Released version 2.5-dev9 with the following main changes :
- head-truc
- REGTESTS: lua: test the httpclient:get() feature
- Revert "head-truc"
- BUG/MEDIUM: httpclient: replace ist0 by istptr
- MINOR: config: use a standard parser for the "nbthread" keyword
- CLEANUP: init: remove useless test against MAX_THREADS in affinity loop
- MEDIUM: init: de-uglify the per-thread affinity setting
- MINOR: init: extract the setup and end of threads to their own functions
- MINOR: log: Try to get the status code when MUX_EXIT_STATUS is retrieved
- MINOR: mux-h1: Set error code if possible when MUX_EXIT_STATUS is returned
- MINOR: mux-h1: Be able to set custom status code on parsing error
- MEDIUM: mux-h1: Reject HTTP/1.0 GET/HEAD/DELETE requests with a payload
- MEDIUM: h1: Force close mode for invalid uses of T-E header
- BUG/MINOR: mux-h1/mux-fcgi: Sanitize TE header to only send "trailers"
- MINOR: http: Add 422-Unprocessable-Content error message
- MINOR: h1: Change T-E header parsing to fail if chunked encoding is found twice
- BUG/MEDIUM: mux-h1/mux-fcgi: Reject messages with unknown transfer encoding
- REGTESTS: Add script to validate T-E header parsing
- REORG: pools: move default settings to defaults.h
- DOC: peers: fix doc "enable" statement on "peers" sections
- MINOR: Makefile: add MEMORY_POOLS to the list of DEBUG_xxx options
- MINOR: ssl: Set connection error code in case of SSL read or write fatal failure
- MINOR: ssl: Rename ssl_bc_hsk_err to ssl_bc_err
- MINOR: ssl: Store the last SSL error code in case of read or write failure
- REGTESTS: ssl: enable show_ssl_ocspresponse.vtc again
- REGTESTS: ssl: enable ssl_crt-list_filters.vtc again
- BUG/MEDIUM: lua: fix wakeup condition from sleep()
- BUG/MAJOR: lua: use task_wakeup() to properly run a task once
- MINOR: arg: Be able to forbid unresolved args when building an argument list
- BUG/MINOR: tcpcheck: Don't use arg list for default proxies during parsing
- BUG/MINOR: tcp-rules: Stop content rules eval on read error and end-of-input
- MINOR: tasks: catch TICK_ETERNITY with BUG_ON() in __task_queue()
- REGTESTS: ssl: show_ssl_ocspresponse w/ freebsd won't use base64
- REGTESTS: ssl: wrong feature cmd in show_ssl_ocspresponse.vtc
- CLEANUP: tasks: remove the long-unused work_lists
- MINOR: task: provide 3 task_new_* wrappers to simplify the API
- MINOR: time: uninline report_idle() and move it to task.c
- REORG: sched: move idle time calculation from time.h to task.h
- REORG: sched: move the stolen CPU time detection to sched_entering_poll()
- BUG/MEDIUM: filters: Fix a typo when a filter is attached blocking the release
- BUG/MEDIUM: http-ana: Clear request analyzers when applying redirect rule
- MINOR: httpclient: destroy() must free the headers and the ists
- MINOR: httpclient: set HTTPCLIENT_F_ENDED only in release
- MINOR: httpclient: stop_and_destroy() ask the applet to autokill
- MINOR: httpclient: test if started during stop_and_destroy()
- MINOR: httpclient/lua: implement garbage collection
- BUG/MEDIUM: httpclient/lua: crash because of b_xfer and get_trash_chunk()
- MINOR: httpclient: destroy checks if a client was started but not stopped
- BUG/MINOR: httpclient/lua: does not process headers when failed
- MINOR: httpclient/lua: supports headers via named arguments
- CLEANUP: server: always include the storage for SSL settings
- CLEANUP: sample: rename sample_conv_var2smp() to *_sint
- CLEANUP: sample: uninline sample_conv_var2smp_str()
- MINOR: sample: provide a generic var-to-sample conversion function
- BUG/MEDIUM: sample: properly verify that variables cast to sample
- BUILD: action: add the relevant structures for function arguments
- BUILD: extcheck: needs to include stream-t.h
- BUILD: hlua: needs to include stream-t.h
- BUILD: stats: define several missing structures in stats.h
- BUILD: resolvers: define missing types in resolvers.h
- BUILD: httpclient: include missing ssl_sock-t
- BUILD: sample: include openssl-compat
- BUILD: http_ana: need to include proxy-t to get redirect_rule
- BUILD: http_rules: requires http_ana-t.h for REDIRECT_*
- BUILD: vars: need to include xxhash
- BUILD: peers: need to include eb{32/mb/pt}tree.h
- BUILD: ssl_ckch: include ebpttree.h in ssl_ckch.c
- BUILD: compiler: add the container_of() and container_of_safe() macros
- BUILD: idleconns: include missing ebmbtree.h at several places
- BUILD: connection: connection.h needs list.h and server.h
- BUILD: tree-wide: add missing http_ana.h from many places
- BUILD: cfgparse-ssl: add missing errors.h
- BUILD: tcp_sample: include missing errors.h and session-t.h
- BUILD: mworker: mworker-prog needs time.h for the 'now' variable
- BUILD: tree-wide: add several missing activity.h
- BUILD: compat: fix -Wundef on SO_REUSEADDR
- CLEANUP: pools: pools-t.h doesn't need to include thread-t.h
- REORG: pools: uninline the UAF allocator and force-inline the rest
- REORG: thread: uninline the lock-debugging code
- MINOR: thread/debug: replace nsec_now() with now_mono_time()
- CLEANUP: remove some unneeded includes from applet-t.h
- REORG: listener: move bind_conf_alloc() and listener_state_str() to listener.c
- CLEANUP: listeners: do not include openssl-compat
- CLEANUP: servers: do not include openssl-compat
- REORG: ssl: move ssl_sock_is_ssl() to connection.h and rename it
- CLEANUP: mux_fcgi: remove dependency on ssl_sock
- CLEANUP: ssl/server: move ssl_sock_set_srv() to srv_set_ssl() in server.c
- REORG: ssl-sock: move the sslconns/totalsslconns counters to global
- REORG: sample: move the crypto samples to ssl_sample.c
- REORG: sched: moved samp_time and idle_time to task.c as well
- REORG: time/ticks: move now_ms and global_now_ms definitions to ticks.h
- CLEANUP: tree-wide: remove unneeded include time.h in ~20 files
- REORG: activity: uninline activity_count_runtime()
- REORG: acitvity: uninline sched_activity_entry()
- CLEANUP: stream: remove many unneeded includes from stream-t.h
- CLEANUP: stick-table: no need to include socket nor in.h
- MINOR: connection: use uint64_t for the hashes
- REORG: connection: move the hash-related stuff to connection.c
- REORG: connection: uninline conn_notify_mux() and conn_delete_from_tree()
- REORG: server: uninline the idle conns management functions
- REORG: ebtree: split structures into their own file ebtree-t.h
- CLEANUP: tree-wide: only include ebtree-t from type files
- REORG: connection: move the largest inlines from connection.h to connection.c
- CLEANUP: connection: do not include http_ana!
