OpenBSD src changes summary for 2016-12-10
==========================================

distrib/sets                            lib/libcrypto
lib/libssl                              share/man
sys/dev/pci                             sys/dev/pv
sys/net80211                            sys/netinet
usr.bin/dc                              

== distrib =========================================================== 01/05 ==

  http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/distrib

sets

  ~ lists/comp/md.octeon                  

  > sync (visa@)

== lib =============================================================== 02/05 ==

  http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/lib

libcrypto

  ~ man/X509_STORE_CTX_get_ex_new_index.3

  > Copyright and license; jmc@ noticed that is forgot this file (schwarze@)

  ~ man/X509_STORE_set_verify_cb_func.3   

  > Add Copyright and license; jmc@ noticed that i missed this file.
  > I'm not merging documentation for X509_STORE_set_verify_func(3)
  > from OpenSSL because their documentations does not appear to match
  > the code we have, and i don't understand what the code does. (schwarze@)

  ~ man/BN_BLINDING_new.3                 ~ man/BN_CTX_new.3
  ~ man/BN_CTX_start.3                    ~ man/BN_add.3
  ~ man/BN_add_word.3                     ~ man/BN_bn2bin.3
  ~ man/BN_cmp.3                          ~ man/BN_copy.3
  ~ man/BN_generate_prime.3               ~ man/BN_get0_nist_prime_521.3
  ~ man/BN_mod_inverse.3                  ~ man/BN_mod_mul_montgomery.3
  ~ man/BN_mod_mul_reciprocal.3           ~ man/BN_num_bytes.3
  ~ man/BN_rand.3                         ~ man/BN_set_bit.3
  ~ man/BN_set_negative.3                 ~ man/BN_swap.3
  ~ man/BN_zero.3                         ~ man/bn_dump.3
  ~ man/crypto.3                          ~ man/dh.3
  ~ man/dsa.3                             ~ man/rsa.3

  > replace all references to bn(3) with BN_new(3),
  > and make sure all BN*(3) pages point back to BN_new(3) (schwarze@)

  - man/bn.3                              ~ man/BN_new.3
  ~ man/Makefile                          

  > Merge bn(3) into BN_new(3).
  > OpenSSL removed bn(3) without replacement, but the introductory
  > text does seem helpful, and it is good for a sub-library to have a
  > central page pointing to all other pages and pointed at from all
  > other pages of the sub-library. (schwarze@)

  ~ man/bn_dump.3                         

  > Add Copyright and license; jmc@ noticed that i missed this file.
  > (schwarze@)

  ~ man/des_read_pw.3                     

  > Add Copyright and license; jmc@ noticed that i forgot this file.
  > Correct the header: It's <openssl/ui_compat.h>, not <openssl/des_old.h>.
  > Delete documentation of des_read_password(3) and des_read_2passwords(3)
  > which no longer exist.
  > Probably, the rest of this ought to be deleted as well... (schwarze@)

  ~ man/DH_generate_key.3                 ~ man/DH_generate_parameters.3
  ~ man/DH_get_ex_new_index.3             ~ man/DH_set_method.3
  ~ man/DH_size.3                         ~ man/DSA_dup_DH.3
  ~ man/RSA_print.3                       ~ man/crypto.3
  ~ man/d2i_DHparams.3                    ~ man/dsa.3
  ~ man/engine.3                          ~ man/rsa.3

  > Replace all references to dh(3) with DH_new(3).
  > Make sure that all DH pages point back to DH_new(3). (schwarze@)

  - man/dh.3                              ~ man/DH_new.3
  ~ man/Makefile                          

  > Merge one sentence and the references from dh(3) to DH_new(3) and
  > discard the rest of the text.  There is no value in documenting
  > some of the fields of a structure and then going on to say that the
  > structure is intended as opaque.  Besides, i doubt that we want
  > such strong marketing of ENGINE support. (schwarze@)

  ~ man/DSA_SIG_new.3                     ~ man/DSA_do_sign.3
  ~ man/DSA_dup_DH.3                      ~ man/DSA_generate_key.3
  ~ man/DSA_generate_parameters.3         ~ man/DSA_get_ex_new_index.3
  ~ man/DSA_set_method.3                  ~ man/DSA_sign.3
  ~ man/DSA_size.3                        ~ man/ECDSA_SIG_new.3
  ~ man/crypto.3                          ~ man/d2i_DSAPublicKey.3

  > Change all references to dsa(3) to point to DSA_new(3).
  > Make sure all DSA pages point back to DSA_new(3). (schwarze@)

  - man/dsa.3                             ~ man/DSA_new.3
  ~ man/Makefile                          

  > Merge one sentence and the references from dsa(3) to DSA_new(3) and
  > discard the rest of the text.  There is no value in documenting
  > some of the fields of a structure and then going on to say that the
  > structure is intended as opaque.  Besides, i doubt that we want
  > such strong marketing of ENGINE support. (schwarze@)

libssl

  ~ man/SSL_get_version.3                 

  > Document SSL_version(3).  It's listed in ssl(3) and <openssl/ssl.h>,
  > so it's clearly a public interface. (schwarze@)

