Willy Tarreau [Fri, 6 Mar 2026 19:22:18 +0000 (20:22 +0100)]
MINOR: filters: set the exec context to the current filter config
Doing this allows to report the allocations/releases performed by filters
when running with memory profiling enabled. The flt_conf pointer is kept
and the report shows the filter name.
Willy Tarreau [Fri, 6 Mar 2026 13:32:51 +0000 (14:32 +0100)]
MINOR: actions: store the location of keywords registered via initcalls
A bit similar to what was done for sample fetch functions and converters,
we now store with each action keyword the location of the initcall when
they're registered this way. Since there are many functions only calling
a LIST_APPEND() (one per ruleset), we now implement a dedicated function
to store the context in all keywords before doing the append.
However that's not sufficient, because keywords are not mandatory for
actions, so we cannot safely rely on rule->kw. Thus we then set the
exec_ctx per rule when they are all scanned in check_action_rules(),
based on the keyword if it exists, otherwise we make a context from
the action_ptr function if it is set (it should).
Finally at all call points we now check rule->exec_ctx.
Willy Tarreau [Wed, 11 Mar 2026 07:53:37 +0000 (08:53 +0100)]
MINOR: tools: support an execution context that is just a function
The purpose here is to be able to spot certain callbacks, such as the
SSL message callbacks, which are difficult to associate to anything.
Thus we introduce a new context type, TH_EX_CTX_FUNC, for which the
context is just the function pointed to by the void *pointer. One
difficulty with callbacks is that the allocation and release contexts
will likely be different, so the code should be properly structured
to allow proper tracking, either by instrumenting all calls, or by
making sure that the free calls are easy to spot in a report.
Willy Tarreau [Fri, 6 Mar 2026 09:59:18 +0000 (10:59 +0100)]
MINOR: sample: also report contexts registered directly
With the two new context types TH_EX_CTX_SMPF/CONV, we can now also
report contexts corresponding to direct calls to sample_register_fetches()
and sample_register_convs(). In this case, the first word of the keyword
list is reported.
Willy Tarreau [Fri, 6 Mar 2026 09:49:46 +0000 (10:49 +0100)]
MINOR: sample: store location for fetch/conv via initcalls
Now keywords are registered with an exec_ctx and this one is passed
when calling ->process. The ctx is of type INITCALL when passed via
an initcall where we know the file name and line number.
This was tested with and extra "malloc(15)" added in smp_fetch_path()
which shows that it works:
$ socat /tmp/sock1 - <<< "show profiling memory"|grep via
Calls | Tot Bytes | Caller and method [via] 1893399 0 60592592 0| 0x78b2ec task_run_applet+0x3339c malloc(32) [via initcall @http_fetch.c:2416]
When the execution context is set to TH_EX_CTX_INITCALL, the pointer
points to a valid initcall, and the decoder will show "kw registered
at %s:%d" with file and line number of the initcall declaration. It's
up to the caller to make the initcall pointer point to the one that was
set during the initcall. The purpose here is to be able to preserve and
pass that knowledge of an initcall down the chain so that future calls
to functions registered via the initcall are still assigned to it.
Willy Tarreau [Tue, 3 Mar 2026 09:08:19 +0000 (10:08 +0100)]
MINOR: initcall: record the file and line declaration of an INITCALL
The INITCALL macros will now store the file and line number where they
are declared into the initcall struct, and RUN_INITCALLS() will assign
them to the global caller_file and caller_line variables, and will even
set caller_initcall to the current initall so that at any instant such
functions know where their caller declared them. This will help with
error messages and traces where a bit of context will be welcome.
Willy Tarreau [Thu, 12 Mar 2026 07:55:04 +0000 (08:55 +0100)]
MINOR: debug: report the execution context on thread dumps
Now we have one extra line saying "exec_ctx: something" in thread dumps
when it's known. It may help with warnings and panics to figure what
is ongoing.
Willy Tarreau [Thu, 12 Mar 2026 06:48:14 +0000 (07:48 +0100)]
MINOR: tools: add a function to write a thread execution context.
The new function chunk_append_thread_ctx() appends to a buffer the given
execution context based on its type and pointer. The goal is to easily
use it in profiling output and thread dumps. For now it only handles
TH_EX_CTX_NONE (which prints nothing) and TH_EX_CTX_OTHER (which indicates
"other ctx" followed by the pointer). It will be extended by new types as
they arrive.
Willy Tarreau [Fri, 6 Mar 2026 18:51:08 +0000 (19:51 +0100)]
MINOR: memprof: also permit to sort output by calling context
By passing "byctx" to "show profiling memory", it's possible to sort by
the calling context first, which could help group certain calls by
subsystem and ease the interpretation of the output.
