There were several problems with rbt hashtable implementation:
1. Our internal hashing function returns uint64_t value, but it was
silently truncated to unsigned int in dns_name_hash() and
dns_name_fullhash() functions. As the SipHash 2-4 higher bits are
more random, we need to use the upper half of the return value.
2. The hashtable implementation in rbt.c was using modulo to pick the
slot number for the hash table. This has several problems because
modulo is: a) slow, b) oblivious to patterns in the input data. This
could lead to very uneven distribution of the hashed data in the
hashtable. Combined with the single-linked lists we use, it could
really hog-down the lookup and removal of the nodes from the rbt
tree[a]. The Fibonacci Hashing is much better fit for the hashtable
function here. For longer description, read "Fibonacci Hashing: The
Optimization that the World Forgot"[b] or just look at the Linux
kernel. Also this will make Diego very happy :).
3. The hashtable would rehash every time the number of nodes in the rbt
tree would exceed 3 * (hashtable size). The overcommit will make the
uneven distribution in the hashtable even worse, but the main problem
lies in the rehashing - every time the database grows beyond the
limit, each subsequent rehashing will be much slower. The mitigation
here is letting the rbt know how big the cache can grown and
pre-allocate the hashtable to be big enough to actually never need to
rehash. This will consume more memory at the start, but since the
size of the hashtable is capped to `1 << 32` (e.g. 4 mio entries), it
will only consume maximum of 32GB of memory for hashtable in the
worst case (and max-cache-size would need to be set to more than
4TB). Calling the dns_db_adjusthashsize() will also cap the maximum
size of the hashtable to the pre-computed number of bits, so it won't
try to consume more gigabytes of memory than available for the
database.
FIXME: What is the average size of the rbt node that gets hashed? I
chose the pagesize (4k) as initial value to precompute the size of
the hashtable, but the value is based on feeling and not any real
data.
For future work, there are more places where we use result of the hash
value modulo some small number and that would benefit from Fibonacci
Hashing to get better distribution.
Notes:
a. A doubly linked list should be used here to speedup the removal of
the entries from the hashtable.
b. https://probablydance.com/2018/06/16/fibonacci-hashing-the-optimization-that-the-world-forgot-or-a-better-alternative-to-integer-modulo/
The rewrite of BIND 9 build system is a large work and cannot be reasonable
split into separate merge requests. Addition of the automake has a positive
effect on the readability and maintainability of the build system as it is more
declarative, it allows conditional and we are able to drop all of the custom
make code that BIND 9 developed over the years to overcome the deficiencies of
autoconf + custom Makefile.in files.
This squashed commit contains following changes:
- conversion (or rather fresh rewrite) of all Makefile.in files to Makefile.am
by using automake
- the libtool is now properly integrated with automake (the way we used it
was rather hackish as the only official way how to use libtool is via
automake
- the dynamic module loading was rewritten from a custom patchwork to libtool's
libltdl (which includes the patchwork to support module loading on different
systems internally)
- conversion of the unit test executor from kyua to automake parallel driver
- conversion of the system test executor from custom make/shell to automake
parallel driver
- The GSSAPI has been refactored, the custom SPNEGO on the basis that
all major KRB5/GSSAPI (mit-krb5, heimdal and Windows) implementations
support SPNEGO mechanism.
- The various defunct tests from bin/tests have been removed:
bin/tests/optional and bin/tests/pkcs11
- The text files generated from the MD files have been removed, the
MarkDown has been designed to be readable by both humans and computers
- The xsl header is now generated by a simple sed command instead of
perl helper
- The <irs/platform.h> header has been removed
- cleanups of configure.ac script to make it more simpler, addition of multiple
macros (there's still work to be done though)
- the tarball can now be prepared with `make dist`
- the system tests are partially able to run in oot build
Here's a list of unfinished work that needs to be completed in subsequent merge
requests:
- `make distcheck` doesn't yet work (because of system tests oot run is not yet
finished)
- documentation is not yet built, there's a different merge request with docbook
to sphinx-build rst conversion that needs to be rebased and adapted on top of
the automake
- msvc build is non functional yet and we need to decide whether we will just
cross-compile bind9 using mingw-w64 or fix the msvc build
- contributed dlz modules are not included neither in the autoconf nor automake
This commit simplifies a bit the lock management within dns_resolver_prime()
and prime_done() functions by means of turning resolver's attribute
"priming" into an atomic_bool and by creating only one dependent object on the
lock "primelock", namely the "primefetch" attribute.
By having the attribute "priming" as an atomic type, it save us from having to
use a lock just to test if priming is on or off for the given resolver context
object, within "dns_resolver_prime" function.
