Currently the H2 mux doesn't have a list of all the streams blocking on
the H2 side. It only knows about those trying to send or waiting for a
connection window update. It is problematic to enforce timeouts because
we never know if a stream has to live as long as the data layer wants or
has to be timed out becase it's waiting for a stream window update. This
patch adds a new list, "blocked_list", to store streams blocking on
stream flow control, or later, dependencies. Streams blocked on sfctl
are now added there. It doesn't modify the rest of the logic.
Released version 2.1-dev2 with the following main changes :
- DOC: management: document reuse and connect counters in the CSV format
- DOC: management: document cache_hits and cache_lookups in the CSV format
- BUG/MINOR: dns: remove irrelevant dependency on a client connection
- MINOR: applet: make appctx use their own pool
- BUG/MEDIUM: checks: Don't attempt to receive data if we already subscribed.
- BUG/MEDIUM: http/htx: unbreak option http_proxy
- BUG/MINOR: backend: do not try to install a mux when the connection failed
- MINOR: mux-h2: Don't adjust anymore the amount of data sent in h2_snd_buf()
- BUG/MINOR: http_fetch: Fix http_auth/http_auth_group when called from TCP rules
- BUG/MINOR: http_htx: Initialize HTX error messages for TCP proxies
- BUG/MINOR: cache/htx: Make maxage calculation HTX aware
- BUG/MINOR: hlua: Make the function txn:done() HTX aware
- MINOR: proto_htx: Directly call htx_check_response_for_cacheability()
- MINOR: proto_htx: Rely on the HTX function to apply a redirect rules
- MINOR: proto_htx: Add the function htx_return_srv_error()
- MINOR: backend/htx: Don't rewind output data to set the sni on a srv connection
- MINOR: proto_htx: Don't stop forwarding when there is a post-connect processing
- DOC: htx: Update comments in HTX files
- CLEANUP: htx: Remove the unsued function htx_add_blk_type_size()
- MINOR: htx: Deduce the number of used blocks from tail and head values
- MINOR: htx: Use an array of char to store HTX blocks
- MINOR: htx: Slightly update htx_dump() to report better messages
- DOC: htx: Add internal documentation about the HTX
- MAJOR: http: Deprecate and ignore the option "http-use-htx"
- MEDIUM: mux-h2: Remove support of the legacy HTTP mode
- CLEANUP: h2: Remove functions converting h2 requests to raw HTTP/1.1 ones
- MINOR: connection: Remove the multiplexer protocol PROTO_MODE_HTX
- MINOR: stream: Rely on HTX analyzers instead of legacy HTTP ones
- MEDIUM: http_fetch: Remove code relying on HTTP legacy mode
- MINOR: config: Remove tests on the option 'http-use-htx'
- MINOR: stream: Remove tests on the option 'http-use-htx' in stream_new()
- MINOR: proxy: Remove tests on the option 'http-use-htx' during H1 upgrade
- MINOR: hlua: Remove tests on the option 'http-use-htx' to reject TCP applets
- MINOR: cache: Remove tests on the option 'http-use-htx'
- MINOR: contrib/prometheus-exporter: Remove tests on the option 'http-use-htx'
- CLEANUP: proxy: Remove the flag PR_O2_USE_HTX
- MINOR: proxy: Don't adjust connection mode of HTTP proxies anymore
- MEDIUM: backend: Remove code relying on the HTTP legacy mode
- MEDIUM: hlua: Remove code relying on the legacy HTTP mode
- MINOR: http_act: Remove code relying on the legacy HTTP mode
- MEDIUM: cache: Remove code relying on the legacy HTTP mode
- MEDIUM: compression: Remove code relying on the legacy HTTP mode
- MINOR: flt_trace: Remove code relying on the legacy HTTP mode
- MINOR: stats: Remove code relying on the legacy HTTP mode
- MAJOR: filters: Remove code relying on the legacy HTTP mode
- MINOR: stream: Remove code relying on the legacy HTTP mode
- MAJOR: http: Remove the HTTP legacy code
- MINOR: hlua: Remove useless test on TX_CON_WANT_* flags
- MINOR: proto_http: Remove unused http txn flags
- MINOR: proto_http: Remove the unused flag HTTP_MSGF_WAIT_CONN
- CLEANUP: proto_http: Group remaining flags of the HTTP transaction
- CLEANUP: channel: Remove the unused flag CF_WAKE_CONNECT
- CLEANUP: proto_http: Remove unecessary includes and comments
- CLEANUP: proto_http: Move remaining code from proto_http.c to proto_htx.c
- REORG: proto_htx: Move HTX analyzers & co to http_ana.{c,h} files
- BUG/MINOR: debug: Remove flags CO_FL_SOCK_WR_ENA/CO_FL_SOCK_RD_ENA
- MINOR: proxy: Remove support of the option 'http-tunnel'
- DOC: config: Update as a result of the legacy HTTP removal
- MEDIUM: config: Remove parsing of req* and rsp* directives
- MINOR: proxy: Remove the unused list of block rules
- MINOR: proxy/http_ana: Remove unused req_exp/rsp_exp and req_add/rsp_add lists
- DOC: config: Remove unsupported req* and rsp* keywords
- MINOR: global: Preset tune.max_http_hdr to its default value
- MINOR: http: Don't store raw HTTP errors in chunks anymore
- BUG/MINOR: session: Emit an HTTP error if accept fails only for H1 connection
- BUG/MINOR: session: Send a default HTTP error if accept fails for a H1 socket
- CLEANUP: mux-h2: Remove unused flags H2_SF_CHNK_*
- BUG/MINOR: checks: do not exit tcp-checks from the middle of the loop
- MINOR: config: Warn only if the option http-use-htx is used with "no" prefix
