The init code of the mworker_proc structs has been moved before the
init of the listeners.
Each socketpair is now connected to a CLI within the workers, which
allows the master to access their CLI.
The inherited flag of the worker side socketpair is removed so the
socket can be closed in the master.
With the new synchronous si_cs_send() at the end of process_stream(),
we're seeing re-appear the I/O layer specific part of the stream interface
which is supposed to deal with I/O event subscription. The only difference
is that now we subscribe to I/Os only after having attempted (and failed)
them.
This patch brings a cleanup in this by reintroducing stream_int_update_conn()
with the send code from process_stream(). However this alone would not be
enough because the flags which are cleared afterwards would result in the
loss of the possible events (write events only at the moment). So the flags
clearing and stream-int state updates are also performed inside si_update()
between the generic code and the I/O specific code. This definitely makes
sense as after this call we can simply check again for channel and SI flag
changes and decide to loop once again or not.
This will supersed channel_alloc_buffer() while relying on it. It will
automatically adjust SI_FL_WAIT_ROOM on the stream-int depending on
success or failure to allocate this buffer.
It's worth noting that it could make sense to also set SI_FL_WANT_PUT
each time we do this to further simplify the code at user places such
as applets, but it would possibly not be easy to clean this flag
everywhere an rx operation stops.
The behaviour of the flag CF_WRITE_PARTIAL was modified by commit
95fad5ba4 ("BUG/MAJOR: stream-int: don't re-arm recv if send fails") due
to a situation where it could trigger an immediate wake up of the other
side, both acting in loops via the FD cache. This loss has caused the
need to introduce CF_WRITE_EVENT as commit c5a9d5bf, to replace it, but
both flags express more or less the same thing and this distinction
creates a lot of confusion and complexity in the code.
Since the FD cache now acts via tasklets, the issue worked around in the
first patch no longer exists, so it's more than time to kill this hack
and to restore CF_WRITE_PARTIAL's semantics (i.e.: there has been some
write activity since we last left process_stream).
This patch mostly reverts the two commits above. Only the part making
use of CF_WROTE_DATA instead of CF_WRITE_PARTIAL to detect the loss of
data upon connection setup was kept because it's more accurate and
better suited.
This patch makes the capable of storing HTTP objects larger than a buffer.
It makes usage of the "block by block shared object allocation" new shctx API.
A new pointer to struct shared_block has been added to the cache applet
context to memorize the next block to be used by the HTTP cache I/O handler
http_cache_io_handler() to emit the data. Another member, named "sent" memorize
the number of bytes already sent by this handler. So, to send an object from cache,
http_cache_io_handler() must be called until "sent" counter reaches the size
of this object.
This patch makes shctx capable of storing objects in several parts,
each parts being made of several blocks. There is no more need to
walk through until reaching the end of a row to append new blocks.
A new pointer to a struct shared_block member, named last_reserved,
has been added to struct shared_block so that to memorize the last block which was
reserved by shctx_row_reserve_hot(). Same thing about "last_append" pointer which
is used to memorize the last block used by shctx_row_data_append() to store the data.
This option makes a proxy use only HTX-compatible muxes instead of the
HTTP-compatible ones for HTTP modes. It must be set on both ends, this
is checked at parsing time.
Some samples representing time will cover more than one sample at once
if they are units of time per time. For this we'd need to have the
ability to loop over swrate_add() multiple times but that would be
inefficient. By developing the function elevated to power N, it's
visible that some coefficients quickly disappear and that those which
remain at the first order more or less compensate each other.
Thus a simplified version of this function was added to provide a single
value for a given number of samples. Tests with multiple values, window
sizes and sample sizes have shown that it is possible to make it remain
surprisingly accurate (typical error < 0.2% over various large window
and sample sizes, even samples representing up to 1/4 of the window).
Avoid using conn_xprt_want_send/recv, and totally nuke cs_want_send/recv,
from the upper layers. The polling is now directly handled by the connection
layer, it is activated on subscribe(), and unactivated once we got the event
and we woke the related task.
Make sure we don't have any subscription when the connection is going in
idle mode, otherwise there's a race condition when the connection is
reused, if there are still old subscriptions, new ones won't be done.
No backport is needed.
When mapping memory with mmap(), we should use a fd of -1, not 0. 0 may
work on linux, but it doesn't work on FreeBSD, and probably other OSes.
It would be nice to backport this to 1.8 to help debugging there.
Commit ac6c880 ("BUILD: memory: fix pointer declaration for atomic CAS")
attemtped to fix a build warning affecting the lock-free version of the
pool allocator. But the fix tried to hide the cause instead of addressing
it, thus clang still complains about (void **) not matching (void ***).
The real solution is to declare free_list (void **) and not to use a cast.
Now this builds fine with gcc/clang with and without threads.
No backport is needed.
The purpose is to detect if threads or processes are competing for the
same CPU. This can happen when threads are incorrectly bound, or after a
reload if the previous process still has an important activity. With
threads this situation is problematic because a preempted thread holding
a lock will block other ones waiting for this lock to be released.
A first attempt consisted in measuring the cumulated lost time more
precisely but the system's scheduler is smart enough to try to limit the
thread preemption rate by mostly context switching during poll()'s blank
periods, so most of the time lost is not seen. In essence this is good
because it means a thread is not preempted with a lock held, and even
regarding the rendez-vous point it cannot prevent the other ones from
making progress. But still it happens tens to hundreds of times per
second that a thread might be preempted, so it's still possible to detect
that the situation is happening, thus it's interesting to measure and
report its frequency.
Each time we enter the poller, we check the CPU time spent working and
see if we've lost time doing something else. To limit false positives,
we're only interested in losses of 500 microseconds or more (i.e. half
a clock tick on a 1 kHz system). If so, it indicates that some time was
stolen by another thread or process. Note that we purposely store some
sub-millisecond counters so that under heavy traffic with a 1 kHz clock,
it's still possible to measure something without being subject to the
risk of rounding errors (i.e. if exactly 1 ms is stolen it's possible
that the time difference could often be slightly lower).
This counter of lost CPU time slots time is reported in "show activity"
in numbers of milliseconds of CPU lost per second, per 15s, and total
over the process' life. By definition, the per-second counter cannot
report values larger than 1000 per thread per second and the 15s one
will be limited to 15000/s in the worst case, but it's possible that
peak values exceed such thresholds after long pauses.
These two functions retrieve respectively the monotonic clock time and
the per-thread CPU time when available on the platform, or return zero.
These syscalls may require to link with -lrt on certain libc, which is
enabled in the Makefile with USE_RT=1 (default on Linux systems).
The calls to HA_ATOMIC_CAS() on the lockfree version of the pool allocator
were mistakenly done on (void*) for the old value instead of (void **).
While this has no impact on "recent" gcc, it does have one for gcc < 4.7
since the CAS was open coded and it's not possible to assign a temporary
variable of type "void".
No backport is needed, this only affects 1.9.
By placing this code into time.h (tv_entering_poll() and tv_leaving_poll())
we can remove the logic from the pollers and prepare for extending this to
offer more accurate time measurements.
The 4 pollers all contain the same code used to compute the poll timeout.
This is pointless, let's centralize this into fd.h. This also gets rid of
the useless SCHEDULER_RESOLUTION macro which used to work arond a very old
linux 2.2 bug causing select() to wake up slightly before the timeout.
Each thread now keeps the last ~512 kB of freed objects into a local
cache. There are some heuristics involved so that a specific pool cannot
use more than 1/8 of the total cache in number of objects. Tests have
shown that 512 kB is an optimal size on a 24-thread test running on a
dual-socket machine, resulting in an overall 7.5% performance increase
and a cache miss ratio reducing from 19.2 to 17.7%. Anyway it seems
pointless to keep more than an L2 cache, which probably explains why
sizes between 256 and 512 kB are optimal.
Cached objects appear in two lists, one per pool and one LRU to help
with fair eviction. Currently there is no way to check each thread's
cache state nor to flush it. This cache cannot be disabled and is
enabled as soon as the lockless pools are enabled (i.e.: threads are
enabled, no pool debugging is in use and the CPU supports a double word
CAS).
For caching it will be convenient to have indexes associated with pools,
without having to dereference the pool itself. One solution could consist
in replacing all pool pointers with integers but this would limit the
number of allocatable pools. Instead here we allocate the 32 first pools
from a pre-allocated array whose base address is known so that it's trivial
to convert a pool to an index in this array. Pools that cannot fit there
will be allocated normally.
Currently we have per-thread arrays of trees and counts, but these
ones unfortunately share cache lines and are accessed very often. This
patch moves the task-specific stuff into a structure taking a multiple
of a cache line, and has one such per thread. Just doing this has
reduced the cache miss ratio from 19.2% to 18.7% and increased the
12-thread test performance by 3%.
It starts to become visible that we really need a process-wide per-thread
storage area that would cover more than just these parts of the tasks.
The code was arranged so that it's easy to move the pieces elsewhere if
needed.
Now we still have a main contention point with the timers in the main
wait queue, but the vast majority of the tasks are pinned to a single
thread. This patch creates a per-thread wait queue and queues a task
to the local wait queue without any locking if the task is bound to a
single thread (the current one) otherwise to the shared queue using
locking. This significantly reduces contention on the wait queue. A
test with 12 threads showed 11 ms spent in the WQ lock compared to
4.7 seconds in the same test without this change. The cache miss ratio
decreased from 19.7% to 19.2% on the 12-thread test, and its performance
increased by 1.5%.
Another indirect benefit is that the average queue size is divided
by the number of threads, which roughly removes log(nbthreads) levels
in the tree and further speeds up lookups.
The vast majority of FDs are only seen by one thread. Currently the lock
on FDs costs a lot because it's touched often, though there should be very
little contention. This patch ensures that the lock is only grabbed if the
FD is shared by more than one thread, since otherwise the situation is safe.
Doing so resulted in a 15% performance boost on a 12-threads test.
peers_init_sync() doesn't check task_new()'s return value and doesn't
return any result to indicate success or failure. Let's make it return
an int and check it from the caller.
This can be backported as far as 1.6.
This statement is used as a hint for the compiler so that it knows that
the location where it's placed cannot be reached. It will mostly be used
after longjmp() or equivalent statements that deal with error processing
and that the compiler doesn't know will not return on certain conditions,
so that it doesn't complain about null dereferences on error paths.
Some pseudo-headers are added during the headers parsing, mainly for the mux
H2. With this flag, it is possible to not add them. This avoid some boring
filtering in the mux H1.
Instead of using offsets relating to the parsed buffer to store start line
infos, we now use indirect strings. So now, these infos remain valid only if the
origin buffer remains untouched. But it's not a real problem because this union
is used during the parsing and never stored to a later use.
This flags will be used by multiplexers to warn a conn-stream (and, by
transitivity, a stream) it is not the first one created by the mux. It will help
mux H1 to handle keep-alive connections.
Since keep-alive mode is the default mode, the passive close has disappeared,
and in the code, httpclose and forceclose options are handled the same way:
connections with the client and the server are closed as soon as the request and
the response are received and missing "Connection: close" header is added in
each direction.
So to make things clearer, forceclose is now an alias for httpclose. And
httpclose is explicitly an active close. So the old passive close does not exist
anymore. Internally, the flag PR_O_HTTP_PCL has been removed and PR_O_HTTP_FCL
has been replaced by PR_O_HTTP_CLO. In HTTP analyzers, the checks done to find
the right mode to use, depending on proxies options and "Connection: " header
value, have been simplified.
This should only be a cleanup and no changes are expected.
To ease the refactoring, the function "http_header_add_tail" have been
remove. Now, "http_header_add_tail2" is always used. And the function
"capture_headers" have been renamed into "http_capture_headers". Finally, some
functions have been exported.
HTTP_FLG_* and HTTP_IS_* were moved from "proto/proto_http.h" to "common/http.h"
but the associated comment was forgotten during the move.
This is 1.9-specific and should not be backported.
Make sure we unsubscribe from events before si_release_endpoint destroys
the conn_stream, or it will be never called. To do so, move the call to
unsubscribe to si_release_endpoint() directly.
This is 1.9-specific and shouldn't be backported.
When subscribing, we don't need to provide a list element, only the h2 mux
needs it. So instead, Add a list element to struct h2s, and use it when a
list is needed.
This forces us to use the unsubscribe method, since we can't just unsubscribe
by using LIST_DEL anymore.
This patch is larger than it should be because it includes some renaming.
As we don't know how subscriptions are handled, we can't just assume we can
use LIST_DEL() to unsubscribe, so introduce a new method to mux and connections
to do so.
This call is now used quite a bit in the fd cache, to decide which cache
to add/remove the fd to/from, when waking up a task for a single thread
in __task_wakeup(), in fd_cant_recv() and in fd_process_cached_events(),
and we can replace it with a single instruction, removing ~30 instructions
and ~80 bytes from the inner loop of some of these functions.
In addition the test for zero value was replaced with a comment saying
that it is illegal and leads to an undefined behaviour. The code does
not make use of this useless case today.
In commit f161d0f51 ("BUG/MINOR: pools/threads: don't ignore DEBUG_UAF
on double-word CAS capable archs") I moved some defines and accidently
messed up with lockfree pools. The problem is that the HA_HAVE_CAS_DW
macro is not defined anymore where the CONFIG_HAP_LOCKLESS_POOLS macro
is set, so this fix implicitly disabled lockfree pools.
This patch fixes this by moving the capabilities definition to config.h
(probably that we'd benefit from having an "arch.h" file to declare the
capabilities offered by the architecture). In a test on a 12-core machine,
we used to measure 19s spent in the pool lock for 1M requests without
this patch, and 0 with it so that's definitely a net saving.
No backport is required, this is only for 1.9.
OpenSSL released support for TLSv1.3. It also added a separate function
SSL_CTX_set_ciphersuites that is used to set the ciphers used in the
TLS 1.3 handshake. This change adds support for that new configuration
option by adding a ciphersuites configuration variable that works
essentially the same as the existing ciphers setting.
Note that it should likely be backported to 1.8 in order to ease usage
of the now released openssl-1.1.1.
In ci_insert_line2() and b_rep_blk(), we can't afford to wrap, so don't use
b_tail() to check if we do, use __b_tail() instead.
This should be backported to previous versions.
The prototypes of functions find_hdr_value_end(), extract_cookie_value()
and http_header_match2() were still in proto_http.h while some of them
don't exist anymore and the others were just moved. Let's remove them.
In addition, da.c was updated to use http_extract_cookie_value() which
is the correct one.
These ones are mostly called from cfgparse.c for the parsing and do
not depend on the HTTP representation. The functions's prototypes
were moved to proto/http_rules.h, making this file work exactly like
tcp_rules. Ideally we should stop calling these functions directly
from cfgparse and register keywords, but there are a few cases where
that wouldn't work (stats http-request) so it's probably not worth
trying to go this far.
The current proto_http.c file is huge and contains different processing
domains making it very difficult to work on an alternative representation.
