Joubert,
By all means, look at the existing PMDs. It also helps if you have an existing
kernel driver to use/modify (preferably BSD based). That way you can have a
baseline to compare against.
At least, that is method I used.
Patrick
> On Jun 13, 2014, at 9:08 AM, Joubert Berger wrote:
>
All,
I see in both the announcements and release notes support for building and
running on FreeBSD. However, there are no instructions, nor does the "Intel
DPDK for FreeBSD* Getting Started Guide" exist.
Is this forthcoming?
I've at least figured out to use 'gmake' instead of make. But is th
General Query, has anyone attempted to use valgrind --tool=callgrind with the
testpmd application? I am seeing valgrind core dumping when EAL attempts to
set up hugetlb pages (2MB size). There seems some collision on the mmap() call.
Thanks,
Patrick
Coming to you from deep inside Fortress M
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 beh
I have a need to keep a timestamp on a piece of global data. When then
timestamp grows too old I want to refresh that data. Is it safe to use,
gettimeofday()?
I thought about using an alarm, but I need to set an alarm from inside the
alarm callback which doesn't look like it will work due to
Has anyone contemplated providing access to the PCI config space
(pci_read_config_[byte|word|dword]) for DPDK? It seems simple to
perform, some preliminary testing by me seems to indicate opening
the /sys/bus/pci/devices//config file, seeking
to the correct offset and reading off the value.
It se
With test-pmd, you need to make sure you allow enough cores to handle each RSS
queue. And as the previous poster mentioned, you must generate traffic to a
wide range of destination addresses.
Patrick
Coming to you from deep inside Fortress Mahan
On Aug 17, 2013, at 8:57 AM, Hamed khanmirza w
Okay,
I have a need to do something similar to
'system("cmd -arg1 -arg2")'
in DPDK? This should (mostly) occur during initialization of hardware, say a
special tool for loading microcode to an intelligent NIC?
Thanks,
Patrick
Coming to you from deep inside Fortress Mahan
Stephen,
When did you generate this diff? It failed to apply to 1.3.1r2 so I am hand
patching instead. But you might want to update to the latest version.
Thanks,
Patrick
On 8/8/13 5:50 PM, Stephen Hemminger wrote:
> Still needs a little cleanup (patch is messy).
>
> Subject: pci: support mu
When?
I scanned the archives for this month and last month. I don't see any patches
regarding allowing other BARs to be mapped.
Thanks,
Patrick
Coming to you from deep inside Fortress Mahan
On Aug 8, 2013, at 4:53 PM, Stephen Hemminger
wrote:
> On Thu, 8 Aug 2013 16:45:42 -0700
&
Before I do this, I wanted to check with the list.
The UIO support code in DPDK only maps in bar 0 for PCI devices, and it does
provide support for mapping any other bar that the device provides.
Has anyone fixed this yet? Or maybe has a patch?
DPDK version: 1.3.1r2
Thanks,
Patrick
Okay, I wadding into the process of writing a poll-mode driver for DPDK.
First question, is the use of the UIO device a requirement (outside of the need
for interrupt support) since it is possible to map the PCI bars via /dev/mem?
Second question, is it expected that the NIC hardware would alre
Dmitry,
Please post the exact blacklist option. I use the '-b' option all the time
with my test box (Dell T620).
Thanks,
Patrick
Sent from my iPad
On Jul 18, 2013, at 8:42 AM, Dmitry Vyal wrote:
> Greetings!
>
> I've been playing with dpdk on my desktop for some time and decided to
> fin
I'm not sure this is a bug or feature.
First, I've been away for a few weeks being on vacation and other stuff. My
client liked my findings on DPDK well enough that I have been loaned a dual-cpu
system (2 Xeon E5-2690 @ 2.9 GHz) and two 82599 cards. This is so I can
eliminate the PCIe bus con
ger :
>> Patrick Mahan wrote:
>>> I am seeing the following issue with the link_status example. This
>>> requires a reboot on my testbox.