- BUG/MEDIUM: mux-h1: Trim excess server data at the end of a transaction
- MINOR: connection: add conn_get_src() and conn_get_dst()
- MINOR: frontend: switch to conn_get_{src,dst}() for logging and debugging
- MINOR: backend: switch to conn_get_{src,dst}() for port and address mapping
- MINOR: ssl: switch to conn_get_dst() to retrieve the destination address
- MINOR: tcp: replace various calls to conn_get_{from,to}_addr with conn_get_{src,dst}
- MINOR: stream-int: use conn_get_{src,dst} in conn_si_send_proxy()
- MINOR: stream/cli: use conn_get_{src,dst} in "show sess" and "show peers" output
- MINOR: log: use conn_get_{dst,src}() to retrieve the cli/frt/bck/srv/ addresses
- MINOR: http/htx: use conn_get_dst() to retrieve the destination address
- MINOR: lua: use conn_get_{src,dst} to retrieve connection addresses
- MINOR: http: check the source address via conn_get_src() in sample fetch functions
- CLEANUP: connection: remove the now unused conn_get_{from,to}_addr()
- MINOR: connection: add new src and dst fields
- MINOR: connection: use conn->{src,dst} instead of &conn->addr.{from,to}
- MINOR: ssl-sock: use conn->dst instead of &conn->addr.to
- MINOR: lua: switch to conn->dst for a connection's target address
- MINOR: peers: use conn->dst for the peer's target address
- MINOR: htx: switch from conn->addr.{from,to} to conn->{src,dst}
- MINOR: stream: switch from conn->addr.{from,to} to conn->{src,dst}
- MINOR: proxy: switch to conn->src in error snapshots
- MINOR: session: use conn->src instead of conn->addr.from
- MINOR: tcp: replace conn->addr.{from,to} with conn->{src,dst}
- MINOR: unix: use conn->dst for the target address in ->connect()
- MINOR: sockpair: use conn->dst for the target address in ->connect()
- MINOR: log: use conn->{src,dst} instead of conn->addr.{from,to}
- MINOR: checks: replace conn->addr.to with conn->dst
- MINOR: frontend: switch from conn->addr.{from,to} to conn->{src,dst}
- MINOR: http: convert conn->addr.from to conn->src in sample fetches
- MEDIUM: backend: turn all conn->addr.{from,to} to conn->{src,dst}
- MINOR: connection: create a new pool for struct sockaddr_storage
- MEDIUM: connection: make sure all address producers allocate their address
- MAJOR: connection: remove the addr field
- MINOR: connection: don't use clear_addr() anymore, just release the address
- MINOR: stream: add a new target_addr entry in the stream structure
- MAJOR: stream: store the target address into s->target_addr
- MINOR: peers: now remove the remote connection setup code
- MEDIUM: lua: do not allocate the remote connection anymore
- MEDIUM: backend: always release any existing prior connection in connect_server()
- MEDIUM: backend: remove impossible cases from connect_server()
- BUG/MINOR: mux-h1: Close server connection if input data remains in h1_detach()
- BUG/MEDIUM: tcp-checks: do not dereference inexisting conn_stream
- BUG/MINOR: http_ana: Be sure to have an allocated buffer to generate an error
- BUG/MINOR: http_htx: Support empty errorfiles
- BUG/CRITICAL: http_ana: Fix parsing of malformed cookies which start by a delimiter
- BUG/MEDIUM: protocols: add a global lock for the init/deinit stuff
- BUG/MINOR: proxy: always lock stop_proxy()
- MEDIUM: mux-h1: Add the support of headers adjustment for bogus HTTP/1 apps
- BUILD: threads: add the definition of PROTO_LOCK
- BUG/MEDIUM: lb-chash: Fix the realloc() when the number of nodes is increased
- BUG/MEDIUM: streams: Don't switch the SI to SI_ST_DIS if we have data to send.
- BUG/MINOR: log: make sure writev() is not interrupted on a file output
- DOC: improve the wording in CONTRIBUTING about how to document a bug fix
- MEDIUM: h1: Don't try to subscribe if we managed to read data.
- MEDIUM: h1: Don't wake the H1 tasklet if we got the whole request.
- REGTESTS: checks: exclude freebsd target for tcp-check_multiple_ports.vtc
- BUG/MINOR: hlua/htx: Reset channels analyzers when txn:done() is called
- BUG/MEDIUM: hlua: Check the calling direction in lua functions of the HTTP class
- MINOR: hlua: Don't set request analyzers on response channel for lua actions
- MINOR: hlua: Add a flag on the lua txn to know in which context it can be used
- BUG/MINOR: hlua: Only execute functions of HTTP class if the txn is HTTP ready
- BUG/MINOR: htx: Fix free space addresses calculation during a block expansion
- MINOR: ssl: merge ssl_sock_load_cert_file() and ssl_sock_load_cert_chain_file()
- MEDIUM: ssl: use cert_key_and_chain struct in ssl_sock_load_cert_file()
- MEDIUM: ssl: split the loading of the certificates
- MEDIUM: ssl: lookup and store in a ckch_node tree
- MEDIUM: ssl: load DH param in struct cert_key_and_chain
- BUG/MAJOR: queue/threads: avoid an AB/BA locking issue in process_srv_queue()
- MINOR: ssl: use STACK_OF for chain certs
- MINOR: ssl: add extra chain compatibility
- MINOR: ssl: check private key consistency in loading
- MINOR: ssl: do not look at DHparam with OPENSSL_NO_DH
- CLEANUP: ssl: ssl_sock_load_crt_file_into_ckch
- MINOR: ssl: clean ret variable in ssl_sock_load_ckchn
- MAJOR: fd: Get rid of the fd cache.
- MEDIUM: pollers: Remember the state for read and write for each threads.