This commit moves some parts to other files :
- ACL registration code => http_acl.c
This code only creates some ACL mappings and doesn't know anything
about HTTP nor about the representation. This code could even have
moved to acl.c but it was not worth polluting it again.
- HTTP sample conversion => http_conv.c
This code doesn't depend on the internal representation but definitely
manipulates some HTTP elements, such as dates. It also has access to
captures.
- HTTP sample fetching => http_fetch.c
This code does depend entirely on the internal representation but is
totally independent on the analysers. Placing it into a different
file will ease the transition to the new representation and the
creation of a wrapper if required. An include file was created due
to CHECK_HTTP_MESSAGE_FIRST() being used at various places.
- HTTP action registration => http_act.c
This code doesn't directly interact with the messages nor the
transaction but it does so via some exported http functions like
http_replace_req_line() or http_set_status() so it will be easier
to change only this after the conversion.
- a few very generic parts were found and moved to http.{c,h} as
relevant.
It is worth noting that the functions moved to these new files are not
referenced anywhere outside of the files and are only called as registered
callbacks, so these files do not even require associated include files.
This ads support for accessing stick tables from Lua. The supported
operations are reading general table info, lookup by string/IP key, and
dumping the table.
Similar to "show table", a data filter is available during dump, and as
an improvement over "show table" it's possible to use up to 4 filter
expressions instead of just one (with implicit AND clause binding the
expressions). Dumping with/without filters can take a long time for
large tables, and should be used sparingly.
Instead of using si_cs_io_cb() in process_stream() use si_cs_send/si_cs_recv
instead, as si_cs_io_cb() may lead to process_stream being woken up when it
shouldn't be, and thus timeout would never get triggered.
Tim Düsterhus found using afl-fuzz that some parts of the HPACK decoder
use incorrect bounds checking which do not catch negative values after
a type cast. The first culprit is hpack_valid_idx() which takes a signed
int and is fed with an unsigned one, but a few others are affected as
well due to being designed to work with an uint16_t as in the table
header, thus not being able to detect the high offset bits, though they
are not exposed if hpack_valid_idx() is fixed.
The impact is that the HPACK decoder can be crashed by an out-of-bounds
read. The only work-around without this patch is to disable H2 in the
configuration.
CVE-2018-14645 was assigned to this bug.
This patch addresses all of these issues at once. It must be backported
to 1.8.
Callers of si_appctx() always use the result without checking it because
they know by construction that it's valid. This results in unchecked null
pointer warnings at -Wextra, so let's remove this test and make it clear
that it's up to the caller to check validity first.
stktable_data_ptr() currently performs null pointer checks but most
callers don't check the result since they know by construction that
it cannot be null. This causes valid warnings when building with
-Wextra which are worth addressing since it will result in better
code. Let's provide an unguarded version of this function for use
where the check is known to be useless and untested.
These two functions were apparently written on the same model as their
parents when added by commit 11bcb6c4f ("[MEDIUM] IPv6 support for syslog")
except that they perform an assignment instead of a return, and as a
result fall through the next case where the assigned value may possibly
be partially overwritten. At least under Linux the port offset is the
same in both sockaddr_in and sockaddr_in6 so the value is written twice
without side effects.
This needs to be backported as far as 1.5.
The transfer-encoding header processing was a bit lenient in this part
because it was made to read messages already validated by haproxy. We
absolutely need to reinstate the strict processing defined in RFC7230
as is currently being done in proto_http.c. That is, transfer-encoding
presence alone is enough to cancel content-length, and must be
terminated by the "chunked" token, except in the response where we
can fall back to the close mode if it's not last.
For this we now use a specific parsing function which updates the
flags and we introduce a new flag H1_MF_XFER_ENC indicating that the
transfer-encoding header is present.
Last, if such a header is found, we delete all content-length header
fields found in the message.
This flag is usefull to handle cases where there is no body, regardless of CL or
TE headers (for instance, responses to HEAD requests). It will not be set by the
parser itself.
The new function h1_parse_connection_header() is called when facing a
connection header in the generic parser, and it will set up to 3 bits
in h1m->flags indicating if at least one "close", "keep-alive" or "upgrade"
tokens was seen.
This will be needed for the mux to know how to process the Connection
header, and will save it from having to re-parse the request line since
it's captured on the fly.
Till now it was very difficult for a mux to know what proxy it was
working for. Let's pass the proxy when the mux is instanciated at
init() time. It's not yet used but the H1 mux will definitely need
it, just like the H2 mux when dealing with backend connections.
The h1 parser used to systematically turn header field names to lower
case because it was designed for H2. Let's add a flag which is off by
default to condition this behaviour so that when using it from an H1
parser it will not affect the message.
This state was only a delimiter between headers and body but it now
causes more harm than good because it requires someone to change it.
Since the H1 parser knows if we're in DATA or CHUNK_SIZE, simply let
it set the right next state so that h1m->state constantly matches
what is expected afterwards.
This will allow the parser to fill some extra fields like the method or
status without having to store them permanently in the HTTP message. At
this point however the parser cannot restart from an interrupted read.
This way we maintain the old mechanism stating that -2 means we block
on errors, -1 means we only capture them, and a positive value indicates
the position of the first error.
Currently the only user of struct h1m is the h2 mux when it has to parse
an H1 message coming from the channel. Unfortunately this is not enough
to efficiently parse HTTP/1 messages like those coming from the network
as we don't want to restart from scratch at every byte received.
This patch reintroduces the "next" offset into the H1 message so that any
H1 parser can use it to restart when called with a state that is not the
initial state.
This is the *parsing* state of an HTTP/1 message. Currently the h1_state
is composite as it's made both of parsing and control (100SENT, BODY,
DONE, TUNNEL, ENDING etc). The purpose here is to have a purely H1 state
that can be used by H1 parsers. For now it's equivalent to h1_state.
Remove the recv() method from mux and conn_stream.
The goal is to always receive from the upper layers, instead of waiting
for the connection later. For now, recv() is still called from the wake()
method, but that should change soon.
For struct connection, struct conn_stream, and for the h2 mux, add 2 new
lists, one that handles waiters for recv, and one that handles waiters for
recv and send. That way we can ask to subscribe for either recv or send.
In tasklet_free(), if we're currently in the runnable task list, don't
forget to decrement taks_list_size, or it'll end up being to big, and we may
not process tasks in the global runqueue.
This protocol is based on the uxst one, but it uses socketpair and FD
passing insteads of a connect()/accept().
The "sockpair@" prefix has been implemented for both bind and server
keywords.
When HAProxy wants to connect through a sockpair@, it creates 2 new
sockets using the socketpair() syscall and pass one of the socket
through the FD specified on the server line.
On the bind side, haproxy will receive the FD, and will use it like it
was the FD of an accept() syscall.
This protocol was designed for internal communication within HAProxy
between the master and the workers, but it's possible to use it
externaly with a wrapper and pass the FD through environment variabls.
It's possible to have several protocols per family which is a problem
with the current way the protocols are stored.
This allows to register a new protocol in HAProxy which is not a
protocol in the strict socket definition. It will be used to register a
SOCK_STREAM protocol using socketpair().
The following functions only deal with header field values and are agnostic
to the HTTP version so they were moved to http.c :
http_header_match2(), find_hdr_value_end(), find_cookie_value_end(),
extract_cookie_value(), parse_qvalue(), http_find_url_param_pos(),
http_find_next_url_param().
Those lacking the "http_" prefix were modified to have it.
These error codes and messages are agnostic to the version, even if
they are represented as HTTP/1.0 messages. Ultimately they will have
to be transformed into internal HTTP messages to be used everywhere.
The HTTP/1.1 100 Continue message was turned to an IST and the local
copy in the Lua code was removed.
This function is purely HTTP once http_txn is put aside. So the original
one was renamed to http_txn_get_path() and it extracts the relevant offsets
from the txn to pass them to http_get_path(). One benefit of the new version
is that it returns the length at the same time so that allowed to slightly
simplify http_get_path_from_string() which had to look up the end pointer
previously and which is not needed anymore.
It's a bit painful to have to deal with HTTP semantics for each protocol
version (H1 and H2), and working on the version-agnostic code further
emphasizes the problem.
This patch creates http.h and http.c which are agnostic to the version
in use, and which borrow a few parts from proto_http and from h1. For
example the once thought h1-specific h1_char_classes array is in fact
dictated by RFC7231 and is used to parse HTTP headers. A few changes
were made to a few files which were including proto_http.h while they
only needed http.h.
Certain string definitions pre-dated the introduction of indirect
strings (ist) so some were used to simplify the definition of the known
HTTP methods. The current lookup code saves 2 kB of a heavily used table
and is faster than the previous table based lookup (typ. 14 ns vs 16
before).
Instead of having a separate area for the captured data, we now have a
contigous block made of the descriptor and the data. At the moment, since
the area is dynamically allocated, we can adjust its size to what is
needed, but the idea is to quickly switch to a pool and an LRU list.
Now upon error we dynamically allocate the snapshot instead of overwriting
it. This way there is no more memory wasted in the proxy to hold the two
error snapshot descriptors. Also an appreciable side effect of this is that
the proxy's lock is only taken during the pointer swap, no more while copying
the buffer's contents. This saves 480 bytes of memory per proxy.
This function now captures an error regardless of its side and protocol.
The caller must pass a number of elements and may pass a protocol-specific
structure and a callback to display it. Later this function may deal with
more advanced allocation techniques to avoid allocating as many buffers
as proxies.
The HTTP dumps are now configurable in the code : "show errors" now
calls a protocol-specific function to emit the decoded output. For
now only HTTP is implemented.
The idea will be to make the error snapshot feature accessible to other
protocols than just HTTP. This patch only introduces an "http_snapshot"
structure and renames a few fields to make things more explicit. The
HTTP part was installed inside a union so that we can easily add more
protocols in the future.
The snapshots have the ability to restart a partial dump and they use
the stream ID as the restart point. Since it's purely HTTP, let's use
the event ID instead.
This function returns the proxy associated to a connection. For front
connections it returns the frontend, and for back connections it
returns the backend. This will be used to retrieve some configuration
parameters from within a mux.
Sometimes a connection is prepared before the target is set, sometimes
after. There's no real rule since the few functions involved operate on
different and independent fields. Soon we'll benefit from knowing the
target at the connection layer, in order to figure the associated proxy
and retrieve the various parameters (timeouts etc). This patch slightly
reorders a few calls to conn_prepare() so that we can make sure that the
target is always known to the mux.
The new function sess_log() only needs a session to emit a log. It will
ignore the parts that depend on the stream. It is usable to emit a log
to report early errors in muxes. These ones will typically mention
"<BADREQ>" for the request and 0 for the HTTP status code.
The current build_logline() can only be used with valid streams, which
means it is not suitable for use from muxes. We start by moving it into
another more generic function which takes the session as an argument,
to avoid complexifying all the internal API for jsut a few use cases.
This new function is not supposed to be called directly from outside so
we'll be able to instrument it to support several calling conventions.
For now the behaviour and conditions remain unchanged.
This function was split in two at commit f7d0447 ("MINOR: buffers:
split b_putblk() into __b_putblk()") but it's wrong, the first half's
length is not adjusted to the requested size so it copies more than
desired.
This is purely 1.9-specific, no backport is needed.
The handshake processing time used to be stored per stream, which was
valid when there was exactly one stream per session. With H2 and
multiplexing it's not the case anymore and the reported handshake times
are wrong in the logs as it's computed between the TCP accept() and the
stream creation. Let's first move the handshake where it belongs, which
is the session.
However, this is not enough because we don't want to report an excessive
idle time either for H2 (since many requests use the connection).
So the solution used here is to have the stream retrieve sess->tv_accept
and the handshake duration when the stream is created, and let the mux
immediately reset them. This way, the handshake time becomes zero for the
second and subsequent requests in H2 (which was already the case in H1),
and the idle time exactly counts how long the connection remained unused
while it could be used, so in H1 it runs from the end of the previous
response and in H2 it runs from the end of the previous request since the
channel is already available.
This patch will need to be backported to 1.8.
We've been missing it several times and now we'll need it to increment
a request counter. Let's do it once for all.
This patch will need to be backported to 1.8 with the associated fix.
Server state file has no indication that a server is currently managed
by a DNS SRV resolution.
And thus, both feature (DNS SRV resolution and server state), when used
together, does not provide the expected behavior: a smooth experience...
This patch introduce the "SRV record name" in the server state file and
loads and applies it if found and wherever required.
This patch applies to haproxy-dev branch only. For backport, a specific patch
is provided for 1.8.
This patch improves the previous fix by implementing the socket draining
code directly in conn_sock_drain() so that it always applies regardless
of the protocol's family. Thus it gets rid of tcp_drain().
Since commit 843b7cb ("MEDIUM: chunks: make the chunk struct's fields
match the buffer struct") a chunk length is unsigned so we can remove
negative size checks.
Since commit 843b7cb ("MEDIUM: chunks: make the chunk struct's fields
match the buffer struct") a chunk length is unsigned so we can remove
negative size checks.
During a test it happened that a connection was deleted before the
stream it's attached to, resulting in a crash related to the fix
18a85fe ("BUG/MEDIUM: streams: Don't forget to remove the si from
the wait list.") during the LIST_DEL(). Make sure to always delete
the list's head in this case so that other elements can safely
detach later.
This is purely 1.9, no backport is needed.
Set the flag for the current thread in active_threads_mask when waking a
tasklet, or we will never run it if no tasks are available.
This is 1.9-specific, no backport is needed.
When we choose to insert a fd in either the global or the local fd update list,
and the thread_mask against all_threads_mask before checking if it's tid_bit,
that way, if we run with nbthreads==1, we will always use the local list,
which is cheaper than the global one.
Instead of just using the conn_stream wait_list, give the stream_interface
its own. When the conn_stream will have its own buffers, the stream_interface
may have to wait on it.
Instead of using si_cs_send() as a task handler, define a new function,
si_cs_io_cb(), and give si_cs_send() its original prototype. Right now
si_cs_io_cb() just handles send, but later it'll handle recv() too.
Modify tasklet_wakeup() so that it handles a task as well, and inserts it
directly into the tasklet list, making it effectively a tasklet.
This should make future developments easier.
This adds the set-priority-class and set-priority-offset actions to
http-request and tcp-request content. At this point they are not used
yet, which is the purpose of the next commit, but all the logic to
set and clear the values is there.
We'll need trees to manage the queues by priorities. This change replaces
the list with a tree based on a single key. It's effectively a list but
allows us to get rid of the list management right now.
We store the queue index in the stream and check it on dequeueing to
figure how many entries were processed in between. This way we'll be
able to count the elements that may later be added before ours.
The current name is misleading as it implies a queue size, but the value
instead indicates a position in the queue.
The value is only the queue size at the exact moment the element is enqueued.
Soon we will gain the ability to insert anywhere into the queue, upon which
clarity of the name is more important.
