[dpdk-dev] [PATCH] pci: fix non-Intel devices probing

[Thomas Monjalon]

16/09/2013 14:42, Patrick Mahan :
> On Sep 16, 2013, at 1:29 PM, Thomas Monjalon <thomas.monjalon at 6wind.com> 
wrote:
> > There is no need to check for bars mapping, especially BAR0 is not
> > required. If bars mapping failed, then pci_uio_map_resource will fail and
> > we won't reach this check. So get rid of BAR0 check.
> > Besides, pci_uio_map_resource should only be called for Intel devices.
> > The flag RTE_PCI_DRV_NEED_IGB_UIO is set for all Intel devices, even when
> > RTE_EAL_UNBIND_PORTS is disabled.
> 
> I totally disagree with this statement.  I am currently working on a
> non-intel device that does need UIO support (I copied the igb UIO to create
> a new UIO driver).  This device not only needs bar0 but also bar1.  I've
> modified the eal PCI support code to support this behavior and change this
> behavior would not be good, IMHO.

I don't know which statement you are talking about.
You are writing a PMD based on UIO. Will you send it on dpdk.org ?
In this release, the only PMDs using UIO are Intel ones.
Also as explained, this check is never needed.
But I think you are talking about support of BAR mapping. This is not changed 
by this patch. Moreover, the BAR mapping is extended in coming release 1.4.1.

Note that it's simpler to discuss about a patch or a clearly pointed sentence. 
So, please be more precise or send your patch proposal.

PS: please don't top-post
-- 
Thomas