- CLEANUP: connection: remove unneeded tcpcheck-t.h and use only session-t.h
- REORG: connection: uninline the rest of the alloc/free stuff
- REORG: task: uninline the loop time measurement code
- CLEANUP: time: move a few configurable defines to defaults.h
- CLEANUP: fd: do not include time.h
- REORG: fd: uninline compute_poll_timeout()
- CLENAUP: wdt: use ha_tkill() instead of accessing pthread directly
- REORG: thread: move the thread init/affinity/stop to thread.c
- REORG: thread: move ha_get_pthread_id() to thread.c
- MINOR: thread: use a dedicated static pthread_t array in thread.c
- CLEANUP: thread: uninline ha_tkill/ha_tkillall/ha_cpu_relax()
- DOC: configuration: add clarification on escaping in keyword arguments
- BUG/MINOR: task: fix missing include with DEBUG_TASK
- MINOR: pools: report the amount used by thread caches in "show pools"
- MINOR: quic: Distinguish packet and SSL read enc. level in traces
- MINOR: quic: Add a function to dump SSL stack errors
- MINOR: quic: BUG_ON() SSL errors.
- MINOR: quic: Fix SSL error issues (do not use ssl_bio_and_sess_init())
- BUG/MEDIUM: mux-quic: reinsert all streams in by_id tree
- BUG/MAJOR: xprt-quic: do not queue qc timer if not set
- MINOR: mux-quic: release connection if no more bidir streams
- BUG/MAJOR: quic: remove qc from receiver cids tree on free
- BUG/MEDIUM: mux_h2: Handle others remaining read0 cases on partial frames
- MINOR: qpack: do not encode invalid http status code
- MINOR: qpack: support non-indexed http status code encoding
- MINOR: qpack: fix memory leak on huffman decoding
- CLEANUP: mux-quic: remove unused code
- BUG/MINOR: quic: fix includes for compilation
- BUILD: connection: avoid a build warning on FreeBSD with SO_USER_COOKIE
- BUILD: init: avoid a build warning on FreeBSD with USE_PROCCTL
- REORG: time: move time-keeping code and variables to clock.c
- REORG: clock: move the updates of cpu/mono time to clock.c
- MINOR: activity: get the run_time from the clock updates
- CLEANUP: clock: stop exporting before_poll and after_poll
- REORG: clock: move the clock_id initialization to clock.c
- REORG: clock/wdt: move wdt timer initialization to clock.c
- MINOR: clock: move the clock_ids to clock.c
- MINOR: wdt: move wd_timer to wdt.c
- CLEANUP: wdt: do not remap SI_TKILL to SI_LWP, test the values directly
- REORG: thread/sched: move the task_per_thread stuff to thread_ctx
- REORG: thread/clock: move the clock parts of thread_info to thread_ctx
- REORG: thread/sched: move the thread_info flags to the thread_ctx
- REORG: thread/sched: move the last dynamic thread_info to thread_ctx
- MINOR: thread: make "ti" a const pointer and clean up thread_info a bit
- MINOR: threads: introduce a minimalistic notion of thread-group
- MINOR: global: add a new "thread-groups" directive
- MINOR: global: add a new "thread-group" directive
- MINOR: threads: make tg point to the current thread's group
- MEDIUM: threads: automatically assign threads to groups
- MINOR: threads: set the group ID and its bit in the thread group
- MINOR: threads: set the tid, ltid and their bit in thread_cfg
- MEDIUM: threads: replace ha_set_tid() with ha_set_thread()
- MINOR: threads: add the current group ID in thread-local "tgid" variable
- MINOR: debug: report the group and thread ID in the thread dumps
- MEDIUM: listeners: support the definition of thread groups on bind lines
- MINOR: threads: add a new function to resolve config groups and masks
- MEDIUM: config: resolve relative threads on bind lines to absolute ones
- MEDIUM: stick-table: never learn the "conn_cur" value from peers
There have been a large number of issues reported with conn_cur
synchronization because the concept is wrong. In an active-passive
setup, pushing the local connections count from the active node to
the passive one will result in the passive node to have a higher
counter than the real number of connections. Due to this, after a
switchover, it will never be able to close enough connections to
go down to zero. The same commonly happens on reloads since the new
process preloads its values from the old process, and if no connection
happens for a key after the value is learned, it is impossible to reset
the previous ones. In active-active setups it's a bit different, as the
number of connections reflects the number on the peer that pushed last.
This patch solves this by marking the "conn_cur" local and preventing
it from being learned from peers. It is still pushed, however, so that
any monitoring system that collects values from the peers will still
see it.
The patch is tiny and trivially backportable. While a change of behavior
in stable branches is never welcome, it remains possible to fix issues
if reports become frequent.
This extends the "thread" statement of bind lines to support an optional
thread group number. When unspecified (0) it's an absolute thread range,
and when specified it's one relative to the thread group. Masks are still
used so no more than 64 threads may be specified at once, and a single
group is possible. The directive is not used for now.
This registers a mapping of threads to groups by enumerating for each thread
what group it belongs to, and marking the group as assigned. It takes care of
checking for redefinitions, overlaps, and holes. It supports both individual
numbers and ranges. The thread group is referenced from the thread config.
Add a more precise description on how backslash escaping is different
than the top-level parser, and give examples of how to handle single
quotes inside arguments.
The ssl_bc_hsk_err sample fetch will need to raise more errors than only
handshake related ones hence its renaming to a more generic ssl_bc_err.
This patch is required because some handshake failures that should have
been caught by this fetch (verify error on the server side for instance)
were missed. This is caused by a change in TLS1.3 in which the
'Finished' state on the client is reached before its certificate is sent
(and verified) on the server side (see the "Protocol Overview" part of
RFC 8446).
This means that the SSL_do_handshake call is finished long before the
server can verify and potentially reject the client certificate.
The ssl_bc_hsk_err will then need to be expanded to catch other types of
errors.
This change is also applied to the frontend fetches (ssl_fc_hsk_err
becomes ssl_fc_err) and to their string counterparts.
In case of a connection error happening after the SSL handshake is
completed, the error code stored in the connection structure would not
always be set, hence having some connection failures being described as
successful in the fc_conn_err or bc_conn_err sample fetches.
The most common case in which it could happen is when the SSL server
rejects the client's certificate. The SSL_do_handshake call on the
client side would be sucessful because the client effectively sent its
client hello and certificate information to the server, but the next
call to SSL_read on the client side would raise an SSL_ERROR_SSL code
(through the SSL_get_error function) which is decribed in OpenSSL
documentation as a non-recoverable and fatal SSL error.
