[openfabrics-ewg] OFED-1.1-rc4 is ready

Tue Sep 12 20:52:09 PDT 2006

The change rolled back to the libibverbs interface from OFED 1.0.

On 9/12/06 11:44 PM, "Betsy Zeller" <betsy at pathscale.com> wrote:

> Scott - I remember the discussion, and know that the intent is to get us
> to a better version of verbs - but didn't realize that this would imply
> what seems to be an interface change from an earlier RC candidate. We'll
> have to send in a patch for an RC5 so we can work with it.
> 
> - Betsy
> 
> On Tue, 2006-09-12 at 20:36 -0700, Scott Weitzenkamp (sweitzen) wrote:
>> You've got it backwards, the change from RC3 to RC4 was to revert to a
>> stabler verbs that was compatible with OFED 1.0.
>> 
>> Scott Weitzenkamp
>> SQA and Release Manager
>> Server Virtualization Business Unit
>> Cisco Systems
>>  
>> 
>>> -----Original Message-----
>>> From: openfabrics-ewg-bounces at openib.org
>>> [mailto:openfabrics-ewg-bounces at openib.org] On Behalf Of Betsy Zeller
>>> Sent: Tuesday, September 12, 2006 8:27 PM
>>> To: vlad at dev.mellanox.co.il
>>> Cc: openfabrics-ewg at openib.org; Robert.walsh at qlogic.com
>>> Subject: Re: [openfabrics-ewg] OFED-1.1-rc4 is ready
>>> 
>>> The change between RC3 and RC4 to the newer implementation of
>>> verbs was
>>> a large change to go into the last release candidate. Amongst other
>>> things, the ibv_driver_init function was changed to
>>> openib_driver_init.
>>> Consequently, the QLogic user verbs support will no longer
>>> even compile.
>>> Given the pushback on simpler changes earlier in the release
>>> cycle, this
>>> is kind of a nasty surprise.
>>> 
>>> We can send in a patch for the QLogic driver, but clearly an RC5, and
>>> quite a bit more testing  are going to be required.
>>> 
>>> Vlad, when do you plan to turn around RC5?
>>> 
>>> - Betsy
>>> 
>>> On Tue, 2006-09-12 at 18:14 +0300, vlad at dev.mellanox.co.il wrote:
>>>> Hi,
>>>> 
>>>> OFED-1.1-rc4 is available on
>>>> https://openib.org/svn/gen2/branches/1.1/ofed/releases/
>>>> File: OFED-1.1-rc4.tgz
>>>> Please report any issues in bugzilla http://openib.org/bugzilla/
>>>> 
>>>> Schedule reminder:
>>>> ==================
>>>> Next milestone:
>>>> Final release is planed for Sep-20.
>>>> 
>>>> Owners - please update release notes for final release not
>>> later then
>>>> Sep-18.
>>>> 
>>>> 
>>>> Tziporet & Vlad
>>>> 
>>> --------------------------------------------------------------
>>> ----------
>>>> -------------
>>>> 
>>>> Release details:
>>>> ================
>>>> Build_id:
>>>> OFED-1.1-rc4
>>>> 
>>>> openib-1.1 (REV=9435)
>>>> # User space
>>>> https://openib.org/svn/gen2/branches/1.1/src/userspace
>>>> Git: git://www.mellanox.co.il/~git/infiniband
>>>> ref: refs/heads/ofed_1_1
>>>> commit 796b6cb83392fd840549e3b6e559dfce022a2c49
>>>> 
>>>> # MPI
>>>> mpi_osu-0.9.7-mlx2.2.0.tgz
>>>> openmpi-1.1.1-1.src.rpm
>>>> mpitests-2.0-0.src.rpm
>>>> 
>>>> OS support:
>>>> ===========
>>>> Novell:
>>>>      - SLES 9.0 SP3
>>>>      - SLES10
>>>> Redhat:
>>>>      - Redhat EL4 up3
>>>> 
>>>>      - Redhat EL4 up4
>>>> kernel.org:
>>>>      - Kernel 2.6.17
>>>> 
>>>> Note: Redhat EL4 up2, Fedora C4 and SuSE Pro 10 were
>>> dropped from the
>>>> list. We keep the backport patches for these OSes and make sure OFED
>>>> compile and loaded properly but will not do full QA cycle.
>>>> 
>>>> Note: Kernel components were updated to 2.6.18-rc6
>>>> 
>>>> Systems:
>>>> ========
>>>>      * x86_64
>>>>      * x86
>>>>      * ia64
>>>>      * ppc64
>>>> 
>>>> Bug fixes from OFED-1.1-rc3:
>>>> ============================
>>>>  1. SDP: Data corruption fix
>>>>  2. libibverbs was reverted to 1.0 version (bug 219)
>>>>  3. libsdp: TCP_RR fix
>>>>  4. Compilation on kernel 2.6.18-rcX is failing
>>>>  5. OSU MPI: fix failure in Intel tests
>>>>  6. SRP: Kernel oops in case of port down
>>>>  7. ib_uverbs fails to load on ia64 (bug 222)
>>>>  8. IPoIB: Spinlock corruption in stress tests
>>>>  9. Added srp_daemon service, enable from
>>> /etc/infiniband/openib.conf
>>>> 10. mthca.so not properly linked to libibverbs (bug 223)
>>>> 11. ipath compilation problem on SLES10 (bug 226)
>>>> 12. problem with MPI: Get_processor_name on MVAPICH (bug 226)
>>>> 13. Add an option to renice the ib_mad thread to highest
>>> priority. Enable
>>>> from /etc/infiniband/openib.conf (workarouund for bug 229)
>>>> 14. Update for ehca driver
>>>> 15. Update for ipath driver
>>>> 16. Madaye installation using OPENIB_PARAMS. To build madeye run:
>>>>     export OPENIB_PARAMS="--with-madeye-mod" (or put it
>>> into ofed.conf file
>>>>     for unattended installation) and run install.sh
>>>> 17. OFED sources: Added kernel include files under:
>>>> <prefix/src/openib/include>.
>>>>     Can be used by kerenl modules, and already include the backport
>>>> pathces for each kernel.
>>>> 18. ibutils - updated with new flags (-P, -pc and -pm)
>>>> 19. SDP: RTU packet is lost Accept call blocks even if
>>> client connected.
>>>> 
>>>> 
>>>> 
>>>> 
>>>> Limitations and known issues:
>>>> =============================
>>>> 1. SDP: For Mellanox Sinai HCAs one must use latest FW
>>> version (1.1.000).
>>>> 2. SDP: Scalability issue when hundreds of connections are opened
>>>> 3. ipath driver is not supported on SLES9 SP3
>>>> 4. ehca driver supports only PPC machines and compiled on
>>> kernel 2.6.18
>>>> 5. OFED installation fails on PPC64 with SLES9.
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> _______________________________________________
>>>> openfabrics-ewg mailing list
>>>> openfabrics-ewg at openib.org
>>>> http://openib.org/mailman/listinfo/openfabrics-ewg
>>>> 
>>> 
>>> 
>>> _______________________________________________
>>> openfabrics-ewg mailing list
>>> openfabrics-ewg at openib.org
>>> http://openib.org/mailman/listinfo/openfabrics-ewg
>>> 

-- 
Jeff Squyres
Server Virtualization Business Unit
Cisco Systems