Monday, March 1, 2004

Sorry! I apologize if you've subscribed to this RSS feed and I haven't posted any new progress in a while. I've been off working on some important stuff for the next OpenDarwin release and didn't focus on this project. I'll try and post some updates later this week! 10:29:53 PM    comment []

SYM much closer SYM selection now works for both 10 and 100 Mb speeds. A few problems remain. 1. Unplugging and plugging the cable results in a fubar interface. It senses the link going down, and sees it come up again, but it won't transmit after this. The workaround is to down/up the interface. Yuck. 2. Unplugging from a 10-half hub port and plugging into a 100-full port does work a little bit. Transmit functions, but duplex is off. 3. Using ifconfig, you cannot override the speed or duplex of the interface. I think I know why (has to do with the capabilities override I do). I'll try and fix this later, but as of right now #1 and #2 are more important to fix. If anyone would like to test this code, I can provide a tarball that will match against DEC2114x and PNIC2 cards. I need feedback on both. As soon as I firm up the support a bit more, I'll commit the changes (and new files) to opendarwin's CVS repository. 10:28:38 PM    comment []

SYM media selection almost working I have SYM media selection mostly working for the DEC 21143. Previously, the DEC 21143 was supported but only if it had MII (media independent interface).  Some cards use the SYM (symbol port) for setting media, and so I had to add this in too. Right now it is working at 10/half. I haven't gotten it to negotiate anything faster yet. I did something interesting. I took all my MII code and dummied up the _readStatus, _readControl and _writeControl methods to write to the 3 different registers necessary for SYM support (CSR12, CSR13 and CSR14). So most of the code looks identical to the MII stuff. I don't see why the 100 Mb and full-duplex settings are so hard to get working, but I'll keep plugging away on it. As soon as this is done, adding in the PNIC2 support should take "minutes." 4:55:56 PM    comment []