This is a part of the works that intends to make the netmgr stable,
testable, maintainable and tested. It contains a numerous changes to
the netmgr code and unfortunately, it was not possible to split this
into smaller chunks as the work here needs to be committed as a complete
works.
NOTE: There's a quite a lot of duplicated code between udp.c, tcp.c and
tcpdns.c and it should be a subject to refactoring in the future.
The changes that are included in this commit are listed here
(extensively, but not exclusively):
* The netmgr_test unit test was split into individual tests (udp_test,
tcp_test, tcpdns_test and newly added tcp_quota_test)
* The udp_test and tcp_test has been extended to allow programatic
failures from the libuv API. Unfortunately, we can't use cmocka
mock() and will_return(), so we emulate the behaviour with #define and
including the netmgr/{udp,tcp}.c source file directly.
* The netievents that we put on the nm queue have variable number of
members, out of these the isc_nmsocket_t and isc_nmhandle_t always
needs to be attached before enqueueing the netievent_<foo> and
detached after we have called the isc_nm_async_<foo> to ensure that
the socket (handle) doesn't disappear between scheduling the event and
actually executing the event.
* Cancelling the in-flight TCP connection using libuv requires to call
uv_close() on the original uv_tcp_t handle which just breaks too many
assumptions we have in the netmgr code. Instead of using uv_timer for
TCP connection timeouts, we use platform specific socket option.
* Fix the synchronization between {nm,async}_{listentcp,tcpconnect}
When isc_nm_listentcp() or isc_nm_tcpconnect() is called it was
waiting for socket to either end up with error (that path was fine) or
to be listening or connected using condition variable and mutex.
Several things could happen:
0. everything is ok
1. the waiting thread would miss the SIGNAL() - because the enqueued
event would be processed faster than we could start WAIT()ing.
In case the operation would end up with error, it would be ok, as
the error variable would be unchanged.
2. the waiting thread miss the sock->{connected,listening} = `true`
would be set to `false` in the tcp_{listen,connect}close_cb() as
the connection would be so short lived that the socket would be
closed before we could even start WAIT()ing
* The tcpdns has been converted to using libuv directly. Previously,
the tcpdns protocol used tcp protocol from netmgr, this proved to be
very complicated to understand, fix and make changes to. The new
tcpdns protocol is modeled in a similar way how tcp netmgr protocol.
Closes: #2194, #2283, #2318, #2266, #2034, #1920
* The tcp and tcpdns is now not using isc_uv_import/isc_uv_export to
pass accepted TCP sockets between netthreads, but instead (similar to
UDP) uses per netthread uv_loop listener. This greatly reduces the
complexity as the socket is always run in the associated nm and uv
loops, and we are also not touching the libuv internals.
There's an unfortunate side effect though, the new code requires
support for load-balanced sockets from the operating system for both
UDP and TCP (see #2137). If the operating system doesn't support the
load balanced sockets (either SO_REUSEPORT on Linux or SO_REUSEPORT_LB
on FreeBSD 12+), the number of netthreads is limited to 1.
* The netmgr has now two debugging #ifdefs:
1. Already existing NETMGR_TRACE prints any dangling nmsockets and
nmhandles before triggering assertion failure. This options would
reduce performance when enabled, but in theory, it could be enabled
on low-performance systems.
2. New NETMGR_TRACE_VERBOSE option has been added that enables
extensive netmgr logging that allows the software engineer to
precisely track any attach/detach operations on the nmsockets and
nmhandles. This is not suitable for any kind of production
machine, only for debugging.
* The tlsdns netmgr protocol has been split from the tcpdns and it still
uses the old method of stacking the netmgr boxes on top of each other.
We will have to refactor the tlsdns netmgr protocol to use the same
approach - build the stack using only libuv and openssl.
* Limit but not assert the tcp buffer size in tcp_alloc_cb
Closes: #2061
When calling the high level netmgr functions, the callback would be
sometimes called synchronously if we catch the failure directly, or
asynchronously if it happens later. The synchronous call to the
callback could create deadlocks as the caller would not expect the
failed callback to be executed directly.
this function sets the read timeout for the socket associated
with a netmgr handle and, if the timer is running, resets it.
for TCPDNS sockets it also sets the read timeout and resets the
timer on the outer TCP socket.
When we are operating on the tcpdns socket, we need to double check
whether the socket or its outerhandle or its listener or its mgr is
still active and when not, bail out early.