Willy Tarreau [Tue, 3 Mar 2026 13:02:38 +0000 (14:02 +0100)]
MINOR: memprof: prepare to consider exec_ctx in reporting
This now allows to report the same function in multiple bins based on the
th_ctx's exec_ctx discriminant. It's also worth noting that the context is
not atomically committed, but this shouldn't be a problem since a single
entry can get it. In the worst case, a second thread trying to create the
same context in parallel would create a different bin just for this call,
which is harmless. The same situation already exists with the caller
pointer.
Willy Tarreau [Tue, 3 Mar 2026 08:37:10 +0000 (09:37 +0100)]
MINOR: tinfo: start to add basic thread_exec_ctx
We have the struct made of a type and a pointer in the th_ctx and a
function to switch it for the current thread. Two macros are provided
to enclose a callee within a temporary context. For now only type OTHER
is supported (only a generic pointer).
Willy Tarreau [Wed, 11 Mar 2026 15:33:33 +0000 (16:33 +0100)]
MINOR: memprof: attempt different retry slots for different hashes on collision
When two pointer hash to the same memprofile bin, we currently try again
with the same bin until we find a spare one or we reach the limit of 16.
Olivier suggested to try with a different step for different pointers so
as to limit the number of bins to visit in such a case, so let's split
the pointer hash calculation so that we keep the raw hash before reduction
and use its lowest bits as the retry step. We force lowest bit to 1 to
avoid integral multiples that would oscillate between only a few positions.
Quick tests with h1+h2 requests show that for ~744 distinct entries, we
used to have 1.17 retries per lookup before and 0.6 now so we're halving
the cost of hash collisions. A heavier workload that used to produce 920
entries with 2.01 retries per lookup now reaches 966 entries (94.3% usage
vs 89.8% before) with only 1.44 retries per lookup.
This should be safe to backport, but depends on this previous commit:
MINOR: tools: extend the pointer hashing code to ease manipulations
Willy Tarreau [Fri, 6 Mar 2026 18:14:06 +0000 (19:14 +0100)]
MINOR: tools: add a new pointer hash function that also takes an argument
The purpose here is to combine two pointers and a long argument instead
of having the caller perform the mixing. Also it's cleaner and more
efficient this was as the arg is mixed after the multiplications, and
modern processors are efficient at multiplying then adding.
Willy Tarreau [Wed, 11 Mar 2026 15:20:23 +0000 (16:20 +0100)]
MINOR: tools: extend the pointer hashing code to ease manipulations
We'll need to further extend the pointer hashing code to pass extra
parameters and to retrieve the dropped bits, so let's first split the
part that hashes the pointer from the part that reduces the hash to
the desired size.
Willy Tarreau [Wed, 11 Mar 2026 14:22:24 +0000 (15:22 +0100)]
MINOR: activity: use dynamic allocation for "show profiling" entries
Historically, the data manipulated by "show profiling" were copied
onto the stack for sorting and aggregating, but not only this limits
the number of entries we can keep, but it also has an impact on CPU
usage (having to redo the whole copy+sort upon each resume) and the
output accuracy (if sorting changes lines, resume may happen from an
incorrect one).
Instead, let's dynamically allocate the work buffer and place it into
the service context. We only allocate it immediately before needing it
and release it immediately afterwards so that it doesn't stay long. It
also requires a release handler to release those allocates by interrupted
dumps, but that's all. The overall result is now much cleaner, more
accurate, faster and safer.
This patch may be backported to older LTS releases.
Willy Tarreau [Thu, 12 Mar 2026 15:35:29 +0000 (16:35 +0100)]
BUG/MINOR: proxy: do not forget to validate quic-initial rules
In check_config_validity() and proxy_finalize() we check the consistency
of all rule sets, but the quic_initial rules were not placed there. This
currently has little to no impact, however we're going to use that to
also finalize certain debugging info so better call the function. This
can be backported to 3.1 (proxy_finalize is 3.4-only).
Willy Tarreau [Wed, 11 Mar 2026 14:14:43 +0000 (15:14 +0100)]
BUG/MINOR: memprof: avoid a small memory leak in "show profiling"
In 3.1, per-DSO statistics were added to the memprofile output by
commit 401fb0e87a ("MINOR: activity/memprofile: show per-DSO stats").
However an strdup() is performed there on the .info field, that is
never freed when leaving the function. Let's do it each time we leave
it. Ironically, this was found thanks to "show profiling" showing
itself as an unbalanced caller of strdup().
This needs to be backported to 3.0 since that commit was backported
there.
Willy Tarreau [Thu, 12 Mar 2026 07:30:30 +0000 (08:30 +0100)]
BUILD: makefile: fix range build without test command
In 3.3, the "make range" target adopted a test command via the TEST_CMD
variable, with commit 90b70b61b1 ("BUILD: makefile: implement support
for running a command in range"). However now it breaks the script when
TEST_CMD is not set due to the shell expansion leaving two '||' operators
side by side. Let's fix this by passing the contents of the makefile
variable in positional arguments before executing them.