- MEDIUM: mux-h2: don't try to read more than needed
- BUG/BUILD: ssl: fix build with openssl < 1.0.2
- BUG/MEDIUM: ssl: does not try to free a DH in a ckch
- BUG/MINOR: debug: fix a small race in the thread dumping code
- MINOR: wdt: also consider that waiting in the thread dumper is normal
- REGTESTS: checks: make 4be_1srv_health_checks more reliable
- BUILD: ssl: BoringSSL add EVP_PKEY_base_id
- BUG/MEDIUM: ssl: don't free the ckch in multi-cert bundle
- BUG/MINOR: ssl: fix ressource leaks on error
- BUG/MEDIUM: lb-chash: Ensure the tree integrity when server weight is increased
- BUG/MAJOR: http/sample: use a static buffer for raw -> htx conversion
- BUG/MINOR: stream-int: make sure to always release empty buffers after sending
- BUG/MEDIUM: ssl: open the right path for multi-cert bundle
- BUG/MINOR: stream-int: also update analysers timeouts on activity
- BUG/MEDIUM: mux-h2: unbreak receipt of large DATA frames
- BUG/MEDIUM: mux-h2: split the stream's and connection's window sizes
- BUG/MEDIUM: proxy: Make sure to destroy the stream on upgrade from TCP to H2
- DOC: Add 'Question.md' issue template, discouraging asking questions
- BUG/MEDIUM: fd: Always reset the polled_mask bits in fd_dodelete().
- BUG/MEDIUM: pollers: Clear the poll_send bits as well.
- BUILD: travis-ci: enable daily Coverity scan
- BUG/MINOR: mux-h2: don't refrain from sending an RST_STREAM after another one
- BUG/MINOR: mux-h2: use CANCEL, not STREAM_CLOSED in h2c_frt_handle_data()
- BUG/MINOR: mux-h2: do not send REFUSED_STREAM on aborted uploads
- BUG/MEDIUM: mux-h2: do not recheck a frame type after a state transition
- BUG/MINOR: mux-h2: always send stream window update before connection's
- BUG/MINOR: mux-h2: always reset rcvd_s when switching to a new frame
- BUG/MEDIUM: checks: make sure to close nicely when we're the last to speak
- BUG/MEDIUM: stick-table: Wrong stick-table backends parsing.
- CLEANUP: mux-h2: move the demuxed frame check code in its own function
- MINOR: cache: add method to cache hash
- MINOR: cache: allow caching of OPTIONS request
- BUG/MINOR: ssl: fix 0-RTT for BoringSSL
- MINOR: ssl: ssl_fc_has_early should work for BoringSSL
- BUG/MINOR: pools: don't mark the thread harmless if already isolated
- BUG/MINOR: buffers/threads: always clear a buffer's head before releasing it
- CLEANUP: buffer: replace b_drop() with b_free()
- CLEANUP: task: move the cpu_time field to the task-only part
- MINOR: cli: add two new states to print messages on the CLI
- MINOR: cli: add cli_msg(), cli_err(), cli_dynmsg(), cli_dynerr()
- CLEANUP: cli: replace all occurrences of manual handling of return messages
- BUG/MEDIUM: proxy: Don't forget the SF_HTX flag when upgrading TCP=>H1+HTX.
- BUG/MEDIUM: proxy: Don't use cs_destroy() when freeing the conn_stream.
- BUG/MINOR: lua: fix setting netfilter mark
- BUG/MINOR: Fix prometheus '# TYPE' and '# HELP' headers
- BUG/MEDIUM: lua: Fix test on the direction to set the channel exp timeout
- BUG/MINOR: stats: Wait the body before processing POST requests
- MINOR: fd: make sure to mark the thread as not stuck in fd_update_events()
- BUG/MEDIUM: mux_pt: Don't call unsubscribe if we did not subscribe.
- BUILD: travis-ci: trigger non-mainstream configurations only on daily builds.
- MINOR: debug: indicate the applet name when the task is task_run_applet()
- MINOR: tools: add append_prefixed_str()
- MINOR: lua: export applet and task handlers
- MEDIUM: debug: make the thread dump code show Lua backtraces
- BUG/MEDIUM: h1: Always try to receive more in h1_rcv_buf().
- MINOR: list: add LIST_SPLICE() to merge one list into another
- MINOR: tools: add a DEFNULL() macro to use NULL for empty args
- REORG: trace: rename trace.c to calltrace.c and mention it's not thread-safe
- MINOR: sink: create definitions a minimal code for event sinks
- MINOR: sink: add a support for file descriptors
- MINOR: trace: start to create a new trace subsystem
- MINOR: trace: add allocation of buffer-sized trace buffers
- MINOR: trace/cli: register the "trace" CLI keyword to list the sources
- MINOR: trace/cli: parse the "level" argument to configure the trace verbosity
- MINOR: trace/cli: add "show trace" to report trace state and statistics
- MINOR: trace: implement a very basic trace() function
- MINOR: trace: add the file name and line number in the prefix
- MINOR: trace: make trace() now also take a level in argument
- MINOR: trace: implement a call to a decode function
- MINOR: trace: add per-level macros to produce traces
- MINOR: trace: add a definition of typed arguments to trace()
- MINOR: trace: make sure to always stop the locking when stopping or pausing
- MINOR: trace: add the possibility to lock on some arguments
- MINOR: trace: parse the "lock" argument to trace
- MINOR: trace: retrieve useful pointers and enforce lock-on
- DOC: management: document the "trace" and "show trace" commands
- BUILD: trace: make the lockon_ptr const to silence a warning without threads
- BUG/MEDIUM: mux-h1: do not truncate trailing 0CRLF on buffer boundary
- BUG/MEDIUM: mux-h1: do not report errors on transfers ending on buffer full
- DOC: fixed typo in management.txt
- BUG/MINOR: mworker: disable SIGPROF on re-exec
- BUG/MEDIUM: listener/threads: fix an AB/BA locking issue in delete_listener()
- BUG/MEDIUM: url32 does not take the path part into account in the returned hash.
