cvs/ccan/ccan/typesafe_cb/test/compile_ok-typesafe_cb-NULL.c:18:2: warning: taking address of expression of type ‘void’
/home/rusty/devel/cvs/ccan/ccan/typesafe_cb/test/compile_ok-typesafe_cb-NULL.c:18:2: warning: taking address of expression of type ‘void’
/home/rusty/devel/cvs/ccan/ccan/typesafe_cb/test/compile_ok-typesafe_cb-NULL.c:18:2: warning: taking address of expression of type ‘void’
/home/rusty/devel/cvs/ccan/ccan/typesafe_cb/test/compile_ok-typesafe_cb-NULL.c:18:2: warning: taking address of expression of type ‘void’
/home/rusty/devel/cvs/ccan/ccan/typesafe_cb/test/compile_ok-typesafe_cb-NULL.c:18:2: warning: taking address of expression of type ‘void’
/home/rusty/devel/cvs/ccan/ccan/typesafe_cb/test/compile_ok-typesafe_cb-NULL.c:18:2: warning: taking address of expression of type ‘void’
/home/rusty/devel/cvs/ccan/ccan/typesafe_cb/test/compile_ok-typesafe_cb-NULL.c:18:2: warning: taking address of expression of type ‘void’
/home/rusty/devel/cvs/ccan/ccan/typesafe_cb/test/compile_ok-typesafe_cb-NULL.c:19:2: warning: taking address of expression of type ‘void’
Rusty Russell [Mon, 13 Dec 2010 08:15:24 +0000 (18:45 +1030)]
ccanlint: use valgrind options when debugging too
If they specify valgrind options in their _info, we should use them when
running with --db-attach=yes, too.
Rusty Russell [Mon, 13 Dec 2010 08:15:25 +0000 (18:45 +1030)]
ccanlint: report valgrind "definite" leaks.
This is complicated by valgrind's limited options, and our desire not to run
valgrind twice (it's already the slowest part of the tests).
Ideally I'd like a different error code for each kind of error. I
could parse and pretty-print the XML output, but instead I just parse
the human-readable (which is fragile).
Rusty Russell [Mon, 13 Dec 2010 08:15:25 +0000 (18:45 +1030)]
ccanlint: fix bogus strerror printing when unknown test handed to --target
Rusty Russell [Mon, 13 Dec 2010 08:23:14 +0000 (18:53 +1030)]
ccanlint: fix error with --target=build
It requires that we build the objects first.
Rusty Russell [Wed, 8 Dec 2010 02:05:46 +0000 (12:35 +1030)]
rbtree: remove unused variable in tests.
Ronnie Sahlberg [Wed, 8 Dec 2010 01:00:35 +0000 (12:00 +1100)]
rb_tree: fix trbt_delete
trbt_delete32() was broken and caused SEGV as soon as you tried to
delete an object from a tree.
Rework trbt_delete32() to instead just call talloc_free() instread of trying
to call delete_node() directly.
This makes the "from_destructor" argument to delete_node() redundant
so that parameter is removed too.
Signed-off-by: Ronnie Sahlberg <sahlberg@lenovo-laptop.(none)>
Rusty Russell [Wed, 8 Dec 2010 00:44:21 +0000 (11:14 +1030)]
configurator: warnings count as failures too.
Unfortunately, gcc only warns if it sees an unknown attribute (in this case, gcc 4.1 vs "cold").
Rusty Russell [Wed, 8 Dec 2010 00:34:03 +0000 (11:04 +1030)]
ccanlint: fix compile on x86-64
cc -g -Wall -Wstrict-prototypes -Wold-style-definition -Wmissing-prototypes -Wmissing-declarations -I. -MD -Werror -c -o tools/ccanlint/tests/examples_run.o tools/ccanlint/tests/examples_run.c
cc1: warnings being treated as errors
tools/ccanlint/tests/examples_run.c: In function ‘scan_forv’:
tools/ccanlint/tests/examples_run.c:37: warning: passing argument 2 of ‘__builtin_va_copy’ discards qualifiers from pointer target type
tools/ccanlint/tests/examples_run.c:43: warning: passing argument 4 of ‘scan_forv’ from incompatible pointer type
tools/ccanlint/tests/examples_run.c:52: warning: passing argument 4 of ‘scan_forv’ from incompatible pointer type
tools/ccanlint/tests/examples_run.c:60: warning: passing argument 4 of ‘scan_forv’ from incompatible pointer type
tools/ccanlint/tests/examples_run.c: In function ‘scan_for’:
tools/ccanlint/tests/examples_run.c:78: warning: passing argument 4 of ‘scan_forv’ from incompatible pointer type
make: *** [tools/ccanlint/tests/examples_run.o] Error 1
It really doesn't like constifying a va_arg, so remove the const declaration.
Rusty Russell [Tue, 7 Dec 2010 05:19:08 +0000 (15:49 +1030)]
rbtree: add tests (which currently fail)
Rusty Russell [Mon, 6 Dec 2010 03:12:13 +0000 (13:42 +1030)]
idtree: fix overflow for v. large ids on allocation and removal
Chris Cowan tracked down a SEGV in sub_alloc: idp->level can actually
be equal to 7 (MAX_LEVEL) there, as it can be in sub_remove.
