[ViewVC] Diff of: src/trunk/UPDATING

Comparing trunk/UPDATING (file contents):
Revision 7338 by laffer1, Wed Sep 30 13:11:05 2015 UTC vs.
Revision 9173 by laffer1, Sat Oct 15 13:48:05 2016 UTC

+Updating Information for MidnightBSD users.
-+
+20161015:
-+
+        libarchive 3.2.1
-+
-+
+20161013:
-+
+        Sync ZFS code with Illuminos/FreeBSD 9.2. Added support for
-+
+        feature flags, pool version 5000. This also includes some
-+
+        bug fixes and performance optimizations.
-+
-+
+20160925:
-+
+        Import NetBSD vis(3) and unvis(3) as well as mtree.
-+
-+
+        one-true-awk 20121220
-+
-+
+        inetd now honors kern.ipc.somaxconn value.
-+
-+
+        netmap synced with FreeBSD 9.2
-+
-+
+        linuxolator now has dtrace probes.
-+
-+
+        bsdgrep now correctly handles -m to exclude only one file.
-+
-+
+        UFS file systems can now be resized in read-write mode due to the new
-+
+        write suspension feature.
-+
-+
+        Basic support added for Intel Raid Recover Technology.
-+
-+
+        GMIRROR & GRAID3 now mark volumes clean on shutdown earlier to help with ZFS issues.
-+
-+
+        Highpoint hpt27xx now in GENERIC kernel.
-+
-+
+20160923:
-+
+        Security update for OpenSSL
-+
-+
+        A malicious client can send an excessively large OCSP Status Request extension.
-+
+        If that client continually requests renegotiation, sending a large OCSP Status
-+
+        Request extension each time, then there will be unbounded memory growth on the
-+
+        server. [CVE-2016-6304]
-+
-+
+        An overflow can occur in MDC2_Update() either if called directly or through
-+
+        the EVP_DigestUpdate() function using MDC2. If an attacker is able to supply
-+
+        very large amounts of input data after a previous call to EVP_EncryptUpdate()
-+
+        with a partial block then a length check can overflow resulting in a heap
-+
+        corruption. [CVE-2016-6303]
-+
-+
+        If a server uses SHA512 for TLS session ticket HMAC it is vulnerable to a
-+
+        DoS attack where a malformed ticket will result in an OOB read which will
-+
+        ultimately crash. [CVE-2016-6302]
-+
-+
+        The function BN_bn2dec() does not check the return value of BN_div_word().
-+
+        This can cause an OOB write if an application uses this function with an
-+
+        overly large BIGNUM. This could be a problem if an overly large certificate
-+
+        or CRL is printed out from an untrusted source. TLS is not affected because
-+
+        record limits will reject an oversized certificate before it is parsed.
-+
+        [CVE-2016-2182]
-+
-+
+        The function TS_OBJ_print_bio() misuses OBJ_obj2txt(): the return value is
-+
+        the total length the OID text representation would use and not the amount
-+
+        of data written. This will result in OOB reads when large OIDs are presented.
-+
+        [CVE-2016-2180]
-+
-+
+        Some calculations of limits in OpenSSL have used undefined pointer arithmetic.
-+
+        This could cause problems with some malloc implementations. [CVE-2016-2177]
-+
-+
+        Operations in the DSA signing algorithm should run in constant time in order to
-+
+        avoid side channel attacks. A flaw in the OpenSSL DSA implementation means that
-+
+        a non-constant time codepath is followed for certain operations. [CVE-2016-2178]
-+
-+
+        In a DTLS connection where handshake messages are delivered out-of-order those
-+
+        messages that OpenSSL is not yet ready to process will be buffered for later
-+
+        use. Under certain circumstances, a flaw in the logic means that those messages
-+
+        do not get removed from the buffer even though the handshake has been completed.
-+
+        An attacker could force up to approx. 15 messages to remain in the buffer when
-+
+        they are no longer required. These messages will be cleared when the DTLS
-+
+        connection is closed. The default maximum size for a message is 100k. Therefore
-+
+        the attacker could force an additional 1500k to be consumed per connection.
-+
+        [CVE-2016-2179]
-+
-+
+        A flaw in the DTLS replay attack protection mechanism means that records that
-+
+        arrive for future epochs update the replay protection "window" before the MAC
-+
+        for the record has been validated. This could be exploited by an attacker by
-+
+        sending a record for the next epoch (which does not have to decrypt or have a
-+
+        valid MAC), with a very large sequence number. This means that all subsequent
-+
+        legitimate packets are dropped causing a denial of service for a specific
-+
+        DTLS connection. [CVE-2016-2181]
-+
-+
+        In OpenSSL 1.0.2 and earlier some missing message length checks can result in
-+
+        OOB reads of up to 2 bytes beyond an allocated buffer. There is a theoretical
-+
