| View previous topic :: View next topic |
| Author |
Message |
snIP3r
l33t
Joined: 21 May 2004
Posts: 853
Location: germany
|
 Posted: Tue Jul 28, 2009 5:41 pm Post subject: install e1000e driver from intel page |
 |
|
hi all!
i am wondering about the two existing drivers for intel pci-e gigabit ethernet nic's:
- the one integrated in kernel sources version 0.3.3.3-k6 in 2.6.29-gentoo-r5
- the one from sourceforge page version 1.0.2.5
both versions seem to come from intel and i wanted to know if its recommended to install the sourceforge driver for higher performance?
does someone have any experience with the two drivers?
any info would be appreciated.
thx & greets
snIP3r
_________________
Intel i3-4130T on ASUS P9D-X
Kernel 5.15.88-gentoo SMP
-----------------------------------------------
if your problem is fixed please add something like [solved] to the topic! |
|
| Back to top |
|
 |
Hu
Administrator
Joined: 06 Mar 2007
Posts: 23091
|
| Posted: Wed Jul 29, 2009 2:46 am Post subject: |
|
|
| I have used the in-kernel e1000e driver ever since it became available, and have had no problems with it. I have not done a performance analysis to see how it compares to the out-of-tree driver. |
|
| Back to top |
|
|
snIP3r
l33t
Joined: 21 May 2004
Posts: 853
Location: germany
|
| Posted: Fri Aug 07, 2009 9:48 am Post subject: |
|
|
| Hu wrote: | | I have used the in-kernel e1000e driver ever since it became available, and have had no problems with it. I have not done a performance analysis to see how it compares to the out-of-tree driver. |
hi hu!
thx for the answer. can you perhaps give me some performance data of your intel gigabit nic with the in-kernel driver?
thx
snIP3r
_________________
Intel i3-4130T on ASUS P9D-X
Kernel 5.15.88-gentoo SMP
-----------------------------------------------
if your problem is fixed please add something like [solved] to the topic! |
|
| Back to top |
|
|
Hu
Administrator
Joined: 06 Mar 2007
Posts: 23091
|
| Posted: Sat Aug 08, 2009 2:38 am Post subject: |
|
|
| The only time I did any measurements, I got throughput of about 40% of the card's theoretical maximum. I strongly suspect cheap hardware somewhere between me and the data source was at fault for the poor performance, but had no reason to investigate or attempt to improve throughput. The systems in question are rarely network bound. |
|
| Back to top |
|
|
|
Kernel & Hardware
