Jump to content

AppleIntelE1000e.kext for 10.8/10.7/10.6/10.5


hnak
 Share

751 posts in this topic

Recommended Posts

  • 2 weeks later...

I compiled the latest version on Mavericks. It's 64-bit only.

 

Hopefully it works ok. I haven't checked as I'm still on 2.4.14 here.

I can confirm that this works with 10.9.2 & a Gigabyte UD7-TH Motherboard. No more issues with large file transfers.

 

If you are running a iSCSI based SAN, this is the driver to use. Very smooth & reliable transfers.

  • Like 1
Link to comment
Share on other sites

I compiled the latest version on Mavericks. It's 64-bit only.

 

Hopefully it works ok. I haven't checked as I'm still on 2.4.14 here.

 

 

I can confirm that this works with 10.9.2 & a Gigabyte UD7-TH Motherboard. No more issues with large file transfers.

 

If you are running a iSCSI based SAN, this is the driver to use. Very smooth & reliable transfers.

 

 

I can also confirm that on 10.9.2 and an Asus P8P67 Pro, that I also have no more issues with large file transfers using Riley Freeman's compiled version.  Thanks!

 

EDIT:  Ends up that it still crashes, more stable than the 2.5.4 option, but alas no go. Back to 2.4.14 here.

Link to comment
Share on other sites

I compiled the latest version on Mavericks. It's 64-bit only.

 

Hopefully it works ok. I haven't checked as I'm still on 2.4.14 here.

 

I'm still experiencing hang errors while transferring 200GB or bigger directories to other (OSX) computers.

 

My mobo is a GA-X79-UP4, I'm on OSX 10.9.2

 

I have repaired permissions, tried to install the kext into the plugin directory and also directly into the extensions directory, tried to boot with -f, nothing seems to work, it also hang randomly, some times after transferring 20gb other when just transferring a few MB.

Link to comment
Share on other sites

looks like disabling TSO might help in the errors when doing large transfers, it can be done with a sysctl switch but i guess that would turn it off on all interfaces, so I'm trying to do this via ifconfig and the -mediaopts switch, i messed around with this before for turning on/off functions like avb and eee (energy efficient ethernet) now it seems it won't let me set any options or I'm using the wrong syntax. If anyone cares the sysctl switch i mentioned is net.inet.tcp.tso (defaults to 1) 

Link to comment
Share on other sites

  • 2 weeks later...

Hello, i'm sorry for my newbish question, but is this driver supposed to enable the intel i218v ? Because every time i enable the Lan in bios it just hangs at the line "changing mtu from 0 to 1500", if i disable the ethernet port in bios it works just fine, very weird...

 

Thank you!

Link to comment
Share on other sites

Hello, i'm sorry for my newbish question, but is this driver supposed to enable the intel i218v ? Because every time i enable the Lan in bios it just hangs at the line "changing mtu from 0 to 1500", if i disable the ethernet port in bios it works just fine, very weird...

 

Thank you!

I forgot to add those new ID values in my last change. I uploaded the new build.

I hope mbuf leak problem is alleviated, too.

  • Like 2
Link to comment
Share on other sites

total noob here but trying to get large transfers to work using this kext....

 

does this look right? have just installed newest version and was checking over the info.plist 

<key>IOProviderClass</key><string>IOPCIDevice</string>
<key>NETIF_F_TSO</key>
<true/>
</dict>

should that be in a <string> ?

 

Link to comment
Share on other sites

  • 2 weeks later...

Hello,

 

Is diddl14's Wake on Lan improvement integrated with the latest release of the driver ?

 

I would have tested the modified kext, but the url does not work anymore.

 

Can someone post it to give a try ?

 

Thanks,

 

Romain

Link to comment
Share on other sites

  • 2 weeks later...

Hello hnak,

I've just installed AppleIntelE1000e driver and observe when I copy large set of files from one hackintosh to another I've got driver crash.

It happens only when I copy files via SMB from my hackintosh to macbook and not happens when I copy files to Windows PC. Also it doesn't happen when I copy files from macbook to hackintosh.

 

While I copy file from my hackintosh to macbook I got log entry:

e1000_tx_map: failed to getphysicalsegment.

And when driver crash: 

AppleIntelE1000e(Err): Detected Hardware Unit Hang:

 

Also I'm viewing the forum and not seeing any problems at all. It seems only happens with SMB connection while transmitting file to another mac.