This patch ensures that in such a case, the connection's error code is
set to a special CO_ERR_SSL_FATAL value.
Checking in code the right keyword is "enabled" and not "enable".
In addition the comment was also completed:
This could appear useless because the "defaults" sections not
yet apply on "peers" sections, but it could be the case in the future.
This statement can currently cancel a previous "disabled" keyword in
the same section.
This patch should be backported in all supported branches (keyword
is present since 1.5)
Released version 2.5-dev8 with the following main changes :
- BUILD: compiler: fixed a missing test on defined(__GNUC__)
- BUILD: halog: fix a -Wundef warning on non-glibc systems
- BUILD: threads: fix -Wundef for _POSIX_PRIORITY_SCHEDULING on libmusl
- BUG/MINOR: compat: make sure __WORDSIZE is always defined
- BUILD: sample: fix format warning on 32-bit archs in sample_conv_be2dec_check()
- CLEANUP: pools: factor all malloc_trim() calls into trim_all_pools()
- MINOR: pools: automatically disable malloc_trim() with external allocators
- MINOR: pools: report it when malloc_trim() is enabled
- DOC: Add .mailmap
- CLEANUP: tree-wide: fix prototypes for functions taking no arguments.
- CLEANUP: Remove prototype for non-existent thread_get_default_count()
- CLEANUP: acl: Remove unused variable when releasing an acl expression
- BUG/MAJOR: mux-h1: Don't eval input data if an error was reported
- DOC: update Tim's address in .mailmap
- MINOR: pools: use mallinfo2() when available instead of mallinfo()
- BUG/MINOR: tcpcheck: Improve LDAP response parsing to fix LDAP check
- DOC: management: certificate files must be sanitized before injection
- BUG/MINOR: connection: prevent null deref on mux cleanup task allocation
- BUILD: ist: prevent gcc11 maybe-uninitialized warning on istalloc
- BUG/MINOR: cli/payload: do not search for args inside payload
- BUILD: sockpair: do not set unused flag
- BUILD: proto_uxst: do not set unused flag
- BUILD: fd: remove unused variable totlen in fd_write_frag_line()
- MINOR: applet: remove the thread mask from appctx_new()
- REORG: threads: move ha_get_pthread_id() to tinfo.h
- CLEANUP: Apply ist.cocci
- DEV: coccinelle: Add ist.cocci
- CLEANUP: Apply bug_on.cocci
- DEV: coccinelle: Add xalloc_size.cocci
- DEV: coccinelle: Add bug_on.cocci
- CLEANUP: Apply xalloc_size.cocci
- DEV: coccinelle: Add xalloc_cast.cocci
- BUG/MINOR: flt-trace: fix an infinite loop when random-parsing is set
- MINOR: httpclient: add the EOH when no headers where provided
- CLEANUP: Include check.h in flt_spoe.c
- CLEANUP: Remove unreachable `break` from parse_time_err()
- BUG/MINOR: server: allow 'enable health' only if check configured
- BUG/MINOR: server: alloc dynamic srv ssl ctx if proxy uses ssl chk rule
- MINOR: server: enable more keywords for ssl checks for dynamic servers
- MINOR: server: enable more check related keywords for dynamic servers
- REORG: server: move slowstart init outside of checks
- MINOR: server: enable slowstart for dynamic server
- MEDIUM: listener: deprecate "process" in favor of "thread" on bind lines
- BUG/MEDIUM: leastconn: fix rare possibility of divide by zero
- BUG/MINOR: quic: Possible NULL pointer dereferencing when dumping streams.
- MINOR: quic: Move transport parmaters to anynomous struct.
- MINOR: mux_quic: Add QUIC mux layer.
- MINOR: connection: Add callbacks definitions for QUIC.
- MINOR: quic: Attach QUIC mux connection objet to QUIC connection.
- MINOR: quic: Add a new definition to store STREAM frames.
- MINOR: h3: Add HTTP/3 definitions.
- MINOR: qpack: Add QPACK compression.
- MINOR: quic_sock: Finalize the QUIC connections.
- MINOR: quic: Disable the action of ->rcv_buf() xprt callback
- MINOR: quic: Add callbacks for (un)scribing to QUIC xprt.
- MINOR: quic: Variable-length integer encoding/decoding into/from buffer struct.
- BUG/MINOR: quic: Wrong ->accept() error handling
- MINOR: quic: Add a wrapper function to update transport parameters.
- MINOR: quic: Update the streams transport parameters.
- MINOR: quic: Avoid header collisions
- MINOR: quic: Replace max_packet_size by max_udp_payload size.
- MINOR: quic: Enable some quic, h3 and qpack modules compilation.
- MINOR: quic: Move an SSL func call from QUIC I/O handler to the xprt init.
- MINOR: quic: Initialize the session before starting the xprt.
- BUG/MINOR: quic: Do not check the acception of a new conn from I/O handler.
- MINOR: quic: QUIC conn initialization from I/O handler
- MINOR: quic: Remove header protection for conn with context
- MINOR: quic: Derive the initial secrets asap
- MINOR: quic: Remove header protection also for Initial packets
- BUG/MINOR: quic: Wrong memory free in quic_update_ack_ranges_list()
- MINOR: quic: quic_update_ack_ranges_list() code factorization
- MINOR: quic: Useless test in quic_update_ack_ranges_list()
- MINOR: quic: Remove a useless variable in quic_update_ack_ranges_list()
- BUG/MINOR: quic: Missing cases treatement when updating ACK ranges
- CLEAUNUP: quic: Usage of a useless variable in qc_treat_rx_pkts()
- BUG/MINOR: quic: Wrong RX packet reference counter usage
- MINOR: quic: Do not stop the packet parsing too early in qc_treat_rx_packets()
- MINOR: quic: Add a lock for RX packets
- MINOR: quic: Move the connection state
- MINOR: quic: Replace quic_conn_ctx struct by ssl_sock_ctx struct
- MINOR: quic: Replace the RX list of packet by a thread safety one.
- MINOR: quic: Replace the RX unprotected packet list by a thread safety one.
- MINOR: quic: Add useful traces for I/O dgram handler
- MINOR: quic: Do not wakeup the xprt task on ACK receipt
- MINOR: quic: Connection allocations rework
- MINOR: quic: Move conn_prepare() to ->accept_conn() callback
- MINOR: quic: Make qc_lstnr_pkt_rcv() be thread safe.
- MINOR: quic: Add a ring buffer implementation for QUIC
- MINOR: quic: Prefer x25519 as ECDH preferred parametes.
- MINOR: quic: Add the QUIC v1 initial salt.