- MINOR: backend: Add srv_queue converter
- MINOR: sink: set the fd-type sinks to non-blocking
- MINOR: tools: add a function varint_bytes() to report the size of a varint
- MINOR: buffer: add functions to read/write varints from/to buffers
- MINOR: fd: add fd_write_frag_line() to send a fragmented line to an fd
- MINOR: sink: now call the generic fd write function
- MINOR: ring: add a new mechanism for retrieving/storing ring data in buffers
- MINOR: ring: add a ring_write() function
- MINOR: ring: add a generic CLI io_handler to dump a ring buffer
- MINOR: sink: add support for ring buffers
- MINOR: sink: implement "show events" to show supported sinks and dump the rings
- MINOR: sink: now report the number of dropped events on output
- MINOR: trace: support a default callback for the source
- MINOR: trace: extend the source location to 13 chars
- MINOR: trace: show thread number and source name in the trace
- MINOR: trace: change the TRACE() calling convention to put the args and cb last
- MINOR: connection: add the fc_pp_authority fetch -- authority TLV, from PROXYv2
- MINOR: tools: add a generic struct "name_desc" for name-description pairs
- MINOR: trace: replace struct trace_lockon_args with struct name_desc
- MINOR: trace: change the "payload" level to "data" and move it
- MINOR: trace: prepend the function name for developer level traces
- MINOR: trace: also report the trace level in the output
- MINOR: trace: change the detail_level to per-source verbosity
- MINOR: mux-h2/trace: register a new trace source with its events
- MINOR: mux-h2/trace: add the default decoding callback
- MEDIUM: mux-h2/trace: add lots of traces all over the code
- MINOR: mux-h2: add functions to convert an h2c/h2s state to a string
- MINOR: mux-h2/trace: add a new verbosity level "clean"
- MINOR: mux-h2/trace: only decode the start-line at verbosity other than "minimal"
- MINOR: mux-h2/trace: always report the h2c/h2s state and flags
- MINOR: mux-h2/trace: report h2s->id before h2c->dsi for the stream ID
- CLEANUP: mux-h2/trace: reformat the "received" messages for better alignment
- CLEANUP: mux-h2/trace: lower-case event names
- MINOR: trace: extend default event names to 12 chars
- BUG/MINOR: ring: fix the way watchers are counted
- MINOR: cli: extend the CLI context with a list and two offsets
- MINOR: mux-h2/trace: report the connection pointer and state before FRAME_H
- MEDIUM: ring: implement a wait mode for watchers
- BUG/MEDIUM: mux-h2/trace: do not dereference h2c->conn after failed idle
- BUG/MEDIUM: mux-h2/trace: fix missing braces added with traces
- BUG/MINOR: ring: b_peek_varint() returns a uint64_t, not a size_t
- CLEANUP: fd: remove leftovers of the fdcache
- MINOR: fd: add a new "initialized" bit in the fdtab struct
- MINOR: fd/log/sink: make the non-blocking initialization depend on the initialized bit
- MEDIUM: log: use the new generic fd_write_frag_line() function
- MINOR: log: add a target type instead of hacking the address family
- MEDIUM: log: add support for logging to a ring buffer
- MINOR: send-proxy-v2: sends authority TLV according to TLV received
- MINOR: build: add linux-glibc-legacy build TARGET
- BUG/MEDIUM: peers: local peer socket not bound.
- BUILD: connection: silence gcc warning with extra parentheses
- BUG/MINOR: http-ana: Reset response flags when 1xx messages are handled
- BUG/MINOR: h1: Properly reset h1m when parsing is restarted
- BUG/MINOR: mux-h1: Fix size evaluation of HTX messages after headers parsing
- BUG/MINOR: mux-h1: Don't stop anymore input processing when the max is reached
- BUG/MINOR: mux-h1: Be sure to update the count before adding EOM after trailers
- BUG/MEDIUM: cache: Properly copy headers splitted on several shctx blocks
- BUG/MEDIUM: cache: Don't cache objects if the size of headers is too big
- BUG/MINOR: mux-h1: Fix a possible null pointer dereference in h1_subscribe()
- MEDIUM: fd: remove the FD_EV_POLLED status bit
- MEDIUM: fd: simplify the fd_*_{recv,send} functions using BTS/BTR
- MINOR: fd: make updt_fd_polling() a normal function
- CONTRIB: debug: add new program "poll" to test poll() events
- BUG/MINOR: checks: stop polling for write when we have nothing left to send
- BUG/MINOR: checks: start sending the request right after connect()
- BUG/MINOR: checks: make __event_chk_srv_r() report success before closing
- BUG/MINOR: checks: do not uselessly poll for reads before the connection is up
- BUG/MINOR: mux-h1: Fix a UAF in cfg_h1_headers_case_adjust_postparser()
- BUILD: CI: add basic CentOS 6 cirrus build
- MINOR: contrib/prometheus-exporter: Report DRAIN/MAINT/NOLB status for servers
- BUG/MINOR: lb/leastconn: ignore the server weights for empty servers
- BUG/MAJOR: ssl: ssl_sock was not fully initialized.
