Re: [dpdk-dev] [PATCH 00/12] DPDK armv8-a support

[Jerin Jacob <jerin.jacob@caviumnetworks.com>]

On Tue, Nov 03, 2015 at 02:17:38PM +0000, Hunt, David wrote:
> On 03/11/2015 13:09, Jerin Jacob wrote:
> >This is the v1 patchset for ARMv8 that now sits on top of the v6 patch
> >of the ARMv7 code by RehiveTech. It adds code into the same arm include
> >directory, reducing code duplication.
> >
> >Tested on an ThunderX arm 64-bit arm server board, with PCI slots. Passes traffic
> >between two physical ports on an Intel 82599 dual-port 10Gig NIC. Should
> >work with many other NICS as long as long as there is no unaligned access to
> >device memory but not yet untested.
> 
> I have your patchset building and running on an X-Gene based 8-core MP30AR0
> system, passing traffic between two ports on and 82599 also.
> 

Thanks.

> >Notes on arm64 kernel configuration:
> >
> >   Tested on using Ubuntu 14.04 LTS with a 3.18 kernel and igb_uio.
> >   ARM64 kernels does not have functional resource mapping of PCI memory
> >   (PCI_MMAP), so the pci driver needs to be patched to enable this. The
> >   symptom of this is when /sys/bus/pci/devices/0000:0X:00.Y directory is
> >   missing the resource0...N files for mmapping the device memory.
> >
> >   Following patch fixes the PCI resource mapping issue om armv8.
> >   Its not yet up streamed.We are in the process of up streaming it.
> >
> >   http://lists.infradead.org/pipermail/linux-arm-kernel/2015-July/358906.html
> 
> Good to see that there's a patch on the way for this. That fix looks almost
> exactly the same as the hack I did to my kernel :)
> 
> I had a couple of small issues when patching/building:
> 
> 1. Three of the files had an extra blank line at the end. Maybe worth
> running checkpatch on the patches. 'git am' was complaining.

I will fix it in next version.

> 
> 2. I had problems compiling two drivers because they were attempting to
> include tmmintrin.h:
> 
> ...dpdk/drivers/net/fm10k/fm10k_rxtx_vec.c:41:23: fatal error: tmmintrin.h:
> No such file or directory
> 
> ...dpdk/drivers/net/i40e/i40e_rxtx_vec.c:43:23: fatal error: tmmintrin.h: No
> such file or directory
>

 
> To avoid this, I added the following two lines into
> defconfig_arm64-armv8a-linuxapp-gcc
> 
> CONFIG_RTE_LIBRTE_FM10K_PMD=n
> CONFIG_RTE_LIBRTE_I40E_PMD=n

the patch was based on 82fb702077f67585d64a07de0080e5cb6a924a72 which
don't have these changes. I will add these in next version.
 
> and then it built fine, and I can run testpmd with my 82599's and run
> autotests.

I ran autotest, "Mbuf autotest" stress failure is due strong vs weak ordering
issue. I will send the next version based on new patch being discussed
on ml.


> 
> Thanks for that.
> Dave.