- BUG/MINOR: quic: Too much reduced computed space to build handshake packets
- MINOR: net_helper: add functions for pointers
- MINOR: quic: Add ring buffer definition (struct qring) for QUIC
- MINOR: proto_quic: Allocate TX ring buffers for listeners
- MINOR: quic: Initialize pointers to TX ring buffer list
- MINOR: quic: Make use of TX ring buffers to send QUIC packets
- MINOR: quic_tls: Make use of the QUIC V1 salt.
- MINOR: quic: Remove old TX buffer implementation
- MINOR: Add function for TX packets reference counting
- MINOR: quic: Add TX packets at the very last time to their tree.
- MINOR: quic: Unitialized mux context upon Client Hello message receipt.
- MINOR: quic: Missing encryption level rx.crypto member initialization and lock.
- MINOR: quic: Rename ->rx.rwlock of quic_enc_level struct to ->rx.pkts_rwlock
- MINOR: quic: Make qc_treat_rx_pkts() be thread safe.
- MINOR: quic: Make ->tx.frms quic_pktns struct member be thread safe
- MINOR: quic: Replace quic_tx_frm struct by quic_frame struct
- MINOR: quic: Add a mask for TX frame builders and their authorized packet types
- MINOR: quic: Add a useful function to compute any frame length.
- MINOR: quic: Add the QUIC connection state to traces
- MINOR: quic: Store post handshake frame in ->pktns.tx.frms MT_LIST
- MINOR: quic: Add the packet type to quic_tx_packet struct
- MINOR: quic: Modify qc_do_build_hdshk_pkt() to accept any packet type
- MINOR: quic: Atomically handle packet number space ->largest_acked_pn variable
- MINOR: quic: Modify qc_build_cfrms() to support any frame
- MINOR: quic: quic_conn_io_cb() task rework
- MINOR: quic: Make qc_build_hdshk_pkt() atomically consume a packet number
- MINOR: quic: qc_do_build_hdshk_pkt() does not need to pass a copy of CRYPTO frame
- MINOR: quic: Remove Application level related functions
- MINOR: quic: Rename functions which do not build only Handshake packets
- MINOR: quic: Make circular buffer internal buffers be variable-sized.
- MINOR: quic: Add a pool for TX ring buffer internal buffer
- MINOR: quic: Make use of the last cbuf API when initializing TX ring buffers
- MINOR: quic: Missing acks encoded size updates.
- MINOR: quic: Evaluate the packet lengths in advance
- MINOR: quic: Update the TLS extension for QUIC transport parameters
- MINOR: quic: Fix handshake state debug strings
- MINOR: quic: Atomically get/set the connection state
- MINOR: quic: Missing QUIC encryption level for qc_build_pkt()
- MINOR: quic: Coalesce Application level packets with Handshake packets.
- MINOR: quic: Wrong flags handling for acks
- MINOR: quic: Missing case when discarding HANDSHAKE secrets
- MINOR: quic: Post handshake packet building improvements
- MINOR: quic: Prepare Application level packet asap.
- MINOR: h3: Send h3 settings asap
- MINOR: quic: Wrong STREAM frame length computing
- MINOR: quic: Wrong short packet minimum length
- MINOR: quic: Prepare STREAM frames to fill QUIC packets
- MINOR: h3: change default settings
- MINOR: quic-enc: fix varint encoding
- MINOR: qpack: fix wrong comment
- MINOR: qpack: generate headers list on decoder
- MINOR: h3: parse headers to htx
- MINOR: h3: allocate stream on headers
- MEDIUM: mux-quic: implement ring buffer on stream tx
- MINOR: mux-quic: send SETTINGS on uni stream
- MINOR: h3: define snd_buf callback and divert mux ops
- MINOR: mux-quic: define FIN stream flag
- MINOR: qpack: create qpack-enc module
- MINOR: qpack: encode headers functions
- MINOR: h3: encode htx headers to QPACK
- MINOR: h3: send htx data
- MINOR: h3/mux: detect fin on last h3 frame of the stream
- MINOR: quic: Shorten some handshakes
- MINOR: quic: Make QUIC-TLS support at least two initial salts
- MINOR: quic: Attach the QUIC connection to a thread.
- MINOR: quic: Missing active_connection_id_limit default value
- MINOR: quic_sock: Do not flag QUIC connections as being set
- MINOR: buf: Add b_force_xfer() function
- MINOR: quic: Make use of buffer structs to handle STREAM frames
- MINOR: mux_quic: move qc_process() code to qc_send()
- MINOR: quic: Add a typedef for unsigned long long
- MINOR: quic: Confusion between TX/RX for the frame builders
- MINOR: quic: Wrong packet flags settings during frame building
- MINOR: quic: Constantness fixes for frame builders/parsers.
- MINOR: quic_tls: Client/serveur state reordering
- MINOR: quic: Wrong packet loss detection due to wrong pktns order
- MINOR: quic: Wrong packet number space selection in quic_loss_pktns()
- MINOR: quic: Initial packet number spaced not discarded
- MINOR: quic: Add useful trace about pktns discarding
- MINOR: mux_quic: Export the mux related flags
- MINOR: quic: Implement quic_conn_subscribe()
- MINOR: quic: Wake up the mux upon ACK receipt
- MINOR: quic: Stream FIN bit fix in qcs_push_frame()
- MINOR: quic: Implement qc_process_mux()
- MINOR: quic: Wake up the xprt from mux
- CLEANUP: quic: Remove useless inline functions
- MINOR: quic: RX packets memory leak
- MINOR: quic: Possible endless loop in qc_treat_rx_pkts()
- MINOR: quic: Crash upon too big packets receipt
- MINOR: quic: define close handler
- MEDIUM: quic: implement mux release/conn free
- MINOR: quic: fix qcc subs initialization
- BUG/MINOR: h1-htx: Fix a typo when request parser is reset
- BUG/MEDIUM: mux-h1: Adjust conditions to ask more space in the channel buffer
- BUG/MEDIUM: stream-int: Notify stream that the mux wants more room to xfer data
- BUG/MEDIUM: stream: Stop waiting for more data if SI is blocked on RXBLK_ROOM
- MINOR: stream-int: Set CO_RFL transient/persistent flags apart in si_cs_rcv()
- MINOR: htx: Add an HTX flag to know when a message is fragmented
- MINOR: htx: Add a function to know if the free space wraps
- BUG/MEDIUM: stream-int: Defrag HTX message in si_cs_recv() if necessary
- MINOR: stream-int: Notify mux when the buffer is not stuck when calling rcv_buf
- BUG/MINOR: http-ana: increment internal_errors counter on response error
- MINOR: stats: Enable dark mode on stat web page
- CLEANUP: stats: Fix some alignment mistakes
- MINOR: httpclient: httpclient_data() returns the available data
- MINOR: httpclient: httpclient_ended() returns 1 if the client ended
- MINOR: httpclient/lua: httpclient:get() API in lua
- MINOR: httpclient/lua: implement the headers in the response object
- BUG/MINOR: httpclient/lua: return an error on argument check
- CLEANUP: slz: Mark `reset_refs` as static
The "process" directive on "bind" lines becomes quite confusing considering
that the only allowed value is 1 for the process, and that threads are
optional and come after the mandatory "1/".
