Permalink
Switch branches/tags
Commits on Jan 22, 2012
  1. kernel: Nuke the old (and deprecated) ata(4) driver and atacontrol(8).

    Sascha Wildner committed Jan 22, 2012
    It's been replaced by nata(4)/natacontrol(8) since more than 4 years.
  2. Shifting master to 3.1.

    fupjack committed Jan 22, 2012
  3. dloader: Remove some unused variables.

    Sascha Wildner committed Jan 22, 2012
  4. crontab(1): Adjust the number of comment lines in front of the tab.

    Sascha Wildner committed Jan 22, 2012
    In 229cb57 I removed the third line
    of comment that crontab(1) puts in front of the tab file (which was
    the CVS ID of crontab.c), but I forgot to adjust the #define saying
    how much lines to skip.
    
    This led to the first line of comment written by the user not being
    visible in crontab -e.
    
    Reported-by: lentferj
  5. kernel/acpi: Fix a reference to an uninitialized ptr.

    Sascha Wildner committed Jan 22, 2012
  6. sendfile: Block or return EAGAIN, iff we are really short of space

    Sepherosa Ziehau committed Jan 22, 2012
    Send sockbuf's lowat could be quite "high" (hiwat/2), the originally
    code actually only uses 1/2 of the send sockbuf space, thus poor
    performance.
  7. sockbuf: Pre-alloc sockbuf space, so asynchronized pru_send could be …

    Sepherosa Ziehau committed Jan 22, 2012
    …used.
    
    When doing asynchronized pru_send, it is quite possible that we could
    break the send sockbuf's upper limit, if we only use ssb_space() to
    calculate the remaining sockbuf space.
    
    Now, we add another two counters into sockbuf, when we send a mbuf
    chain to the netisr, we update the these two counters to reflect how
    much mbufs will be pending on the send sockbuf.  When calculating the
    remaining space of the send sockbuf we also take these two counters
    into consideration, so we will not break the upper limit of the send
    sockbuf.  These two counter will also be updated when the mbufs in
    the sockbuf is to be dropped/freed.
    
    Reenable the asynchronized pru_send in sendfile and sosendtcp
  8. pru_send: Disable asynchornized pru_send usage by default

    Sepherosa Ziehau committed Jan 22, 2012
    It may cause problem, e.g. too many mbufs sitting in the TCP send buffer,
    when facing with certain attacks, e.g. FIN_WAIT_1 DoS.
    
    Reported-by: pavalos@
  9. carp: BGL -> CARP global token

    Sepherosa Ziehau committed Jan 22, 2012
Commits on Jan 21, 2012
  1. Merge commit 'crater/master'

    Thomas Nikolajsen committed Jan 21, 2012
  2. etc/auth.conf: Add sha512/sha256/blf

    Thomas Nikolajsen committed Jan 21, 2012
  3. login.conf.5: Add sha256/sha512

    Thomas Nikolajsen committed Jan 21, 2012
  4. pam_unix.c: Change default hash

    Thomas Nikolajsen committed Jan 21, 2012
    Change to sha512; same as in etc/login.conf.
  5. login.conf.5: Fix mdoc problems

    Thomas Nikolajsen committed Jan 21, 2012
    Fix problems identified by mandoc(1).
    
    To get both groff(1) and mandoc(1) to format nice
    headlines changed from TAB changed to Ta use.
    This seems like a groff(1) bug.
  6. crypt.3: Update algorithm numbers ($N$)

    Thomas Nikolajsen committed Jan 21, 2012
    Fix algorithm number for Blowfish, SHA256 and SHA512,
    note that existing SHA256 and SHA512 implementations
    are deprecated and add new implementations.
    
    While here add missing line breaks at end of sentence.
  7. Fix buildkernel -j N

    Thomas Nikolajsen committed Jan 21, 2012
    Work around build ordering issue which caused
    `make -j N buildkernel' to fail on some systems.
    
    DragonFly-bug: <http://bugs.dragonflybsd.org/issue2279>
    Submitted-by:  YONETANI Tomokazu
  8. nataraid(4): Add devstat support.

    Sascha Wildner committed Jan 21, 2012
  9. installer: Warn <10GB HAMMER filesystems are unsupported

    Francois Tigeot committed Jan 21, 2012
Commits on Jan 20, 2012
  1. Add a copyright notice to cpputil.h

    Aggelos Economopoulos committed Jan 20, 2012
    Claim copyright inasmuch as there is any to claim for this trivial file.
    
    Reminded-by: swildner@
  2. kernel: Remove some more old major number related stuff.

    Sascha Wildner committed Jan 20, 2012
  3. crypt(3) - Paper over sizeof()/strlen() bug causing 32/64-bit issues

    thesjg committed Jan 20, 2012
    * In one place sizeof() was used instead of strlen(), causing the password
      to be muxed with 4 bytes of random stack data on 64-bit machines. Codify
      this behavior for reverse compatibility.
  4. kernel - Fix AHCI callout timer race (2)

    Matthew Dillon committed Jan 20, 2012
    * Fix minor bug in last commit.  If the serial number does not match
      after the timeout is stopped we don't want to mess with the xa at
      all.
  5. kernel - Fix race in callout_stop_sync()

    Matthew Dillon committed Jan 20, 2012
    * When callout_stop_sync() races a timeout callback which re-arms the
      callout it can hit an assertion.  The assertion is incorrect.  Instead
      of asserting we have to loop up and try again.
  6. kernel - Fix AHCI callout timer race

    Matthew Dillon committed Jan 20, 2012
    * callout_stop_sync() can block.  If this occurs a race can cause a
      CCB to be processed for completion twice.
    
    * Add a serial number to detect this situation.
    
    Reported-by: "G. Isenmann via Redmine"
  7. kernel - Fix pmap_release() bug for i386

    Matthew Dillon committed Jan 20, 2012
    * Fix 'pmap_release: page should already be gone' panic
    
    * There is a race condition which can result in the case.  Release the
      page normally instead of panicing.
    
    Reported-by: tuxillo
  8. crypt(3) - Switch SHA256/512 to the Linux implementation

    thesjg committed Jan 20, 2012
    * The existing SHA backends have been shown to be more susceptible to brute-
      force attacks than we would prefer --
      http://www.openwall.com/lists/oss-security/2012/01/16/2
    
    * Bring in the reference implementation used in Linux, code is in the public
      domain.
    
    * Add required (standard) functions mempcpy and stpcpy.
    
    * Change default for future installs to SHA512, this is the default on at
      least archlinux and fedora.
    
    * Add some minor hacks to libcrypt/crypt.c to ensure that
      a) All existing passwords continue to work
      b) All future passwords will be more secure with no changes required
    
    * To update passwords to the new format use passwd(1) for each user, and to
      change your default password type to SHA512 (default for new installations)
      change the passwd_format setting under default to "sha512".
  9. icu_abi/x86_64: {msi,msix}_release returns nothing

    Sepherosa Ziehau committed Jan 20, 2012
  10. i386: Implement MSI-X allocation and release support

    Sepherosa Ziehau committed Jan 20, 2012