There were more races that could happen while connecting to a
socket while closing or shutting down the same socket. This
commit introduces a .closing flag to guard the socket from
being closed twice.
There was a data race where a new event could be scheduled after
isc__nm_async_shutdown() had cleaned up all the dangling UDP/TCP
sockets from the loop.
- don't bother closing sockets that are already closing.
- UDP read timeout timer was not stopped after reading.
- improve handling of TCP connection failures.
- isc_nm_tcpdnsconnect() sets up up an outgoing TCP DNS connection.
- isc_nm_tcpconnect(), _udpconnect() and _tcpdnsconnect() now take a
timeout argument to ensure connections time out and are correctly
cleaned up on failure.
- isc_nm_read() now supports UDP; it reads a single datagram and then
stops until the next time it's called.
- isc_nm_cancelread() now runs asynchronously to prevent assertion
failure if reading is interrupted by a non-network thread (e.g.
a timeout).
- isc_nm_cancelread() can now apply to UDP sockets.
- added shim code to support UDP connection in versions of libuv
prior to 1.27, when uv_udp_connect() was added
all these functions will be used to support outgoing queries in dig,
xfrin, dispatch, etc.
When client disconnects before the connection can be accepted, the named
would log a spurious log message:
error: Accepting TCP connection failed: socket is not connected
We now ignore the ISC_R_NOTCONNECTED result code and log only other
errors
1. The isc__nm_tcp_send() and isc__nm_tcp_read() was not checking
whether the socket was still alive and scheduling reads/sends on
closed socket.
2. The isc_nm_read(), isc_nm_send() and isc_nm_resumeread() have been
changed to always return the error conditions via the callbacks, so
they always succeed. This applies to all protocols (UDP, TCP and
TCPDNS).
There were two problems how tcp_send_direct() was used:
1. The tcp_send_direct() can return ISC_R_CANCELED (or translated error
from uv_tcp_send()), but the isc__nm_async_tcpsend() wasn't checking
the error code and not releasing the uvreq in case of an error.
2. In isc__nm_tcp_send(), when the TCP send is already in the right
netthread, it uses tcp_send_direct() to send the TCP packet right
away. When that happened the uvreq was not freed, and the error code
was returned to the caller. We need to return ISC_R_SUCCESS and
rather use the callback to report an error in such case.
When closing the socket that is actively reading from the stream, the
read_cb() could be called between uv_close() and close callback when the
server socket has been already detached hence using sock->statichandle
after it has been already freed.
When networking statistics was added to the netmgr (in commit
5234a8e00a6ae1df738020f27544594ccb8d5215), two lines were added that
increment the 'STATID_RECVFAIL' statistic: One if 'uv_read_start'
fails and one at the end of the 'read_cb'. The latter happens
if 'nread < 0'.
According to the libuv documentation, I/O read callbacks (such as for
files and sockets) are passed a parameter 'nread'. If 'nread' is less
than 0, there was an error and 'UV_EOF' is the end of file error, which
you may want to handle differently.
In other words, we should not treat EOF as a RECVFAIL error.
If we clone the csock (children socket) in TCP accept_connection()
instead of passing the ssock (server socket) to the call back and
cloning it there we unbreak the assumption that every socket is handled
inside it's own worker thread and therefore we can get rid of (at least)
callback locking.
The isc__nm_tcpdns_stoplistening() would call isc__nmsocket_clearcb()
that would clear the .accept_cb from non-netmgr thread. Change the
tcpdns_stoplistening to enqueue ievent that would get processed in the
right netmgr thread to avoid locking.
On POSIX based systems both uv_os_sock_t and uv_os_fd_t are both typedef
to int. That's not true on Windows, where uv_os_sock_t is SOCKET and
uv_os_fd_t is HANDLE and they differ in level of indirection.
The isc__nm_socket_freebind() has been refactored to match other
isc__nm_socket_...() helper functions and take uv_os_fd_t and
sa_family_t as function arguments.
Attaching and detaching handle pointers will make it easier to
determine where and why reference counting errors have occurred.