  ~ man/Makefile                          ~ man/SSL_renegotiate.3
  + man/SSL_get_certificate.3             + man/SSL_get_state.3
  + man/SSL_num_renegotiations.3          

  > New manual pages SSL_get_certificate(3), SSL_get_state(3), and
  > SSL_num_renegotiations(3) written from scratch.  These functions
  > are listed in ssl(3) and <openssl/ssl.h>, so they are clearly public.
  > (schwarze@)

  ~ man/Makefile                          + man/SSL_get_shared_ciphers.3

  > Write an SSL_get_shared_ciphers(3) manual from scratch; another one
  > where BUGS is longer than DESCRIPTION.  The function is listed in
  > ssl(3) and <openssl/ssl.h>, so it's clearly public.
  > The code looks slightly mysterious to me, so it would be welcome if
  > somebody more familiar with TLS protocols could check factual accuracy.
  > (schwarze@)

  ~ man/ssl.3                             

  > Add Copyright and license.
  > Delete all the function prototypes.
  > They are all available from their individual manual pages.
  > Here, they were incomplete and nothing but a maintenance nightmare.
  > Add several missing cross reference, such that
  > this page now references all libssl manual pages.
  > Delete a sentence that said nothing and correct a typo.
  > Now all libssl manuals have proper Copyright notices and licenses,
  > and i have merged all improvements from OpenSSL that i could find.
  > (schwarze@)

== share ============================================================= 03/05 ==

  http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/share

man

  ~ man4/iwm.4                            

  > Update iwm(4) CAVEATS section; MIMO is now supported (stsp@)

== sys =============================================================== 04/05 ==

  http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys

dev/pci

  ~ if_iwm.c                              ~ if_iwn.c

  > Make retry-heavy rates less attractive to mira.
  > The number of retried frames shrinks significantly.
  > ok tb@ (stsp@)

  ~ if_iwm.c                              

  > In iwm(4), do not enable HT protection (RTS) unconditionally if the AP
  > requires
  > protection from 20MHz-only STAs on a 40MHz channel. We do not support 40MHz
  > channels yet so there is nothing to protect ourselves from. It is the 40MHz
  > users who need to start using RTS when a 20MHz-only user shows up. (stsp@)

  ~ if_iwm.c                              ~ if_iwmreg.h

  > Add support for MIMO Tx rates (MCS 8-15) to iwm(4).
  > The nominal maximum Tx rate is now 144Mbit/s (MCS 15, 20MHz channel, SGI).
  > In practice it's more around 30Mbit/s, though. The most significant
  > limiting
  > factor is probably lack of Tx aggregation support. (Due to protocol
  > overhead,
  > reaching the nominal max is pretty much impossible anyway. When shopping
  > wifi
  > devices, do not believe what stickers on the box are saying; full of lies!)
  > But APs will send us aggregates, so Rx should be faster than Tx (which was
  > already the case, just less pronounced).
  > Tested by phessler@, bmercer@, tb@
  > ok tb@ (stsp@)

dev/pv

  ~ xbf.c                                 

  > Don't fail if optional feature properties aren't found
  > Nathanael Rensen has noticed that the driver would fail to attach if
  > optional "feature-barrier" or "feature-flush-cache" properties cannot
  > be fetched and has provided a patch to solve the issue, thanks! (mikeb@)

  ~ xbf.c                                 

  > Bring back the sector count variable as it's used in the debug printf
  > (mikeb@)

net80211

  ~ ieee80211_mira.c                      ~ ieee80211_mira.h

  > Make mira rate scaling actually work with MIMO Tx rates.
  > The previous approach with an extra timeout was dumb (and of course
  > untested).
  > Additional fixes include:
  > - Take HT protection settings into account when determining whether RTS is
  > used.
  > - Stop probing the current rateset as soon as measurements become worse.
  > - Properly move probing timeouts for bad rates further into the future.
  > Tested with MIMO-enabled iwm(4) (MCS 0-15) and also iwn(4) (MCS 0-7 only).
  > Early versions also tested by phessler@ and bmercer@, and more testing by
  > tb@.
  > ok tb@ (stsp@)

netinet

  ~ udp_usrreq.c                          

  > Pass the correct offset of the next protocol field for the given IP
  > version to the upper layers.  The stack will set this field to IPIP
  > later on, so make sure it doesn't overwrite the packet in the wrong
  > location.
  > ok bluhm@ deraadt@ stsp@ (patrick@)

== usr.bin =========================================================== 05/05 ==

  http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin

dc

  ~ dc.1                                  

  > replace all references to bn(3) with BN_new(3),
  > and make sure all BN*(3) pages point back to BN_new(3) (schwarze@)

===============================================================================
_______________________________________________
odc mailing list
[email protected]
http://www.squish.net/mailman/listinfo/odc

Reply via email to