Posts by Verdeoro

webif: fix mismatched HTML tags in reader and config templates

readerconfig.html / readerconfigaio.html:
  - Saveemm/Blockemm rows opened <TR> but only closed inner </TD>,
    relying on browser auto-close at the next <TR>. Add explicit
    </TR> after each </TD>.

config/dvbapi.html:
  - Four checkbox rows (Enabled, AU, Enable ecm.info file,
    Write detected prov name to srvid) had <TR><TD>...<label></label>
    without the closing </TD></TR>. Complete the row markup.

config/gbox.html:
  - Line 46: missing </A> in 'Message type:' label
    (<A ...>Message type:</TD> -> Message type:</A></TD>).
  - Line 71: stray </TR> directly after a fully inline
    <TR ...></TR> in the previous line; remove it.

All affected files now have balanced <TR>, <TD> and <A> tags.

*11958

webif: include WITH_SSL in is_defined.txt when USE_SSL flag is set

The webif template generator filters template blocks based on
webif/is_defined.txt, which config.sh's write_enabled() builds. With
WITH_SSL no longer in the addons list, it was never written to that
file even when USE_SSL=1 is passed to make (or -DWITH_SSL=1 to cmake,
which propagates as USE_SSL via --use-flags).

Effect: the SSL Settings block on the WebIf config page disappeared.

Mirror the existing USE_PCSC -> CARDREADER_PCSC handling and emit
WITH_SSL when USE_SSL is in USE_FLAGS.
--------------------------------------------
*11957

cscrypt: fix DES symbol collision in mdc2 header when WITH_SSL is on

Including <openssl/des.h> from cscrypt/mdc2.h pulled OpenSSL's DES
backward-compatibility macros (e.g. des_ecb_encrypt -> DES_ecb_encrypt
with different argument counts) into every translation unit that uses
mdc2.h. tests.c includes both mdc2.h and cscrypt/des.h, which made the
test build fail with 'macro requires N arguments, but only M given'
errors on toolchains shipping OpenSSL 1.0.x.

The header only needed DES_cblock for two MDC2_CTX struct fields, both
of which are simply 8-byte buffers. Replace them with unsigned char[8]
and move the OpenSSL include (and the local DES typedef fallback) into
mdc2.c, where the DES API is actually used.
---------------------------------------------
*11956

build: remove obsolete WITH_SSL config entry

WITH_SSL was added to config.sh in 2010 (commit 6c2b2c49) when there was
no separation between make and cmake. Today the convention is:
  - make uses USE_SSL=1
  - cmake uses -DWITH_SSL=1

WITH_SSL in config.sh's addons list violated this and only acted as a
config-file fallback that translated to USE_SSL in the Makefile. Both
build systems work fine without it.

Removed:
  - WITH_SSL from addons list and defconfig in config.sh
  - WITH_SIGNING -> WITH_SSL auto-dependency (Makefile and CMakeLists
    already handle this on their own)
  - WITH_SSL menuconfig entry and help examples
  - config.sh --enabled WITH_SSL fallback in Makefile
  - CONFIG_WITH_SSL fallback logic in CMakeLists.txt

The GitLab CI pipeline previously passed the make-style flag -DUSE_SSL
to cmake, which cmake silently ignored — the dropped config.sh fallback
used to paper over this. Switch the cmake invocations to -DWITH_SSL=1
(the variable cmake actually reads) and drop the obsolete
DISABLE_OPT WITH_SSL entries since WITH_SSL is no longer a config.sh
option.

tests: gate run_aes_key_list_tests with READER_VIACCESS

The AES key-list helpers it exercises — add_aes_entry,
aes_clear_entries, aes_decrypt_from_list, aes_present — are
defined in oscam-aes.c under #ifdef READER_VIACCESS, since
Viaccess is the only consumer. Calling them unconditionally
from run_crypto_tests() makes tests.bin fail to link on every
build without the Viaccess reader:

  oscam-aes.c: undefined reference to add_aes_entry
                                       aes_clear_entries
                                       aes_decrypt_from_list
                                       aes_present

Reproduces e.g. on the cmake test build for the VuZero4k target
with MODULE_CCCAM enabled (so WITH_LIB_AES is on) but no readers.

