The axe interface when using a USB Netgear FA120 will get multiple read PHY failed and then become inoperative. Unloading if_axe.ko and then reloading brings the adapter back up. The OS is running on a VIA ML1000 mini itx motherboard. Unit boots from flash so root is read only. I have the same problem on the via CL1000 also. I have tried other FA120s and have ruled out bad adpater hardware. How-To-Repeat: Unknown cause ??? Seems to be random. If if_axe.ko is buit into the kernel the unit will not completely boot. Maybe this is somehow related ?
I had the same problem on RELENG_5 but since updating to RELENG_6 it works without a problem. Can you try updating to RELENG_6 ?
Nope it is still a problem in 6.0 for me. Seems to only want to do it when compiling a port from a telnet session.... Thanks Mike
I had the same problem on an EPIA-based box running a recent RELENG_6 equipped with both a D-Link and an SMC USB Ethernet adapter (both using the axe driver). The symptoms were identical; after a period of time one of the interfaces locked up, spewing read PHY failed messages. Tried to limit it by forcing the interfaces to 10baseT/UTP, and this made them recover automatically: Aug 20 11:33:09 epia kernel: axe0: link state changed to DOWN Aug 20 11:33:09 epia kernel: axe0: read PHY failed Aug 20 11:33:09 epia kernel: axe0: read PHY failed Aug 20 11:33:09 epia kernel: axe0: link state changed to UP Aug 20 11:33:09 epia kernel: axe0: read PHY failed Aug 20 11:33:09 epia kernel: axe0: read PHY failed Aug 20 11:33:09 epia kernel: axe0: link state changed to DOWN Aug 20 11:33:09 epia kernel: axe0: read PHY failed Aug 20 11:33:09 epia kernel: axe0: link state changed to UP However, forcing the interface to full-duplex made the problem go away, even at 100baseTX and heavy load. The driver could perhaps handle duplex mismatches a bit more gracefully?
For bugs matching the following criteria: Status: In Progress Changed: (is less than) 2014-06-01 Reset to default assignee and clear in-progress tags. Mail being skipped