- MEDIUM: fd: mark the FD as ready when it's inserted
- MINOR: fd: add two new calls fd_cond_{recv,send}()
- MEDIUM: connection: enable reading only once the connection is confirmed
- MINOR: fd: add two flags ERR and SHUT to describe FD states
- MEDIUM: fd: do not use the FD_POLL_* flags in the pollers anymore
- BUG/MEDIUM: connection: don't keep more idle connections than ever needed
- MINOR: stats: report the number of idle connections for each server
- BUILD: CI: skip reg-tests/connection/proxy_protocol_random_fail.vtc on CentOS 6
- BUILD/MINOR: auth: enabling for osx
- BUG/MINOR: listener: Fix a possible null pointer dereference
- BUG/MINOR: ssl: always check for ssl connection before getting its XPRT context
- MINOR: stats: Add JSON export from the stats page
- BUG/MINOR: filters: Properly set the HTTP status code on analysis error
- MINOR: sample: Add UUID-fetch
- CLEANUP: mux-h2: Remove unused flag H2_SF_DATA_CHNK
- BUG/MINOR: acl: Fix memory leaks when an ACL expression is parsed
- BUG/MINOR: backend: Fix a possible null pointer dereference
- BUG/MINOR: Missing stat_field_names (since f21d17bb)
- BUG/MEDIUM: stick-table: Properly handle "show table" with a data type argument
- BUILD: CI: temporarily disable ASAN
- MINOR: htx: Add a flag on HTX message to report processing errors
- MINOR: mux-h1: Report a processing error during output processing
- MINOR: http-ana: Handle HTX errors first during message analysis
- MINOR: http-ana: Remove err_state field from http_msg
- MINOR: config: Support per-proxy and per-server deinit functions callbacks
- MINOR: config: Support per-proxy and per-server post-check functions callbacks
- MINOR: http_fetch: Add sample fetches to get auth method/user/pass
- MINOR: istbuf: Add the function b_isteqi()
- MINOR: log: Provide a function to emit a log for an application
- MINOR: http: Add function to parse value of the header Status
- MEDIUM: mux-h1/h1-htx: move HTX convertion of H1 messages in dedicated file
- MINOR: h1-htx: Use the same function to copy message payload in all cases
- MINOR: muxes/htx: Ignore pseudo header during message formatting
- MINOR: fcgi: Add code related to FCGI protocol
- MEDIUM: fcgi-app: Add FCGI application and filter
- MEDIUM: mux-fcgi: Add the FCGI multiplexer
- MINOR: doc: Add documentation about the FastCGI support
- BUG/MINOR: build: Fix compilation of mux_fcgi.c when compiled without SSL
- BUILD: CI: install golang-1.13 when building BoringSSL
- BUG/MINOR: mux-h2: Be sure to have a connection to unsubcribe
- BUG/MINOR: mux-fcgi: Be sure to have a connection to unsubcribe
- CLEANUP: fcgi-app: Remove useless test on fcgi_conf pointer
- BUG/MINOR: mux-fcgi: Don't compare the filter name in its parsing callback
- BUG/MAJOR: mux-h2: Handle HEADERS frames received after a RST_STREAM frame
- BUG/MEDIUM: check/threads: make external checks run exclusively on thread 1
- MEDIUM: list: Separate "locked" list from regular list.
- MINOR: mt_lists: Add new macroes.
- MEDIUM: servers: Use LIST_DEL_INIT() instead of LIST_DEL().
- MINOR: mt_lists: Do nothing in MT_LIST_ADD/MT_LIST_ADDQ if already in list.
- MINOR: mt_lists: Give MT_LIST_ADD, MT_LIST_ADDQ and MT_LIST_DEL a return value.
- MEDIUM: tasklets: Make the tasklet list a struct mt_list.
- TESTS: Add a stress-test for mt_lists.
- BUILD: travis-ci: add PCRE2, SLZ build
- BUG/MINOR: build: fix event ports (Solaris)
- BUG/MEDIUM: namespace: fix fd leak in master-worker mode
- OPTIM: listeners: use tasklets for the multi-queue rings
- BUILD: makefile: work around yet another GCC fantasy (-Wstring-plus-int)
- BUG/MINOR: stream-int: Process connection/CS errors first in si_cs_send()
- BUG/MEDIUM: stream-int: Process connection/CS errors during synchronous sends
- BUG/MEDIUM: checks: make sure the connection is ready before trying to recv
- CLEANUP: task: remove impossible test
- CLEANUP: task: cache the task_per_thread pointer
- MINOR: task: split the tasklet vs task code in process_runnable_tasks()
- MINOR: task: introduce a thread-local "sched" variable for local scheduler stuff
- CLEANUP: mux-fcgi: Remove the unused function fcgi_strm_id()
- BUG/MINOR: mux-fcgi: Use a literal string as format in app_log()
- BUG/MEDIUM: tasklets: Make sure we're waking the target thread if it sleeps.
- MINOR: h2/trace: indicate 'F' or 'B' to locate the side of an h2c in traces
- MINOR: h2/trace: report the frame type when known
- BUG/MINOR: mux-h2: do not wake up blocked streams before the mux is ready
- BUG/MEDIUM: namespace: close open namespaces during soft shutdown
- MINOR: time: add timeofday_as_iso_us() to return instant time as ISO
- MINOR: sink: finally implement support for SINK_FMT_{TIMED,ISO}
- MINOR: sink: change ring buffer "buf0"'s format to "timed"
- BUG/MEDIUM: mux-h2: don't reject valid frames on closed streams
- BUG/MINOR: mux-fcgi: silence a gcc warning about null dereference
- BUG/MINOR: mux-h2: Fix missing braces because of traces in h2_detach()
- BUG/MINOR: mux-h2: Use the dummy error when decoding headers for a closed stream
- BUG/MAJOR: mux_h2: Don't consume more payload than received for skipped frames
- BUG/MINOR: mux-h1: Do h2 upgrade only on the first request
- BUG/MEDIUM: spoe: Use a different engine-id per process
- MINOR: spoe: Improve generation of the engine-id
- MINOR: spoe: Support the async mode with several threads
- MINOR: http: Add server name header from HTTP multiplexers
- CLEANUP: http-ana: Remove the unused function http_send_name_header()
- MINOR: stats: Add the support of float fields in stats
- BUG/MINOR: contrib/prometheus-exporter: Return the time averages in seconds
- DOC: Fix documentation about the cli command to get resolver stats
- BUG/MEDIUM: fcgi: fix missing list tail in sample fetch registration
- BUG/MINOR: stats: Add a missing break in a switch statement
- BUG/MINOR: lua: Properly initialize the buffer's fields for string samples in hlua_lua2(smp|arg)
- CLEANUP: lua: Get rid of obsolete (size_t *) cast in hlua_lua2(smp|arg)
- BUG/MEDIUM: lua: Store stick tables into the sample's `t` field
- CLEANUP: proxy: Remove `proxy_tbl_by_name`
- BUILD: ssl: fix a warning when built with openssl < 1.0.2
- DOC: replace utf-8 quotes by ascii ones
- BUG/MEDIUM: fd: HUP is an error only when write is active
- BUG/MINOR: action: do-resolve does not yield on requests with body
- Revert "MINOR: cache: allow caching of OPTIONS request"
This reverts commit 1263540fe8.