Commit 7ce0c89 ("MEDIUM: mux: Use the mux protocol specified on
bind/server lines") assumed a bit too strongly that we could only have
servers on the connect side :-) It segfaults under this config :
defaults
contimeout 5s
clitimeout 5s
srvtimeout 5s
mode http
listen test1
bind :8001
dispatch 127.0.0.1:8002
frontend test2
mode http
bind :8002
redirect location /
No backport needed.
To do so, mux choices are split to handle incoming and outgoing connections in a
different way. The protocol specified on the bind/server line is used in
priority. Then, for frontend connections, the ALPN is retrieved and used to
choose the best mux. For backend connection, there is no ALPN. Finaly, if no
protocol is specified and no protocol matches the ALPN, we fall back on a
default mux, choosing in priority the first mux with exactly the same mode.
Because there can be several default multiplexers (without name), they are now
reported with the name "<default>". And a message warns they cannot be
referenced with the "proto" keyword on a bind line or a server line.
Now we try to synchronously push updates as they come using the new rdv
point, so that the call to the server update function from the main poll
loop is not needed anymore.
It further reduces the apparent latency in the health checks as the response
time almost always appears as 0 ms, resulting in a slightly higher check rate
of ~1960 conn/s. Despite this, the CPU consumption has slightly dropped again
to ~32% for the same test.
The only trick is that the checks code is built with a bit of recursivity
because srv_update_status() calls server_recalc_eweight(), and the latter
needs to signal srv_update_status() in case of updates. Thus we added an
extra argument to this function to indicate whether or not it must
propagate updates (no if it comes from srv_update_status).
Multiplexers are not necessarily associated to an ALPN. ALPN is a TLS extension,
so it is not always defined or used. Instead, we now rather speak of
multiplexer's protocols. So in this patch, there are no significative changes,
some structures and functions are just renamed.
Now, a multiplexer can specify if it can be install on incoming connections
(ALPN_SIDE_FE), on outgoing connections (ALPN_SIDE_BE) or both
(ALPN_SIDE_BOTH). These flags are compatible with proxies' ones.
This function is generic and is able to automatically transfer data from a
buffer to the conn_stream's tx buffer. It does this automatically if the mux
doesn't define another snd_buf() function.
It cannot yet be used as-is with the conn_stream's txbuf without risking to
lose data on close since conn_streams need to be orphaned for this.
To be symmetrical with the recv() part, we no handle retryable and partial
transmission using a intermediary buffer in the conn_stream. For now it's only
set to BUF_NULL and never allocated nor used.
It cannot yet be used as-is without risking to lose data on close since
conn_streams need to be orphaned for this.
This is a partial revert of the commit deccd1116 ("MEDIUM: mux: make
mux->snd_buf() take the byte count in argument"). It is a requirement to do
zero-copy transfers. This will be mandatory when the TX buffer of the
conn_stream will be used.
So, now, data are consumed by mux->snd_buf() and not only sent. So it needs to
update the buffer state. On its side, the caller must be aware the buffer can be
replaced y an empty or unallocated one.
As a side effet of this change, the function co_set_data() is now only responsible
to update the channel set, by update ->output field.
When b_slow_realign is called with the <output> parameter equal to 0, the
buffer's head, after the realign, must be set to 0. It was errornously set to
the buffer's size, because there was no test on the value of <output>.
The current synchronization point enforces certain restrictions which
are hard to workaround in certain areas of the code. The fact that the
critical code can only be called from the sync point itself is a problem
for some callback-driven parts. The "show fd" command for example is
fragile regarding this.
Also it is expensive in terms of CPU usage because it wakes every other
thread just to be sure all of them join to the rendez-vous point. It's a
problem because the sleeping threads would not need to be woken up just
to know they're doing nothing.
Here we implement a different approach. We keep track of harmless threads,
which are defined as those either doing nothing, or doing harmless things.
The rendez-vous is used "for others" as a way for a thread to isolate itself.
A thread then requests to be alone using thread_isolate() when approaching
the dangerous area, and then waits until all other threads are either doing
the same or are doing something harmless (typically polling). The function
only returns once the thread is guaranteed to be alone, and the critical
section is terminated using thread_release().
When threads are disabled, some variables such as tid and tid_bit are
still checked everywhere, the MAX_THREADS_MASK macro is ~0UL while
MAX_THREADS is 1, and the all_threads_mask variable is replaced with a
macro forced to zero. The compiler cannot optimize away all this code
involving checks on tid and tid_bit, and we end up in special cases
where all_threads_mask has to be specifically tested for being zero or
not. It is not even certain the code paths are always equivalent when
testing without threads and with nbthread 1.
Let's change this to make sure we always present a single thread when
threads are disabled, and have the relevant values declared as constants
so that the compiler can optimize all the tests away. Now we have
MAX_THREADS_MASK set to 1, all_threads_mask set to 1, tid set to zero
and tid_bit set to 1. Doing just this has removed 4 kB of code in the
no-thread case.
A few checks for all_threads_mask==0 have been removed since it never
happens anymore.
An offsetof() macro was introduced with commit 928fbfa ("MINOR: compiler:
introduce offsetoff().") with a fallback for older compilers. But this
breaks gcc 3.4 because __size_t and __uintptr_t are not defined there.
However size_t and uintptr_t are, so let's fix it this way. No backport
needed.
The purpose is to make sure that all variables which directly depend
on this nbthread argument are set at the right moment. For now only
all_threads_mask needs to be set. It used to be set while calling
thread_sync_init() which is called too late for certain checks. The
same function handles threads and non-threads, which removes the need
for some thread-specific knowledge from cfgparse.c.
If nbthread is MAX_THREADS, the shift operation needed to compute
all_threads_mask fails in thread_sync_init(). Instead pass a number
of threads to this function and let it compute the mask without
overflowing.
This should be backported to 1.8.
Since BoringSSL 3b2ff028, API now correctly match OpenSSL 1.1.0.
The patch revert part of haproxy 019f9b10: "Fix BoringSSL call and
openssl-compat.h/#define occordingly.".
This will not break openssl/libressl compat.
Add a new pipe, one per thread, so that we can write on it to wake a thread
sleeping in a poller, and use it to wake threads supposed to take care of a
task, if they are all sleeping.
This lock was necessary to manipulate the pendconn element between
concurrent places, but was causing great difficulties in the list walk
by having to iterate over multiple entries instead of being able to
safely pick the first one (in fact the first element was always the
right one but the locking model was hard to prove).
Here since we know we can always rely on the queue's locks, we take
the queue's lock every time we need to modify the element. In practice
it was already the case everywhere except in pendconn_dequeue() which
only works on an element that was already detached. This function had
to be protected against the risk of meeting an incompletely detached
element (which could be unlinked but not yet assigned). By taking the
queue lock around the LIST_ISEMPTY test, it's enough to ensure that a
concurrent thread either didn't begin or had completed the operation.
The true benefit really is in pendconn_process_next_strm() where we
can again safely work with the first element of each queue. This will
significantly simplify next updates to this code.
The pendconn struct uses ->px and ->srv to designate where the element is
queued. There is something confusing regarding threads though, because we
have to lock the appropriate queue before inserting/removing elements, and
this queue may only be determined by looking at ->srv (if it's not NULL
it's the server, otherwise use the proxy). But pendconn_grab_from_px() and
pendconn_process_next_strm() both assign this ->srv field, making it
complicated to know what queue to lock before manipulating the element,
which is exactly why we have the pendconn_lock in the first place.
This commit introduces pendconn->target which is the target server that
the two aforementioned functions will set when assigning the server.
Thanks to this, the server pointer may always be relied on to determine
what queue to use.
Now pendconn_free() takes a stream, checks that pend_pos is set, clears
it, and uses pendconn_unlink() to complete the job. It's cleaner and
centralizes all the bookkeeping work in pendconn_unlink() only and
ensures that there's a single place where the stream's position in the
queue is manipulated.
For now the pendconns may be dequeued at two places :
- pendconn_unlink(), which operates on a locked queue
- pendconn_free(), which operates on an unlocked queue and frees
everything.
Some changes are coming to the queue and we'll need to be able to be a
bit stricter regarding the places where we dequeue to keep the accounting
accurate. This first step renames the locked function __pendconn_unlink()
as it's for use by those aware of it, and introduces a new general purpose
pendconn_unlink() function which automatically grabs the necessary locks
before calling the former, and pendconn_cond_unlink() which additionally
checks the pointer and the presence in the queue.
As __task_wakeup() is responsible for increasing
rqueue_local[tid]/global_rqueue_size, make __task_unlink_rq responsible for
decreasing it, as process_runnable_tasks() isn't the only one that removes
tasks from runqueues.
By removing the reason code for the wakeup we can gain 8 extra bits to
encode the task's state. The reason code was never used at all and is
wrong by design since subsequent calls will OR this value anyway. Let's
say it goodbye and leave the room for more precious bits. The woken bits
were moved to the higher byte so that the most important bits can stay
grouped together.
Whenever it's possible to avoid a copy, b_xfer() will simply swap the
buffer's heads without touching the data. This has brought the performance
back from 140 kH/s to 202 kH/s on the test case.
This function is generic and is able to automatically transfer data
from a conn_stream's rx buffer to the destination buffer. It does this
automatically if the mux doesn't define another rcv_buf() function.
In order to reorganize the connection layers, recv() operations will
need to be retryable and to support partial transfers. This requires
an intermediary buffer to hold the data coming from the mux. After a
few attempts, it turns out that this buffer is best placed inside the
conn_stream itself. For now it's only set to buf_empty and it will be
up to the caller to allocate it if required.
The latter function is more suited to operations that don't require any
check because the check has already been performed. It will be used by
other b_* functions.
This function is used a lot in block copies and is needlessly
complicated since it still uses pointer arithmetic. Let's fall
back to regular offsets and simplify it. This removed around
23 bytes from b_putblk() and it removed any conditional jump.
In thread_sync_barrier, we exit when all threads have set their own bit in the
barrier mask. It is done by comparing it to all_threads_mask. But we must not
use a simple equality to do so, becaue all_threads_mask may change. Since commit
ba86c6c25 ("MINOR: threads: Be sure to remove threads from all_threads_mask on
exit"), when a thread exit, its bit is removed from all_threads_mask. Instead,
we must use a bitwise AND to test is all bits of all_threads_mask are set.
This also requires that all_threads_mask is set to volatile if we want to
catch changes.
This patch must be backported in 1.8.
This new function wl_set_waitcb() prepopulates a wait_list with a tasklet
and a context and returns it so that it can be passed to ->subscribe() to
be added to a connection or conn_stream's wait_list. The caller doesn't
need to know all the insiders details anymore this way.
Totally nuke the "send" method, instead, the upper layer decides when it's
time to send data, and if it's not possible, uses the new subscribe() method
to be called when it can send data again.
Add a new "subscribe" method for connection, conn_stream and mux, so that
upper layer can subscribe to them, to be called when the event happens.
Right now, the only event implemented is "SUB_CAN_SEND", where the upper
layer can register to be called back when it is possible to send data.
The connection and conn_stream got a new "send_wait_list" entry, which
required to move a few struct members around to maintain an efficient
cache alignment (and actually this slightly improved performance).
Now all the code used to manipulate chunks uses a struct buffer instead.
The functions are still called "chunk*", and some of them will progressively
move to the generic buffer handling code as they are cleaned up.
Chunks are only a subset of a buffer (a non-wrapping version with no head
offset). Despite this we still carry a lot of duplicated code between
buffers and chunks. Replacing chunks with buffers would significantly
reduce the maintenance efforts. This first patch renames the chunk's
fields to match the name and types used by struct buffers, with the goal
of isolating the code changes from the declaration changes.
Most of the changes were made with spatch using this coccinelle script :
@rule_d1@
typedef chunk;
struct chunk chunk;
@@
- chunk.str
+ chunk.area
@rule_d2@
typedef chunk;
struct chunk chunk;
@@
- chunk.len
+ chunk.data
@rule_i1@
typedef chunk;
struct chunk *chunk;
@@
- chunk->str
+ chunk->area
@rule_i2@
typedef chunk;
struct chunk *chunk;
@@
- chunk->len
+ chunk->data
Some minor updates to 3 http functions had to be performed to take size_t
ints instead of ints in order to match the unsigned length here.
Now the buffers only contain the header and a pointer to the storage
area which can be anywhere. This will significantly simplify buffer
swapping and will make it possible to map chunks on buffers as well.
The buf_empty variable was removed, as now it's enough to have size==0
and area==NULL to designate the empty buffer (thus a non-allocated head
is the empty buffer by default). buf_wanted for now is indicated by
size==0 and area==(void *)1.
The channels and the checks now embed the buffer's head, and the only
pointer is to the storage area. This slightly increases the unallocated
buffer size (3 extra ints for the empty buffer) but considerably
simplifies dynamic buffer management. It will also later permit to
detach unused checks.
The way the struct buffer is arranged has proven quite efficient on a
number of tests, which makes sense given that size is always accessed
and often first, followed by the othe ones.
It used to be called 'len' during the reorganisation but strictly speaking
it's not a length since it wraps. Also we already use '_data' as the suffix
to count available data, and data is also what we use to indicate the amount
of data in a pipe so let's improve consistency here. It was important to do
this in two operations because data used to be the name of the pointer to
the storage area.
This one is more generic and designed to work on a random block. It
may later get a b_rep_ist() variant since many strings are already
available as (ptr,len).
There was no point keeping that function in the buffer part since it's
exclusively used by HTTP at the channel level, since it also automatically
appends the CRLF. This further cleans up the buffer code.
The new file istbuf.h links the indirect strings (ist) with the buffers.
The purpose is to encourage addition of more standard buffer manipulation
functions that rely on this in order to improve the overall ease of use
along all the code. Just like ist.h and buf.h, this new file is not
expected to depend on anything beyond these two files.
A few functions were added and/or converted from buffer.h :
- b_isteq() : indicates if a buffer and a string match
- b_isteat() : consumes a string from the buffer if it matches
- b_istput() : appends a small string to a buffer (all or none)
- b_putist() : appends part of a large string to a buffer
The equivalent functions were removed from buffer.h and changed at the
various call places.
The two variants now do exactly the same (appending at the tail of the
buffer) so let's not keep the distinction between these classes of
functions and have generic ones for this. It's also worth noting that
b{i,o}_putchk() wasn't used at all and was removed.
There's no distinction between in and out data now. The latter covers
the needs of the former and supports wrapping. The extra cost is
negligible given the locations where it's used.
Since we never access this field directly anymore, but only through the
channel's wrappers, it can now move to the channel. The buffers are now
completely free from the distinction between input and output data.
Since we use "_data" for the amount of data at many places, as opposed to
"_space" for the amount of space, let's rename the "data" field to "area"
so that we can reuse "data" later for the amount of data in the buffer
(currently called "len" despite not being contigous).
b_set_data() is used :
- in proto_http and hlua to trim input data (b_set_data(co_data()))
- in SPOE to append data to a buffer while building a message
In no case will this truncate a buffer so we can safely remove the
test for len < b->output.
b_del() is used in :
- mux_h2 with the demux buffer : always processes input data
- checks with output data though output is not considered at all there
- b_eat() which is not used anywhere
- co_skip() where the len is always <= output
Thus the distinction for output data is not needed anymore and the
decrement can be made inconditionally in co_skip().
