Don't install MD5.3 for now because it is obsolete (conflicts with md5.3 on
case insensitive filesystems). Perhaps we should rename it openssl_MD5.3
and install it as such.
arm/imx23_olinuxino: fix audio playback warnings
We now get a signal instead of kasserts and dmesg spam, but you might want to protect your ears before listening to the output.
Import OpenSSL-3.5.5 (previous was 3.5.1)
Changes between 3.5.4 and 3.5.5 [27 Jan 2026]
Fixed Improper validation of PBMAC1 parameters in PKCS#12 MAC verification.
Severity: Moderate
Issue summary: PBMAC1 parameters in PKCS#12 files are missing
validation which can trigger a stack-based buffer overflow, invalid
pointer or NULL pointer dereference during MAC verification.
Impact summary: The stack buffer overflow or NULL pointer dereference
may cause a crash leading to Denial of Service for an application
that parses untrusted PKCS#12 files. The buffer overflow may also
potentially enable code execution depending on platform mitigations.
Reported by: Stanislav Fort (Aisle Research) and Petr #ime#ek (Aisle
Research) and Hamza (Metadust)
[346 lines not shown]
Pull up following revision(s) (requested by riastradh in ticket #2000):
external/cddl/osnet/dev/sdt/sdt.c: revision 1.25
dtrace_sdt: Don't ignore the first stack frame in sdt probes.
The aframes argument to dtrace_probe_create gives the number of stack
frames that are ignored in stack(), the first frame of which is
`caller'.
This is for, e.g., profiler interrupts where the interrupt frame
itself isn't interesting; it's the code that was interrupted that was
interesting (hence profile_aframes = PROF_ARTIFICIAL_FRAMES is often
nonzero, or even several frames deep).
But for sdt, the direct caller is the interesting part -- when we
have something like:
foo()
[16 lines not shown]
Pull up following revision(s) (requested by riastradh in ticket #1229):
external/cddl/osnet/dev/sdt/sdt.c: revision 1.25
dtrace_sdt: Don't ignore the first stack frame in sdt probes.
The aframes argument to dtrace_probe_create gives the number of stack
frames that are ignored in stack(), the first frame of which is
`caller'.
This is for, e.g., profiler interrupts where the interrupt frame
itself isn't interesting; it's the code that was interrupted that was
interesting (hence profile_aframes = PROF_ARTIFICIAL_FRAMES is often
nonzero, or even several frames deep).
But for sdt, the direct caller is the interesting part -- when we
have something like:
foo()
[16 lines not shown]