As discussed in issues #214 and #251, this is not the correct way to
cache CORS responses, since it relies on hacking the cache to cache
the OPTIONS method which is explicitly non-cacheable and for which
we cannot rely on any standard caching semantics (cache headers etc
are not expected there). Let's roll this back for now and keep that
for a more reliable and flexible CORS-specific solution later.
@davidmogar reported a github issue (#227) about problems with
do-resolve action when the request contains a body.
The variable was never populated in such case, despite tcpdump shows a
valid DNS response coming back.
The do-resolve action is a task in HAProxy and so it's waken by the
scheduler each time the scheduler think such task may have some work to
do.
When a simple HTTP request is sent, then the task is called, it sends
the DNS request, then the scheduler will wake up the task again later
once the DNS response is there.
Now, when the client send a PUT or a POST request (or any other type)
with a BODY, then the do-resolve action if first waken up once the
headers are processed. It sends the DNS request. Then, when the bytes
for the body are processed by HAProxy AND the DNS response has not yet
been received, then the action simply terminates and cleans up all the
data associated to this resolution...
This patch detect such behavior and if the action is now waken up while
a DNS resolution is in RUNNING state, then the action will tell the
scheduler to wake it up again later.
Backport status: 2.0 and above
William reported that since commit 6b3089856f ("MEDIUM: fd: do not use
the FD_POLL_* flags in the pollers anymore") the master's CLI often
fails to access sub-processes. There are two causes to this. One is
that we did report FD_POLL_ERR on an FD as soon as FD_EV_SHUT_W was
seen, which is automatically inherited from POLLHUP. And since we do
not store the current shutdown state of an FD we can't know if the
poller reports a sudden close resulting from an error or just a
byproduct of a previous shutdown(WR) followed by a read0. The current
patch addresses this by only considering this when the FD was active,
since a shutdown FD is not active. The second issue is that *somewhere*
down the chain, channel data are ignored if an error is reported on a
channel. This results in content truncation, but this cause was not
figured yet.
No backport is needed.
src/ssl_sock.c:2928:12: warning: ‘ssl_sock_is_ckch_valid’ defined but not used [-Wunused-function]
static int ssl_sock_is_ckch_valid(struct cert_key_and_chain *ckch)
This function is only used with openssl >= 1.0.2, this patch adds a
condition to build the function.
`size` is used in conditional jumps and valgrind complains:
==24145== Conditional jump or move depends on uninitialised value(s)
==24145== at 0x4B3028: smp_is_safe (sample.h:98)
==24145== by 0x4B3028: smp_make_safe (sample.h:125)
==24145== by 0x4B3028: smp_to_stkey (stick_table.c:936)
==24145== by 0x4B3F2A: sample_conv_in_table (stick_table.c:1113)
==24145== by 0x420AD4: hlua_run_sample_conv (hlua.c:3418)
==24145== by 0x54A308F: ??? (in /usr/lib/x86_64-linux-gnu/liblua5.3.so.0.0.0)
==24145== by 0x54AFEFC: ??? (in /usr/lib/x86_64-linux-gnu/liblua5.3.so.0.0.0)
==24145== by 0x54A29F1: ??? (in /usr/lib/x86_64-linux-gnu/liblua5.3.so.0.0.0)
==24145== by 0x54A3523: lua_resume (in /usr/lib/x86_64-linux-gnu/liblua5.3.so.0.0.0)
==24145== by 0x426433: hlua_ctx_resume (hlua.c:1097)
==24145== by 0x42D7F6: hlua_action (hlua.c:6218)
==24145== by 0x43A414: http_req_get_intercept_rule (http_ana.c:3044)
==24145== by 0x43D946: http_process_req_common (http_ana.c:500)
==24145== by 0x457892: process_stream (stream.c:2084)
Found while investigating issue #306.
A variant of this issue exists since 55da165301,
which was using the old `chunk` API instead of the `buffer` API thus this patch
must be backported to HAProxy 1.6 and higher.
A break is missing in the switch statement in the function
stats_emit_json_data_field(). This bug was introduced in the commit 88a0db28a
("MINOR: stats: Add the support of float fields in stats").
This patch fixes the issue #302 and #303. It must be backported to 2.0.
Ilya reported in bug #300 that ASAN found a read overflow during startup
in the fcgi code due to a missing empty element at the end of the list
of sample fetches. The effect is that will randomly either work or crash
on startup.
No backport is needed, this is solely for 2.1-dev.
In the management guide, this command was still referenced as "show stat
resolvers" instead of "show resolvers". The cli command was fixed by the commit
ff88efbd7 ("BUG/MINOR: dns: Fix CLI keyword declaration").
This patch fixes the issue #296. It can be backported as fas as 1.7.
The metrics QTIME, CTIME, RTIME and TTIME are now returned in seconds using a
float representation instead of in milliseconds. So these metrics are now
consistent with their announced type and respect Prometheus naming conventions.
This patch fixes the issue #288. It may be backported to 2.0. If so, the
previous patch, introducing the support for float fields in stats is mantatory
and should be backported first.
It is now possible to format stats counters as floats. But the stats applet does
not use it.
This patch is required by the Prometheus exporter to send the time averages in
seconds. If the promex change is backported, this patch must be backported
first.
the option "http-send-name-header" is an eyesore. It was responsible of several
bugs because it is handled after the message analysis. With the HTX
representation, the situation is cleaner because no rewind on forwarded data is
required. But it remains ugly.
With recent changes in HAProxy, we have the opportunity to make it fairly
better. The message formatting in now done in the HTTP multiplexers. So it seems
to be the right place to handle this option. Now, the server name is added by
the HTTP multiplexers (h1, h2 and fcgi).
A different engine-id is now generated for each thread. So, it is possible to
enable the async mode with several threads.
This patch may be backported to older versions.
Use the same algo than the sample fetch uuid(). This one was added recently. So
it is better to use the same way to generate UUIDs.
This patch may be backported to older versions.
SPOE engine-id is the same for all processes when nbproc is more than 1. So, in
async mode, an agent receiving a NOTIFY frame from a process may send the ACK to
another process. It is abviously wrong. A different engine-id must be generated
for each process.