This is intentionally the minimal and safest set of changes, some cleanups
area still required. These changes are quite tricky and cannot be
independantly tested, so it's important to keep this patch as bisectable
as possible.
buf_empty and buf_wanted were changed and are now exactly similar since
there's no <p> member in the structure anymore. Given that no test is
ever made in the code to check that buf == &buf_wanted, it may be possible
that we don't need to have two anymore, unless some buf_empty tests have
precedence. This will have to be investigated.
A significant part of this commit affects the HTTP compression code,
which used to deeply manipulate the input and output buffers without
any reasonable solution for a better abstraction. For this reason, if
any regression is met and designates this patch as the culprit, it is
important to run tests which specifically involve compression or which
definitely don't use it in order to spot the issue.
Cc: Olivier Houchard <ohouchard@haproxy.com>
For the same consistency reasons, let's use b_empty() at the few places
where an empty buffer is expected, or c_empty() if it's done on a channel.
Some of these places were there to realign the buffer so
{b,c}_realign_if_empty() was used instead.
We used to have variations around buffer_total_space() and
size-buffer_len() or size-b_data(). Let's simplify all this. buffer_len()
was also removed as not used anymore.
Now the new API functions are being used everywhere, we can get rid
of b_ptr(). A few last users like bi_istput() and bo_istput() appear
to only differ by what part of the buffer they're increasing, but
that should quickly be merged.
With this flag we introduce the notion of "dry" vs "wet" buffers : some
demultiplexers like the H2 mux require as much room as possible for some
operations that are not retryable like decoding a headers frame. For this
they need to know if the buffer is congested with data scheduled for
leaving soon or not. Since the new API will not provide this information
in the buffer itself, the caller must indicate it. We never need to know
the amount of such data, just the fact that the buffer is not in its
optimal condition to be used for receipt. This "CO_RFL_BUF_WET" flag is
used to mention that such outgoing data are still pending in the buffer
and that a sensitive receiver should better let it "dry" before using it.
The mux and transport rcv_buf() now takes a "flags" argument, just like
the snd_buf() one or like the equivalent syscall lower part. The upper
layers will use this to pass some information such as indicating whether
the buffer is free from outgoing data or if the lower layer may allocate
the buffer itself.
It also returns a size_t. This is in order to clean the API. Note
that the H2 mux still uses some ints in the functions called from
h2_rcv_buf(), though it's not really a problem given that H2 frames
are smaller. It may deserve a general cleanup later though.
Just like we have a size_t for xprt->snd_buf(), we adjust to use size_t
for rcv_buf()'s count argument and return value. It also removes the
ambiguity related to the possibility to see a negative value there.
This way the mux doesn't need to modify the buffer's metadata anymore
nor to know the output's size. The mux->snd_buf() function now takes a
const buffer and it's up to the caller to update the buffer's state.
The return type was updated to return a size_t to comply with the count
argument.
This way the senders don't need to modify the buffer's metadata anymore
nor to know about the output's split point. This way the functions can
take a const buffer and it's clearer who's in charge of updating the
buffer after a send. That's why the buffer realignment is now performed
by the caller of the transport's snd_buf() functions.
The return type was updated to return a size_t to comply with the count
argument.
Now that there are no more users requiring to modify the buffer anymore,
switch these ones to const char and const buffer. This will make it more
obvious next time send functions are tempted to modify the buffer's output
count. Minor adaptations were necessary at a few call places which were
using char due to the function's previous prototype.
Till now the callers had to know which one to call for specific use cases.
Let's fuse them now since a single one will remain after the API migration.
Given that bi_del() may only be used where o==0, just combine the two tests
by first removing output data then only input.
This will be important so that we can parse a buffer without touching it.
Now we indicate where from the buffer's head we plan to start to copy, and
for how many bytes. This will be used by send functions to loop at the end
of the buffer without having to update the buffer's output byte count.
This new functoin limits itself to the amount of data available in the
buffer and doesn't care about the direction anymore. It's only called
from co_getblk() which already checks that no more than the available
output bytes is requested.
These ones were merged into a single b_contig_space() that covers both
(the bo_ case was a simplified version of the other one). The function
doesn't use ->i nor ->o anymore.
This function was sometimes used from a channel and sometimes from a buffer.
In both cases it requires knowledge of the size of the output data (to skip
them). Here the split ensures the channel can deal with this point, and that
other places not having output data can continue to work.
These ones manipulate the output data count which will be specific to
the channel soon, so prepare the call points to use the channel only.
The b_* functions are now unused and were removed.
The few call places where it's used can use the trash as a swap buffer,
which is made for this exact purpose. This way we can rely on the
generic b_slow_realign() call.
Where relevant, the channel version is used instead. The buffer version
was ported to be more generic and now takes a swap buffer and the output
byte count to know where to set the alignment point. The H2 mux still
uses buffer_slow_realign() with buf->o but it will change later.
This adds :
- c_orig() : channel buffer's origin
- c_size() : channel buffer's size
- c_wrap() : channel buffer's wrapping location
- c_data() : channel buffer's total data count
- c_room() : room left in channel buffer's
- c_empty() : true if channel buffer is empty
- c_full() : true if channel buffer is full
- c_ptr() : pointer to an offset relative to input data in the buffer
- c_adv() : advances the channel's buffer (bytes become part of output)
- c_rew() : rewinds the channel's buffer (output bytes not output anymore)
- c_realign_if_empty() : realigns the buffer if it's empty
- co_data() : # of output data
- co_head() : beginning of output data
- co_tail() : end of output data
- ci_data() : # of input data
- ci_head() : beginning of input data
- ci_tail() : end of input data
- ci_stop() : location after ci_tail()
- ci_next() : pointer to next input byte
And for the ci_* / co_* functions above, the "__*" variants which disable
wrapping checks, and the "_ofs" variants which return an offset relative to
the buffer's origin instead.
Many places deal with buffer realignment after data removal. The method
is always the same : if the buffer is empty, set its pointer to the origin.
Let's have a function for this so that we have less code to change with the
new API.
Add a new function that lets you set the amount of input in a buffer.
For now it extends/truncates b->i except if the total length is
below b->o in which case it clears i and adjusts o.
Instead of doing b->i -= directly, introduce b_sub(), that does the job, to
make it easier to switch to the future API.
Also add b_add(), that increases b->i, instead of using it directly, and
bo_add(), that does increase b->o.
Here's the list of newly introduced functions :
- b_data(), returning the total amount of data in the buffer (currently i+o)
- b_orig(), returning the origin of the storage area, that is, the place of
position 0.
- b_wrap(), pointer to wrapping point (currently data+size)
- b_size(), returning the size of the buffer
- b_room(), returning the amount of bytes left available
- b_full(), returning true if the buffer is full, otherwise false
- b_stop(), pointer to end of data mark (currently p+i), used to compute
distances or a stop pointer for a loop.
- b_peek(), this one will help make the transition to the new buffer model.
It returns a pointer to a position in the buffer known from an offest
relative to the beginning of the data in the buffer. Thus, we can replace
the following occurrences :
bo_ptr(b) => b_peek(b, 0);
bo_end(b) => b_peek(b, b->o);
bi_ptr(b) => b_peek(b, b->o);
bi_end(b) => b_peek(b, b->i + b->o);
b_ptr(b, ofs) => b_peek(b, b->o + ofs);
- b_head(), pointer to the beginning of data (currently bo_ptr())
- b_tail(), pointer to first free place (currently bi_ptr())
- b_next() / b_next_ofs(), pointer to the next byte, taking wrapping
into account.
- b_dist(), returning the distance between two pointers belonging to a buffer
- b_reset(), which resets the buffer
- b_space_wraps(), indicating if the free space wraps around the buffer
- b_almost_full(), indicating if 3/4 or more of the buffer are used
Some of these are provided with the unchecked variants using the "__"
prefix, or with the "_ofs" suffix indicating they return a relative
position to the buffer's origin instead of a pointer.
Cc: Olivier Houchard <ohouchard@haproxy.com>
Passing unsigned ints everywhere is painful, and will cause some headache
later when we'll want to integrate better with struct ist which already
uses size_t. Let's switch buffers to use size_t instead.
The buffer code currently depends on pools and other stuff and is not
really autonomous anymore. The rewrite of the new API is an opportunity
to clean this up. This patch creates a new file (buf.h) which does not
depend on other elements and which will only contain what is needed to
perform the most basic buffer operations. The new API will be introduced
in this file and the conversion will be finished once buffer.h is empty.
The definition of struct buffer was moved to this new file, using more
explicity stdint types for the sizes and offsets.
Most new functions will be implemented in two variants :
__b_something() : unchecked variant, no wrapping is expected
b_something() : wrapping-checked variant
This way callers will be able to select which one to use depending on
the use cases.
Commit 200b0fa ("MEDIUM: Add support for updating TLS ticket keys via
socket") introduced support for updating TLS ticket keys from the CLI,
but missed a small corner case : if multiple bind lines reference the
same tls_keys file, the same reference is used (as expected), but during
the clean shutdown, it will lead to a double free when destroying the
bind_conf contexts since none of the lines knows if others still use
it. The impact is very low however, mostly a core and/or a message in
the system's log upon old process termination.
Let's introduce some basic refcounting to prevent this from happening,
so that only the last bind_conf frees it.
Thanks to Janusz Dziemidowicz and Thierry Fournier for both reporting
the same issue with an easy reproducer.
This fix needs to be backported from 1.6 to 1.8.
By default, HAProxy's DNS resolution at runtime ensure that there is no
IP address duplication in a backend (for servers being resolved by the
same hostname).
There are a few cases where people want, on purpose, to disable this
feature.
This patch introduces a couple of new server side options for this purpose:
"resolve-opts allow-dup-ip" or "resolve-opts prevent-dup-ip".
Up until now, a tasklet couldn't be free'd while it was in the list, it is
no longer the case, so make sure we remove it from the list before freeing it.
To do so, we have to make sure we correctly initialize it, so use LIST_INIT,
instead of setting the pointers to NULL.
The behavior of sigprocmask in an multithreaded environment is
undefined.
The new macro ha_sigmask() calls either pthreads_sigmask() or
sigprocmask() if haproxy was built with thread support or not.
This should be backported to 1.8.
To make sure we don't inadvertently insert task in the global runqueue,
while only the local runqueue is used without threads, make its definition
and usage conditional on USE_THREAD.
When building without threads enabled, instead of just using the global
runqueue, just use the local runqueue associated with the only thread, as
that's what is now expected for a single thread in prcoess_runnable_tasks().
This should fix haproxy when built without threads.
When an applet is created, let's assign it the same nice value as the task
of the stream which owns it. It ensures that fairness is properly propagated
to applets, and that the CLI can regain a low latency behaviour again. Huge
differences have been seen under extreme loads, with the CLI being called
every 200 microseconds instead of 11 milliseconds.
This function returns true is some notifications are registered.
This function is usefull for the following patch
BUG/MEDIUM: lua/socket: Sheduling error on write: may dead-lock
It should be backported in 1.6, 1.7 and 1.8
Don't forget to increase tasks_run_queue when we're adding a task to the
tasklet list, and to decrease it when we remove a task from a runqueue,
or its value won't be accurate, and could lead to tasks not being executed
when put in the global run queue.
1.9-dev only, no backport is needed.
This patch adds a warning if an http-(request|reponse) (add|set)-header
rewrite fails to change the respective header in a request or response.
This usually happens when tune.maxrewrite is not sufficient to hold all
the headers that should be added.
There's no real reason to have a specific scheduler for applets anymore, so
nuke it and just use tasks. This comes with some benefits, the first one
being that applets cannot induce high latencies anymore since they share
nice values with other tasks. Later it will be possible to configure the
applets' nice value. The second benefit is that the applet scheduler was
not very thread-friendly, having a big lock around it in prevision of this
change. Thus applet-intensive workloads should now scale much better with
threads.
Some more improvement is possible now : some applets also use a task to
handle timers and timeouts. These ones could now be simplified to use only
one task.
Introduce tasklets, lightweight tasks. They have no notion of priority,
they are just run as soon as possible, and will probably be used for I/O
later.
For the moment they're used to replace the temporary thread-local list
that was used in the scheduler. The first part of the struct is common
with tasks so that tasks can be cast to tasklets and queued in this list.
Once a task is in the tasklet list, it has its leaf_p set to 0x1 so that
it cannot accidently be confused as not in the queue.
Pure tasklets are identifiable by their nice value of -32768 (which is
normally not possible).
A lot of tasks are run on one thread only, so instead of having them all
in the global runqueue, create a per-thread runqueue which doesn't require
any locking, and add all tasks belonging to only one thread to the
corresponding runqueue.
The global runqueue is still used for non-local tasks, and is visited
by each thread when checking its own runqueue. The nice parameter is
thus used both in the global runqueue and in the local ones. The rare
tasks that are bound to multiple threads will have their nice value
used twice (once for the global queue, once for the thread-local one).
In preparation for thread-specific runqueues, change the task API so that
the callback takes 3 arguments, the task itself, the context, and the state,
those were retrieved from the task before. This will allow these elements to
change atomically in the scheduler while the application uses the copied
value, and even to have NULL tasks later.
A few users reported that building without threads was accidently broken
after commit 6b96f72 ("BUG/MEDIUM: pollers: Use a global list for fd
shared between threads.") due to all_threads_mask not being defined.
It's OK to set it to zero as other code parts do when threads are
enabled but only one thread is used.
This needs to be backported to 1.8.
The function hlua_ctx_resume return less text message and more error
code. These error code allow the caller to return appropriate
message to the user.
The polled_mask is only used in the pollers, and removing it from the
struct fdtab makes it fit in one 64B cacheline again, on a 64bits machine,
so make it a separate array.
With the old model, any fd shared by multiple threads, such as listeners
or dns sockets, would only be updated on one threads, so that could lead
to missed event, or spurious wakeups.
To avoid this, add a global list for fd that are shared, using the same
implementation as the fd cache, and only remove entries from this list
when every thread as updated its poller.
[wt: this will need to be backported to 1.8 but differently so this patch
must not be backported as-is]
Modify fd_add_to_fd_list() and fd_rm_from_fd_list() so that they take an
offset in the fdtab to the list entry, instead of hardcoding the fd cache,
so we can use them with other lists.
While running a task, we may try to delete and free a task that is about to
be run, because it's part of the local tasks list, or because rq_next points
to it.
So flag any task that is in the local tasks list to be deleted, instead of
run, by setting t->process to NULL, and re-make rq_next a global,
thread-local variable, that is modified if we attempt to delete that task.
Many thanks to PiBa-NL for reporting this and analysing the problem.
This should be backported to 1.8.
For large farms where servers are regularly added or removed, picking
a random server from the pool can ensure faster load transitions than
when using round-robin and less traffic surges on the newly added
servers than when using leastconn.