A handle needs to be referenced more than once when multiple
asynchronous operations are in flight, so callers must now maintain
multiple handle pointers for each pending operation. For example,
ns_client objects now contain:
- reqhandle: held while waiting for a request callback (query,
notify, update)
- sendhandle: held while waiting for a send callback
- fetchhandle: held while waiting for a recursive fetch to
complete
- updatehandle: held while waiting for an update-forwarding
task to complete
control channel connection objects now contain:
- readhandle: held while waiting for a read callback
- sendhandle: held while waiting for a send callback
- cmdhandle: held while an rndc command is running
httpd connections contain:
- readhandle: held while waiting for a read callback
- sendhandle: held while waiting for a send callback
- rename isc_nmsocket_t->tcphandle to statichandle
- cancelread functions now take handles instead of sockets
- add a 'client' flag in socket objects, currently unused, to
indicate whether it is to be used as a client or server socket
Each worker has a receive buffer with space for 20 DNS messages of up
to 2^16 bytes each, and the allocator function passed to uv_read_start()
or uv_udp_recv_start() will reserve a portion of it for use by sockets.
UDP can use recvmmsg() and so it needs that entire space, but TCP reads
one message at a time.
This commit introduces separate allocator functions for TCP and UDP
setting different buffer size limits, so that libuv will provide the
correct buffer sizes to each of them.
When a new IPv6 interface/address appears it's first in a tentative
state - in which we cannot bind to it, yet it's already being reported
by the route socket. Because of that BIND9 is unable to listen on any
newly detected IPv6 addresses. Fix it by setting IP_FREEBIND option (or
equivalent option on other OSes) and then retrying bind() call.
There's a possibility of race in isc__nm_tcpconnect if the asynchronous
connect operation finishes with all the callbacks before we exit the
isc__nm_tcpconnect itself we might access an already freed memory.
Fix it by creating an additional reference to the socket freed at the
end of isc__nm_tcpconnect.
the blackhole ACL was accidentally disabled with respect to client
queries during the netmgr conversion.
in order to make this work for TCP, it was necessary to add a return
code to the accept callback functions passed to isc_nm_listentcp() and
isc_nm_listentcpdns().
The isc_nm_cancelread() function cancels reading on a connected
socket and calls its read callback function with a 'result'
parameter of ISC_R_CANCELED.
the isc_nm_tcpconnect() function establishes a client connection via
TCP. once the connection is esablished, a callback function will be
called with a newly created network manager handle.
A TCPDNS socket creates a handle for each complete DNS message.
Previously, when all the handles were disconnected, the socket
would be closed, but the wrapped TCP socket might still have
more to read.
Now, when a connection is established, the TCPDNS socket creates
a reference to itself by attaching itself to sock->self. This
reference isn't cleared until the connection is closed via
EOF, timeout, or server shutdown. This allows the socket to remain
open even when there are no active handles for it.
- isc__nmhandle_get() now attaches to the sock in the nmhandle object.
the caller is responsible for dereferencing the original socket
pointer when necessary.
- tcpdns listener sockets attach sock->outer to the outer tcp listener
socket. tcpdns connected sockets attach sock->outerhandle to the handle
for the tcp connected socket.
- only listener sockets need to be attached/detached directly. connected
sockets should only be accessed and reference-counted via their
associated handles.
there is no need for a caller to reference-count socket objects.
they need tto be able tto close listener sockets (i.e., those
returned by isc_nm_listen{udp,tcp,tcpdns}), and an isc_nmsocket_close()
function has been added for that. other sockets are only accessed via
handles.
There's a possibility of a race in TCP accepting code:
T1 accepts a connection C1
T2 accepts a connection C2
T1 tries to accept a connection C3, but we hit a quota,
isc_quota_cb_init() sets quota_accept_cb for the socket,
we return from accept_connection
T2 drops C2, but we race in quota_release with accepting C3 so
we don't see quota->waiting is > 0, we don't launch the callback
T1 accepts a connection C4, we are able to get the quota we clear
the quota_accept_cb from sock->quotacb
T1 drops C1, tries to call the callback which is zeroed, sigsegv.
Instead of using bind() and passing the listening socket to the children
threads using uv_export/uv_import use one thread that does the accepting,
and then passes the connected socket using uv_export/uv_import to a random
worker. The previous solution had thundering herd problems (all workers
waking up on one connection and trying to accept()), this one avoids this
and is simpler.
The tcp clients quota is simplified with isc_quota_attach_cb - a callback
is issued when the quota is available.
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
tcpdns used transport-specific functions to operate on the outer socket.
Use generic ones instead, and select the proper call in netmgr.c.
Make the missing functions (e.g. isc_nm_read) generic and add type-specific
calls (isc__nm_tcp_read). This is the preparation for netmgr TLS layer.
In tcp and udp stoplistening code we accessed libuv structures
from a different thread, which caused a shutdown crash when named
was under load. Also added additional DbC checks making sure we're
in a proper thread when accessing uv_ functions.