>>
>> This is an example of why I posted the RFC patch to igb_uio to not use
>> pci_lock. -
>> Subject: igb_uio: rework
on DUT(which run l3fwd)
> Please refer the attachment.
>
> thanks
>
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Patrick Mahan
> Sent: Friday, June 14, 2013 1:10 PM
> To: dev at dpdk.org
> Subject: [dpdk-dev] l3fwd doesn'
All,
I am seeing the following issue with the link_status example. This requires a
reboot on my testbox.
Ideas or suggestions?
Thanks,
Patrick
>From /var/log/messages:
Jun 12 11:08:22 dpdk-eval kernel: ixgbe :03:00.0: removed PHC on p2p1
Jun 12 11:08:22 dpdk-eval kernel: ixgbe :03:00
All,
I am trying to run l3fwd example.
The system is a single chip E5-2690 (8 core) with 64 GB of DDR3 memory and
a 82599 2 port NIC connected to an Ixia traffic generator.
I have successfully run test, testpmd and l2fwd. Both testpmd and l2fwd seem to
work as documented.
However, l3fwd doesn'
On May 29, 2013, at 7:07 AM, Damien Millescamps wrote:
> On 05/28/2013 09:15 PM, Patrick Mahan wrote:
>> So the overhead cost is almost 70%?
>>
>> Can this ever do line rate? Under what conditions? It has been my
>> experience that the industry standard is testin
On May 25, 2013, at 1:59 PM, Damien Millescamps wrote:
> On 05/25/2013 09:23 PM, Damien Millescamps wrote:
>> Hi Patrick,
>>
>> If you are using both ports of the same Niantics at the same time then
>> you won't be able to reach the line-rate on both port.
>
> For a better explanation, you can
Yes it is a typo. I'm away from the box and was referring to my notes.
Thanks,
Patrick
Sent from my iPad
On May 24, 2013, at 1:03 PM, Olivier MATZ wrote:
> Hello Patrick,
>
>> sudo build/app/testpmd -b :03:00.0 -b :03:00.1 -c -n3
>> -- --nb-cores= --nb-ports=2 --rxd=2048 --rxd=2048
On May 24, 2013, at 8:45 AM, Thomas Monjalon
wrote:
> Adding other questions about packet generator:
>
> 24/05/2013 16:41, Thomas Monjalon :
>> 24/05/2013 16:11, Patrick Mahan :
>>> Intel Xeon E5-2690 (8 physical, 16 virtual)
>>
>> How many CPU sockets hav
On May 24, 2013, at 7:41 AM, Thomas Monjalon
wrote:
> Hello,
>
> 24/05/2013 16:11, Patrick Mahan :
>> Intel Xeon E5-2690 (8 physical, 16 virtual)
>
> How many CPU sockets have you ?
This is a Dell PowerEdge T620, it has two sockets, but only one has a CPU in it.
>
Good morning,
I have been playing with this code now for about 2 weeks. I posted
earlier about being unable to get it to work on Fedora 14, but have
it working on CentOS 6.4. Here is my hardware -
Intel Xeon E5-2690 (8 physical, 16 virtual)
64 Gbyte DDR3 memory
Intel 82599EB-SPF dual port 10GE
This is an update to my last email requesting help with the DPDK sources and
Fedora 14.
After talking with a colleague who was investigating DPDK as well, I removed the
Fedora 14 install and installed instead CentOS 6.4. After installing the
release, I did a 'yum update' on the kernel sources to
Good Morning,
I am doing a bit of work investigating the DPDK code for a prospective client.
I have the following hardware setup:
Dell PowerEdge T620 running Fedora 14 with the 2.6.35.14-106 kernel. I had to
build a custom version to enable memmap for HPET.
Uname -a output:
pmahan at dpdk-eva
26 matches
Mail list logo