Let's introduce a new "thread" directive to directly configure thread
numbers, and mark "process" as deprecated. Now "process" will emit a
warning and will suggest how to be replaced with "thread" instead.
The doc was updated accordingly (mostly a copy-paste of the previous
description which was already up to date).
This is marked as MEDIUM as it will impact users having "zero-warning"
and "process" specified.
Enable the 'slowstart' keyword for dynamic servers. The slowstart task
is allocated in 'add server' handler if slowstart is used.
As the server is created in disabled state, there is no need to start
the task. The slowstart task will be automatically started on the first
'enable server' invocation.
Allow to use the check related keywords defined in server.c. These
keywords can be enabled now that checks have been implemented for
dynamic servers.
Here is the list of the new keywords supported :
- error-limit
- observe
- on-error
- on-marked-down
- on-marked-up
Allow to configure ssl support for dynamic server checks independently
of the ssl server configuration. This is done via the keyword
"check-ssl". Also enable to configure the sni/alpn used for the check
via "check-sni/alpn".
A lot of people encounter problems when trying to inject a certificate
file which contains extra informations or empty lines.
This patch adds a paragraph and a sanitizing example.
Must be backported as far as 2.1.
Released version 2.5-dev7 with the following main changes :
- BUG/MINOR: config: reject configs using HTTP with bufsize >= 256 MB
- CLEANUP: htx: remove comments about "must be < 256 MB"
- BUG/MAJOR: htx: fix missing header name length check in htx_add_header/trailer
- Revert "BUG/MINOR: stream-int: Don't block reads in si_update_rx() if chn may receive"
- MINOR: proxy: add a global "grace" directive to postpone soft-stop
- MINOR: vars: rename vars_init() to vars_init_head()
- CLEANUP: vars: rename sample_clear_stream() to var_unset()
- REORG: vars: remerge sample_store{,_stream}() into var_set()
- MEDIUM: vars: make the ifexist variant of set-var only apply to the proc scope
- MINOR: vars: add a VF_CREATEONLY flag for creation
- MINOR: vars: support storing empty sample data with a variable
- MINOR: vars: store flags into variables and add VF_PERMANENT
- MEDIUM: vars: make var_clear() only reset VF_PERMANENT variables
- MEDIUM: vars: pre-create parsed SCOPE_PROC variables as permanent ones
- MINOR: vars: preset a random seed to hash variables names
- MEDIUM: vars: replace the global name index with a hash
- CLEANUP: vars: remove the now unused var_names array
- MINOR: vars: centralize the lock/unlock into static inlines
- OPTIM: vars: only takes the variables lock on shared entries
- OPTIM: vars: remove internal bookkeeping for vars_global_size
- OPTIM: vars: do not keep variables usage stats if no limit is set
- BUILD: fix dragonfly build again on __read_mostly
- CI: Github Actions: temporarily disable Opentracing
- BUG/MEDIUM: mux-h1: Remove "Upgrade:" header for requests with payload
- MINOR: htx: Skip headers with no value when adding a header list to a message
- CLEANUP: mux-h1: Remove condition rejecting upgrade requests with payload
- BUG/MEDIUM: stream-int: Don't block SI on a channel policy if EOI is reached
- BUG/MEDIUM: http-ana: Reset channels analysers when returning an error
- BUG/MINOR: filters: Set right FLT_END analyser depending on channel
- CLEANUP: Add haproxy/xxhash.h to avoid modifying import/xxhash.h
- CLEANUP: ebmbtree: Replace always-taken elseif by else
- CLEANUP: Move XXH3 macro from haproxy/compat.h to haproxy/xxhash.h
- BUILD: opentracing: exclude the use of haproxy variables for the OpenTracing context
- BUG/MINOR: opentracing: enable the use of http headers without a set value
- CLEANUP: opentracing: use the haproxy function to generate uuid
- MINOR: opentracing: change the scope of the variable 'ot.uuid' from 'sess' to 'txn'
- CI: Github Actions: re-enable Opentracing
- CLEANUP: opentracing: simplify the condition on the empty header
- BUG/MEDIUM lua: Add missing call to RESET_SAFE_LJMP in hlua_filter_new()
When setting variables, there are currently two variants, one which will
always create the variable, and another one, "ifexist", which will only
create or update a variable if a similarly named variable in any scope
already existed before.
The goal was to limit the risk of injecting random names in the proc
scope, but it was achieved by making use of the somewhat limited name
indexing model, which explains the scope-agnostic restriction.
With this change, we're moving the check downwards in the chain, at the
variable level, and only variables under the scope "proc" will be subject
to the restriction. A new set of VF_* flags was added to adjust how
variables are set, and VF_UPDATEONLY is used to mention this restriction.
In this exact state of affairs, this is not completely exact, as if a
similar name was not known in any scope, the variable will continue to
be rejected like before, but this will change soon.
In ticket #1348 some users expressed some concerns regarding the removal
of the "grace" directive from the proxies. Their use case very closely
mimmicks the original intent of the grace keyword, which is, let haproxy
accept traffic for some time when stopping, while indicating an external
LB that it's stopping.
This is implemented here by starting a task whose expiration triggers
the soft-stop for real. The global "stopping" variable is immediately
set however. For example, this below will be sufficient to instantly
notify an external check on port 9999 that the service is going down,
while other services remain active for 10s:
global
grace 10s
frontend ext-check
bind :9999
monitor-uri /ext-check
monitor fail if { stopping }
Released version 2.5-dev6 with the following main changes :
- BUG/MINOR threads: Use get_(local|gm)time instead of (local|gm)time
- BUG/MINOR: tools: Fix loop condition in dump_text()
- BUILD: ssl: next round of build warnings on LIBRESSL_VERSION_NUMBER
- BUILD: ssl: fix two remaining occurrences of #if USE_OPENSSL
- BUILD: tools: properly guard __GLIBC__ with defined()
- BUILD: globally enable -Wundef
- MINOR: log: Remove log-error-via-logformat option
- MINOR: log: Add new "error-log-format" option
- BUG/MAJOR: queue: better protect a pendconn being picked from the proxy
- CLEANUP: Add missing include guard to signal.h
- MINOR: ssl: Add new ssl_bc_hsk_err sample fetch
- MINOR: connection: Add a connection error code sample fetch for backend side
- REGTESTS: ssl: Add tests for bc_conn_err and ssl_bc_hsk_err sample fetches
- MINOR: http-rules: add a new "ignore-empty" option to redirects.