Olivier Houchard [Thu, 12 Mar 2026 16:31:43 +0000 (17:31 +0100)]
BUG/MEDIUM: ssl: Don't report read data as early data with AWS-LC
To read early data with AWS-LC (and BoringSSL), we have to use
SSL_read(). But SSL_read() will also try to do the handshake if it
hasn't been done yet, and at some point will do the handshake and will
return data that are actually not early data. So use SSL_in_early_data()
to make sure that the data we received are actually early data, and only
if so add the CO_FL_EARLY_DATA flag. Otherwise any data first received will be
considered early, and a Early-data header will be added.
As this bug was introduced by 76ba026548975a6d1bc23d1344807c64d994bf1e,
it should be backported with it.
BUG/MINOR: mworker: always stop the receiving listener
Upon _send_status, always stop the listener from which the request
was received, rather than looking it up from the proc_list entry via
fdtab[proc->ipc_fd[0]].owner.
A BUG_ON is added to verify that the listener which received the
request is the one expected for the reported PID.
This means it is no longer possible to send "_send_status READY XXX"
manually through the master CLI for testing, as that would trigger
the BUG_ON.
BUG/MEDIUM: ssl: Handle receiving early data with BoringSSL/AWS-LC
The API for early data is a bit different with BoringSSL and AWS-LC than
it is for OpenSSL. As it was implemented, early data would be accepted,
but would not be processed until the handshake is done. Change that by
doing something similar to what OpenSSL does, and, if 0RTT has been
enabled on the listener, use SSL_read() to try to get early data before
starting the handshake, and if there's any, provide them to the mux the
same way it is done for OpenSSL.
That replaces a bunch of #ifdef SSL_READ_EARLY_DATA_SUCCESS by
something specific to OpenSSL has to be done.
This should be backported to 3.3.
BUILD: ssl: make X509_NAME usage OpenSSL 4.0 ready
Starting with OpenSSL 4.0, X509_get_subject_name(), X509_get_issuer_name(),
and X509_CRL_get_issuer() return a const-qualified X509_NAME pointer.
Similarly, X509_NAME_get_entry() returns a const X509_NAME_ENTRY *, and
X509_NAME_ENTRY_get_data() returns a const ASN1_STRING *.
Introduce the __X509_NAME_CONST__ macro (defined to 'const' for OpenSSL
>= 4.0.0, empty for WolfSSL and older OpenSSL version which lacks const
on these APIs) and use it to qualify X509_NAME * variables and the
parameters of the three DN helper functions ssl_sock_get_dn_entry(),
ssl_sock_get_dn_formatted(), and ssl_sock_get_dn_oneline(). This avoids
both const-qualifier warnings on OpenSSL 4.0 and discarded-qualifier
warnings on WolfSSL, without needing explicit casts at call sites.
In ssl_sock.c (ssl_get_client_ca_file) and ssl_gencert.c
(ssl_sock_do_create_cert), a __X509_NAME_CONST__ X509_NAME * variable was
being reused to store the result of X509_NAME_dup() and then passed to
mutating functions (X509_NAME_add_entry_by_txt, X509_NAME_free). Introduce
separate X509_NAME * variables (xn_dup, subject) to hold the mutable
duplicate.
Original patch from Alexandr Nedvedicky <sashan@openssl.org>:
https://www.mail-archive.com/haproxy@formilux.org/msg46696.html
BUILD: ssl: use ASN1_STRING accessors for OpenSSL 4.0 compatibility
In OpenSSL 4.0, the ASN1_STRING struct was made opaque and direct access
to its members (->data, ->length, ->type) no longer compiles. Replace
these accesses in ssl_sock_get_serial(), ssl_sock_get_time(), and
asn1_generalizedtime_to_epoch() with the proper accessor functions
ASN1_STRING_get0_data(), ASN1_STRING_length(), and ASN1_STRING_type().
The old direct access is preserved under USE_OPENSSL_WOLFSSL since
WolfSSL does not provide these accessor functions.
Original patch from Alexandr Nedvedicky <sashan@openssl.org>:
https://www.mail-archive.com/haproxy@formilux.org/msg46696.html
MEDIUM: mworker: exiting when couldn't find the master mworker_proc element
When a master process is reloading, the HAPROXY_PROCESSES variable is
deserialized. In older version of the master-worker (< 1.9), no master
element was existing in this variable.
This is not suppose to happen anymore, and could have provoked problems
in the master anyway.
This patch changes the behavior by exiting the master with an alert if
mp master element was found in this variable.
DEBUG: stconn: Add a CHECK_IF() when I/O are performed on a orphan SC
When no endpoint is attached to a SC, it is unexpected to have I/O (receive
or send). But we honestly don't know if it happens or not. So a CHECK_IF()
is added to be able to track such calls.
MINOR: stconn: Simplify sc_abort/sc_shutdown by merging calls to se_shutdown
Calls to se_shutdown were no the same between applets and mux endpoints.