This commit introduces "balance random". It internally uses a random as
the key to the consistent hashing mechanism, thus all features available
in consistent hashing such as weights and bounded load via hash-balance-
factor are usable. It is extremely convenient because one common concern
when using random is what happens when a server is hammered a bit too
much. Here that can trivially be avoided, like in the configuration below :
backend bk0
balance random
hash-balance-factor 110
server-template s 1-100 127.0.0.1:8000 check inter 1s
Note that while "balance random" internally relies on a hash algorithm,
it holds the same properties as round-robin and as such is compatible with
reusing an existing server connection with "option prefer-last-server".
In order to use arbitrary data in the CLI (multiple lines or group of words
that must be considered as a whole, for example), it is now possible to add a
payload to the commands. To do so, the first line needs to end with a special
pattern: <<\n. Everything that follows will be left untouched by the CLI parser
and will be passed to the commands parsers.
Per-command support will need to be added to take advantage of this
feature.
Signed-off-by: Aurélien Nephtali <aurelien.nephtali@corp.ovh.com>
We'll need this in order to support uploading chunks. The h2 to h1
converter checks for the presence of the content-length header field
as well as the CONNECT method and returns these information to the
caller. The caller indicates whether or not a body is detected for
the message (presence of END_STREAM or not). No transfer-encoding
header is emitted yet.
In some cases, we call cs_destroy() very early, so early the connection
doesn't yet have a mux, so we can't call mux->detach(). In this case,
just destroy the associated connection.
This should be backported to 1.8.
With gcc < 4.7, when HAProxy is built with threads, the macros
HA_ATOMIC_CAS/XCHG/STORE relies on the legacy __sync builtins. These macros
are slightly complicated than the versions relying on the '_atomic'
builtins. Internally, some local variables are defined, prefixed with '__' to
avoid name clashes with the caller.
On the other hand, the macros HA_ATOMIC_UPDATE_MIN/MAX call HA_ATOMIC_CAS. Some
local variables are also definied in these macros, following the same naming
rule as below. The problem is that '__new' variable is used in
HA_ATOMIC_MIN/_MAX and in HA_ATOMIC_CAS. Obviously, the behaviour is undefined
because '__new' in HA_ATOMIC_CAS is left uninitialized. Unfortunatly gcc fails
to detect this error.
To fix the problem, all internal variables to macros are now suffixed with name
of the macros to avoid clashes (for instance, '__new_cas' in HA_ATOMIC_CAS).
This patch must be backported in 1.8.
In addition to metrics about time spent in the SPOE, following counters have
been added:
* applets : number of SPOE applets.
* idles : number of idle applets.
* nb_sending : number of streams waiting to send data.
* nb_waiting : number of streams waiting for a ack.
* nb_processed : number of events/groups processed by the SPOE (from the
stream point of view).
* nb_errors : number of errors during the processing (from the stream point of
view).
Log messages has been updated to report these counters. Following pattern has
been added at the end of the log message:
... <idles>/<applets> <nb_sending>/<nb_waiting> <nb_error>/<nb_processed>
Now it is possible to configure a logger in a spoe-agent section using a "log"
line, as for a proxy. "no log", "log global" and "log <address> ..." syntaxes
are supported.
With "log global" line, the global list of loggers are copied into the proxy's
struct. The list coming from the default section is also copied when a frontend
or a backend section is parsed. So it is possible to have duplicate entries in
the proxy's list. For instance, with this following config, all messages will be
logged twice:
global
log 127.0.0.1 local0 debug
daemon
defaults
mode http
log global
option httplog
frontend front-http
log global
bind *:8888
default_backend back-http
backend back-http
server www 127.0.0.1:8000
Now, the function parse_logsrv should be used to parse a "log" line. This
function will update the list of loggers passed in argument. It can release all
log servers when "no log" line was parsed (by the caller) or it can parse "log
global" or "log <address> ... " lines. It takes care of checking the caller
context (global or not) to prohibit "log global" usage in the global section.
"set-process-time" and "set-total-time" options have been added to store
processing times in the transaction scope, at each event and group processing,
the current one and the total one. So it is possible to get them.
TODO: documentation
Following metrics are added for each event or group of messages processed in the
SPOE:
* processing time: the delay to process the event or the group. From the
stream point of view, it is the latency added by the SPOE
processing.
* request time : It is the encoding time. It includes ACLs processing, if
any. For fragmented frames, it is the sum of all fragments.
* queue time : the delay before the request gets out the sending queue. For
fragmented frames, it is the sum of all fragments.
* waiting time: the delay before the reponse is received. No fragmentation
supported here.
* response time: the delay to process the response. No fragmentation supported
here.
* total time: (unused for now). It is the sum of all events or groups
processed by the SPOE for a specific threads.
Log messages has been updated. Before, only errors was logged (status_code !=
0). Now every processing is logged, following this format:
SPOE: [AGENT] <TYPE:NAME> sid=STREAM-ID st=STATUC-CODE reqT/qT/wT/resT/pT
where:
AGENT is the agent name
TYPE is EVENT of GROUP
NAME is the event or the group name
STREAM-ID is an integer, the unique id of the stream
STATUS_CODE is the processing's status code
reqT/qT/wT/resT/pT are delays descrive above
For all these delays, -1 means the processing was interrupted before the end. So
-1 for the queue time means the request was never dequeued. For fragmented
frames it is harder to know when the interruption happened.
For now, messages are logged using the same logger than the backend of the
stream which initiated the request.
Clearing the update_mask bit in fd_insert may lead to duplicate insertion
of fd in fd_updt, that could lead to a write past the end of the array.
Instead, make sure the update_mask bit is cleared by the pollers no matter
what.
This should be backported to 1.8.
[wt: warning: 1.8 doesn't have the lockless fdcache changes and will
require some careful changes in the pollers]
This function will be called from the CLI's "show fd" command to append some
extra mux-specific information that only the mux handler can decode. This is
supposed to help collect various hints about what is happening when facing
certain anomalies.
This patch add option crc32c (PP2_TYPE_CRC32C) to proxy protocol v2.
It compute the checksum of proxy protocol v2 header as describe in
"doc/proxy-protocol.txt".
Commit 4815c8c ("MAJOR: fd/threads: Make the fdcache mostly lockless.")
made the fd cache lockless, but after a few iterations, a subtle part was
lost, consisting in setting the bit on the fd_cache_mask immediately when
adding an event. Now it was done only when the cache started to process
events, but the problem it causes is that fd_cache_mask isn't reliable
anymore as an indicator of presence of events to be processed with no
delay outside of fd_process_cached_events(). This results in some spurious
delays when processing inter-thread wakeups between tasks. Just restoring
the flag when the event is added is enough to fix the problem.
Kudos to Christopher for spotting this one!
No backport is needed as this is only in the development version.
The management of the servers and the proxies queues was not thread-safe at
all. First, the accesses to <strm>->pend_pos were not protected. So it was
possible to release it on a thread (for instance because the stream is released)
and to use it in same time on another one (because we redispatch pending
connections for a server). Then, the accesses to stream's information (flags and
target) from anywhere is forbidden. To be safe, The stream's state must always
be updated in the context of process_stream.
So to fix these issues, the queue module has been refactored. A lock has been
added in the pendconn structure. And now, when we try to dequeue a pending
connection, we start by unlinking it from the server/proxy queue and we wake up
the stream. Then, it is the stream reponsibility to really dequeue it (or
release it). This way, we are sure that only the stream can create and release
its <pend_pos> field.
However, be careful. This new implementation should be thread-safe
(hopefully...). But it is not optimal and in some situations, it could be really
slower in multi-threaded mode than in single-threaded one. The problem is that,
when we try to dequeue pending connections, we process it from the older one to
the newer one independently to the thread's affinity. So we need to wait the
other threads' wakeup to really process them. If threads are blocked in the
poller, this will add a significant latency. This problem happens when maxconn
values are very low.
This patch must be backported in 1.8.
When a listener is temporarily disabled, we start by locking it and then we call
.pause callback of the underlying protocol (tcp/unix). For TCP listeners, this
is not a problem. But listeners bound on an unix socket are in fact closed
instead. So .pause callback relies on unbind_listener function to do its job.
Unfortunatly, unbind_listener hold the listener's lock and then call an internal
function to unbind it. So, there is a deadlock here. This happens during a
reload. To fix the problemn, the function do_unbind_listener, which is lockless,
is now exported and is called when a listener bound on an unix socket is
temporarily disabled.
This patch must be backported in 1.8.
This patch implement proxy protocol v2 options related to crypto information:
ssl-cipher (PP2_SUBTYPE_SSL_CIPHER), cert-sig (PP2_SUBTYPE_SSL_SIG_ALG) and
cert-key (PP2_SUBTYPE_SSL_KEY_ALG).
ssl_sock_get_pkey_algo can be used to report pkey algorithm to log
and ppv2 (RSA2048, EC256,...).
Extract pkey information is not free in ssl api (lock/alloc/free):
haproxy can use the pkey information computed in load_certificate.
Store and use this information in a SSL ex_data when available,
compute it if not (SSL multicert bundled and generated cert).
Private key information is used in switchctx to implement native multicert
selection (ecdsa/rsa/anonymous). This patch extract and store full pkey
information: dsa type and pkey size in bits. This can be used for switchctx
or to report pkey informations in ppv2 and log.
When the block of data need to be split to support the wrapping, the start of
the second block of data was wrong. We must be sure to skup data copied during
the first memcpy.
This patch must be backported to 1.8.
When the block of data need to be split to support the wrapping, the start of
the second block of data was wrong. We must be sure to skip data copied during
the first memcpy.
This patch must be backported to 1.8, 1.7, 1.6 and 1.5.
Since we use padding before the allocated page, it's trivial to place
the allocated address there and see if it gets mangled once we release
it.
This may be backported to stable releases already using DEBUG_UAF.
Commit 158fa75 ("MINOR: pools: implement DEBUG_UAF to detect use after free")
implemented pool use-after-free detection, but the mmap() return value isn't
properly checked, preventing the call to pool_alloc_area() from returning
NULL. So on out-of-memory a mangled pointer is returned, causing a crash on
the pool_alloc() site instead of forcing a GC. It doesn't affect regular
operations however, just complicates complex bug investigations.
This fix should be backported to 1.8 and to 1.7.
Since commit cf975d4 ("MINOR: pools/threads: Implement lockless memory
pools."), we support lockless pools. However the parts dedicated to
detecting use-after-free are not present in this part, making DEBUG_UAF
useless in this situation.
The present patch sets a new define CONFIG_HAP_LOCKLESS_POOLS when such
a compatible architecture is detected, and when pool debugging is not
requested, then makes use of this everywhere in pools and buffers
functions. This way enabling DEBUG_UAF will automatically disable the
lockless version.
No backport is needed as this is purely 1.9-dev.
This removes the end label from memory.h.
The labels are unused as of cf975d46bc
which is unreleased (and incidentally the first commit containing
those labels, thus they never have been used).
A TLS ticket keys file can be updated on the CLI and used in same time. So we
need to protect it to be sure all accesses are thread-safe. Because updates are
infrequent, a R/W lock has been used.
This patch must be backported in 1.8
Commit f61f0cb ("MINOR: threads: Introduce double-width CAS on x86_64
and arm.") introduced the double CAS. But the ARMv7 version is bogus,
it uses the value of the pointers instead of dereferencing them. When
lucky, it simply doesn't build due to impossible registers combinations.
Otherwise it will immediately crash at run time when facing traffic.
No backport is needed, this bug was introduced in 1.9-dev.
Each fd_{may|cant|stop|want}_{recv|send} function sets or resets a
single bit at once, then recomputes the need for updates, and then
the new cache state. Later, pollers will compute the new polling
state based on the resulting operations here. In fact the conditions
are so simple that they can be performed by a single "if", or sometimes
even optimized away.
This means that in practice a simple compare-and-swap operation if often
enough to set the new value inluding the new polling state, and that only
the cache and fdupdt have to be performed under the lock. Better, for the
most common operations (fd_may_{recv,send}, used by the pollers), a simple
atomic OR is needed.
This patch does this for the fd_* functions above and it doesn't yet
remove the now useless fd_compute_new_polling_status() because it's still
used by other pollers. A pure connection rate test shows a 1% performance
increase.
An fd cache entry might be removed and added at the end of the list, while
another thread is parsing it, if that happens, we may miss fd cache entries,
to avoid that, add a new field in the struct fdtab, "added_mask", which
contains a mask for potentially affected threads, if it is set, the
corresponding thread will set its bit in fd_cache_mask, to avoid waiting in
poll while it may have more work to do.
Create a local, per-thread, fdcache, for file descriptors that only belongs
to one thread, and make the global fd cache mostly lockless, as we can get
a lot of contention on the fd cache lock.
Instead of looking for CO_FL_EARLY_DATA to know if we have to try to wake
up a stream, because it is waiting for a SSL handshake, instead add a new
conn_stream flag, CS_FL_WAIT_FOR_HS. This way we don't have to rely on
CO_FL_EARLY_DATA, and we will only wake streams that are actually waiting.
Instead of using a list of applets with idle ones in front, we now use an
ebtree. Aapplets in the tree are idle by definition. And the key is the applet's
weight. When a new frame is queued, the first idle applet (with the lowest
weight) is woken up and its weight is increased by one. And when an applet sends
a frame to a SPOA, its weight is decremented by one.
This is empirical, but it should avoid to overuse a very few number of applets
and increase the balancing between idle applets.
So it is easier to respect the max_fpa value. This is no more the maximum frames
processed by an applet at each loop but the maximum frames waiting for an ack
for a specific applet.
The function spoe_handle_processing_appctx has been rewritten accordingly.
sending_rate was a counter used to evaluate the SPOE capacity to process
frames. Because it was not really accurrate, it has been replaced by a frequency
counter representing the number of frames handled by the SPOE per second. We
just check this counter is higher than the number of streams waiting for a
reply. If not, a new applet is created.
The calculation of a minimal number of active applets was really empirical and
finally useless. On heavy load, there are always many active applets (most of
time, more than the minimal required) and when the load is low, there is no
reason to keep unused applets opened.
Because of this change, the flag SPOE_APPCTX_FL_PERSIST is now unused. So it has
been removed.
Recent changes to the enum were not synchronized with the lock debugging
code. Now we use a switch/case instead of an array so that the compiler
throws a warning if there is any inconsistency.
To be backported to 1.8 (at least to add the START entry).
fd_insert() is currently called just after setting the owner and iocb,
but proceeding like this prevents the operation from being atomic and
requires a lock to protect the maxfd computation in another thread from
meeting an incompletely initialized FD and computing a wrong maxfd.
Fortunately for now all fdtab[].owner are set before calling fd_insert(),
and the first lock in fd_insert() enforces a memory barrier so the code
is safe.
This patch moves the initialization of the owner and iocb to fd_insert()
so that the function will be able to properly arrange its operations and
remain safe even when modified to become lockless. There's no other change
beyond the internal API.