- CI: Github Actions: temporarily disable BoringSSL builds
- BUG/MINOR: vars: fix set-var/unset-var exclusivity in the keyword parser
- BUG/MINOR: vars: improve accuracy of the rules used to check expression validity
- MINOR: sample: add missing ARGC_ entries
- BUG/MINOR: vars: properly set the argument parsing context in the expression
- DOC: configuration: remove wrong tcp-request examples in tcp-response
- MEDIUM: vars: add a new "set-var-fmt" action
- BUG/MEDIUM: vars: run over the correct list in release_store_rules()
- BUG/MINOR: vars: truncate the variable name in error reports about scope.
- BUG/MINOR: vars: do not talk about global section in CLI errors for set-var
- CLEANUP: vars: name the temporary proxy "CFG" instead of "CLI" for global vars
- MINOR: log: make log-format expressions completely usable outside of req/resp
- MINOR: vars: add a "set-var-fmt" directive to the global section
- MEDIUM: vars: also support format strings in CLI's "set var" command
- CLEANUP: vars: factor out common code from vars_get_by_{desc,name}
- MINOR: vars: make vars_get_by_* support an optional default value
- MINOR: vars: make the vars() sample fetch function support a default value
- BUILD: ot: add argument for default value to vars_get_by_name()
It is quite common to see in configurations constructions like the
following one:
http-request set-var(txn.bodylen) 0
http-request set-var(txn.bodylen) req.hdr(content-length)
...
http-request set-header orig-len %[var(txn.bodylen)]
The set-var() rules are almost always duplicated when manipulating
integers or any other value that is mandatory along operations. This is
a problem because it makes the configurations complicated to maintain
and slower than needed. And it becomes even more complicated when several
conditions may set the same variable because the risk of forgetting to
initialize it or to accidentally reset it is high.
This patch extends the var() sample fetch function to take an optional
argument which contains a default value to be returned if the variable
was not set. This way it becomes much simpler to use the variable, just
set it where needed, and read it with a fall back to the default value:
http-request set-var(txn.bodylen) req.hdr(content-length)
...
http-request set-header orig-len %[var(txn.bodylen,0)]
The default value is always passed as a string, thus it will experience
a cast to the output type. It doesn't seem userful to complicate the
configuration to pass an explicit type at this point.
The vars.vtc regtest was updated accordingly.
Most often "set var" on the CLI is used to set a string, and using only
expressions is not always convenient, particularly when trying to
concatenate variables sur as host names and paths.
Now the "set var" command supports an optional keyword before the value
to indicate its type. "expr" takes an expression just like before this
patch, and "fmt" a format string, making it work like the "set-var-fmt"
actions.
The VTC was updated to include a test on the format string.
Just like the set-var-fmt action for tcp/http rules, the set-var-fmt
directive in global sections allows to pre-set process-wide variables
using a format string instead of a sample expression. This is often
more convenient when it is required to concatenate multiple fields,
or when emitting just one word.
The set-var() action is convenient because it preserves the input type
but it's a pain to deal with when trying to concatenate values. The
most recurring example is when it's needed to build a variable composed
of the source address and the source port. Usually it ends up like this:
tcp-request session set-var(sess.port) src_port
tcp-request session set-var(sess.addr) src,concat(":",sess.port)
This is even worse when trying to aggregate multiple fields from stick-table
data for example. Due to this a lot of users instead abuse headers from HTTP
rules:
http-request set-header(x-addr) %[src]:%[src_port]
But this requires some careful cleanups to make sure they won't leak, and
it's significantly more expensive to deal with. And generally speaking it's
not clean. Plus it must be performed for each and every request, which is
expensive for this common case of ip+port that doesn't change for the whole
session.
This patch addresses this limitation by implementing a new "set-var-fmt"
action which performs the same work as "set-var" but takes a format string
in argument instead of an expression. This way it becomes pretty simple to
just write:
tcp-request session set-var-fmt(sess.addr) %[src]:%[src_port]
It is usable in all rulesets that already support the "set-var" action.
It is not yet implemented for the global "set-var" directive (which already
takes a string) and the CLI's "set var" command, which would definitely
benefit from it but currently uses its own parser and engine, thus it
must be reworked.
The doc and regtests were updated.
There is a massive abuse of copy-paste in the doc that is visible in
the examples and arguments declaration. Let's at least remove irrelevant
examples for now.
Sometimes it is convenient to remap large sets of URIs to new ones (e.g.
after a site migration for example). This can be achieved using
"http-request redirect" combined with maps, but one difficulty there is
that non-matching entries will return an empty response. In order to
avoid this, duplicating the operation as an ACL condition ending in
"-m found" is possible but it becomes complex and error-prone while it's
known that an empty URL is not valid in a location header.
This patch addresses this by improving the redirect rules to be able to
simply ignore the rule and skip to the next one if the result of the
evaluation of the "location" expression is empty. However in order not
to break existing setups, it requires a new "ignore-empty" keyword.
There used to be an ACT_FLAG_FINAL on redirect rules that's used during
the parsing to emit a warning if followed by another rule, so here we
only set it if the option is not there. The http_apply_redirect_rule()
function now returns a 3rd value to mention that it did nothing and
that this was not an error, so that callers can just ignore the rule.
The regular "redirect" rules were not modified however since this does
not apply there.
The map_redirect VTC was completed with such a test and updated to 2.5
and an example was added into the documentation.
The bc_conn_err and bc_conn_err_str sample fetches give the status of
the connection on the backend side. The error codes and error messages
are the same than the ones that can be raised by the fc_conn_err fetch.
This new sample fetch along the ssl_bc_hsk_err_str fetch contain the
last SSL error of the error stack that occurred during the SSL
handshake (from the backend's perspective).
This option can be used to define a specific log format that will be
used in case of error, timeout, connection failure on a frontend... It
will be used for any log line concerned by the log-separate-errors
option. It will also replace the format of specific error messages
decribed in section 8.2.6.
If no "error-log-format" is defined, the legacy error messages are still
emitted and the other error logs keep using the regular log-format.
This option will be replaced by a "error-log-format" that enables to use
a dedicated log-format for connection error messages instead of the
regular log-format (in which most of the fields would be invalid in such
a case).
The "log-error-via-logformat" mechanism will then be replaced by a test
on the presence of such an error log format or not. If a format is
defined, it is used for connection error messages, otherwise the legacy
error log format is used.