This patch must be backported to 2.0, 1.9 and 1.8.
When a request is received, if the h2 preface is matched, an implicit upgrade
from h1 to h2 is performed. This must only be done for the first request on a
connection. But a test was missing to unsure it is really the first request.
This patch must be backported to 2.0.
When a frame is received for a unknown or already closed stream, it must be
skipped. This also happens when a stream error is reported. But we must be sure
to only skip received data. In the loop in h2_process_demux(), when such frames
are handled, all the frame lenght is systematically skipped. If the frame
payload is partially received, it leaves the demux buffer in an undefined
state. Because of this bug, all sort of errors may be observed, like crash or
intermittent freeze.
This patch must be backported to 2.0, 1.9 and 1.8.
Since the commit 6884aa3e ("BUG/MAJOR: mux-h2: Handle HEADERS frames received
after a RST_STREAM frame"), HEADERS frames received for an unknown or already
closed stream are decoded. Once decoded, an error is reported for the
stream. But because it is a dummy stream (h2_closed_stream), its state cannot be
changed. So instead, we must return the dummy error stream (h2_error_stream).
This patch must be backported to 2.0 and 1.9.
Consecutive to commit 6884aa3eb0 ("BUG/MAJOR: mux-h2: Handle HEADERS frames
received after a RST_STREAM frame") some valid frames on closed streams
(RST_STREAM, PRIORITY, WINDOW_UPDATE) were now rejected. It turns out that
the previous condition was in fact intentional to catch only sensitive
frames, which was indeed a mistake since these ones needed to be decoded
to keep HPACK synchronized. But we must absolutely accept WINDOW_UPDATES
or we risk to stall some transfers. And RST/PRIO definitely are valid.
Let's adjust the condition to reflect that and update the comment to
explain the reason for this unobvious condition.
This must be backported to 2.0 and 1.9 after the commit above is brought
there.
This way we now always have the events date which were really missing,
especially when used with traces :
<0>2019-09-26T07:57:25.183845 [00|h2|1|mux_h2.c:3024] receiving H2 HEADERS frame : h2c=0x1ddcad0(B,FRP) h2s=0x1dde9e0(3,HCL)
<0>2019-09-26T07:57:25.183845 [00|h2|4|mux_h2.c:2505] h2c_bck_handle_headers(): entering : h2c=0x1ddcad0(B,FRP) h2s=0x1dde9>
<0>2019-09-26T07:57:25.183846 [00|h2|4|mux_h2.c:4096] h2c_decode_headers(): entering : h2c=0x1ddcad0(B,FRP)
<0>2019-09-26T07:57:25.183847 [00|h2|4|mux_h2.c:4298] h2c_decode_headers(): leaving : h2c=0x1ddcad0(B,FRH)
<0>2019-09-26T07:57:25.183848 [00|h2|0|mux_h2.c:2559] rcvd H2 response : h2c=0x1ddcad0(B,FRH) : [3] H2 RES: HTTP/2.0 200
<0>2019-09-26T07:57:25.183849 [00|h2|4|mux_h2.c:2560] h2c_bck_handle_headers(): leaving : h2c=0x1ddcad0(B,FRH) h2s=0x1dde9e>
<0>2019-09-26T07:57:25.183849 [00|h2|4|mux_h2.c:2866] h2_process_demux(): no more Rx data : h2c=0x1ddcad0(B,FRH)
<0>2019-09-26T07:57:25.183849 [00|h2|4|mux_h2.c:3123] h2_process_demux(): notifying stream before switching SID : h2c=0x1dd>
<0>2019-09-26T07:57:25.183850 [00|h2|4|mux_h2.c:1014] h2s_notify_recv(): in : h2c=0x1ddcad0(B,FRH) h2s=0x1dde9e0(3,HCL)
<0>2019-09-26T07:57:25.183850 [00|h2|4|mux_h2.c:3135] h2_process_demux(): leaving : h2c=0x1ddcad0(B,FRH)
<0>2019-09-26T07:57:25.183851 [00|h2|4|mux_h2.c:3319] h2_send(): entering : h2c=0x1ddcad0(B,FRH)
<0>2019-09-26T07:57:25.183851 [00|h2|4|mux_h2.c:3145] h2_process_mux(): entering : h2c=0x1ddcad0(B,FRH)
<0>2019-09-26T07:57:25.183851 [00|h2|4|mux_h2.c:3234] h2_process_mux(): leaving : h2c=0x1ddcad0(B,FRH)
<0>2019-09-26T07:57:25.183852 [00|h2|4|mux_h2.c:3428] h2_send(): leaving with everything sent : h2c=0x1ddcad0(B,FRH)
<0>2019-09-26T07:57:25.183852 [00|h2|4|mux_h2.c:3319] h2_send(): entering : h2c=0x1ddcad0(B,FRH)
It looks like some format options could finally be separate from the sink,
or maybe enforced. For example we could imagine making the date optional
or its resolution configurable within a same buffer.
Similarly, maybe trace events would like to always emit the date even
on stdout, while traffic logs would prefer not to emit the date in the
ring buffer given that there's already one in the message.
We often need ISO time + microseconds in traces and ring buffers, thus
function does this by calling gettimeofday() and keeping a cached value
of the part representing the tv_sec value, and only rewrites the microsecond
part. The cache is per-thread so it's lockless and safe to use as-is.
Some tests already show that it's easy to see 3-4 events in a single
microsecond, thus it's likely that the nanosecond version will have to
be implemented as well. But certain comments on the net suggest that
some parsers are having trouble beyond microsecond, thus for now let's
stick to the microsecond only.
When doing a soft shutdown, we won't be making new connections anymore so
there's no point in keeping the namespace file descriptors open anymore.
Keeping these open effectively makes it impossible to properly clean up
namespaces which are no longer used in the new configuration until all
previously opened connections are closed in the old worker process.
This change introduces a cleanup function that is called during soft shutdown
that closes all namespace file descriptors by iterating over the namespace
ebtree.
