[openib-general] 2 SLES 10 backport directories
Erez Zilber
erezz at voltaire.com
Wed Sep 20 05:45:13 PDT 2006
Michael S. Tsirkin wrote:
> Quoting r. Erez Zilber <erezz at voltaire.com>:
>
>> Subject: 2 SLES 10 backport directories
>>
>> Michael,
>>
>> I saw that there are 2 SLES 10 backport directories in the svn:
>>
>> https://openib.org/svn/gen2/branches/backport/sles10/ - this one
>> contains patches that we added for SLES 10
>>
>> https://openib.org/svn/gen2/branches/backport/2.6.16_sles10/ - this one
>> was added later by you.
>>
>> Can we unite them?
>>
>> Here's my motivation: I want to be able to install SLES 10, replace its
>> infiniband dir with infiniband from openib's svn, apply all SLES 10
>> patches (from a single directory) and then it should work.
>>
>> This should help us in future OFED releases.
>>
>
> I'd like that too, but there's a difficulty here.
>
> The rest of the backport patches make it possible to build
> IB support out of kernel, without patching the kernel code itself.
> This is an explicit requirement of some users, so we made an effort
> to preserve this ability, and so far it works with the rest of the IB stack -
> assuming that user has built infiniband support as a module or disabled it -
> but that's what most people currenty have, anyway.
>
> Unfortunately sles10 patches for iser that you mention violate this rule - they
> patch the iscsi support that is already there as part of the kernel.
> So unless this can be fixed somehow, we need the iscsi stuff separate, so that
> 1. we know to apply it in kernel source directory, not where we unpacked IB code
> 2. it can be applied conditionally when the user has enabled iser, so that
> others still have the ability not to touch their kernel
>
I think that we can throw away
https://openib.org/svn/gen2/branches/backport/sles10/. These patches
apply to SLES 10 beta 8. They are no longer needed. As for
https://openib.org/svn/gen2/branches/backport/2.6.16_sles10/, it
contains 2 iSER patches. Both affect only iSER code (nothing in
open-iscsi or any other kernel code). Therefore, I think that it's ok.
What do you think?
Erez
More information about the general
mailing list