Only the SHUTW flag was not the same. However, on the multiplexers are
sensitive to the true SHUTW flag. The applets handle all of them the same
way. So calls to se_shutdown() from sc_abort() and sc_shutdown() can be
merged to always use the multiplexer version.
wake_srv_chk() function is now only used by srv_chk_io_cb(), the
health-checl I/O callback function. So let's remove it. The code of the
function was moved in srv_chk_io_cb().
MINOR: connection: Call sc_conn_process() instead of .wake() callback function
At we fail to create a mux, in conn_create_mux(), instead of calling the
app_ops .wake() callback function, we can directly call sc_conn_process().
At this stage, we know we are using an connection, so it is safe to do so.
MINOR: applet: Call sc_applet_process() instead of .wake() callback function
At the end of task_run_applet() and task_process_applet(), instead of
calling the app_ops .wake() callback function, we can directly call
sc_applet_process(). At this stage, we know we are using an applet, so it is
safe to do so.
MAJOR: muxes: No longer use app_ops .wake() callback function from muxes
Thanks to previous commits, it is now possible to wake the data layer up,
via a tasklet_wakeup, instead of using the app_ops .wake() callback
function.
When a data layer must be notified of a mux event (an error for instance),
we now always perform a tasklet_wakeup(). TASK_WOKEN_MSG state is used by
default. TASK_WOKEN_IO is eventually added if the data layer was subscribed
to receives or sends.
Changes are not trivial at all. We replaced a synchronous call to the
sc_conn_process() function by a tasklet_wakeup().
MINOR: muxes: Wakup the data layer from a mux stream with TASK_WOKEN_IO state
Now, when a mux stream is waking its data layer up for receives or sends, it
uses the TASK_WOKEN_IO state. The state is not used by the stconn I/O
callback function for now.
MINOR: mux-spop: Rely on spop_strm_notify_send() when resuming streams for sending
In spop_resume_each_sending_spop_strm(), there was exactly the same code
than spop_strm_notify_send(). So let's use spop_strm_notify_send() instead
of duplicating code.
MINOR: stconn: Call sc_conn_process from the I/O callback if TASK_WOKEN_MSG state was set
It is the first commit of a series to refactor the SC app_ops. The first
step is to remove the .wake() callback function from the app_ops to replace
all uses by a wakeup of the SC tasklet.
Here, when the SC is woken up, the state is now tested and if TASK_WOKEN_MSG
is set, sc_conn_process() is called.
MINOR: jwt: Manage ECDH-ES algorithm in jwt_decrypt_jwk function
When ECDH-ES algorithm is used in a JWE token, no cek is provided and
one must be built in order to decrypt the contents of the token. The
decrypting key is built by deriving a temporary key out of a public key
provided in the token and the private key provided by the user and
performing a concatKDF operation.
MINOR: jwt: Parse ec-specific fields in jose header
When the encoding is of the ECDH family, the optional "apu" and "apv"
fields of the JOSE header must be parsed, as well as the mandatory "epk"
field that contains an EC public key used to derive a key that allows
either to decrypt the contents of the token (in case of ECDH-ES) or to
decrypt the content encoding key (cek) when using ECDH-ES+AES Key Wrap.
Convert a JWK with the "EC" key type ("kty") into an EVP_PKEY. The JWK
can either represent a public key if it only contains the "x" and "y"
fields, or a private key if it also contains the "d" field.
The 'jwt_tokenize' function that can be used to split a JWT token into
its subparts can either fully process the token (from beginning to end)
when we need to check its signature, or only partially when using the
jwt_header_query or jwt_member_query converters. In this case we relied
on the fact that the return value of the 'jwt_tokenize' function was not
checked because a '-1' was returned (which was not actually an error).
In order to make this logic more explicit, the 'jwt_tokenize' function
now has a way to warn the caller that the token was invalid (less
subparts than the specified 'item_num') or that the token was not
processed in full (enough subparts found without parsing the token all
the way).
The function will now only return 0 if we found strictly the same number
of subparts as 'item_num'.
Willy Tarreau [Mon, 9 Mar 2026 14:20:35 +0000 (15:20 +0100)]
SCRIPTS: git-show-backports: add a restart-from-last option
It's always a bit tricky to avoid already backported patches when they
just got a different ID (e.g. a critical fix in a topic branch). Most
often with stable topic branches we just want to pick all stable commits
since the last backported one. New option -L instead of -m does exactly
this: it enumerates only commits that were added to the reference branch
after its most recent backport.
BUG/MINOR: admin: haproxy-reload rename -vv long option
The -vv option used --verbose as its long form, which was identical to
the long form of -v. Since the case statement matches top-to-bottom,
--verbose would always trigger -v (VERBOSE=2), making -vv unreachable
via its long option. The long form is renamed to --verbose=all to avoid
the conflict, and the usage string is updated accordingly.