+        DoS risk but this has not been observed in practice on common platforms.
-+
+        [CVE-2016-6306]
-+
-+
+20160918:
-+
+        With the addition of auditdistd(8), a new auditdistd user is now
-+
+        depended on during installworld.  "mergemaster -p" can be used to add
-+
+        the user prior to installworld.
-+
-+
+        The VFS KBI was changed with the merge of several nullfs
-+
+        optimizations and fixes.  All filesystem modules must be
-+
+        recompiled.
-+
-+
+20160916:
-+
+        The random(4) support for the VIA hardware random number
-+
+        generator (`PADLOCK') is no longer enabled unconditionally.
-+
+        Add the PADLOCK_RNG option in the custom kernel config if
-+
+        needed.  The GENERIC kernels on i386 and amd64 do include the
-+
+        option, so the change only affects the custom kernel
-+
+        configurations.
-+
-+
+        A new version of ZFS (pool version 5000) has been merged.
-+
+        Starting with this version the old system of ZFS pool versioning
-+
+        is superseded by "feature flags". This concept enables forward
-+
+        compatibility against certain future changes in functionality of ZFS
-+
+        pools. The first two read-only compatible "feature flags" for ZFS
-+
+        pools are "com.delphix:async_destroy" and "com.delphix:empty_bpobj".
-+
+        For more information read the new zpool-features(7) manual page.
-+
+        Please refer to the "ZFS notes" section of this file for information
-+
+        on upgrading boot ZFS pools.
-+
-+
+20160906:
-+
+        Add support for the MosChip MCS9904 four serial ports
-+
+        controller.
-+
-+
+        Add support for walltimestamp in DTrace.
-+
-+
+        Various gdb improvments.
-+
-+
+        ZFS
-+
+        Import the zio nop-write improvement from Illumos. To reduce I/O,
-+
+        nop-write omits overwriting data if the checksum (cryptographically
-+
+        secure) of new data matches the checksum of existing data.
-+
+        It also saves space if snapshots are in use.
-+
-+
+        It currently works only on datasets with enabled compression, disabled
-+
+        deduplication and sha256 checksums.
-+
-+
+        Add loader(8) tunable to enable/disable nopwrite functionality:
-+
+        vfs.zfs.nopwrite_enabled
-+
-+
+        Introduce a new dataset aclmode setting "restricted" to protect ACL's
-+
+        being destroyed or corrupted by a drive-by chmod.
-+
-+
+        New loader-only tunables:
-+
+        vfs.zfs.sync_pass_deferred_free
-+
+        vfs.zfs.sync_pass_dont_compress
-+
+        vfs.zfs.sync_pass_rewrite
-+
-+
+        chkgrp(8) add support for q flag
-+
-+
+        Fix problem with the Samsung 840 PRO series SSD detection.
-+
+        The device reports support for SATA Asynchronous Notification in its
-+
+        IDENTIFY data, but returns error on attempt to enable that feature.
-+
+        Make SATA XPT of CAM only report these errors, but not fail the device.
-+
-+
+20160905:
-+
+        Add a resource limit for the total number of kqueues
-+
+        available to the user. Kqueue now saves the ucred of the
-+
+        allocating thread, to correctly decrement the counter on close.
-+
+        Based on FreeBSD SVN 256849
-+
-+
+        Import netcat from OpenBSD 5.2
-+
-+
+20160904:
-+
+        Introduced experimental TCP sysctls starting with
-+
+        net.inet.tcp.experimental.initcwnd10
-+
-+
+20160814:
-+
+        switched default desktop port to midnightbsd-desktop. This gives us flexibility to change it
-+
+        in the release after the fact.
-+
-+
+        tzdata 2016a
-+
-+
+20160811:
-+
+        libdispatch 210
-+
-+
+        Added quirks for several models of SSDs to enable advanced format/4k mode. List includes
-+
+        Samsung 830, 840, 850 and 750 series, Intel x25 and a few Toshiba models. Also
-+
+        added WD Red drives.
-+
-+
+        Updated list of pci device vendors.
-+
-+
+        Updated list of usb devices.
-+
-+
+20160807:
-+
+        Implement several changes to libmport to fix some memory corruption issues.
-+
-+
+20160806:
-+
+        sqlite3 3.13.0
-+
-+
+20160805:
-+
+        Merged fixes for libmport that improve error handling when installing packages. Also
-+
+        support mkdir -p like behavior for plist entries.
-+
-+
+20160531:
-+
+        Fix four security issues with MidnightBSD.
-+
-+
+        The implementation of TIOCGSERIAL ioctl(2) does not clear the output
-+
+        struct before sending to userland in the linux emulation layer.
-+
-+
+        The compat 43 stat(2) system call exposes kernel stack to userland.
-+
-+
+        libarchive - CVE-2015-2304 and CVE-2013-0211 fix issues with
-+
+        cpio directory traversal and an integer signedness error in the archive
-+
+        write zip data routine.
-+
-+
-+
+20160528:
-+
+        Fixed minor issues with mined(1) and msearch(1).
-+
-+
+20160526:
-+
+        Add support for Ivybridge and Haswell Intel CPUs to hwpmc(4).
