#tyndur · 2009-11-09

54 Nachrichten · Seite 1 von 1 · Original-CSV herunterladen

  1. pcmattmankwolf: With respect to the e1000 driver: I implemented the alternative EEPROM read methods, and all they did is return zeroes
  2. kwolfBoth in VBox and qemu?
  3. pcmattmanIn QEMU it works fine, in VBox it doesn't work at all
  4. kwolfOk, so now we have two methods that work for qemu but none for VBox?
  5. pcmattmanI'm going to have to take a look at the Linux source, if it's available, and determine what's happening
  6. kwolfYou also could try the VBox source
  7. pcmattmanPerhaps, but I'd prefer reading a driver that is known to work rather than read the code of an emulation
  8. XanClicIf you're looking for that what I'm thinking at, have a look at /usr/src/linux/drivers/net/e1000
  9. XanClicSo I think it is free
  10. kwolfWell, if you like to guess once more what method VBox uses, go ahead with Linux
  11. kwolfFreakyPenguin: Do you remember what the problem with your real e1000 was? Was it reading the EEPROM?
  12. FreakyPenguinI think so, yes
  13. FreakyPenguinReading the mac address iirc
  14. pcmattmanThat's what I'm hitting in VBox
  15. XanClicDidn't you just had to remove a while?
  16. XanClicI think I remember something like that and it worked...
  17. XanClicBut I'm not sure at all
  18. kwolfFreakyPenguin: But that box/card is in Langnau, I guess, so you can't test pcmattmans patch?
  19. kwolf+'
  20. pcmattmanI have tried something like that, and ended up with an all-zero MAC, which is just as bad
  21. FreakyPenguinkwolf: I think i have one card here too...
  22. pcmattmanhttp://slexy.org/view/s2kctiAsrW <-- That updated eerpom_read function returns an all-zero MAC in VBox (and probably on real HW)
  23. kwolfCan't remember if the driver did hang when we tried it. Quite possible
  24. kwolfBut FreakyPenguin volunteers to test it, so we'll know soon ;-)
  25. pcmattmanAlso, cdi_sleep_ms - is that micro, or milli, seconds?
  26. kwolfIsn't there a comment?
  27. FreakyPenguinWhat do you want me to test? ;-)
  28. XanClicIt's milli, afair
  29. FreakyPenguinI think it should be milli ;-)
  30. pcmattmanThat's what I though, just verifying :)
  31. kwolf"Unterbricht die Ausfuehrung fuer mehrere Millisekunden"
  32. kwolfpcmattman: You can understand that without having studied German for some years, I hope? ;-)
  33. kwolfAt least the milli vs. micro part
  34. pcmattmanI can :)
  35. kwolfFreakyPenguin: What happens with the tyndur e1000 driver. Does it hang or doesn't read a wrong MAC?
  36. kwolf-doesn't +does it
  37. FreakyPenguinHum no floppy drive here... Do I really want to trust ata...? ;-)
  38. FreakyPenguinAnd no ps2 keyboard... XanClic? *g*
  39. XanClicHu?
  40. XanClicI do have one, if you wanted to know that ;-)
  41. kwolfHe wants a USB driver with HID that works on real hardware ;-)
  42. FreakyPenguinexactly *g*
  43. XanClicOuch *g*
  44. DerHartmut*g*
  45. kwolfFreakyPenguin: Serial console?
  46. kwolfOr even netconsole... wait... scratch that. *g*
  47. FreakyPenguin*g*
  48. FreakyPenguinNo null modem cable here...
  49. FreakyPenguinWell, what about a little shell script? ;-)
  50. kwolfGood luck ;-)
  51. pcmattmanHmm, I found this: http://www.koders.com/c/fid614AFE4A3AF8BEDDB131D690010328BBCED1F817.aspx
  52. pcmattmanUnfortunately GPL'd, but it is in software which works on most real e1000 NICs
  53. pcmattmanI'm off, 1:30 AM - time for sleep! :)
  54. FreakyPenguingn8 pc[away]