HAProxy 2.1.4 was released on 2020/04/02. It added 99 new commits
after version 2.1.3.

The main driver for this release is that it contains a fix for a serious
vulnerability that was responsibly reported last week by Felix Wilhelm
from Google Project Zero, affecting the HPACK decoder used for HTTP/2.
CVE-2020-11100 was assigned to this issue.

There is no configuration-based workaround for 2.1 and above.

This vulnerability makes it possible under certain circumstances to write
to a wide range of memory locations within the process' heap, with the
limitation that the attacker doesn't control the absolute address, so the
most likely result and by a far margin will be a process crash, but it is
not possible to completely rule out the faint possibility of a remote code
execution, at least in a lab-controlled environment. Felix was kind enough
to agree to delay the publication of his findings to the 20th of this month
in order to leave enough time to haproxy users to apply updates. But please
do not wait, as it is not very difficult to figure how to exploit the bug
based on the fix. Distros were notified and will also have fixes available
very shortly.

Three other important fixes are present in this version:
  - a non-portable way of calculating a list pointer that breaks with
    gcc 10 unless using -fno-tree-pta. This bug results in infinite loops
    at random places in the code depending how the compiler decides to
    optimize the code.

  - a bug in the way TLV fields are extracted from the PROXY protocol, as
    they could be mistakenly looked up in the subsequent payload, even
    though these would have limited effects since these ones would generally
    be meaningless for the transported protocol, but could be used to hide a
    source address from logging for example.

  - the "tarpit" rules were partially broken in that since 1.9 they wouldn't
    prevent a connection from being sent to a server while the 500 response
    is delivered to the client. Given that they are often used to block
    suspicious activity it's problematic.

The rest is less important, but still relevant to some users. Among those
noticeable I can enumerate:
  - the O(N^2) ACL unique-id allocator that could take several minutes to
    boot on certain very large configs was reworked to follow O(NlogN)

  - the default global maxconn setting when not set in the configuration was
    incorrectly set to the process' soft limit instead of the hard limit,
    resulting in much lower connection counts on some setups after upgrade
    from 1.x to 2.x. It now properly follows the hard limit.

  - a new thread-safe random number generator that will avoid the risk that
    the "uuid" sample fetch function returns the exact same UUID in several

  - issues in HTX mode affecting filters, namely cache and compression, that
    could lead to data corruption.

  - alignment issues causing bus error on Sparc64 were addressed

  - fixed a rare case of possible segfault on soft-stop when a finishing thread
    flushes its pools while another one is freeing some elements.

Please have a look at the changelog below for a more detailed list of fixes,
and do not forget to update, either from the sources or from your regular
distro channels.

Please find the usual URLs below :
   Site index       : http://www.haproxy.org/
   Discourse        : http://discourse.haproxy.org/
   Slack channel    : https://slack.haproxy.org/
   Issue tracker    : https://github.com/haproxy/haproxy/issues
   Sources          : http://www.haproxy.org/download/2.1/src/
   Git repository   : http://git.haproxy.org/git/haproxy-2.1.git/
   Git Web browsing : http://git.haproxy.org/?p=haproxy-2.1.git
   Changelog        : http://www.haproxy.org/download/2.1/src/CHANGELOG
   Cyril's HTML doc : http://cbonte.github.io/haproxy-dconv/

Complete changelog :
Balvinder Singh Rawat (1):
      DOC: correct typo in alert message about rspirep

Bjoern Jacke (1):
      DOC: fix typo about no-tls-tickets

Björn Jacke (1):
      DOC: improve description of no-tls-tickets

Carl Henrik Lunde (1):
      OPTIM: startup: fast unique_id allocation for acl.

Christopher Faulet (26):
      BUG/MINOR: mux-fcgi: Forbid special characters when matching PATH_INFO 
      MINOR: mux-fcgi: Make the capture of the path-info optional in pathinfo 
      MINOR: http-htx: Add a function to retrieve the headers size of an HTX 
      MINOR: filters: Forward data only if the last filter forwards something
      BUG/MINOR: filters: Count HTTP headers as filtered data but don't forward 
      BUG/MINOR: http-htx: Don't return error if authority is updated without 
      BUG/MINOR: http-ana: Matching on monitor-uri should be case-sensitive
      MINOR: http-ana: Match on the path if the monitor-uri starts by a /
      BUG/MAJOR: http-ana: Always abort the request when a tarpit is triggered
      BUG/MINOR: http-htx: Do case-insensive comparisons on Host header name
      MINOR: contrib/prometheus-exporter: Add heathcheck status/code in server 
      MINOR: contrib/prometheus-exporter: Add the last heathcheck duration 
      BUG/MINOR: filters: Use filter offset to decude the amount of forwarded 
      BUG/MINOR: filters: Forward everything if no data filters are called
      MINOR: htx: Add a function to return a block at a specific offset
      BUG/MEDIUM: cache/filters: Fix loop on HTX blocks caching the response 
      BUG/MEDIUM: compression/filters: Fix loop on HTX blocks compressing the 
      BUG/MINOR: http-ana: Reset request analysers on a response side error
      BUG/MINOR: lua: Ignore the reserve to know if a channel is full or not
      BUG/MINOR: http-rules: Preserve FLT_END analyzers on reject action
      BUG/MINOR: http-rules: Fix a typo in the reject action function
      BUG/MINOR: rules: Preserve FLT_END analyzers on silent-drop action
      BUG/MINOR: rules: Increment be_counters if backend is assigned for a 
      MINOR: http-rules: Add a flag on redirect rules to know the rule direction
      MINOR: http-rules: Handle the rule direction when a redirect is evaluated
      BUG/MINOR: http-ana: Reset request analysers on error when waiting for 

Daniel Corbett (1):
      BUG/MINOR: stats: Fix color of draining servers on stats page

David Carlier (1):
      BUILD: on ARM, must be linked to libatomic.