-+
-+
+        Fix libpmc(3) build with clang compiler.
-+
-+
+20160519:
-+
+        Kernel Security updates
-+
-+
+        atkbd(4) - Incorrect signedness comparison in the ioctl(2) handler allows a malicious
-+
+        local user to overwrite a portion of the kernel memory.
-+
-+
+        Incorrect argument handling in sendmsg(2)
-+
-+
+        Incorrect argument handling in the socket code allows malicious local
-+
+        user to overwrite large portion of the kernel memory.
-+
-+
+20160505:
-+
+        OpenSSL security patch
-+
-+
+        The padding check in AES-NI CBC MAC was rewritten to be in constant time
-+
+        by making sure that always the same bytes are read and compared against
-+
+        either the MAC or padding bytes. But it no longer checked that there was
-+
+        enough data to have both the MAC and padding bytes. [CVE-2016-2107]
-+
-+
+        An overflow can occur in the EVP_EncodeUpdate() function which is used for
-+
+        Base64 encoding of binary data. [CVE-2016-2105]
-+
-+
+        An overflow can occur in the EVP_EncryptUpdate() function, however it is
-+
+        believed that there can be no overflows in internal code due to this problem.
-+
+        [CVE-2016-2106]
-+
-+
+        When ASN.1 data is read from a BIO using functions such as d2i_CMS_bio()
-+
+        a short invalid encoding can casuse allocation of large amounts of memory
-+
+        potentially consuming excessive resources or exhausting memory.
-+
+        [CVE-2016-2109]
-+
-+
+20160412:
-+
+.8 stable branch created. Continue development as 0.9.
-+
-+
+        Fix several issues with wait6 system call addition.
-+
-+
+20160409:
-+
+        libmport now supports two new plist formats:
-+
+        @(root,wheel,4775) myfile
-+
+        @dir(root,wheel,775) mydir
-+
-+
+        On delete, absoluate paths are now handled properly.
-+
-+
+20160317:
-+
+        OpenSSH doesn't have the luck of the Irish.
-+
-+
+        Fix a security issue with OpenSSH X11 forwarding that can allow an attacker
-+
+        run shell commands on the call to xauth.
-+
-+
+        Incorrect argument validation in sysarch(2)
-+
-+
+        A special combination of sysarch(2) arguments, specify a request to
-+
+        uninstall a set of descriptors from the LDT.  The start descriptor
-+
+        is cleared and the number of descriptors are provided.  Due to invalid
-+
+        use of a signed intermediate value in the bounds checking during argument
-+
+        validity verification, unbound zero'ing of the process LDT and adjacent
-+
+        memory can be initiated from usermode.
-+
-+
+        Patch obtained from FreeBSD.
-+
-+
+20160229:
-+
+        top now displays information on ZFS arc cache.
-+
-+
+20160228:
-+
+        llvm + clang 3.3 is now the default compiler in MidnightBSD.
-+
-+
+20160222:
-+
+        Introduce pipe2 to linux emulation layer.
-+
-+
+20160114:
-+
+        OpenSSL
-+
-+
+        The signature verification routines will crash with a NULL pointer dereference
-+
+        if presented with an ASN.1 signature using the RSA PSS algorithm and absent
-+
+        mask generation function parameter. [CVE-2015-3194]
-+
-+
+        When presented with a malformed X509_ATTRIBUTE structure, OpenSSL will leak
-+
+        memory. [CVE-2015-3195]
-+
-+
+        If PSK identity hints are received by a multi-threaded client then the values
-+
+        are incorrectly updated in the parent SSL_CTX structure.  [CVE-2015-3196]
-+
-+
+        Fix security on bsnmpd configuration file during installation.
-+
-+
+        TCP MD5 signature denial of service
-+
-+
+        A programming error in processing a TCP connection with both TCP_MD5SIG
-+
+        and TCP_NOOPT socket options may lead to kernel crash.
-+
-+
+        SCTP
-+
-+
+        A lack of proper input checks in the ICMPv6 processing in the SCTP stack
-+
+        can lead to either a failed kernel assertion or to a NULL pointer
-+
+        dereference.  In either case, a kernel panic will follow.
-+
-+
+20160102:
-+
+        Happy New Year
-+
-+
+20151101:
-+
+        Increase kern.ipc.somaxconn default to 256.
-+
-+
+20151017:
-+
+        Add initial statistics api to libmport and a driver to print
-+
+        it in mport(1).
-+
-+
+20151002:
-+
+        Revised rpcbind(8) patch to fix issues with NIS
-+
+20150930:
+        In rpcbind(8), netbuf structures are copied directly, which would result in
+        two netbuf structures that reference to one shared address buffer.  When one

Diff Legend

-–
+Removed lines
-+
+Added lines
-<
+Changed lines
->
+Changed lines

Comparing trunk/UPDATING (file contents): Revision 7338 by laffer1, Wed Sep 30 13:11:05 2015 UTC vs. Revision 9173 by laffer1, Sat Oct 15 13:48:05 2016 UTC

Diff Legend

Comparing trunk/UPDATING (file contents):
Revision 7338 by laffer1, Wed Sep 30 13:11:05 2015 UTC vs.
Revision 9173 by laffer1, Sat Oct 15 13:48:05 2016 UTC