MEDIUM: admin: haproxy-reload conversion to POSIX sh
The script relied on a bash-specific process substitution (< <(...)) to
feed socat's output into the read loop. This is replaced with a standard
POSIX pipe into a command group.
The response parsing is also simplified: instead of iterating over each
line with a while loop and echoing them individually, the status line is
read first, the "--" separator consumed, and the remaining output is
streamed to stderr or discarded as a whole depending on the verbosity
level.
Could be backported to 3.3 as it makes it more portable, but introduce a
slight change in the error format.
BUG/MINOR: admin: haproxy-reload use explicit socat address type
socat was used with the ${MASTER_SOCKET} variable directly, letting it
auto-detect the network protocol. However, when given a plain filename
that does not point to a UNIX socket, socat would create a file at that
path instead of reporting an error.
To fix this, the address type is now determined explicitly: if
MASTER_SOCKET points to an existing UNIX socket file (checked with -S),
UNIX-CONNECT: is used; if it matches a <host>:<port> pattern, TCP: is
used; otherwise an error is reported. The socat_addr variable is also
properly scoped as local to the reload() function.
CLEANUP: flt_http_comp: comp_state doesn't bother about the direction anymore
no need to have duplicated comp_ctx and comp_algo for request vs response
in comp_state struct, because thanks to previous commit compression filter
is either oriented on the request or the response, and 2 distinct filters
are instanciated when we need to handle both requests and responses
compression.
Thus we can save us from duplicated struct members and related operations.
MEDIUM: flt_http_comp: split "compression" filter in 2 distinct filters
Existing "compression" filter is a multi-purpose filter that will try
to compress both requests and responses according to "compression"
settings, such as "compression direction".
One of the pre-requisite work identified to implement decompression
filter is that we needed a way to manually define the sequence of
enabled filters to chain them in the proper order to make
compression and decompression chains work as expected in regard
to the intended use-case.
Due to the current nature of the "compression" filter this was not
possible, because the filter has a combined action as it will try
to compress both requests and responses, and as we are about to
implement "filter-sequence" directive, we will not be able to
change the order of execution of the compression filter between
requests and responses.
A possible solution we identified to solve this issue is to split the
existing "compression" filter into 2 distinct filters, one which is
request-oriented, "comp-req", and another one which is response-oriented
"comp-res". This is what we are doing in this commit. Compression logic
in itself is unchanged, "comp-req" will only aim to compress the request
while "comp-res" will try to compress the response. Both filters will
still be invoked on request and responses hooks, but they only do their
part of the job.
From now on, to compress both requests and responses, both filters have
to be enabled on the proxy. To preserve original behavior, the "compression"
filter is still supported, what it does is that it instantiates both
"comp-req" and "comp-res" filters implicitly, as the compression filter is
now effectively split into 2 separate filters under the hood.
When using "comp-res" and "comp-req" filters explicitly, the use of the
"compression direction" setting is not relevant anymore. Indeed, the
compression direction is assumed as soon as one or both filters are
enabled. Thus "compression direction" is kept as a legacy option in
order to configure the "compression" generic filter.
BUG/MINOR: backend: Don't get proto to use for webscoket if there is no server
In connect_server(), it is possible to have no server defined (dispatch mode
or transparent backend). In that case, we must be carefull to check the srv
variable in all calls involving the server. It was not perform at one place,
when the protocol to use for websocket is retrieved. This must not be done
when there is no server.
This patch should fix the first report in #3144. It must be backported to
all stable version.
BUG/MINOR: ssl-sample: Fix sample_conv_sha2() by checking EVP_Digest* failures
In sample_conv_sha2(), calls to EVP_Digest* can fail. So we must check
return value of each call and report a error on failure and release the
digest context.
This patch should fix the issue #3274. It should be backported as far as
2.6.