These functions were created for poll() in 1.5-dev18 (commit 80da05a4) to
replace the previous FD_{CLR,SET,ISSET} that were shared with select()
because some libcs enforce a limit on FD_SET. But FD_SET doesn't seem
to be universally MT-safe, requiring locks in the select() code that
are not needed in the poll code. So let's move back to the initial
situation where we used to only use bit fields, since that has been in
use since day one without a problem, and let's use these hap_fd_*
functions instead of FD_*.
This patch only moves the functions to fd.h and revives hap_fd_isset()
that was recently removed to kill an "unused" warning.
Since only select() and poll() still make use of maxfd, let's move
its computation right there in the pollers themselves, and only
during each fd update pass. The computation doesn't need a lock
anymore, only a few atomic ops. It will be accurate, be done much
less often and will not be required anymore in the FD's fast patch.
This provides a small performance increase of about 1% in connection
rate when using epoll since we get rid of this computation which was
performed under a lock.
The incorrect comment was introduced in commit:
2ac5718dbd
v1.5-dev9 is the first tag containing this comment, the fix
should be backported to haproxy 1.5 and newer.
Marc Fournier reported an interesting case when using threads with the
master-worker mode : sometimes, a listener would have its FD closed
during startup. Sometimes it could even be health checks seeing this.
What happens is that after the threads are created, and the pollers
enabled on each threads, the master-worker pipe is registered, and at
the same time a close() is performed on the write side of this pipe
since the children must not use it.
But since this is replicated in every thread, what happens is that the
first thread closes the pipe, thus releases the FD, and the next thread
starting a listener in parallel gets this FD reassigned. Then another
thread closes the FD again, which this time corresponds to the listener.
It can also happen with the health check sockets if they're started
early enough.
This patch splits the mworker_pipe_register() function in two, so that
the close() of the write side of the FD is performed very early after the
fork() and long before threads are created (we don't need to delay it
anyway). Only the pipe registration is done in the threaded code since
it is important that the pollers are properly allocated for this.
The mworker_pipe_register() function now takes care of registering the
pipe only once, and this is guaranteed by a new surrounding lock.
The call to protocol_enable_all() looks fragile in theory since it
scans the list of proxies and their listeners, though in practice
all threads scan the same list and take the same locks for each
listener so it's not possible that any of them escapes the process
and finishes before all listeners are started. And the operation is
idempotent.
This fix must be backported to 1.8. Thanks to Marc for providing very
detailed traces clearly showing the problem.
Some pollers like epoll() need to know if the fd is already known or
not in order to compute the operation to perform (add, mod, del). For
now this is performed based on the difference between the previous FD
state and the new state but this will not be usable anymore once threads
become responsible for their own polling.
Here we come with a different approach : a bitmask is stored with the
fd to indicate which pollers already know it, and the pollers will be
able to simply perform the add/mod/del operations based on this bit
combined with the new state.
This patch only adds the bitmask declaration and initialization, it
is it not yet used. It will be needed by the next two fixes and will
need to be backported to 1.8.
Since the fd update tables are per-thread, we need to have a bit per
thread to indicate whether an update exists, otherwise this can lead
to lost update events every time multiple threads want to update the
same FD. In practice *for now*, it only happens at start time when
listeners are enabled and ask for polling after facing their first
EAGAIN. But since the pollers are still shared, a lost event is still
recovered by a neighbor thread. This will not reliably work anymore
with per-thread pollers, where it has been observed a few times on
startup that a single-threaded listener would not always accept
incoming connections upon startup.
It's worth noting that during this code review it appeared that the
"new" flag in the fdtab isn't used anymore.
This fix should be backported to 1.8.
A bitfield has been added to know if there are some FDs processable by a
specific thread in the FD cache. When a FD is inserted in the FD cache, the bits
corresponding to its thread_mask are set. On each thread, the bitfield is
updated when the FD cache is processed. If there is no FD processed, the thread
is removed from the bitfield by unsetting its tid_bit.
Note that this bitfield is updated but not checked in
fd_process_cached_events. So, when this function is called, the FDs cache is
always processed.
[wt: should be backported to 1.8 as it will help fix a design limitation]
A number of counters have been added at special places helping better
understanding certain bug reports. These counters are maintained per
thread and are shown using "show activity" on the CLI. The "clear
counters" commands also reset these counters. The output is sent as a
single write(), which currently produces up to about 7 kB of data for
64 threads. If more counters are added, it may be necessary to write
into multiple buffers, or to reset the counters.
To backport to 1.8 to help collect more detailed bug reports.
This one allows not to inflate some structures when threads are
disabled. Now struct global is 1.4 kB instead of 33 kB.
Should be backported to 1.8 for ease of backporting of upcoming
patches.
The "thread" part is 32kB long, better move it at the end of the
structure since it's only used during initialization, to keep the
rest grouped together.
Should be backported to 1.8 to ease backporting of upcoming patches,
no functional impact.
In addition to "option force-set-var", recently added, this directive can be
used to selectivelly register unknown variable names, without totally relaxing
their registration during the runtime, like "option force-set-var" does.
So there is no way for a malicious agent to exhaust memory by defining a too
high number of variable names. In other hand, you need to enumerate all
variable names. This could be painfull in some circumstances.
Remember, this directive is only usefull when the variable names are not
referenced anywhere in the HAProxy configuration or the SPOE one.
Thanks to Etienne Carrière for his help on this part.
Till now the use of __atomic_* gcc builtins required gcc >= 4.7. Since
some supported and quite common operating systems like CentOS 6 still
come with older versions (4.4) and the mapping to the older builtins
is reasonably simple, let's implement it.
This code is only used for gcc < 4.7. It has been quickly tested on a
machine using gcc 4.4.4 and provided expected results.
This patch should be backported to 1.8.
A SRV record weight can range from 0 to 65535, while haproxy weight goes
from 0 to 256, so we have to divide it by 256 before handing it to haproxy.
Also, a SRV record with a weight of 0 doesn't mean the server shouldn't be
used, so use a minimum weight of 1.
This should probably be backported to 1.8.
Since commit f9ce57e ("MEDIUM: connection: make conn_sock_shutw() aware
of lingering"), we refrain from performing the shutw() on the socket if
there is no lingering risk. But there is a problem with this in tunnel
and in TCP modes where a client is explicitly allowed to send a shutw
to the server, eventhough it it risky.
Not doing it creates this situation reported by Ricardo Fraile and
diagnosed by Christopher : a typical HTTP client (eg: curl) connecting
via the config below to an HTTP server would receive its response,
immediately close while the server remains in keep-alive mode. The
shutr() received by haproxy from the client is "propagated" to the
server side but not acted upon because fdtab[fd].linger_risk is set,
so we expect that the next close will immediately complete this
operation.
listen proxy-tcp
bind 127.0.0.1:8888
mode tcp
timeout connect 5s
timeout server 10s
timeout client 10s
server server1 127.0.0.1:8000
But since the whole stream will not end until the server closes in
turn, the server doesn't close and haproxy expires on server timeout.
This problem has already struck by waking up an older bug and was
partially fixed with commit 8059351 ("BUG/MEDIUM: http: don't disable
lingering on requests with tunnelled responses") though it was not
enough.
The problem is that linger_risk is not suited here. In fact we need to
know whether or not it is desired to close normally or silently, and
whether or not a shutr() has already been received on this connection.
This is the approach this patch takes, and it solves the problem for
the various difficult modes (tcp, http-server-close, pretend-keepalive).
This fix needs to be backported to 1.8. Many thanks to Ricardo for
providing very detailed traces and configurations.
The new function check_request_for_cacheability() is used to check if
a request may be served from the cache, and/or allows the response to
be stored into the cache. For this it checks the cache-control and
pragma header fields, and adjusts the existing TX_CACHEABLE and a new
TX_CACHE_IGNORE flags.
For now, just like its response side counterpart, it only checks the
first value of the header field. These functions should be reworked to
improve their parsers and validate all elements.
By copying the info in the stream interface that the mux cleanly reports
aborts, we'll have the ability to check this flag wherever needed regardless
of the presence of a mux or not.
This new field will be used to describe certain properties of some
muxes. For now we only add MX_FL_CLEAN_ABRT to indicate that a mux
is able to unambiguously report aborts using CS_FL_ERROR contrary
to others who may only report it via a read0. This will be used to
improve handling of the abortonclose option with H2. Other flags
may come later to report multiplexing capabilities or not, support
of client/server sides etc.
For security reasons, the spoe filter was only able to change values of
existing variables. In specific cases (ex : with LUA code), the name of
variables are unknown at the configuration parsing phase.
The force-set-var option can be enabled to register all variables.
Due to the nature of multiplexed protocols, it will often happen that
some operations are only performed on full frames, preventing any partial
operation from being performed. HTTP/2 is one such example. The current
MUX API causes a problem here because the rcv_buf() function has no way
to let the stream layer know that some data could not be read due to a
lack of room in the buffer, but that data are definitely present. The
problem with this is that the stream layer might not know it needs to
call the function again after it has made some room. And if the frame
in the buffer is not followed by any other, nothing will move anymore.
This patch introduces a new conn_stream flag CS_FL_RCV_MORE whose purpose
is to indicate on the stream that more data than what was received are
already available for reading as soon as more room will be available in
the buffer.
This patch doesn't make use of this flag yet, it only declares it. It is
expected that other similar flags may come in the future, such as reports
of pending end of stream, errors or any such event that might save the
caller from having to poll, or simply let it know that it can take some
actions after having processed data.
The thread patches adds refcount for notifications. The notifications are
used with the Lua cosocket. These refcount free the notifications when
the session is cleared. In the Lua task case, it not have sessions, so
the nofications are never cleraed.
This patch adds a garbage collector for signals. The garbage collector
just clean the notifications for which the end point is disconnected.
This patch should be backported in 1.8
The number of async fd is computed considering the maxconn, the number
of sides using ssl and the number of engines using async mode.
This patch should be backported on haproxy 1.8
In hpack_dht_make_room(), we try to fulfill this rule form RFC7541#4.4 :
"It is not an error to attempt to add an entry that is larger than the
maximum size; an attempt to add an entry larger than the maximum size
causes the table to be emptied of all existing entries and results in
an empty table."
Unfortunately it is not consistent with the way it's used in
hpack_dht_insert() as this last one will consider a success as a
confirmation it can copy the header into the table, and a failure as
an indexing error. This results in the two following issues :
- if a client sends too large a header into an empty table, this
header may overflow the table. Fortunately, most clients send
small headers like :authority first, and never mark headers that
don't fit into the table as indexable since it is counter-productive ;
- if a client sends too large a header into a populated table, the
operation fails after the table is totally flushed and the request
is not processed.
This patch fixes the two issues at once :
- a header not fitting into an empty table is always a sign that it
will never fit ;
- not fitting into the table is not an error
Thanks to Yves Lafon for reporting detailed traces demonstrating this
issue. This fix must be backported to 1.8.
If the hpack decoder sees an invalid header index, it emits value
"### ERR ###" that was used during debugging instead of rejecting the
block. This is harmless, and was detected by h2spec.
To backport to 1.8.
This BUG was introduced with:
'MEDIUM: threads/stick-tables: handle multithreads on stick tables'
The API was reviewed to handle stick table entry updates
asynchronously and the caller must now call a 'stkable_touch_*'
function each time the content of an entry is modified to
register the entry to be synced.
There was missing call to stktable_touch_* resulting in
not propagated entries to remote peers (or local one during reload)
server.h needs checks.h since it references the struct check, but depending
on the include order it will fail if check.h is included first due to this
one including server.h in turn while it doesn't need it.
Released version 1.9-dev0 with the following main changes :
- BUG/MEDIUM: stream: don't automatically forward connect nor close
- BUG/MAJOR: stream: ensure analysers are always called upon close
- BUG/MINOR: stream-int: don't try to read again when CF_READ_DONTWAIT is set
- MEDIUM: mworker: Add systemd `Type=notify` support
- BUG/MEDIUM: cache: free callback to remove from tree
- CLEANUP: cache: remove unused struct
- MEDIUM: cache: enable the HTTP analysers
- CLEANUP: cache: remove wrong comment
- MINOR: threads/atomic: rename local variables in macros to avoid conflicts
- MINOR: threads/plock: rename local variables in macros to avoid conflicts
- MINOR: threads/atomic: implement pl_mb() in asm on x86
- MINOR: threads/atomic: implement pl_bts() on non-x86
- MINOR: threads/build: atomic: replace the few inlines with macros
- BUILD: threads/plock: fix a build issue on Clang without optimization
- BUILD: ebtree: don't redefine types u32/s32 in scope-aware trees
- BUILD: compiler: add a new type modifier __maybe_unused
- BUILD: h2: mark some inlined functions "unused"
- BUILD: server: check->desc always exists
- BUG/MEDIUM: h2: properly report connection errors in headers and data handlers
- MEDIUM: h2: add a function to emit an HTTP/1 request from a headers list
- MEDIUM: h2: change hpack_decode_headers() to only provide a list of headers
- BUG/MEDIUM: h2: always reassemble the Cookie request header field
- BUG/MINOR: systemd: ignore daemon mode
- CONTRIB: spoa_example: allow to compile outside HAProxy.
- CONTRIB: spoa_example: remove bref, wordlist, cond_wordlist
- CONTRIB: spoa_example: remove last dependencies on type "sample"
- CONTRIB: spoa_example: remove SPOE enums that are useless for clients
- CLEANUP: cache: reorder includes
- MEDIUM: shctx: use unsigned int for len and block_count
- MEDIUM: cache: "show cache" on the cli
- BUG/MEDIUM: cache: use key=0 as a condition for freeing
- BUG/MEDIUM: cache: refcount forbids to free the objects
- BUG/MEDIUM: cache fix cli_kws structure
- BUG/MEDIUM: deinit: correctly deinitialize the proxy and global listener tasks
- BUG/MINOR: ssl: Always start the handshake if we can't send early data.
- MINOR: ssl: Don't disable early data handling if we could not write.
- MINOR: pools: prepare functions to override malloc/free in pools
- MINOR: pools: implement DEBUG_UAF to detect use after free
- BUG/MEDIUM: threads/time: fix time drift correction
- BUG/MEDIUM: threads/time: maintain a common time reference between all threads
- MINOR: sample: Add "thread" sample fetch
- BUG/MINOR: Use crt_base instead of ca_base when crt is parsed on a server line
- BUG/MINOR: stream: fix tv_request calculation for applets
- BUG/MAJOR: h2: always remove a stream from the send list before freeing it
- BUG/MAJOR: threads/task: dequeue expired tasks under the WQ lock
- MINOR: ssl: Handle reading early data after writing better.
- MINOR: mux: Make sure every string is woken up after the handshake.