The "primelock" lock is still necessary, since dns_resolver_prime() function
internally calls dns_resolver_createfetch(), and whenever this function
succeeds it registers an event in the task manager which could be called by
another thread, namely the "prime_done" function, and this function is
responsible for disposing the "primefetch" attribute in the resolver object,
also for resetting "priming" attribute to false.
It is important that the invariant "priming == false AND primefetch == NULL"
remains constant, so that any thread calling "dns_resolver_prime" knows for sure
that if the "priming" attribute is false, "primefetch" attribute should also be
NULL, so a new fetch context could be created to fulfill this purpose, and
assigned to "primefetch" attribute under the lock protection.
To honor the explanation above, dns_resolver_prime is implemented as follow:
1. Atomically checks the attribute "priming" for the given resolver context.
2. If "priming" is false, assumes that "primefetch" is NULL (this is
ensured by the "prime_done" implementation), acquire "primelock"
lock and create a new fetch context, update "primefetch" pointer to
point to the newly allocated fetch context.
3. If "priming" is true, assumes that the job is already in progress,
no locks are acquired, nothing else to do.
To keep the previous invariant consistent, "prime_done" is implemented as follow:
1. Acquire "primefetch" lock.
2. Keep a reference to the current "primefetch" object;
3. Reset "primefetch" attribute to NULL.
4. Release "primefetch" lock.
5. Atomically update "priming" attribute to false.
6. Destroy the "primefetch" object by using the temporary reference.
This ensures that if "priming" is false, "primefetch" was already reset to NULL.
It doesn't make any difference in having the "priming" attribute not protected
by a lock, since the visible state of this variable would depend on the calling
order of the functions "dns_resolver_prime" and "prime_done".
As an example, suppose that instead of using an atomic for the "priming" attribute
we employed a lock to protect it.
Now suppose that "prime_done" function is called by Thread A, it is then preempted
before acquiring the lock, thus not reseting "priming" to false.
In parallel to that suppose that a Thread B is scheduled and that it calls
"dns_resolver_prime()", it then acquires the lock and check that "priming" is true,
thus it will consider that this resolver object is already priming and it won't do
any more job.
Conversely if the lock order was acquired in the other direction, Thread B would check
that "priming" is false (since prime_done acquired the lock first and set "priming" to false)
and it would initiate a priming fetch for this resolver.
An atomic variable wouldn't change this behavior, since it would behave exactly the
same, depending on the function call order, with the exception that it would avoid
having to use a lock.
There should be no side effects resulting from this change, since the previous
implementation employed use of the more general resolver's "lock" mutex, which
is used in far more contexts, but in the specifics of the "dns_resolver_prime"
and "prime_done" it was only used to protect "primefetch" and "priming" attributes,
which are not used in any of the other critical sections protected by the same lock,
thus having zero dependency on those variables.
Both clang-tidy and uncrustify chokes on statement like this:
for (...)
if (...)
break;
This commit uses a very simple semantic patch (below) to add braces around such
statements.
Semantic patch used:
@@
statement S;
expression E;
@@
while (...)
- if (E) S
+ { if (E) { S } }
@@
statement S;
expression E;
@@
for (...;...;...)
- if (E) S
+ { if (E) { S } }
@@
statement S;
expression E;
@@
if (...)
- if (E) S
+ { if (E) { S } }
Previously, the dns_name API used isc_thread_key API for TLS, which is
fairly complicated and requires initialization of memory contexts, etc.
This part of code was refactored to use a ISC_THREAD_LOCAL pointer which
greatly simplifies the whole code related to storing TLS variables.
The dns_name_copy() function followed two different semanitcs that was driven
whether the last argument was or wasn't NULL. This commit splits the function
in two where now third argument to dns_name_copy() can't be NULL and
dns_name_copynf() doesn't have third argument.
- this enables memory to be allocated and freed in dyndb modules
when named is linked statically. when we standardize on libtool,
this should become unnecessary.
- also, simplified the isc_mem_create/createx API by removing
extra compatibility functions
Since idn_output_filter() no longer uses its 'absolute' argument and no
other callback is used with dns_name_settotextfilter(), remove the
'absolute' argument from the dns_name_totextfilter_t prototype.
Replace dns_fixedname_init() calls followed by dns_fixedname_name()
calls with calls to dns_fixedname_initname() where it is possible
without affecting current behavior and/or performance.
This patch was mostly prepared using Coccinelle and the following
semantic patch:
@@
expression fixedname, name;
@@
- dns_fixedname_init(&fixedname);
...
- name = dns_fixedname_name(&fixedname);
+ name = dns_fixedname_initname(&fixedname);
The resulting set of changes was then manually reviewed to exclude false
positives and apply minor tweaks.
It is likely that more occurrences of this pattern can be refactored in
an identical way. This commit only takes care of the low-hanging fruit.