Willy Tarreau [Thu, 5 Mar 2026 14:55:28 +0000 (15:55 +0100)]
[RELEASE] Released version 3.4-dev6
Released version 3.4-dev6 with the following main changes :
- CLEANUP: acme: remove duplicate includes
- BUG/MINOR: proxy: detect strdup error on server auto SNI
- BUG/MINOR: server: set auto SNI for dynamic servers
- BUG/MINOR: server: enable no-check-sni-auto for dynamic servers
- MINOR: haterm: provide -b and -c options (RSA key size, ECDSA curves)
- MINOR: haterm: add long options for QUIC and TCP "bind" settings
- BUG/MINOR: haterm: missing allocation check in copy_argv()
- BUG/MINOR: quic: fix counters used on BE side
- MINOR: quic: add BUG_ON() on half_open_conn counter access from BE
- BUG/MINOR: quic/h3: display QUIC/H3 backend module on HTML stats
- BUG/MINOR: acme: acme_ctx_destroy() leaks auth->dns
- BUG/MINOR: acme: wrong labels logic always memprintf errmsg
- MINOR: ssl: clarify error reporting for unsupported keywords
- BUG/MINOR: acme: fix incorrect number of arguments allowed in config
- CLEANUP: haterm: remove unreachable labels hstream_add_data()
- CLEANUP: haterm: avoid static analyzer warnings about rand() use
- CLEANUP: ssl: Remove a useless variable from ssl_gen_x509()
- CI: use the latest docker for QUIC Interop
- CI: remove redundant "halog" compilation
- CLENAUP: cfgparse: accept-invalid-http-* does not support "no"/"defaults"
- BUG/MEDIUM: spoe: Acquire context buffer in applet before consuming a frame
- MINOR: traces: always mark trace_source as thread-aligned
- MINOR: ncbmbuf: improve itbmap_next() code
- MINOR: proxy: improve code when checking server name conflicts
- MINOR: quic: add a new metric for ncbuf failures
- BUG/MINOR: haterm: cannot reset default "haterm" mode
- BUG/MEDIUM: cpu-topo: Distribute CPUs fairly across groups
- BUG/MINOR: quic: missing app ops init during backend 0-RTT sessions
- CLEANUP: ssl: remove outdated comments
- MINOR: mux-h2: also count glitches on invalid trailers
- MINOR: mux-h2: add a new setting, "tune.h2.log-errors" to tweak error logging
- BUG/MEDIUM: mux-h2: make sure to always report pending errors to the stream
- BUG/MINOR: server: adjust initialization order for dynamic servers
- CLEANUP: tree-wide: drop a few useless null-checks before free()
- CLEANUP: quic-stats: include counters from quic_stats
- REORG: stats/counters: move extra_counters to counters not stats
- CLEANUP: stats: drop stats.h / stats-t.h where not needed
- MEDIUM: counters: change the fill_stats() API to pass the module and extra_counters
- CLEANUP: counters: only retrieve zeroes for unallocated extra_counters
- MEDIUM: counters: add a dedicated storage for extra_counters in various structs
- MINOR: counters: store a tgroup step for extra_counters to access multiple tgroups
- MEDIUM: counters: store the number of thread groups accessing extra_counters
- MINOR: counters: add EXTRA_COUNTERS_BASE() to retrieve extra_counters base storage
- MEDIUM: counters: return aggregate extra counters in ->fill_stats()
- MEDIUM: counters: make EXTRA_COUNTERS_GET() consider tgid
- BUG/MINOR: call EXTRA_COUNTERS_FREE() before srv_free_params() in srv_drop()
- MINOR: promex: test applet resume in stress mode
- BUG/MINOR: promex: fix server iteration when last server is deleted
- BUG/MINOR: proxy: add dynamic backend into ID tree
- MINOR: proxy: convert proxy flags to uint
- MINOR: server: refactor srv_detach()
- MINOR: proxy: define a basic "del backend" CLI
- MINOR: proxy: define proxy watcher member
- MINOR: stats: protect proxy iteration via watcher
- MINOR: promex: use watcher to iterate over backend instances
- MINOR: lua: use watcher for proxies iterator
- MINOR: proxy: add refcount to proxies
- MINOR: proxy: rename default refcount to avoid confusion
- MINOR: server: take proxy refcount when deleting a server
- MINOR: lua: handle proxy refcount
- MINOR: proxy: prevent backend removal when unsupported
- MINOR: proxy: prevent deletion of backend referenced by config elements
- MINOR: proxy: prevent backend deletion if server still exists in it
- MINOR: server: mark backend removal as forbidden if QUIC was used
- MINOR: cli: implement wait on be-removable
- MINOR: proxy: add comment for defaults_px_ref/unref_all()
- MEDIUM: proxy: add lock for global accesses during proxy free
- MEDIUM: proxy: add lock for global accesses during default free
- MINOR: proxy: use atomic ops for default proxy refcount
- MEDIUM: proxy: implement backend deletion
- REGTESTS: add a test on "del backend"
- REGTESTS: complete "del backend" with unnamed defaults ref free
- BUG/MINOR: hlua: fix return with push nil on proxy check
- BUG/MEDIUM: stream: Handle TASK_WOKEN_RES as a stream event
- MINOR: quic: use signed char type for ALPN manipulation
- MINOR: quic/h3: reorganize stream reject after MUX closure
- MINOR: mux-quic: add function for ALPN to app-ops conversion
- MEDIUM: quic/mux-quic: adjust app-ops install
- MINOR: quic: use server cache for ALPN on BE side
- BUG/MEDIUM: hpack: correctly deal with too large decoded numbers
- BUG/MAJOR: qpack: unchecked length passed to huffman decoder
- BUG/MINOR: qpack: fix 1-byte OOB read in qpack_decode_fs_pfx()
- BUG/MINOR: quic: fix OOB read in preferred_address transport parameter