Rusty Russell [Mon, 6 Dec 2010 03:12:00 +0000 (13:42 +1030)]
idtree: fix right shift of signed ints
(Imported from SAMBA commit
2db1987f5a3a)
Right-shifting signed integers in undefined; indeed it seems that on
AIX with their compiler, doing a 30-bit shift on (INT_MAX-200) gives
0, not 1 as we might expect (THIS WAS WRONG, REAL FIX LATER).
The obvious fix is to make id and oid unsigned: l (level count) is also
logically unsigned.
(Note: Samba doesn't generally get to ids > 1 billion, but ctdb does)
Reported-by: Chris Cowan <cc@us.ibm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Rusty Russell [Mon, 6 Dec 2010 03:00:59 +0000 (13:30 +1030)]
idtree: import fix from SAMBA source
(Imported from
b53f8c187de8)
Author: Rusty Russell <rusty@rustorp.com.au>
Date: Thu Jun 10 13:27:51 2010 -0700
Since idtree assigns sequentially, it rarely reaches high numbers.
But such numbers can be forced with idr_get_new_above(), and that
reveals two bugs:
1) Crash in sub_remove() caused by pa array being too short.
2) Shift by more than 32 in _idr_find(), which is undefined, causing
the "outside the current tree" optimization to misfire and return NULL.
Rusty Russell [Mon, 6 Dec 2010 02:22:19 +0000 (12:52 +1030)]
idtree: add unit test for ids around INT_MAX
This causes a SEGV on my laptop.
Rusty Russell [Mon, 6 Dec 2010 02:21:40 +0000 (12:51 +1030)]
ccanlint: ignore comments when parsing examples.
We insert comments when we massage or combine examples; don't let these
throw off our analysis (as happened for idtree.h).
Rusty Russell [Mon, 6 Dec 2010 02:20:45 +0000 (12:50 +1030)]
ccanlint: show example we actually compiled when we report warnings.
Rusty Russell [Mon, 6 Dec 2010 02:19:31 +0000 (12:49 +1030)]
ccanlint: examples_compile depends on build, which depends on build_objs
We might as well use the compiled .o rather than all the little .o files.
Rusty Russell [Mon, 6 Dec 2010 02:18:18 +0000 (12:48 +1030)]
ccanlint: fix usage of static funcs
Out-by-one error had us using character prior to declaration, eg, in
"static int *foo" we use "*foo". This seems to compile, but is weird.
Rusty Russell [Fri, 3 Dec 2010 12:23:52 +0000 (22:53 +1030)]
array: fix type introduced in handling !HAVE_TYPEOF
Commit
da72623aec30 added a typo; ccanlint caught it, but doesn't consider
test compile failing to be fatal (it should!).
Rusty Russell [Wed, 1 Dec 2010 13:24:33 +0000 (23:54 +1030)]
Merge branch 'tdb2'
Rusty Russell [Wed, 1 Dec 2010 13:16:26 +0000 (23:46 +1030)]
tdb2: update documentation
Specifically the linked free tables, and reflect on the status of each
point of the design document.
Rusty Russell [Wed, 1 Dec 2010 13:15:56 +0000 (23:45 +1030)]
tdb2: use separate magic constants for chain, htable and ftable entries
Rather than overloading TDB_USED_MAGIC and the hash value as we do now.
We also rename "free list" to the more-accurate "free table" everywhere.
Rusty Russell [Wed, 1 Dec 2010 13:14:54 +0000 (23:44 +1030)]
tdb2: direct access during transactions.
Currently we fall back to copying data during a transaction, but we don't
need to in many cases. Grant direct access in those cases.
Before:
$ ./speed --transaction
1000000
Adding
1000000 records: 2409 ns (
59916680 bytes)
Finding
1000000 records: 1156 ns (
59916680 bytes)
Missing
1000000 records: 604 ns (
59916680 bytes)
Missing
1000000 records: 604 ns (
59916680 bytes)
Traversing
1000000 records: 1226 ns (
59916680 bytes)
Deleting
1000000 records: 1556 ns (
119361928 bytes)
Re-adding
1000000 records: 2326 ns (
119361928 bytes)
Appending
1000000 records: 3269 ns (
246656880 bytes)
Churning
1000000 records: 5613 ns (
338235248 bytes)
After:
$ ./speed --transaction
1000000
Adding
1000000 records: 1902 ns (
59916680 bytes)
Finding
1000000 records: 1032 ns (
59916680 bytes)
Missing
1000000 records: 606 ns (
59916680 bytes)
Missing
1000000 records: 606 ns (
59916680 bytes)
Traversing
1000000 records: 741 ns (
59916680 bytes)
Deleting
1000000 records: 1347 ns (
119361928 bytes)
Re-adding
1000000 records: 1727 ns (
119361928 bytes)
Appending
1000000 records: 2561 ns (
246656880 bytes)
Churning
1000000 records: 4403 ns (
338235248 bytes)