Goodbye from our Newsletter, sorry to see you go.
You have been unsubscribed from our newsletters.
This is the last email you will receive from us. Our newsletter system,
phpList,
will refuse to send you any further messages, without manual intervention
by our administrator.
If the
No, I generate traffic with random src ip and src port (but same dst ip
and dst port). So I have many flows.
12.12.2014 2:10, Skidmore, Donald C пишет:
> If you have ATR disabled we will spread the flows to different queues based
> on their RSS hash. Could it be that you are only seeing one fl
>>>
>>> I'm not running Ubuntu, I'm looking at the 3.16.4-based kernel
>>> source in our local LXR database reportedly from the Ubuntu
>>> 14.10 linux-source package. IIRC, I ran 'make modules_prepare'
>>> using the 14.10 kernel config file which is supposed to generate
>>> the appropriate heade
> -Original Message-
> From: Chris J Arges [mailto:chris.j.ar...@canonical.com]
> Sent: Thursday, December 11, 2014 4:05 PM
> To: Allan, Bruce W; Fujinaka, Todd
> Cc: e1000-devel@lists.sourceforge.net
> Subject: Re: [E1000-devel] [PATCH] Issues with compiling ixgbevf-2.15.3 on
> Ubuntu 3.13
On 12/11/2014 05:17 PM, Allan, Bruce W wrote:
>> -Original Message- From: Chris J Arges
>> [mailto:chris.j.ar...@canonical.com] Sent: Thursday, December 11,
>> 2014 2:45 PM To: Allan, Bruce W; Fujinaka, Todd Cc:
>> e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel]
>> [PATCH] Is
> -Original Message-
> From: Chris J Arges [mailto:chris.j.ar...@canonical.com]
> Sent: Thursday, December 11, 2014 2:45 PM
> To: Allan, Bruce W; Fujinaka, Todd
> Cc: e1000-devel@lists.sourceforge.net
> Subject: Re: [E1000-devel] [PATCH] Issues with compiling ixgbevf-2.15.3 on
> Ubuntu 3.13
On 12/11/2014 04:33 PM, Allan, Bruce W wrote:
>> -Original Message-
>> From: Chris J Arges [mailto:chris.j.ar...@canonical.com]
>> Sent: Thursday, December 11, 2014 2:01 PM
>> To: Fujinaka, Todd
>> Cc: e1000-devel@lists.sourceforge.net
>> Subject: Re: [E1000-devel] [PATCH] Issues with com
> -Original Message-
> From: Chris J Arges [mailto:chris.j.ar...@canonical.com]
> Sent: Thursday, December 11, 2014 2:01 PM
> To: Fujinaka, Todd
> Cc: e1000-devel@lists.sourceforge.net
> Subject: Re: [E1000-devel] [PATCH] Issues with compiling ixgbevf-2.15.3 on
> Ubuntu 3.13.0-30
>
> On Th
On Thu, Dec 11, 2014 at 09:46:15PM +, Fujinaka, Todd wrote:
> First, thanks for the patch. You'll have to include it in the text of the
> message because it gets stripped here.
>
> Unfortunately, we may not be able to take it (depending on what you did, the
> attachment gets stripped) becaus
On Thu, Dec 11, 2014 at 09:46:15PM +, Fujinaka, Todd wrote:
> First, thanks for the patch. You'll have to include it in the text of the
> message because it gets stripped here.
>
Todd,
I've inlined the patch below.
> Unfortunately, we may not be able to take it (depending on what you did, th
First, thanks for the patch. You'll have to include it in the text of the
message because it gets stripped here.
Unfortunately, we may not be able to take it (depending on what you did, the
attachment gets stripped) because you guys don't put any identifiers for your
kernel (like a define that
Attached is a patch that fixes compile issues where
57a7744e09867ebcfa0ccf1d6d529caa7728d552 was backported to kernel
versions eariler than 3.15.
Please accept this patch for your next stable update.
Thanks,
--chris j arges
--
From: David Miller
Date: Thu, 11 Dec 2014 15:24:17 -0500 (EST)
> From: Sowmini Varadhan
> Date: Thu, 11 Dec 2014 15:21:00 -0500
>
>> All this may be true, but it would also be true for Solaris, which
>> manages to do line-speed (for the exact same setup), so there must be
>> some other bottlene
From: Sowmini Varadhan
Date: Thu, 11 Dec 2014 15:21:00 -0500
> All this may be true, but it would also be true for Solaris, which
> manages to do line-speed (for the exact same setup), so there must be
> some other bottleneck going on?
They have DMA mapping interfaces which pre-allocate large b
On (12/11/14 15:09), David Miller wrote:
>
> The real overhead is unavoidable due to the way the hypervisor access
> to the IOMMU is implemented in sun4v.
>
> If we had direct access to the hardware, we could avoid all of the
> real overhead in %99 of all IOMMU mappings, as we do for pre-sun4v
>
If you have ATR disabled we will spread the flows to different queues based on
their RSS hash. Could it be that you are only seeing one flow, in which case
it would always be directed to one queue?
Thanks,
-Don Skidmore
> -Original Message-
> From: Viktor Khomyuk [mailto:v.khom...@off
From: Sowmini Varadhan
Date: Thu, 11 Dec 2014 14:45:42 -0500
> 1. lockstat and perf report that iommu->lock is the hot-lock (in a typical
>instance, I get about 21M contentions out of 27M acquisitions, 25 us
>avg wait time). Even if I fix this issue (see below), I see:
The real overhead
e1000-developers,
[Cc-ing sparclinux due to the iommu observations..]
I'm looking at an iperf issue running over ixgbe on linux
on a sparc T5-2 platform (64 cpu) where we cannot get to line-speed
(peaks at 3 Gbps on a 10Gbps link) and I'm trying to get to the bottom
of this.
I've run iperf with
Hi Bernhard,
+ This query shows me only your mail as result, so AFAICS it's "bit-banged
I2C", not
+ bit-bagned (the "n" must be before the "g") and a grep in the Kernel source
only
+ finds "banged":
+
+ https://www.google.com/search?q=%22bit-bagned%22+I2c
Quite right! I'll change that in the re
19 matches
Mail list logo