In state match error cases, we don't know what frame type was received
because we don't reach the frame parsers. Let's add the demuxed frame
type and flags in the trace when it's known. For this we make sure to
always reset h2c->dsi when switching back to FRAME_H. Only one location
was missing. The state transitions were not always clear (sometimes
reported before, sometimes after), these were clarified by being
reported only before switching.
It was difficult in traces showing h2-to-h2 communications to figure the
connection side solely based on the pointer. With this patch we prepend
'F' or 'B' before the state to make this more explicit:
[06|h2|4|mux_h2.c:5487] h2_rcv_buf(): entering : h2c=0x7f6acc026440(F,FRH) h2s=0x7f6acc021720(1,CLO)
[06|h2|4|mux_h2.c:5547] h2_rcv_buf(): leaving : h2c=0x7f6acc026440(F,FRH) h2s=0x7f6acc021720(1,CLO)
[06|h2|4|mux_h2.c:4040] h2_shutw(): entering : h2c=0x7f6acc026440(F,FRH) h2s=0x7f6acc021720(1,CLO)
Now that we can wake tasklet for other threads, make sure that if the thread
is sleeping, we wake it up, or the tasklet won't be executed until it's
done sleeping.
That also means that, before going to sleep, and after we put our bit
in sleeping_thread_mask, we have to check that nobody added a tasklet for
us, just checking for global_tasks_mask isn't enough anymore.
The aim is to rassemble all scheduler information related to the current
thread. It simply points to task_per_thread[tid] without having to perform
the operation at each time. We save around 1.2 kB of code on performance
sensitive paths and increase the request rate by almost 1%.
There are a number of tests there which are enforced on tasklets while
they will never apply (various handlers, destroyed task or not, arguments,
results, ...). Instead let's have a single TASK_IS_TASKLET() test and call
the tasklet processing function directly, skipping all the rest.
It now appears visible that the only unneeded code is the update to
curr_task that is never used for tasklets, except for opportunistic
reporting in the debug handler, which can only catch si_cs_io_cb,
which in practice doesn't appear in any report so the extra cost
incurred there is pointless.
This change alone removes 700 bytes of code, mostly in
process_runnable_tasks() and increases the performance by about
1%.
In process_runnable_tasks() we perform a lot of dereferences to
task_per_thread[tid] but tid is thread_local and the compiler cannot
know that it doesn't change so this results in making lots of thread
local accesses and array dereferences. By just keeping a copy pointer
of this, we let the compiler optimize the code. Just doing this has
reduced process_runnable_tasks() by 124 bytes in the fast path. Doing
the same in wake_expired_tasks() results in 16 extra bytes saved.
In process_runnable_task(), after the task's process() function returns,
we used to check if the return is not NULL and is not a tasklet, to update
profiling measurements. This is useless since only tasks can return non-null
here. Let's remove this useless test.
As identified in issue #278, the backport of commit c594039225 ("BUG/MINOR:
checks: do not uselessly poll for reads before the connection is up")
introduced a regression in 2.0 when default checks are enabled (not
"option tcp-check"), but it did not affect 2.1.
What happens is that in 2.0 and earlier we have the fd cache which makes
a speculative call to the I/O functions after an attempt to connect, and
the __event_srv_chk_r() function was absolutely not designed to be called
while a connection attempt is still pending. Thus what happens is that the
test for success/failure expects the verdict to be final before waking up
the check task, and since the connection is not yet validated, it fails.
It will usually work over the loopback depending on scheduling, which is
why it doesn't fail in reg tests.
In 2.1 after the failed connect(), we subscribe to polling and usually come
back with a validated connection, so the function is not expected to be
called before it completes, except if it happens as a side effect of some
spurious wake calls, which should not have any effect on such a check.
The other check types are not impacted by this issue because they all
check for a minimum data length in the buffer, and wait for more data
until they are satisfied.
This patch fixes the issue by explicitly checking that the connection
is established before trying to read or to give a verdict. This way the
function becomes safe to call regardless of the connection status (even
if it's still totally ugly).
This fix must be backported to 2.0.
If an error occurred on the connection or the conn-stream, no syncrhonous send
is performed. If the error was not already processed and there is no more I/O,
it will never be processed and the stream will never be notified of this
error. This may block the stream until a timeout is reached or infinitly if
there is no timeout.
Concretly, this bug can be triggered time to time with h2spec, running the test
"http2/5.1.1/2".
This patch depends on the commit 328ed220a "BUG/MINOR: stream-int: Process
connection/CS errors first in si_cs_send()". Both must be backported to 2.0 and
probably to 1.9. In 1.9, the code is totally different, so this patch would have
to be adapted.
Errors on the connections or the conn-stream must always be processed in
si_cs_send(), even if the stream-interface is already subscribed on
sending. This patch does not fix any concrete bug per-se. But it is required by
the following one to handle those errors during synchronous sends.
This patch must be backported with the following one to 2.0 and probably to 1.9
too, but with caution because the code is really different.
According to issue #294 some gcc versions suspect that developers are
having trouble dealing with string offsets and now emit another new
childish warning when mapping indexes to characters. Instead of annoying
developers each time it happens and ask them to modify their valid code,
let's just get rid of this absurd warning.
Now that we can wake up a remote thread's tasklet, it's way more
interesting to use a tasklet than a task in the accept queue, as it
will avoid passing through all the scheduler. Just doing this increases
the accept rate by about 4%, overall recovering the slight loss
introduced by the tasklet change. In addition it makes sure that
even a heavily loaded scheduler (e.g. many very fast checks) will
not delay a connection accept.
When namespaces are used in the configuration, the respective namespace handles
are opened during config parsing and stored in an ebtree for lookup later.
Unfortunately, when the master process re-execs itself these file descriptors
were not closed, effectively leaking the fds and preventing destruction of
namespaces no longer present in the configuration.
This change fixes this issue by opening the namespace file handles as
close-on-exec, making sure that they will be closed during re-exec.