- MEDIUM: cache: store sha1 for hashing the cache key
- MINOR: http: implement the "http-request reject" rule
- MINOR: h2: send RST_STREAM before GOAWAY on reject
- MEDIUM: h2: don't gracefully close the connection anymore on Connection: close
- MINOR: h2: make use of client-fin timeout after GOAWAY
- MEDIUM: config: ensure that tune.bufsize is at least 16384 when using HTTP/2
- MINOR: ssl: Handle early data with BoringSSL
- BUG/MEDIUM: stream: always release the stream-interface on abort
- BUG/MEDIUM: cache: free ressources in chn_end_analyze
- MINOR: cache: move the refcount decrease in the applet release
- BUG/MINOR: listener: Allow multiple "process" options on "bind" lines
- MINOR: config: Support a range to specify processes in "cpu-map" parameter
- MINOR: config: Slightly change how parse_process_number works
- MINOR: config: Export parse_process_number and use it wherever it's applicable
- MINOR: standard: Add my_ffsl function to get the position of the bit set to one
- MINOR: config: Add auto-increment feature for cpu-map
- MINOR: config: Support partial ranges in cpu-map directive
- MINOR:: config: Remove thread-map directive
- MINOR: config: Add the threads support in cpu-map directive
- MINOR: config: Add threads support for "process" option on "bind" lines
- MEDIUM: listener: Bind listeners on a thread subset if specified
- CLEANUP: debug: Use DPRINTF instead of fprintf into #ifdef DEBUG_FULL/#endif
- CLEANUP: log: Rename Alert/Warning in ha_alert/ha_warning
- MINOR/CLEANUP: proxy: rename "proxy" to "proxies_list"
- CLEANUP: pools: rename all pool functions and pointers to remove this "2"
- DOC: update the roadmap file with the latest changes merged in 1.8
- DOC: fix mangled version in peers protocol documentation
- DOC: add initial peers protovol v2.0 documentation.
- DOC: mention William as maintainer of the cache and master-worker
- DOC: add Christopher and Emeric as maintainers of the threads
- MINOR: cache: replace a fprint() by an abort()
- MEDIUM: cache: max-age configuration keyword
- DOC: explain HTTP2 timeout behavior
- DOC: cache: configuration and management
- MAJOR: mworker: exits the master on failure
- BUG/MINOR: threads: don't drop "extern" on the lock in include files
- MINOR: task: keep a pointer to the currently running task
- MINOR: task: align the rq and wq locks
- MINOR: fd: cache-align fdtab and fdcache locks
- MINOR: buffers: cache-align buffer_wq_lock
- CLEANUP: server: reorder some fields in struct server to save 40 bytes
- CLEANUP: proxy: slightly reorder the struct proxy to reduce holes
- CLEANUP: checks: remove 16 bytes of holes in struct check
- CLEANUP: cache: more efficiently pack the struct cache
- CLEANUP: fd: place the lock at the beginning of struct fdtab
- CLEANUP: pools: align pools on a cache line
- DOC: config: add a few bits about how to configure HTTP/2
- BUG/MAJOR: threads/queue: avoid recursive locking in pendconn_get_next_strm()
- BUILD: Makefile: reorder object files by size
pendconn_get_next_strm() is called from process_srv_queue() under the
server lock, and calls stream_add_srv_conn() with this lock held, while
the latter tries to take it again. This results in a deadlock when
a server's maxconn is reached and haproxy is built with thread support.
There are just a few pools, and they're stressed a lot, so it makes
sense to dedicate them a cache line to avoid contention and to place
the lock at the beginning.
The struct is not cache line aligned but at least, every time the lock
will appear in the same cache line as the fd it will benefit from being
accessed first. This improves the performance by about 2% on fd-intensive
workloads with 4 threads.
Commit 9dcf9b6 ("MINOR: threads: Use __decl_hathreads to declare locks")
accidently lost a few "extern" in certain lock declarations, possibly
causing certain entries to be declared at multiple places. Apparently
it hasn't caused any harm though.
The offending ones were :
- fdtab_lock
- fdcache_lock
- poll_lock
- buffer_wq_lock
This patch changes the behavior of the master during the exit of a
worker.
When a worker exits with an error code, for example in the case of a
segfault, all workers are now killed and the master leaves.
If you don't want this behavior you can use the option
"master-worker no-exit-on-failure".
During the migration to the second version of the pools, the new
functions and pool pointers were all called "pool_something2()" and
"pool2_something". Now there's no more pool v1 code and it's a real
pain to still have to deal with this. Let's clean this up now by
removing the "2" everywhere, and by renaming the pool heads
"pool_head_something".
Rename the global variable "proxy" to "proxies_list".
There's been multiple proxies in haproxy for quite some time, and "proxy"
is a potential source of bugs, a number of functions have a "proxy" argument,
and some code used "proxy" when it really meant "px" or "curproxy". It worked
by pure luck, because it usually happened while parsing the config, and thus
"proxy" pointed to the currently parsed proxy, but we should probably not
rely on this.
[wt: some of these are definitely fixes that are worth backporting]
It is now possible on a "bind" line (or a "stats socket" line) to specify the
thread set allowed to process listener's connections. For instance:
# HTTPS connections will be processed by all threads but the first and HTTP
# connection will be processed on the first thread.
bind *:80 process 1/1
bind *:443 ssl crt mycert.pem process 1/2-
Now, it is possible to bind CPU at the thread level instead of the process level
by defining a thread set in "cpu-map" directives. Thus, its format is now:
cpu-map [auto:]<process-set>[/<thread-set>] <cpu-set>...
where <process-set> and <thread-set> must follow the format:
all | odd | even | number[-[number]]
Having a process range and a thread range in same time with the "auto:" prefix
is not supported. Only one range is supported, the other one must be a fixed
number. But it is allowed when there is no "auto:" prefix.
Because it is possible to define a mapping for a process and another for a
thread on this process, threads will be bound on the intersection of their
mapping and the one of the process on which they are attached. If the
intersection is null, no specific binding will be set for the threads.
The prefix "auto:" can be added before the process set to let HAProxy
automatically bind a process to a CPU by incrementing process and CPU sets. To
be valid, both sets must have the same size. No matter the declaration order of
the CPU sets, it will be bound from the lower to the higher bound.
Examples:
# all these lines bind the process 1 to the cpu 0, the process 2 to cpu 1
# and so on.
cpu-map auto:1-4 0-3
cpu-map auto:1-4 0-1 2-3
cpu-map auto:1-4 3 2 1 0
# bind each process to exaclty one CPU using all/odd/even keyword
cpu-map auto:all 0-63
cpu-map auto:even 0-31
cpu-map auto:odd 32-63
# invalid cpu-map because process and CPU sets have different sizes.
cpu-map auto:1-4 0 # invalid
cpu-map auto:1 0-3 # invalid
The cache was relying on the txn->uri for creating its key, which was a
big problem when there was no log activated.
This patch does a sha1 of the host + uri, and stores it in the txn.
When a object is stored, the eb32node uses the first 32 bits of the hash
as a key, and the whole hash is stored in the cache entry.
During a lookup, the truncated hash is used, and when it matches an
entry we check the real sha1.
It can happen that we want to read early data, write some, and then continue
reading them.
To do so, we can't reuse tmp_early_data to store the amount of data sent,
so introduce a new member.
If we read early data, then ssl_sock_to_buf() is now the only responsible
for getting back to the handshake, to make sure we don't miss any early data.
This code has been used successfully a few times in the past to detect
that a pool was used after being freed. Its main goal is to allocate a
full page for each object so that they are always released individually
and unmapped from memory. This way if any part of the code reference the
object after is was freed and before it is reallocated, a segv occurs at
the exact offending location. It does a few extra things such as writing
to the memory area before freeing to detect double-frees and free of
read-only areas, and placing the data at the end of the page instead of
the beginning so that out of bounds accesses are easier to spot. The
amount of memory used with this is huge (about 10 times the regular
usage) but it can be useful sometimes.
Allows bigger objects to be cached in the shctx, the first
implementation was only storing small ssl session, but we want to store
bigger HTTP response.
The current H2 to H1 protocol conversion presents some issues which will
require to perform some processing on certain headers before writing them
so it's not possible to convert HPACK to H1 on the fly.
This commit modifies the headers decoding so that it now works in two
phases : hpack_decode_headers() only decodes the HPACK stream in the
HEADERS frame and puts the result into a list. Headers which require
storage (huffman-compressed or from the dynamic table) are stored in
a chunk allocated by the H2 demuxer. Then once the headers are properly
decoded into this list, h2_make_h1_request() is called with this list
to produce the HTTP/1.1 request into the destination buffer. The list
necessarily enforces a limit. Here we use 2*MAX_HTTP_HDR, which means
that we can have as many individual cookies as we have regular headers
if a client decides to break their cookies into multiple values. This
seams reasonable and will allow the H1 parser to decide whether it's
too much or not.
Thus the output stream is not produced on the fly anymore and this will
permit to deal with certain corner cases like reparing the Cookie header
(which for now is not done).
In order to limit header duplication and parsing, the known pseudo headers
continue to be passed by their index : the name element in the list then
has a NULL pointer and the value is the pseudo header's index. Given that
these ones represent about half of the incoming requests and need to be
found quickly, it maintains an acceptable level of performance.
The code was significantly reduced by doing this because the orignal code
had to deal with HPACK and H1 combinations (eg: index vs not indexed, etc)
and now the HPACK decoding is totally focused on the decompression, and
the H1 encoding doesn't have to deal with the issue of wrapping input for
example.
One bug was addressed here (though it couldn't happen at the moment). The
H2 demuxer used to detect a failure to write the request into the H1 buffer
and would then detect if the output buffer wraps, realign it and try again.
The problem by doing so was that the HPACK context was already modified and
not rewindable. Thus the size check is now performed first and a failure is
reported if it doesn't fit.
The current H2 to H1 protocol conversion presents some issues which will
require to perform some processing on certain headers before writing them
so it's not possible to convert HPACK to H1 on the fly.
Here we introduce a function which performs half of what hpack_decode_header()
used to do, which is to take a list of headers on input and emit the
corresponding request in HTTP/1.1 format. The code is the same and functions
were renamed to be prefixed with "h2" instead of "hpack", though it ends
up being simpler as the various HPACK-specific cases could be fused into
a single one (ie: add header).
Moving this part here makes a lot of sense as now this code is specific to
what is documented in HTTP/2 RFC 7540 and will be able to deal with special
cases related to H2 to H1 conversion enumerated in section 8.1.
Various error codes which were previously assigned to HPACK were never
used (aside being negative) and were all replaced by -1 with a comment
indicating what error was detected. The code could be further factored
thanks to this but this commit focuses on compatibility first.
This code is not yet used but builds fine.
While gcc only emits warnings about unused static functions, Clang also
emits such a warning when the functions are inlined. This is a bit
annoying at certain places where functions are provided to manipulate
multiple data types and are not yet used. Let's have a type modifier
"__maybe_unused" which sets the "unused" attribute like the Linux kernel
does. It's elegant as it allows the code author to indicate that it knows
that this element might be unused. It works on variables as well, which
is convenient to remove ifdefs around local variables in certain functions,
but doesn't work on labels.
[ plock commit 4c53fd3a0b2b1892817cebd0db012a52f4087850 ]
Pieter Baauw reported a build issue affecting haproxy after plock was
included. It happens that expressions of the form :
if ((const) ? (expr1) : (expr2))
do_something()
always produce code for both expr1 and expr2 on Clang when building
without optimization. The resulting asm code is even funny, basically
doing :
mov reg, 1
cmp reg, 1
...
This causes our sizeof() tests to fail to build because we purposely
dereference a fake function that reports the location and nature of the
inconsistency, but this fake function appears in the object code despite
all conditions being there to avoid it.
However the compiler is still smart enough to optimize away code doing
if (const)
do_something()
So we simply repeat the condition before do_something(), and the dummy
function is not referenced anymore unless really required.
[ plock commit 61e255286ae32e83e1a3174dd7c49eda99880a8b]
There are a few inlines such as pl_barrier() and pl_cpu_relax() which
are used a lot. Unfortunately, while building test code at -O0, inlining
is disabled and these ones are called a lot and show up a lot in any
profile, are traced into when single-stepping with a debugger, etc, thus
they are polluting the landscape. Since they're single-asm statements,
there is no reason for not turning them into macros.
The result becomes fairly visible here at -O0 :
$ size latency.inline latency.macro
text data bss dec hex filename
11431 692 656 12779 31eb treelock.inline
10967 692 656 12315 301b treelock.macro
And it was verified that regularly optimized code remains strictly identical.
[ plock commit 44081ea493dd78dab48076980e881748e9b33db5 ]
Older compilers (eg: gcc 3.4) don't provide __sync_synchronize() so let's
do it by hand on this platform.
[ plock commit b155d5c762fb9a9793911881f80e61faa6b0e889 ]
Local variables "l", "i" and "ret" were renamed "__pl_l", "__pl_i" and
"__pl_r" respectively, to limit the risk of conflicts with existing
variables in application code.
[ plock commit bfac5887ebabb8ef753b0351f162265767eb219b ]
Local variable "t" was renamed "__pl_t" to limit the risk of conflicts
with existing variables in application code.
This patch adds support for `Type=notify` to the systemd unit.
Supporting `Type=notify` improves both starting as well as reloading
of the unit, because systemd will be let known when the action completed.
See this quote from `systemd.service(5)`:
> Note however that reloading a daemon by sending a signal (as with the
> example line above) is usually not a good choice, because this is an
> asynchronous operation and hence not suitable to order reloads of
> multiple services against each other. It is strongly recommended to
> set ExecReload= to a command that not only triggers a configuration
> reload of the daemon, but also synchronously waits for it to complete.
By making systemd aware of a reload in progress it is able to wait until
the reload actually succeeded.
This patch introduces both a new `USE_SYSTEMD` build option which controls
including the sd-daemon library as well as a `-Ws` runtime option which
runs haproxy in master-worker mode with systemd support.
When haproxy is running in master-worker mode with systemd support it will
send status messages to systemd using `sd_notify(3)` in the following cases:
- The master process forked off the worker processes (READY=1)
- The master process entered the `mworker_reload()` function (RELOADING=1)
- The master process received the SIGUSR1 or SIGTERM signal (STOPPING=1)
Change the unit file to specify `Type=notify` and replace master-worker
mode (`-W`) with master-worker mode with systemd support (`-Ws`).
Future evolutions of this feature could include making use of the `STATUS`
feature of `sd_notify()` to send information about the number of active
connections to systemd. This would require bidirectional communication
between the master and the workers and thus is left for future work.
Instead of storing the SSL_SESSION pointer directly in the struct server,
store the ASN1 representation, otherwise, session resumption is broken with
TLS 1.3, when multiple outgoing connections want to use the same session.
a bitfield has been added to know if there are runnable applets for a
thread. When an applet is woken up, the bits corresponding to its thread_mask
are set. When all active applets for a thread is get to be processed, the thread
is removed from active ones by unsetting its tid_bit from the bitfield.
a bitfield has been added to know if there are runnable tasks for a thread. When
a task is woken up, the bits corresponding to its thread_mask are set. When all
tasks for a thread have been evaluated without any wakeup, the thread is removed
from active ones by unsetting its tid_bit from the bitfield.