Emeric Brun (1):
      BUG/MEDIUM: peers: resync ended with RESYNC_PARTIAL in wrong cases.

Frédéric Lécaille (1):
      BUG/MINOR: peers: Use after free of "peers" section.

Ilya Shipitsin (4):
      DOC: configuration.txt: fix various typos
      DOC: assorted typo fixes in the documentation and Makefile
      DOC: assorted typo fixes in the documentation
      DOC: assorted typo fixes in the documentation

Jerome Magnin (4):
      MINOR: ist: add an iststop() function
      BUG/MINOR: http: http-request replace-path duplicates the query string
      MINOR: listener: add so_name sample fetch
      BUG/MINOR: http_ana: make sure redirect flags don't have overlapping bits

Lukas Tribus (2):
      BUG/MINOR: dns: ignore trailing dot
      DOC: ssl: clarify security implications of TLS tickets

Miroslav Zagorac (1):
      DOC: internals: Fix spelling errors in filters.txt

Olivier Houchard (8):
      BUG/MEDIUM: muxes: Use the right argument when calling the destroy method.
      BUG/MEDIUM: mt_lists: Make sure we set the deleted element to NULL;
      MINOR: mt_lists: Appease gcc.
      BUG/MEDIUM: pools: Always update free_list in pool_gc().
      MINOR: wdt: Move the definitions of WDTSIG and DEBUGSIG into 
      BUG/MEDIUM: wdt: Don't ignore WDTSIG and DEBUGSIG in 
      MINOR: memory: Change the flush_lock to a spinlock, and don't get it in 
      BUG/MINOR: connections: Make sure we free the connection on failure.

Tim Duesterhus (5):
      CLEANUP: cfgparse: Fix type of second calloc() parameter
      BUG/MINOR: sample: Make sure to return stable IDs in the unique-id fetch
      BUG/MINOR: pattern: Do not pass len = 0 to calloc()
      BUG/MAJOR: proxy_protocol: Properly validate TLV lengths
      DOC: proxy_protocol: Reserve TLV type 0x05 as PP2_TYPE_UNIQUE_ID

William Dauchy (1):
      BUG/MINOR: namespace: avoid closing fd when socket failed in my_socketat

William Lallemand (2):
      BUG/MINOR: peers: init bind_proc to 1 if it wasn't initialized
      BUG/MINOR: peers: avoid an infinite loop with peers_fe is NULL

Willy Tarreau (38):
      SCRIPTS: make announce-release executable again
      SCRIPTS: announce-release: use mutt -H instead of -i to include the draft
      BUG/MEDIUM: shctx: make sure to keep all blocks aligned
      MINOR: compiler: move CPU capabilities definition from config.h and 
complete them
      BUG/MEDIUM: ebtree: don't set attribute packed without unaligned access 
      BUILD: fix recent build failure on unaligned archs
      BUG/MINOR: sample: fix the json converter's endian-sensitivity
      BUG/MEDIUM: ssl: fix several bad pointer aliases in a few sample fetch 
      BUG/MINOR: connection: make sure to correctly tag local PROXY connections
      MINOR: compiler: add new alignment macros
      BUILD: ebtree: improve architecture-specific alignment
      BUG/MINOR: h2: reject again empty :path pseudo-headers
      BUG/MEDIUM: random: initialize the random pool a bit better
      MINOR: tools: add 64-bit rotate operators
      BUG/MEDIUM: random: implement a thread-safe and process-safe PRNG
      MINOR: backend: use a single call to ha_random32() for the random LB algo
      BUG/MINOR: checks/threads: use ha_random() and not rand()
      BUG/MAJOR: list: fix invalid element address calculation
      MINOR: debug: report the task handler's pointer relative to main
      BUG/MEDIUM: debug: make the debug_handler check for the thread in 
      MINOR: haproxy: export main to ease access from debugger
      BUILD: tools: remove obsolete and conflicting trace() from standard.c
      BUG/MINOR: wdt: do not return an error when the watchdog couldn't be 
      DOC: fix incorrect indentation of http_auth_*
      BUG/MINOR: init: make the automatic maxconn consider the max of soft/hard 
      REGTEST: make the PROXY TLV validation depend on version 2.2
      BUILD: wdt: only test for SI_TKILL when compiled with thread support
      BUG/MEDIUM: random: align the state on 2*64 bits for ARM64
      BUG/MINOR: haproxy: always initialize sleeping_thread_mask
      BUG/MINOR: listener/mq: do not dispatch connections to remote threads 
when stopping
      BUG/MINOR: haproxy/threads: try to make all threads leave together
      BUILD: makefile: fix regex syntax in ARM platform detection
      BUILD: makefile: fix expression again to detect ARM platform
      REGTESTS: use "command -v" instead of "which"
      REGTEST: increase timeouts on the seamless-reload test
      BUG/MINOR: haproxy/threads: close a possible race in soft-stop detection
      BUILD: ssl: only pass unsigned chars to isspace()
      BUG/CRITICAL: hpack: never index a header into the headroom after wrapping


Reply via email to