291 replies to this topic

[rals2007]

@ hnak

Up the version and/or IOProbeScore of AppleIntelE1000e perhaps ?

[hnak]

rals2007, on 23 September 2012 - 01:47 AM, said:

@ hnak

Up the version and/or IOProbeScore of AppleIntelE1000e perhaps ?

IOProbeScore may help prevent AppleIntel8254XEthernet.kext from loading, though I have never used it.

[maleorderbride]

sea_man, on 21 September 2012 - 03:49 AM, said:

I updated to 10.8.2 and the new kext , youtube and all is working very fine!

Thanks hnak !

dmesg show me this :

AppleIntelE1000e(Info): Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
AppleIntelE1000e(Info): changing MTU from 0 to 1500
AppleIntelE1000e(Info): Link is Up 1000 Mbps Full Duplex, Flow Control: Rx/Tx
AppleIntelE1000e(Info): AppleIntelE1000e:WOL = 2

I noticed that you have WoL (or I assume that is what it means).  I do not have a WoL choice in Energy Saver in System Preferences.  Is this a problem with my DSDT or is WoL only enabled for some of the chipsets that E1000E supports?

I have an X79 P9x79 Pro with 82579V running 10.7.5.    


(Thanks Hnak!  This driver is working great for me)

[hnak]

maleorderbride, on 23 September 2012 - 11:15 PM, said:

I noticed that you have WoL (or I assume that is what it means).  I do not have a WoL choice in Energy Saver in System Preferences.  Is this a problem with my DSDT or is WoL only enabled for some of the chipsets that E1000E supports?

I have an X79 P9x79 Pro with 82579V running 10.7.5.


(Thanks Hnak!  This driver is working great for me)

It is a WOL setting value set by BIOS. I display the value just for reference.

[Mull7965]

hnak, on 08 September 2012 - 05:50 AM, said:

The issue was  (probably) that Mac OS X did not check the actual value of UDP  checksum once the driver reports the sum is verified - until 10.8.
Intel hardware "complements the payload checksum" ( as written in the original source file) , so it must be undo-ed for the host which re-checks the value for consistency.
I simply disabled hardware UDP checksum and let the OS do it.

Thanks for the change hnak. One question: did you commit the code? I'm looking at SourceForge (http://sourceforge.n...rs/code/45/log/) and the last commit is from late June. Am i missing something?

[hnak]

Sorry, I have not updated sourceforge.net repo since I upgraded to sourceforge 2.0.
I have just moved my old repo to the new one.

[rals2007]

hnak

This is off-topic. But i was just wondering if you can port this Ralink 2860 linux driver to work on OSX x32 and x64.
Last time they updated this was for Leopard.

Thank You

Attached Files

•  Ralink RT2860Â Linux.zip   2.85MB   6 downloads

[Mr. Blonde]

Even with the latest kext, I'm stilling having the same problem. I've installed the kext with a program and manually in IONetworkingFamily.kext. Did I do something wrong? (Mountain Lion)

[hnak]

rals2007, on 01 October 2012 - 06:08 AM, said:

hnak

This is off-topic. But i was just wondering if you can port this Ralink 2860 linux driver to work on OSX x32 and x64.
Last time they updated this was for Leopard.

Thank You

It is practically impossible, as :
1. I do not have the hardware
2. There's no sample source for USB network driver ( not like PCI).

If the source for old Mac OS X driver is available, it should not be too difficult...

[Mr. Blonde]

Do you have the correct installation instructions? Thanks

[rals2007]

hnak, on 05 October 2012 - 12:54 PM, said:

It is practically impossible, as :
1. I do not have the hardware
2. There's no sample source for USB network driver ( not like PCI).

If the source for old Mac OS X driver is available, it should not be too difficult...

hnak

It is a PCI card

[Mr. Blonde]

To specify it further:

Short Youtube clips cause NO problems.
Long clips start buffer half way, and never play again
Clips with an advertisement in front of the clip, cause most of the problems

(Using Mountain Lion 10.8.2)

[hnak]

rals2007, on 06 October 2012 - 05:28 AM, said:

hnak

It is a PCI card

Sorry, I mistook  it  for something dieffrent. The term 1 is still valid, however.

Mr. Blonde, on 06 October 2012 - 11:27 AM, said:

To specify it further:

Short Youtube clips cause NO problems.
Long clips start buffer half way, and never play again
Clips with an advertisement in front of the clip, cause most of the problems

(Using Mountain Lion 10.8.2)

Please specify reproducible "long clip" URLs so that I can check if it really happens in my environment.

[Mr. Blonde]

Happens after several minutes. Thanks!

[hnak]

Mr. Blonde, on 07 October 2012 - 08:46 PM, said:

Happens after several minutes. Thanks!

I could play through the movie without stall, on my 82572.

[Mr. Blonde]

Would you like to pm me proper installation instructions? Thanks in advance!

[hnak]

Mr. Blonde, on 08 October 2012 - 06:50 AM, said:

Would you like to pm me proper installation instructions? Thanks in advance!

You should be able to find such a instruction everywhere.
Open terminal.app and go to the directory and execute  the following commands:

If you have an old one in /System/Library/Extensions :
1. sudo kextunload /System/Library/Extensions/AppleIntelE1000e.kext
2. sudo rm -rf /System/Library/Extensions/AppleIntelE1000e.kext

Then install a new one:
1. cd [directory where AppleIntelE1000e.kext exists]
1. sudo chown -R root:wheel AppleIntelE1000e.kext
2. sudo mv AppleIntelE1000e.kext /System/Library/Extensions/

The next command loads the kext immediately.
1. sudo kextload /System/Library/Extensions/AppleIntelE1000e.kext

[Mr. Blonde]

I think I've found the problem: the player-mode was switched to HTML5. In flash it doesn't stall. Hope the problems stays away!

Thanks for the instructions!

[hnak]

Mr. Blonde, on 08 October 2012 - 09:43 PM, said:

I think I've found the problem: the player-mode was switched to HTML5. In flash it doesn't stall. Hope the problems stays away!

It is exactly the problem introduced in 10.8. Will you show me any html5-playable URL ? The  above one seems to require Flash.

[khe91]

Mr. Blonde, on 08 October 2012 - 09:43 PM, said:

I think I've found the problem: the player-mode was switched to HTML5. In flash it doesn't stall.

I don't have Flash installed and didn't experience any Youtube problems.
Version 2.1.7., Intel 82579V (8086 1503)
Maybe the problem comes only for some Intel ethernet chipsets that are supported by the driver.