On Thu, May 12, 2005 3:27 am, Marian Durkovic said:
>> Seems like i am getting half the performance when sending to the fbsd
>> box. Also, enabling jumbo frames does not help, and sometimes even
>> yields slightly slower results.
>
> Yes, that's exactly the problem my patch is addressing - for lar
On Thu, May 12, 2005 3:27 am, Marian Durkovic said:
> Hi,
>
>
> On Wed, May 11, 2005 at 06:38:48PM -0400, Mike Jakubik wrote:
>
>> Any luck submitting the patch for this?
>>
>
> Yes, it's kern/80932
Good stuff, ill test it when i get a chance.
>
>> I looked at Intels website, and the latest drive
Hi,
On Wed, May 11, 2005 at 06:38:48PM -0400, Mike Jakubik wrote:
> Any luck submitting the patch for this?
Yes, it's kern/80932
> I looked at Intels website, and the latest drive for FreeBSD 4.7 is 1.7.35.
1.7.41 could be found at:
http://downloadfinder.intel.com/scripts-df-external/filter_re
On Wed, May 11, 2005 5:24 pm, Mike Jakubik said:
> Any luck submitting the patch for this? I looked at Intels website, and
> the latest drive for FreeBSD 4.7 is 1.7.35. Which is what is also used on
> -CURRENT now. They also state "Development is no longer taking place on
> this driver. For the la
On Sat, May 7, 2005 12:37 am, Kris Kennaway said:
> On Fri, May 06, 2005 at 08:59:50AM +0200, Marian Durkovic wrote:
>
>> Hi all,
>>
>>
>>
>> seems we've found the problem. The performance degradation was
>> happening it the TX path, due to insufficient setting of TX packet
>> buffer FIFO on the ch
On Sat, May 7, 2005 5:35 am, Marian Durkovic said:
>>> To achieve wirespeed performance, the TX FIFO must be large enough to
>>> accomodate 2 jumbo packets (not just 1 as the driver was assuming).
>>> There was also a typo in the driver, causing the PBA tuning on most
>>> cards to be non-functiona
> > To achieve wirespeed performance, the TX FIFO must be large enough to
> > accomodate 2 jumbo packets (not just 1 as the driver was assuming).
> > There was also a typo in the driver, causing the PBA tuning on most
> > cards to be non-functional.
> >
> > Please be sure to use em driver vers
On Fri, May 06, 2005 at 08:59:50AM +0200, Marian Durkovic wrote:
> Hi all,
>
>
> seems we've found the problem. The performance degradation was happening
> it the TX path, due to insufficient setting of TX packet buffer FIFO on the
> chip.
>
> To achieve wirespeed performance, the TX FIFO mu