- BUG/MEDIUM: qpack: correctly deal with too large decoded numbers
- BUG/MINOR: hlua: Properly enable/disable line receives from HTTP applet
- BUG/MEDIUM: hlua: Fix end of request detection when retrieving payload
- BUG/MINOR: hlua: Properly enable/disable receives for TCP applets
- MINOR: htx: Add a function to retrieve the HTTP version from a start-line
- MINOR: h1-htx: Reports non-HTTP version via dedicated flags
- BUG/MINOR: h1-htx: Be sure that H1 response version starts by "HTTP/"
- MINOR: http-ana: Save the message version in the http_msg structure
- MEDIUM: http-fetch: Rework how HTTP message version is retrieved
- MEDIUM: http-ana: Use the version of the opposite side for internal messages
- DEBUG: stream: Display the currently running rule in stream dump
- MINOR: filters: Use filter API as far as poissible to break loops on filters
- MINOR: filters: Set last_entity when a filter fails on stream_start callback
- MINOR: stream: Display the currently running filter per channel in stream dump
- DOC: config: Use the right alias for %B
- BUG/MINOR: channel: Increase the stconn bytes_in value in channel_add_input()
- BUG/MINOR: sample: Fix sample to retrieve the number of bytes received and sent
- BUG/MINOR: http-ana: Increment scf bytes_out value if an haproxy error is sent
- BUG/MAJOR: fcgi: Fix param decoding by properly checking its size
- BUG/MAJOR: resolvers: Properly lowered the names found in DNS response
- BUG/MEDIUM: mux-fcgi: Use a safe loop to resume each stream eligible for sending
- MINOR: mux-fcgi: Use a dedicated function to resume streams eligible for sending
- CLEANUP: qpack: simplify length checks in qpack_decode_fs()
- MINOR: counters: Introduce COUNTERS_UPDATE_MAX()
- MINOR: listeners: Update the frequency counters separately when needed
- MINOR: proxies: Update beconn separately
- MINOR: stats: Add an option to disable the calculation of max counters
MINOR: stats: Add an option to disable the calculation of max counters
Add a new option, "stats calculate-max-counters [on|off]".
It makes it possible to disable the calculation of max counters, as they
can have a performance cost.
Update beconn separately from the call to COUNTERS_UPDATE_MAX(), as soon
there will be an option to get COUNTERS_UPDATE_MAX() to do nothing, and
we still want beconn to be properly updated, as it is used for other
purposes.
MINOR: listeners: Update the frequency counters separately when needed
Update the frequency counters that are exported to the stats page
outside of the call to COUNTERS_UPDATE_MAX(), so that they will
happen even if COUNTERS_UPDATE_MAX() ends up doing nothing.
Introduce COUNTERS_UPDATE_MAX(), and use it instead of using
HA_ATOMIC_UPDATE_MAX() directly.
For now it just calls HA_ATOMIC_UPDATE_MAX(), but will later be modified
so that we can disable max calculation.
This can be backported up to 2.8 if the usage of COUNTERS_UPDATE_MAX()
generates too many conflicts.
CLEANUP: qpack: simplify length checks in qpack_decode_fs()
This patch simplifies the decoding loop by merging the variable-length
integer truncation check (len == -1) with the subsequent buffer
availability check (len < length).
This removes redundant code blocks and improves readability without
changing the decoding logic.
Note that the second removal is correct, as the check was duplicate and
unnecessary."
BUG/MEDIUM: mux-fcgi: Use a safe loop to resume each stream eligible for sending
At the end of fcgi_send(), if the connection is not full anymore, we loop on
the send list to resume FCGI stream for sending. But a streams may be
removed from the this list during the loop. So a safe loop must be used.
This patch should be backported to all stable versions.
BUG/MAJOR: resolvers: Properly lowered the names found in DNS response
Names found in DNS responses are lowered to be compared. A name is composed
of several labels, strings precedeed by their length on one byte. For
instance:
3www7haproxy3org
There is an bug when labels are lowered. The label length is not skipped and
tolower() function is called on it. So for label length in the range [65-90]
(uppercase char), 32 is added to the label length due to the conversion of a
uppercase char to lowercase. This bugs can lead to OOB read later in the
resolvers code.
The fix is quite obvious, the label length must be skipped when the label is
lowered.
Thank you to Kamil Frankowicz for having reported this.
This patch must be backported to all stable versions.
BUG/MAJOR: fcgi: Fix param decoding by properly checking its size
In functions used to decode a FCGI parameter, the test on the data length
before reading the parameter's name and value did not consider the offset
value used to skip already parsed data. So it was possible to read more data
than available (OOB read). To do so, a malicious FCGI server must send a
forged GET_VALUES_RESULT record containing a parameter with wrong name/value
length.
Thank you to Kamil Frankowicz for having reported this.
This patch must be backported to all stable versions.
BUG/MINOR: sample: Fix sample to retrieve the number of bytes received and sent
There was an issue in the if/else statement in smp_fetch_bytes() function.