Wrap both the run_aes_key_list_tests() definition and its
caller in run_crypto_tests() in #ifdef READER_VIACCESS, matching
the gating of the helpers themselves.

tests: enable reliable standalone test builds and proper CI failure reporting

* build: keep test binaries unsigned

Skip the link-stage signing step when BUILD_TESTS is enabled in the
Makefile. The tests target already avoids strip/sign/upx when copying
tests.bin from tests.bin.debug, but tests.bin.debug itself was still
passed through SIGN_COMMAND_OSCAM. With signing enabled this could still
reassemble stale UPX split files into the test binary before the copy
step, reintroducing the failure the tests build path was meant to avoid.

* build: add BUILD_TESTS support to CMake and skip strip/sign/upx

Makefile: SIGN_COMMAND_OSCAM reassembles any leftover UPX split parts
(upx.aa / upx.ab from a previous build) onto the current target — on
the second invocation this re-creates a UPX-compressed
tests.bin.debug, so the subsequent strip fails with 'input file has no
sections' and aborts the build. The test binary does not need to be
stripped, signed or compressed. Skip all three steps when
BUILD_TESTS=1 and just copy .debug → bin.

CMakeLists.txt previously had no tests target at all — 'make tests'
only worked via the Makefile. Add a BUILD_TESTS option (OFF by
default) that mirrors the Makefile path:
  - pulls tests.c into the csoscam library
  - adds -DBUILD_TESTS=1 to the global compile definitions
  - renames the executable to tests.bin and links it without the
    strip/sign/upx POST_BUILD commands
  - exposes a 'run-tests' custom target that builds the binary and
    executes it in the build tree

Standard usage:
    mkdir build-tests && cd build-tests
    cmake -DBUILD_TESTS=ON -DUSE_SSL=1 -DUSE_LIBCRYPTO=1 ..
    make run-tests

* tests: add crypto test vectors

Extend run_all_tests() with known-answer tests against the crypto API
surface oscam uses:

  - MD5        RFC 1321 Appendix A.5 + __md5_crypt ($1$ WebIf hashes)
  - SHA1       FIPS 180-1 (short vectors, one-shot and streaming)
  - SHA256     FIPS 180-2 + 1M 'a' long-stream via mbedtls_sha256
  - MDC2       OpenSSL-1.0 mdc2test.c KAT + streaming split
  - AES ECB    NIST SP 800-38A F.1.1/F.1.3/F.1.5 (AES-128/192/256)
  - AES CBC    NIST SP 800-38A F.2.1/F.2.5, single + chunked (IV chain)
  - AesCtx     cscrypt/fast_aes ECB + CBC
  - AES void*  oscam-aes.c wrappers (aes_cbc_encrypt/decrypt,
               aes_encrypt_idx, aes_decrypt)
  - AES DB     aes_decrypt_from_list / aes_present round-trip
  - IDEA       Ascom KAT + CBC reversal
  - RC6        RC6 paper all-zero KAT
  - DES        ECB (Stallings) + single-block via schedule
  - DES-CBC    verified against openssl -des-cbc
  - 3DES-EDE2-CBC KAT verified against openssl -des-ede-cbc
  - 3DES-ECB3  reversal
  - BN         mod_exp, bn2bin, num_bytes, mod_inverse, mul,
               add_word, sub_word, cmp, copy
  - DES parity helpers (des_set_odd_parity / _all)

Exit status reflects failure count so CI can gate on it.

* tests: return failure count and exit with correct status

The test binary previously always exited with status 0, regardless of
whether any tests failed. This made it useless in CI: a regression would
print errors to stdout but the process would still report success.

This commit fixes that end-to-end, and addresses several related issues
found while doing so.