At the end of the master initialisation, a call to protocol_unbind_all()
was made, in order to close all the FDs.
Unfortunately, this function closes the inherited FDs (fd@), upon reload
the master wasn't able to reload a configuration with those FDs.
The create_listeners() function now store a flag to specify if the fd
was inherited or not.
Replace the protocol_unbind_all() by mworker_cleanlisteners() +
deinit_pollers()
Now we can show in dotted red the node being removed or surrounded in red
a node having been inserted, and add a description on the graph related to
the operation in progress for example.
b_alloc_margin is, strickly speeking, thread-safe. It will not crash
HAproxy. But its contract is not respected anymore in a multithreaded
environment. In this function, we need to be sure to have <margin> buffers
available in the pool after the allocation. So to have this guarantee, we must
lock the memory pool during all the operation. This also means, we must call
internal and lockless memory functions (prefixed with '__').
For the record, this patch fixes a pernicious bug happens after a soft reload
where some streams can be blocked infinitly, waiting for a buffer in the
buffer_wq list. This happens because, during a soft reload, pool_gc2 is called,
making some calls to b_alloc_fast fail.
This is specific to threads, no backport is needed.
This macro should be used to declare variables or struct members depending on
the USE_THREAD compile option. It avoids the encapsulation of such declarations
between #ifdef/#endif. It is used to declare all lock variables.
At a number of places, bitmasks are used for process affinity and to map
listeners to processes. Every time 1UL<<(relative_pid-1) is used. Let's
create a "pid_bit" variable corresponding to this value to clean this up.
In commit 53a4766 ("MEDIUM: connection: start to introduce a mux layer
between xprt and data") we introduced a release() function which ends
up never being used. Let's get rid of it now.
This small inline function causes some pain to the compiler when used
inside other functions due to its use of the unlikely() hint for non-digits.
It causes the letters to be processed far away in the calling function and
makes the code less efficient. Removing these unlikely() hints has increased
the chunk size parsing by around 5%.
The HTTP/1 code always has the reserve left available so the buffer is
never full there. But with HTTP/2 we have to deal with full buffers,
and it happens that the chunk size parser cannot tell the difference
between a full buffer and an empty one since it compares the start and
the stop pointer.
Let's change this to instead deal with the number of bytes left to process.
As a side effect, this code ends up being about 10% faster than the previous
one, even on HTTP/1.
When a write activity is reported on a channel, it is important to keep this
information for the stream because it take part on the analyzers' triggering.
When some data are written, the flag CF_WRITE_PARTIAL is set. It participates to
the task's timeout updates and to the stream's waking. It is also used in
CF_MASK_ANALYSER mask to trigger channels anaylzers. In the past, it was cleared
by process_stream. Because of a bug (fixed in commit 95fad5ba4 ["BUG/MAJOR:
stream-int: don't re-arm recv if send fails"]), It is now cleared before each
send and in stream_int_notify. So it is possible to loss this information when
process_stream is called, preventing analyzers to be called, and possibly
leading to a stalled stream.
Today, this happens in HTTP2 when you call the stat page or when you use the
cache filter. In fact, this happens when the response is sent by an applet. In
HTTP1, everything seems to work as expected.
To fix the problem, we need to make the difference between the write activity
reported to lower layers and the one reported to the stream. So the flag
CF_WRITE_EVENT has been added to notify the stream of the write activity on a
channel. It is set when a send succedded and reset by process_stream. It is also
used in CF_MASK_ANALYSER. finally, it is checked in stream_int_notify to wake up
a stream and in channel_check_timeouts.
This bug is probably present in 1.7 but it seems to have no effect. So for now,
no needs to backport it.
The H1 parser used by the H2 gateway was a bit lax and could validate
non-numbers in the status code. Since it computes the code on the fly
it's problematic, as "30:" is read as status code 310. Let's properly
check that it's a number now. No backport needed.
This adds a new keyword on the "server" line, "allow-0rtt", if set, we'll try
to send early data to the server, as long as the client sent early data, as
in case the server rejects the early data, we no longer have them, and can't
resend them, so the only option we have is to send back a 425, and we need
to be sure the client knows how to interpret it correctly.
The spin locks used to rely on W locks, which involve a loop waiting
for readers to leave, and this doesn't happen here. It's more efficient
to use S locks instead, which are also mutually exclusive and do not
have this loop. This saves one test per spinlock and a few tens of
bytes allowing certain functions to be inlined.
Currently the task scheduler suffers from an O(n) lookup when
skipping tasks that are not for the current thread. The reason
is that eb32_lookup_ge() has no information about the current
thread so it always revisits many tasks for other threads before
finding its own tasks.
This is particularly visible with HTTP/2 since the number of
concurrent streams created at once causes long series of tasks
for the same stream in the scheduler. With only 10 connections
and 100 streams each, by running on two threads, the performance
drops from 640kreq/s to 11.2kreq/s! Lookup metrics show that for
only 200000 task lookups, 430 million skips had to be performed,
which means that on average, each lookup leads to 2150 nodes to
be visited.
This commit backports the principle of scope lookups for ebtrees
from the ebtree_v7 development tree. The idea is that each node
contains a mask indicating the union of the scopes for the nodes
below it, which is fed during insertion, and used during lookups.
Then during lookups, branches that do not contain any leaf matching
the requested scope are simply ignored. This perfectly matches a
thread mask, allowing a thread to only extract the tasks it cares
about from the run queue, and to always find them in O(log(n))
instead of O(n). Thus the scheduler uses tid_bit and
task->thread_mask as the ebtree scope here.
Doing this has recovered most of the performance, as can be seen on
the test below with two threads, 10 connections, 100 streams each,
and 1 million requests total :
Before After Gain
test duration : 89.6s 4.73s x19
HTTP requests/s (DEBUG) : 11200 211300 x19
HTTP requests/s (PROD) : 15900 447000 x28
spin_lock time : 85.2s 0.46s /185
time per lookup : 13us 40ns /325
Even when going to 6 threads (on 3 hyperthreaded CPU cores), the
performance stays around 284000 req/s, showing that the contention
is much lower.
A test showed that there's no benefit in using this for the wait queue
though.
The __appctx_wakeup() function already does it. It matters with threads
enabled because it simplifies the code in appctx_res_wakeup() to get rid
of this test.
unbind_listener() takes the listener lock, which is already held by
enable_listener(). This situation happens when starting with nbproc > 1
with some bind lines limited to a certain process, because in this case
enable_listener() tries to stop unneeded listeners.
This commit introduces __do_unbind_listeners() which must be called with
the lock held, and makes enable_listener() use this one. Given that the
only return code has never been used and that it starts to make the code
more complicated to propagate it before throwing it to the trash, the
function's return type was changed to void.
This function incorrectly dealt with the case where data doesn't
wrap but lies at the end of the buffer, resulting in Lukas' reported
data corruption with HTTP/2. No backport is needed, it was introduced
for HTTP/2 in 1.8-dev.
For now it only supports literals and a bit of static header table
references for the 9 most common header field names (date, server,
content-type, content-length, last-modified, accept-ranges, etag,
cache-control, location).
A previous incarnation of this commit used to strip the forbidden H2
header names (connection, proxy-connection, upgrade, transfer-encoding,
keep-alive) but this is no longer the case as this filtering is irrelevant
to HPACK encoding and is specific to H2, so this will have to be done by
the caller.
It's quite not optimal but works fine enough to prepare some valid and
partially compressed responses during development.
The decoder is now fully functional. It makes use of the dynamic header
table. Dynamic header table size updates are currently ignored, as our
initially advertised value is the highest we support. Strictly speaking,
the impact is that a client referencing a header field after such an
update wouldn't observe an error instead of the connection being dropped
if it was implemented.
Decoded header fields are copied into a target buffer in HTTP/1 format
using HTTP/1.1 as the version. The Host header field is automatically
appended if a ":authority" header field is present.
All decoded header fields can be displayed if the file is compiled with
DEBUG_HPACK.
This code deals with header insertion, retrieval and eviction, as well
as with dynamic header table defragmentation. It is functional for use
as a decoder and was heavily tested in this context. There's still some
room for optimization (eg: the defragmentation code currently does it
in place using a memcpy).
Also for now the dynamic header table is allocated using malloc() while
a pool needs to be created instead.
This code was mostly imported from https://github.com/wtarreau/http2-exp
with "hpack_" prepended in front of most names to avoid risks of conflicts.
Some small cleanups and renamings were applied during the import. This
version must be considered more recent.
Some HPACK error codes were placed here (HPACK_ERR_*), not exactly because
they're needed by the decoder but they'll be needed by all callers. Maybe
a different location should be found.
The code was borrowed from the HPACK experimental implementations
available here :
https://github.com/wtarreau/http2-exp
It contains the Huffman table as specified in RFC7541 Appendix B, and a
set of reverse tables used to decode a Huffman byte stream, and produced
by contrib/h2/gen-rht. The encoder is not finalized, it doesn't emit the
byte stream but this is not needed for now.
This callback will be used to release upper layers when a mux is in
use. Given that the mux can be asynchronously deleted, we need a way
to release the extra information such as the session.
This callback will be called directly by the mux upon releasing
everything and before the connection itself is released, so that
the callee can find its information inside the connection if needed.
The way it currently works is not perfect, and most likely this should
instead become a mux release callback, but for now we have no easy way
to add mux-specific stuff, and since there's one mux per connection,
it works fine this way.
For H2, only the mux's timeout or other conditions might cause a
release of the mux and the connection, no stream should be allowed
to kill such a shared connection. So a stream will only detach using
cs_destroy() which will call mux->detach() then free the cs.
For now it's only handled by mux_pt. The goal is that the data layer
never has to care about the connection, which will have to be released
depending on the mux's mood.
This basically calls cs_shutw() followed by cs_shutr(). Both of them
are called in the most conservative mode so that any previous call is
still respected. The CS flags are cleared so that it can be reused
(this is important for connection retries when conn and CS are reused
without being reallocated).
In order to support all shutdown modes on the CS, we introduce the
following flags :
CS_FL_SHRD : shut read, drain extra data
CS_FL_SHRR : shut read, reset extra data
CS_FL_SHWN : shut write, normal notification
CS_FL_SHWS : shut write, silent mode (no notification)
And the following modes for shutr/shutw :
CS_SHR_DRAIN, CS_SHR_RESET, CS_SHW_NORMAL, CS_SHW_SILENT.
Note: it's possible that we won't need to distinguish the two shutw
above as they're only an action.
For now they are not used.
All the references to connections in the data path from streams and
stream_interfaces were changed to use conn_streams. Most functions named
"something_conn" were renamed to "something_cs" for this. Sometimes the
connection still is what matters (eg during a connection establishment)
and were not always renamed. The change is significant and minimal at the
same time, and was quite thoroughly tested now. As of this patch, all
accesses to the connection from upper layers go through the pass-through
mux.
Most of the functions dealing with conn_streams are here. They act at
the data layer and interact with the mux. For now they are not used yet
but everything builds.
This patch introduces a new struct conn_stream. It's the stream-side of
a multiplexed connection. A pool is created and destroyed on exit. For
now the conn_streams are not used at all.
When an incoming connection is made on an HTTP mode frontend, the
session now looks up the mux to use based on the ALPN token and the
proxy mode. This will allow easier mux registration, and we don't
need to hard-code the mux_pt_ops anymore.
Selecting a mux based on ALPN and the proxy mode will quickly become a
pain. This commit provides new functions to register/lookup a mux based
on the ALPN string and the proxy mode to make this easier. Given that
we're not supposed to support a wide range of muxes, the lookup should
not have any measurable performance impact.
For HTTP/2 and QUIC, we'll need to deal with multiplexed streams inside
a connection. After quite a long brainstorming, it appears that the
connection interface to the existing streams is appropriate just like
the connection interface to the lower layers. In fact we need to have
the mux layer in the middle of the connection, between the transport
and the data layer.
A mux can exist on two directions/sides. On the inbound direction, it
instanciates new streams from incoming connections, while on the outbound
direction it muxes streams into outgoing connections. The difference is
visible on the mux->init() call : in one case, an upper context is already
known (outgoing connection), and in the other case, the upper context is
not yet known (incoming connection) and will have to be allocated by the
mux. The session doesn't have to create the new streams anymore, as this
is performed by the mux itself.
This patch introduces this and creates a pass-through mux called
"mux_pt" which is used for all new connections and which only
calls the data layer's recv,send,wake() calls. One incoming stream
is immediately created when init() is called on the inbound direction.
There should not be any visible impact.
Note that the connection's mux is purposely not set until the session
is completed so that we don't accidently run with the wrong mux. This
must not cause any issue as the xprt_done_cb function is always called
prior to using mux's recv/send functions.
This is needed in the H2->H1 gateway so that we know how long the trailers
block is in chunked encoding. It returns the number of bytes, or 0 if some
are missing, or -1 in case of parse error.
It was a leftover from the last cleaning session; this mask applies
to threads and calling it process_mask is a bit confusing. It's the
same in fd, task and applets.
srv_set_fqdn() may be called with the DNS lock already held, but tries to
lock it anyway. So, add a new parameter to let it know if it was already
locked or not;
Commit 819fc6f ("MEDIUM: threads/stick-tables: handle multithreads on
stick tables") introduced a valid warning about an uninitialized return
value in stksess_kill_if_expired(). It just happens that this result is
never used, so let's turn the function back to void as previously.
The wrong bit was set to keep the lock on freq counter update. And the read
functions were re-worked to use volatile.
Moreover, when a freq counter is updated, it is now rotated only if the current
counter is in the past (now.tv_sec > ctr->curr_sec). It is important with
threads because the current time (now) is thread-local. So, rounded to the
second, the time may vary by more or less 1 second. So a freq counter rotated by
one thread may be see 1 second in the future. In this case, it is updated but
not rotated.
There was a flaw in the way the threads was created. the main one was just used
to create all the others and just wait to exit. Now, it is used to run a poll
loop. So we only create nbthread-1 threads.
This also fixes a bug about the compression filter when there is only 1 thread
(nbthread == 1 or no threads support). The bug was in the way thread-local
resources was initialized. per-thread init/deinit callbacks were never called
for the main process. So, with nthread set to 1, some buffers remained
uninitialized.
By default, no affinity is set for threads. To bind threads on CPU, you must
define a "thread-map" in the global section. The format is the same than the
"cpu-map" parameter, with a small difference. The process number must be
defined, with the same format than cpu-map ("all", "even", "odd" or a number
between 1 and 31/63).
A thread will be bound on the intersection of its mapping and the one of the
process on which it is attached. If the intersection is null, no specific bind
will be set for the thread.
Because there is not migration mechanism yet, all runtime information about an
SPOE agent are thread-local and async exchanges with agents are disabled when we
have serveral threads. Howerver, pipelining is still available. So for now, the
thread part of the SPOE is pretty simple.