When req.bytes_in or req.bytes_out was requested, res.bytes_in was always
returned. It is now fixed.
BUG/MINOR: channel: Increase the stconn bytes_in value in channel_add_input()
This function is no longer used. So it is not really an bug. But it is still
available and could be used by legacy applets. In that case, we must take
care to increment the stconn bytes_in value accordingly when input data are
inserted.
MINOR: stream: Display the currently running filter per channel in stream dump
Since the 3.1, when stream's info are dump, it is possible to print the
yielding filter on each channel, if any. It was useful to detect buggy
filter on spinning loop. But it is not possible to detect a filter consuming
too much CPU per-execution. We can see a filter was executing in the
backtrace reported by the watchdog, but we are unable to spot the specific
one.
Thanks to this patch, it is now possible. When a dump is emitted, the
running or yield filters on each channel are now displayed with their
current state (RUNNING or YIELDING).
This patch could be backported as far as 3.2 because it could be useful to
spot issues. But the filter API was slightly refactored in 3.4, so this
patch should be adapted.
MINOR: filters: Set last_entity when a filter fails on stream_start callback
On the stream, the last_entity should reference the last rule or the last
filter evaluated during the stream processing. However, this info was not
saved when a filter failed on strem_start callback function. It is now
fixed.
MINOR: filters: Use filter API as far as poissible to break loops on filters
When the filters API was refactored to improve loops on filters, some places
were not updated (or not fully updated). Some loops were not relying on
resume_filter_list_break() while it was possible. So let's do so with this
patch.
DEBUG: stream: Display the currently running rule in stream dump
Since the 2.5, when stream's info are dump, it is possible to print the
yielding rule, if any. It was useful to detect buggy rules on spinning
loop. But it is not possible to detect a rule consuming too much CPU
per-execution. We can see a rule was executing in the backtrace reported by
the watchdog, but we are unable to spot the specific rule.
Thanks to this patch, it is now possible. When a dump is emitted, the
running or yield rule is now displayed with its current state (RUNNING or
YIELDING).
This patch could be backported as far as 3.2 because it could be useful to
spot issues.
MEDIUM: http-ana: Use the version of the opposite side for internal messages
When the response is send by HAProxy, from a applet or for the analyzers,
The request version is used for the response. The main reason is that there
is not real version for the response when this happens. "HTTP/1.1" is used,
but it is in fact just an HTX response. So the version of the request is
used.
In the same manner, when the request is sent from an applet (httpclient),
the response version is used, once available.
The purpose of this change is to return the most accurate version from the
user point of view.
MEDIUM: http-fetch: Rework how HTTP message version is retrieved
Thanks to previous patches, we can now rely on the version stored in the
http_msg structure to get the request or the response version.
"req.ver" and "res.ver" sample fetch functions returns the string
representation of the version, without the prefix, so "<major>.<minor>", but
only if the version is valid. For the response, "res.ver" may be added from
a health-check context, in that case, the HTX message is used.
"capture.req.ver" and "capture.res.ver" does the same but the "HTTP/" prefix
is added to the result. And "capture.res.ver" cannot be called from a
health-check.
To ease the version formatting and avoid code duplication, an helper
function was added. So these samples are now relying on "get_msg_version()".
MINOR: http-ana: Save the message version in the http_msg structure
When the request or the response is received, the numerical value of the
message version is now saved. To do so, the field "vsn" was added in the
http_msg structure. It is an unsigned char. The 4 MSB bits are used for the
major digit and the 4 LSB bits for the minor one.
Of couse, the version must be valid. the HTX_SL_F_NOT_HTTP flag of the
start-line is used to be sure the version is valid. But because this flag is
quite new, we also take care the string representation of the version is 8
bytes length. 0 means the version is not valid.
BUG/MINOR: h1-htx: Be sure that H1 response version starts by "HTTP/"
When the response is parsed, we test the version to be sure it is
valid. However, the protocol was not tested. Now we take care that the
response version starts by "HTTP/", otherwise an error is returned.
Of course, it is still possible to by-pass this test with
"accept-unsafe-violations-in-http-response" option.
This patch could be backported to all stable versions.
MINOR: h1-htx: Reports non-HTTP version via dedicated flags
Now, when the HTTP version format is not strictly valid, flags are set on
the h1 parser and the HTX start-line. H1_MF_NOT_HTTP is set on the H1 parser
and HTX_SL_F_NOT_HTTP is set on the HTX start-line. These flags were
introduced to avoid parsing again and again the version to know if it is a
valid version or not, escpecially because it is most of time valid.
MINOR: htx: Add a function to retrieve the HTTP version from a start-line
htx_sl_vsn() function can now be used to retrieve the ist string
representing the HTTP version from a start-line passed as parameter. This
function takes care to return the right part of the start-line, depending on
its type (request or response).