Exit status
-----------
run_all_tests() now returns int (total failure count) instead of void.
run_tests() in oscam.c passes that value to exit():

    exit(run_all_tests() ? EXIT_FAILURE : EXIT_SUCCESS);

A non-zero failure count now produces a non-zero exit code. A summary
line is printed at the end of the run:

    Summary: N failure(s)

Type-safe adapter layer
-----------------------
The test_type struct previously stored function pointers cast to generic
types (CHK_FN *, MK_T_FN *, etc.) in order to accommodate different
concrete data types. Calling a function through a mismatched pointer
type is undefined behavior in C, even when the signatures are similar.

This is replaced with a DEFINE_TEST_ADAPTERS macro that generates a set
of correctly-typed wrapper functions for each data type:

    DEFINE_TEST_ADAPTERS(ecm_whitelist, ECM_WHITELIST, ewnum);
    DEFINE_TEST_ADAPTERS(ecm_hdr_whitelist, ECM_HDR_WHITELIST, ehnum);
    ...

Each expansion produces chk_*, mk_t_*, clear_*, clone_*, and has_data_*
adapters that cast the void * argument to the correct concrete type and
forward the call. The struct fields now hold pointers to these adapters
with no casts required.

Clone failure detection
-----------------------
A new HAS_DATA_FN callback is added to test_type. It reports whether a
parsed data structure contains any entries (by checking the count field
named in the macro, e.g. ewnum, ehnum). This enables a meaningful
distinction in the test loop:

- If parsing produced data and clone_fn returns false, that is a clone
  error and is reported as such with a dedicated "=== CLONE ERROR ==="
  message.
- If parsing produced no data and clone returns false, that is expected
  behavior and is not counted as a failure.

Previously a failed clone would silently fall through, potentially
producing misleading output from stale data.

Co-authored-by: default avatarWXbet <4-WXbet@users.noreply.git.streamboard.tv>

*11952

cmake: switch library detection from auto-detect to opt-in

* cmake: declare STATIC_* and LIBCRYPTO_LIB/SSL_LIB as cache variables

When a user passes -DSTATIC_PCSC=0 while HAVE_PCSC is also 0 (or
vice-versa for libcrypto/libssl/libusb/libdvbcsa), cmake prints:

  CMake Warning:
    Manually-specified variables were not used by the project:
      STATIC_PCSC

The warning is harmless but confusing: the STATIC_* flag is only
read inside if (HAVE_*) blocks, so when the feature is disabled
the -D value is indeed unreferenced.

Declare the flags via option() (and LIBCRYPTO_LIB/SSL_LIB via
set(... CACHE STRING ...)) at the top of the file so cmake
considers them known even when the feature guard is inactive.
Placed before system detection so they don't override the
macOS-specific set(STATIC_LIBUSB True).

* cmake: add LIBCRYPTO_LIB and SSL_LIB overrides for Makefile parity

The Makefile allows passing a full library path to override the
default detection:

  make USE_LIBCRYPTO=1 LIBCRYPTO_LIB=/usr/lib/libcrypto.a
  make USE_SSL=1 SSL_LIB=/usr/lib/libssl.a

Mirror this in cmake by honouring LIBCRYPTO_LIB and SSL_LIB when
set. The supplied path is used as-is for OPENSSL_CRYPTO_LIBRARIES
and OPENSSL_SSL_LIBRARIES, bypassing the static auto-detection.

Both static and dynamic archives work. Existing flags
(STATIC_LIBCRYPTO, STATIC_SSL, OPENSSL_*_LIBRARY) keep working.

* cmake: simplify libusb and libdvbcsa status summary

Apply the same pattern used for PCSC to libusb and libdvbcsa:
report the final resolved link mode instead of mixing request
flags with result flags.

  LIBUSBDIR/LIBDVBCSADIR -> custom location selected
  STATICLIBUSB/STATICLIBDVBCSA -> static linking was selected

Removes the redundant "You selected to enable static" messages
(the result message already covers that case) and the duplicate
LIBUSBDIR branches.

* cmake: simplify PCSC status summary

Restructure the PCSC summary block to describe the final resolved
link mode instead of mixing request flags with result flags:

  PCSCDIR     -> user selected a custom PCSC location
  STATICPCSC  -> static linking was actually selected

STATIC_PCSC (user request) and STATICPCSC (actual result) are not
always the same, e.g. when libpcsclite.a cannot be found despite
STATIC_PCSC=1. Using only STATICPCSC in the summary keeps the
status message aligned with the real build result.