Released version 2.5-dev5 with the following main changes :
- MINOR: httpclient: initialize the proxy
- MINOR: httpclient: implement a simple HTTP Client API
- MINOR: httpclient/cli: implement a simple client over the CLI
- MINOR: httpclient/cli: change the User-Agent to "HAProxy"
- MEDIUM: ssl: Keep a reference to the client's certificate for use in logs
- BUG/MEDIUM: h2: match absolute-path not path-absolute for :path
- BUILD/MINOR: ssl: Fix compilation with OpenSSL 1.0.2
- MINOR: server: check if srv is NULL in free_server()
- MINOR: proxy: check if p is NULL in free_proxy()
- BUG/MEDIUM: cfgparse: do not allocate IDs to automatic internal proxies
- BUG/MINOR: http_client: make sure to preset the proxy's default settings
- REGTESTS: http_upgrade: fix incorrect expectation on TCP->H1->H2
- REGTESTS: abortonclose: after retries, 503 is expected, not close
- REGTESTS: server: fix agent-check syntax and expectation
- BUG/MINOR: httpclient: fix uninitialized sl variable
- BUG/MINOR: httpclient/cli: change the appctx test in the callbacks
- BUG/MINOR: httpclient: check if hdr_num is not 0
- MINOR: httpclient: cleanup the include files
- MINOR: hlua: take the global Lua lock inside a global function
- MINOR: tools: add FreeBSD support to get_exec_path()
- BUG/MINOR: systemd: ExecStartPre must use -Ws
- MINOR: systemd: remove the ExecStartPre line in the unit file
- MINOR: ssl: add an openssl version string parser
- MINOR: cfgcond: implements openssl_version_atleast and openssl_version_before
- CLEANUP: ssl: remove useless check on p in openssl_version_parser()
- BUG/MINOR: stick-table: fix the sc-set-gpt* parser when using expressions
- BUG/MINOR: httpclient: remove deinit of the httpclient
- BUG/MEDIUM: base64: check output boundaries within base64{dec,urldec}
- MINOR: httpclient: set verify none on the https server
- MINOR: httpclient: add the server to the proxy
- BUG/MINOR: httpclient: fix Host header
- BUILD: httpclient: fix build without OpenSSL
- CI: github-actions: remove obsolete options
- CLEANUP: assorted typo fixes in the code and comments
- MINOR: proc: setting the process to produce a core dump on FreeBSD.
- BUILD: adopt script/build-ssl.sh for OpenSSL-3.0.0beta2
- MINOR: server: return the next srv instance on free_server
- BUG/MINOR: stats: use refcount to protect dynamic server on dump
- MEDIUM: server: extend refcount for all servers
- MINOR: server: define non purgeable server flag
- MINOR: server: mark referenced servers as non purgeable
- MINOR: server: mark servers referenced by LUA script as non purgeable
- MEDIUM: server: allow to remove servers at runtime except non purgeable
- BUG/MINOR: base64: base64urldec() ignores padding in output size check
- REGTEST: add missing lua requirements on server removal test
- REGTEST: fix haproxy required version for server removal test
- BUG/MINOR: proxy: don't dump servers of internal proxies
- REGTESTS: Use `feature cmd` for 2.5+ tests
- REGTESTS: Remove REQUIRE_VERSION=1.5 from all tests
- BUG/MINOR: resolvers: mark servers with name-resolution as non purgeable
- MINOR: compiler: implement an ONLY_ONCE() macro
- BUG/MINOR: lua: use strlcpy2() not strncpy() to copy sample keywords
- MEDIUM: ssl: Capture more info from Client Hello
- MINOR: sample: Expose SSL captures using new fetchers
- MINOR: sample: Add be2dec converter
- MINOR: sample: Add be2hex converter
- MEDIUM: config: Deprecate tune.ssl.capture-cipherlist-size
- BUG/MINOR: time: fix idle time computation for long sleeps
- MINOR: time: add report_idle() to report process-wide idle time
- BUG/MINOR: ebtree: remove dependency on incorrect macro for bits per long
- BUILD: activity: use #ifdef not #if on USE_MEMORY_PROFILING
- BUILD/MINOR: defaults: eliminate warning on MAXHOSTNAMELEN with -Wundef
- BUILD/MINOR: ssl: avoid a build warning on LIBRESSL_VERSION with -Wundef
- IMPORT: slz: silence a build warning with -Wundef
- BUILD/MINOR: regex: avoid a build warning on USE_PCRE2 with -Wundef
To be able to provide JA3 compatible TLS Fingerprints we need to expose
all Client Hello captured data using fetchers. Patch provides new
and modifies existing fetchers to add ability to filter out GREASE values:
- ssl_fc_cipherlist_*
- ssl_fc_ecformats_bin
- ssl_fc_eclist_bin
- ssl_fc_extlist_bin
- ssl_fc_protocol_hello_id
When we set tune.ssl.capture-cipherlist-size to a non-zero value
we are able to capture cipherlist supported by the client. To be able to
provide JA3 compatible TLS fingerprinting we need to capture more
information from Client Hello message:
- SSL Version
- SSL Extensions
- Elliptic Curves
- Elliptic Curve Point Formats
This patch allows HAProxy to capture such information and store it for
later use.
Relax the condition on "delete server" CLI handler to be able to remove
all servers, even non dynamic, except if they are flagged as non
purgeable.
This change is necessary to extend the use cases for dynamic servers
with reload. It's expected that each dynamic server created via the CLI
is manually commited in the haproxy configuration by the user. Dynamic
servers will be present on reload only if they are present in the
configuration file. This means that non-dynamic servers must be allowed
to be removable at runtime.
The dynamic servers removal reg-test has been updated and renamed to
reflect its purpose. A new test is present to check that non-purgeable
servers cannot be removed.
