[PATCH] staging: ipath: fix build failure

[Jim Davis]

On Tue, Sep 1, 2015 at 7:52 AM, Doug Ledford <dledford at redhat.com> wrote:
> On 09/01/2015 09:01 AM, Sudip Mukherjee wrote:
>> building of ipath depends on infiniband. And if ipath is selected and
>> infiniband is not then build fails with:
>> ERROR: "ib_alloc_device" [drivers/staging/ipath/ib_ipath.ko] undefined!
>> ERROR: "ib_dealloc_device" [drivers/staging/ipath/ib_ipath.ko] undefined!
>> ERROR: "ib_dispatch_event" [drivers/staging/ipath/ib_ipath.ko] undefined!
>> ERROR: "ib_register_device" [drivers/staging/ipath/ib_ipath.ko] undefined!
>> ERROR: "ib_unregister_device" [drivers/staging/ipath/ib_ipath.ko] undefined!
>> ERROR: "ib_modify_qp_is_ok" [drivers/staging/ipath/ib_ipath.ko] undefined!
>> ERROR: "ib_wq" [drivers/staging/ipath/ib_ipath.ko] undefined!
>>
>> Config at: https://lkml.org/lkml/2015/8/31/328
>>
>> Reported-by: Jim Davis <jim.epost at gmail.com>
>> Signed-off-by: Sudip Mukherjee <sudip at vectorindia.org>
>> ---
>>  drivers/staging/ipath/Kconfig | 1 +
>>  1 file changed, 1 insertion(+)
>>
>> diff --git a/drivers/staging/ipath/Kconfig b/drivers/staging/ipath/Kconfig
>> index 041ce06..1edc21b 100644
>> --- a/drivers/staging/ipath/Kconfig
>> +++ b/drivers/staging/ipath/Kconfig
>> @@ -1,6 +1,7 @@
>>  config INFINIBAND_IPATH
>>       tristate "QLogic HTX HCA support"
>>       depends on 64BIT && NET && HT_IRQ
>> +     depends on INFINIBAND
>>       ---help---
>>       This is a driver for the deprecated QLogic Hyper-Transport
>>       IB host channel adapter (model QHT7140),
>>
>
> The ipath driver was moved to staging/rdma/ipath and has proper
> protection against being built without the InfiniBand subsystem.  Where
> are you seeing this tree?  I'm curious because I no longer have this
> driver in this location in my tree and I never pushed it anywhere other
> than one of my trees.  Is this maybe in linux-next?

Yes; http://marc.info/?l=linux-kernel&m=144104080331664&w=3
-- 
Jim