* cmake: add static libcrypto and libssl support

Add STATIC_LIBCRYPTO and STATIC_SSL flags to independently control
static linking of libcrypto and libssl, matching the Makefile where
LIBCRYPTO_LIB and SSL_LIB can be set to .a paths separately.

When either flag is set, OPENSSL_USE_STATIC_LIBS is enabled for
find_package(OpenSSL), then find_library locates the static archives
relative to the OpenSSL include directory. The native CMake variables
OPENSSL_CRYPTO_LIBRARY and OPENSSL_SSL_LIBRARY can also be used to
pass explicit paths directly.

Usage:
  cmake -DHAVE_LIBCRYPTO=1 -DSTATIC_LIBCRYPTO=1 ..
  cmake -DHAVE_LIBCRYPTO=1 -DSTATIC_SSL=1 ..
  cmake -DHAVE_LIBCRYPTO=1 -DSTATIC_LIBCRYPTO=1 -DSTATIC_SSL=1 ..
  cmake -DHAVE_LIBCRYPTO=1 -DOPENSSL_CRYPTO_LIBRARY=/pfad/libcrypto.a ..

* cmake: add static PCSC and custom PCSCDIR support

Add PCSCDIR and STATIC_PCSC support for PCSC linking, bringing it
on par with libusb (LIBUSBDIR/STATIC_LIBUSB) and libdvbcsa
(LIBDVBCSADIR). Previously cmake could only link PCSC dynamically
with the bare -lpcsclite flag.

PCSCDIR uses find_library with IMPORTED targets for both static
and dynamic linking. Without PCSCDIR, -DSTATIC_PCSC=1 searches
the system for libpcsclite.a and falls back to dynamic if not found.

Usage:
  cmake -DHAVE_PCSC=1 ..
  cmake -DHAVE_PCSC=1 -DSTATIC_PCSC=1 ..
  cmake -DHAVE_PCSC=1 -DPCSCDIR=/opt/pcsc ..
  cmake -DHAVE_PCSC=1 -DPCSCDIR=/opt/pcsc -DSTATIC_PCSC=1 ..

* cmake: switch library detection from auto-detect to opt-in

Libraries like pcsclite, libusb and OpenSSL were auto-detected via
check_include_file, causing unwanted dynamic linking when the toolchain
sysroot contained these libraries. This changes cmake to match the
Makefile behavior where libraries must be explicitly requested.

Use -DHAVE_PCSC=1, -DHAVE_LIBUSB=1 or -DHAVE_LIBCRYPTO=1 to enable.
OpenSSL/libcrypto is auto-enabled when WITH_SSL is active in config.

Wrap the static library link step in --start-group/--end-group
(GNU ld) to resolve circular dependencies between csoscam, csmodules
and csreaders. Since 28f2598b (cleanup reader macros) the
READER_VIACCESS ifdef in oscam-aes.c caused oscam-aes.o to not be
pulled from libcsoscam.a when READER_VIACCESS was disabled, leaving
aes_encrypt_idx/aes_decrypt/aes_set_key_alloc unresolved for modules
like camd35 and camd33. Simple reordering is not possible since
csoscam and csmodules have mutual dependencies. macOS is excluded
as Apple ld rescans archives automatically.
--------------------------------------------------
*11951

net: remove obsolete gethostbyname resolver and use getaddrinfo exclusively

gethostbyname() is not thread-safe and has been marked obsolescent by
POSIX. Newer glibc versions could emit a linker warning for any binary
referencing this symbol, even if the code path is gated behind a
runtime config option.

The modern replacement getaddrinfo() was already fully implemented as
the default resolver (resolvegethostbyname=0). This removes the legacy
gethostbyname() fallback path along with:

- the resolvegethostbyname config option (oscam.conf [global])
- the Resolver dropdown from the webinterface
- the gethostbyname_lock mutex (no longer needed since getaddrinfo
  is thread-safe)