 

My System Configuration:

OS: Mac OX X 10.9.3, No DSDT

MB: Asus Sabertooth Z77 (Intel® 82579V)

CPU: Intel Core i5 3570K

Video: Gtx 670 2GB

Storage: 128Gb SSD Plextor, 1.5TB HDD Seagate (I copied from both drives, the same result)  

Driver version: 3.0.1 (2014/6/26)

 

Macbook Configuration:

Mac OS X 10.9.3

Macbook Pro Retina Late 2013 13" 128GB SSD

 

Could you investigate this issue.

 

Thanks.

 

You could see my system log in attachment. If you'll need additional tests or traces please let me know.

AppleIntel1000eHang.txt

Link to comment
Share on other sites

I have the same problem as k1x0r, except that:

 

The errors are for any protocol (not just SMB or AFP).  In particular, for me, the protocol used by Chronosync.

 

I get "e1000_tx_map: failed to getphysicalsegment" frequently whenever doing any transfer.

 

The "AppleIntelE1000e(Err): Detected Hardware Unit Hang:" are rare, but when they happen they repeat endlessly and require a reboot.

 

The detail I have for mine is:

Jul  2 16:42:40 Adam kernel[0]: AppleIntelE1000e(Err): Detected Hardware Unit Hang:
Jul  2 16:42:40 Adam kernel[0]: TDH                  <21>
Jul  2 16:42:40 Adam kernel[0]: TDT                  <f1>
Jul  2 16:42:40 Adam kernel[0]: next_to_use          <f1>
Jul  2 16:42:40 Adam kernel[0]: next_to_clean        <1b>
Jul  2 16:42:40 Adam kernel[0]: buffer_info[next_to_clean]:
Jul  2 16:42:40 Adam kernel[0]: time_stamp           <2cb74>
Jul  2 16:42:40 Adam kernel[0]: next_to_watch        <21>
Jul  2 16:42:40 Adam kernel[0]: next_to_watch.status <0>
Jul  2 16:42:40 Adam kernel[0]: MAC Status             <80043>
Jul  2 16:42:40 Adam kernel[0]: PHY Status             <796d>
Jul  2 16:42:40 Adam kernel[0]: PHY 1000BASE-T Status  <0>
Jul  2 16:42:40 Adam kernel[0]: PHY Extended Status    <3000>
Jul  2 16:42:40 Adam kernel[0]: PCI Status             <10>

 

I have a GA-Z87X-UD3H and the Device ID is 0x153b.

Link to comment
Share on other sites

i keep trying these newer (3.x) versions and i am getting the same errors/issues listed above. i have read on other forums (linux and/or intel dev site etc) that these "hangs" maybe related to the driver and multi-port cards i.e. not able to get at or set a 2nd or non-existing port - in my case i have a single port on my laptop. FWIW, I rollback to v2.4.14 and i do not see any of these problems.

Link to comment
Share on other sites

i keep trying these newer (3.x) versions and i am getting the same errors/issues listed above. i have read on other forums (linux and/or intel dev site etc) that these "hangs" maybe related to the driver and multi-port cards i.e. not able to get at or set a 2nd or non-existing port - in my case i have a single port on my laptop. FWIW, I rollback to v2.4.14 and i do not see any of these problems.

 

Have you tried turning off NETIF_F_TSO in the kext's info.plist? I think that was one of the big changes from 2.4.14.

Link to comment
Share on other sites

I disabled this property, hang seems is gone, but I still get "e1000_tx_map: failed to getphysicalsegment" while copying file. 

Also I noticed slow transmit speed (6-10MB/s) in version 3.0.1 in compare to version 3.0.4.1a (30MB/s). 

Update: downgraded to 3.0.4.1a, set NETIF_F_TSO to false, don't see any errors at all and got 30-40Mb transmission speed.

Link to comment
Share on other sites

I have (for now) stayed with v3.1.0 because it seemed to be faster than older versions - though I admit I have not done a thorough comparison.  But I have set NET_F_TSO to false, as Riley suggested.  The "e1000_tx_map: failed to getphysicalsegment" errors have gone away :)   The "AppleIntelE1000e(Err): Detected Hardware Unit Hang" were rare (though fatal) for me so will have to wait and see if they are solved too.

Link to comment
Share on other sites

 Share

×
×
  • Create New...