[openfabrics-ewg] FW: [Bug 219] OFED 1.1rc3 contains prerelease unstable libibverbs code

[Scott Weitzenkamp (sweitzen)]

I'd agree with Roland we should not use unstable, unreleased libibverbs.

Scott Weitzenkamp
SQA and Release Manager
Server Virtualization Business Unit
Cisco Systems
 

> -----Original Message-----
> From: bugzilla-daemon at openib.org [mailto:bugzilla-daemon at openib.org] 
> Sent: Wednesday, September 06, 2006 1:41 PM
> To: Scott Weitzenkamp (sweitzen)
> Subject: [Bug 219] OFED 1.1rc3 contains prerelease unstable 
> libibverbs code
> 
> http://openib.org/bugzilla/show_bug.cgi?id=219
> 
> 
> rolandd at cisco.com changed:
> 
>            What    |Removed                     |Added
> --------------------------------------------------------------
> --------------
>              Status|RESOLVED                    |REOPENED
>          Resolution|WONTFIX                     |
> 
> 
> 
> 
> ------- Comment #3 from rolandd at cisco.com  2006-09-06 13:41 -------
> If current librdmacm needs symbols from unstable, unreleased 
> libibverbs then
> perhaps librdmacm should not be in OFED.  Or librdmacm needs to have
> compatibility code to work with older versions of libibverbs 
> 1.0 (look at how
> libmthca handles ibv_read_sysfs_file() for a hint how this 
> could work).
> 
> I absolutely don't think it's acceptable to pull code from 
> unreleased, unstable
> trees into OFED -- OFED is supposed to be a stable release.
> 
> And this particular case _does_ have an impact: the ABI of 
> libibverbs changed
> immediately after the libibverbs-1.0 tree was forked for 
> maintenance.  For
> example, HP MPI built against OFED 1.0 does not work with 
> OFED 1.1-rc3, and I
> don't think it's reasonable for HP to have to support so many
> binary-incompatible releases.
> 
> 
> 
> 
> ------- You are receiving this mail because: -------
> You are on the CC list for the bug, or are watching someone who is.
>