Hi!

 The next -uc for i386 is already available at:

http://distro2.conectiva.com.br/~lcapitulino/mdv_kernels/uc/2.6.12.23.uc3mdk/

 and will be available for x86_64 in the next hours at:

http://tmb.kkc.fi/Kernels/2006.0/x86_64/uc/2.6.12.23.uc3mdk/

 Changelog:

* Tue Jun 20 2006 Luiz Capitulino <[EMAIL PROTECTED]> 2.6.12-23.uc3mdk
  o Luiz Capitulino <[EMAIL PROTECTED]>
   - Fixes kernel-doc package name: For some reason I don't know, the fix for
    #21345 broke the kernel-doc package (it became kernel-doc-1-1mdk.i586.rpm),
    It's strange, because it didn't happen on cooker (where the patch came
    from). For now we change the kernel-doc package to be consistent with the
    rest of the kernels (patch from Thomas Backlund <[EMAIL PROTECTED]>) but
    would be better to know why the kernel-doc package has broke
   - Removes 'final' marks from changelog: we don't need them, since the
     release dates tells us the same thing
   - Adds a new provide: drbd-api. This helps to prevent that not compatible
     drbd's user-space tools get installed
   - Fixes NULL pointer dereference in USB-Serial's serial_open() function

  o Arnaud Patard <[EMAIL PROTECTED]>
   - Add sysfs support in ide-tape

  o Flavio Bruno Leitner <[EMAIL PROTECTED]>
   - Security fixes:
     * ZZB1_CVE-2006-2274_sctp_fix_infinite_recursion.patch    (#22442)
     * ZZB2_CVE-2006-1858_sctp_respect_real_chunk_length.patch (#22833)
     * ZZB3_CVE-2006-1857_sctp_fix_parse_HB_ACK_chunk.patch    (#22832)
     * ZZB4_CVE-2006-2445_fix_timer_race_and_deadlock.patch    (#23220)

 The most important change is the drbd one: now the drbd-utils package
requires the drbd API it depends on, and the kernel's drbd API version is in
the kernel package's provides list. The end result is that you can't install
the utils package without installing the kernel it needs (thanks for Marcelo
Leitner for this idea).

 You can find the drbd user-space tools at:

http://distro2.conectiva.com.br/~lcapitulino/mdv_kernels/uc/extra/

 If you're a drbd user, please, give it a try.

 Also note that the kernel-doc change description is harmless: Thomas has
already clarified why it was needed.

 And, as a final note, for those who knew him, yes, Flavio is helping us
with some kernel issues.

 Thanks for reading this and happy testing!

-- 
Luiz Fernando N. Capitulino

Reply via email to