Released version 2.5-dev4 with the following main changes :
- MINOR: log: rename 'dontloglegacyconnerr' to 'log-error-via-logformat'
- MINOR: doc: rename conn_status in `option httsplog`
- MINOR: proxy: disabled takes a stopping and a disabled state
- MINOR: stats: shows proxy in a stopped state
- BUG/MINOR: server: fix race on error path of 'add server' CLI if track
- CLEANUP: thread: fix fantaisist indentation of thread_harmless_till_end()
- MINOR: threads: make thread_release() not wait for other ones to complete
- MEDIUM: threads: add a stronger thread_isolate_full() call
- MEDIUM: servers: make the server deletion code run under full thread isolation
- BUG/MINOR: server: remove srv from px list on CLI 'add server' error
- MINOR: activity/fd: remove the dead_fd counter
- MAJOR: fd: get rid of the DWCAS when setting the running_mask
- CLEANUP: fd: remove the now unused fd_set_running()
- CLEANUP: fd: remove the now unneeded fd_mig_lock
- BUG/MINOR: server: update last_change on maint->ready transitions too
- MINOR: spoe: Add a pointer on the filter config in the spoe_agent structure
- BUG/MEDIUM: spoe: Create a SPOE applet if necessary when the last one is released
- BUG/MEDIUM: spoe: Fix policy to close applets when SPOE connections are queued
- MINOR: server: unmark deprecated on enable health/agent cli
- MEDIUM: task: implement tasklet kill
- MINOR: server: initialize fields for dynamic server check
- MINOR: check: allocate default check ruleset for every backends
- MINOR: check: export check init functions
- MINOR: check: do not increment global maxsock at runtime
- MINOR: server: implement a refcount for dynamic servers
- MEDIUM: check: implement check deletion for dynamic servers
- MINOR: check: enable safe keywords for dynamic servers
- MEDIUM: server: implement check for dynamic servers
- MEDIUM: server: implement agent check for dynamic servers
- REGTESTS: server: add dynamic check server test
- MINOR: doc: specify ulimit-n usage for dynamic servers
- REGTESTS: server: fix dynamic server with checks test
- CI: travis-ci: temporarily disable arm64 builds
- BUG/MINOR: check: test if server is not null in purge
- MINOR: global: define MODE_STOPPING
- BUG/MINOR: server: do not use refcount in free_server in stopping mode
- ADMIN: dyncookie: implement a simple dynamic cookie calculator
- BUG/MINOR: check: do not reset check flags on purge
- BUG/MINOR: check: fix leak on add dynamic server with agent-check error
- BUG/MEDIUM: check: fix leak on agent-check purge
- BUG/MEDIUM: server: support both check/agent-check on a dynamic instance
- BUG/MINOR: buffer: fix buffer_dump() formatting
- MINOR: channel: remove an htx block from a channel
- BUG/MINOR: tcpcheck: Properly detect pending HTTP data in output buffer
- BUG/MINOR: stream: Don't release a stream if FLT_END is still registered
- MINOR: lua: Add a flag on lua context to know the yield capability at run time
- BUG/MINOR: lua: Yield in channel functions only if lua context can yield
- BUG/MINOR: lua: Don't yield in channel.append() and channel.set()
- MINOR: filters/lua: Release filters before the lua context
- MINOR: lua: Add a function to get a reference on a table in the stack
- MEDIUM: lua: Process buffer data using an offset and a length
- MEDIUM: lua: Improve/revisit the lua api to manipulate channels
- DOC: Improve the lua documentation
- MEDIUM: filters/lua: Add support for dummy filters written in lua
- MINOR: lua: Add a function to get a filter attached to a channel class
- MINOR: lua: Add flags on the lua TXN to know the execution context
- MEDIUM: filters/lua: Be prepared to filter TCP payloads
- MEDIUM: filters/lua: Support declaration of some filter callback functions in lua
- MEDIUM: filters/lua: Add HTTPMessage class to help HTTP filtering
- MINOR: filters/lua: Add request and response HTTP messages in the lua TXN
- MINOR: filters/lua: Support the HTTP filtering from filters written in lua
- DOC: config: Fix 'http-response send-spoe-group' documentation
- BUG/MINOR: lua: Properly check negative offset in Channel/HttpMessage functions
- BUG/MINOR: lua: Properly catch alloc errors when parsing lua filter directives
- BUG/MEDIUM: cfgcheck: verify existing log-forward listeners during config check
- MINOR: cli: delare the CLI frontend as an internal proxy
- MINOR: proxy: disable warnings for internal proxies
- BUG/MINOR: filters: Always set FLT_END analyser when CF_FLT_ANALYZE flag is set
- BUG/MINOR: lua/filters: Return right code when txn:done() is called
- DOC: lua-api: Add documentation about lua filters
- CI: Remove obsolete USE_SLZ=1 CI job
- CLEANUP: assorted typo fixes in the code and comments
- CI: github actions: relax OpenSSL-3.0.0 version comparision
- BUILD: tools: get the absolute path of the current binary on NetBSD.
- DOC: Minor typo fix - 'question mark' -> 'exclamation mark'
- DOC/MINOR: fix typo in management document
- MINOR: http: add a new function http_validate_scheme() to validate a scheme
- BUG/MAJOR: h2: verify early that non-http/https schemes match the valid syntax
- BUG/MAJOR: h2: verify that :path starts with a '/' before concatenating it
- BUG/MAJOR: h2: enforce stricter syntax checks on the :method pseudo-header
- BUG/MEDIUM: h2: give :authority precedence over Host
- REGTESTS: add a test to prevent h2 desync attacks
Lua filter support is highly experimental. The documentation was added to
allow first lua filter implementations. The API is not stabilized and must
be improved to be fully usable. This docuementation is quite light for
now. But more will be added.
First of all, following functions are now considered deprecated:
* Channel:dup()
* Channel:get()
* Channel:getline()
* Channel:get_in_len()
* Cahnnel:get_out_len()
It is just informative, there is no warning and functions may still be
used. Howver it is recommended to use new functions. New functions are more
flexible and use a better naming pattern. In addition, the same names will
be used in the http_msg class to manipulate http messages from lua filters.
The new API is:
* Channel:data()
* Channel:line()
* Channel:append()
* Channel:prepend()
* Channel:insert()
* Channel:remove()
* Channel:set()
* Channel:input()
* Channel:output()
* Channel:send()
* Channel:forward()
* Channel:is_resp()
* Channel:is_full()
* Channel:may_recv()
The lua documentation was updated accordingly.
This commit is the counterpart for agent check of
"MEDIUM: server: implement check for dynamic servers".
The "agent-check" keyword is enabled for dynamic servers. The agent
check must manually be activated via "enable agent" CLI. This can
enable the dynamic server if the agent response is "ready" without an
explicit "enable server" CLI.
Implement check support for dynamic servers. The "check" keyword is now
enabled for dynamic servers. If used, the server check is initialized
and the check task started in the "add server" CLI handler. The check is
explicitely disabled and must be manually activated via "enable health"
CLI handler.
The dynamic server refcount is incremented if a check is configured. On
"delete server" handler, the check is purged, which decrements the
refcount.
Right now we're using a DWCAS to atomically set the running_mask while
being constrained by the thread_mask. This DWCAS is annoying because we
may seriously need it later when adding support for thread groups, for
checking that the running_mask applies to the correct group.
It turns out that the DWCAS is not strictly necessary because we never
need it to set the thread_mask based on the running_mask, only the other
way around. And in fact, the running_mask is always cleared alone, and
the thread_mask is changed alone as well. The running_mask is only
relevant to indicate a takeover when the thread_mask matches it. Any
bit set in running and not present in thread_mask indicates a transition
in progress.
As such, it is possible to re-arrange this by using a regular CAS around a
consistency check between running_mask and thread_mask in fd_update_events
and by making a CAS on running_mask then an atomic store on the thread_mask
in fd_takeover(). The only other case is fd_delete() but that one already
sets the running_mask before clearing the thread_mask, which is compatible
with the consistency check above.
This change has happily survived 10 billion takeovers on a 16-thread
machine at 800k requests/s.
The fd-migration doc was updated to reflect this change.
