Comment 7 for bug 45014

Another vote for supporting tp_smapi. I've reported the same problems as others in the various related bugs (each with a different model or two), and have it working on my T61p following:

  http://www.thinkwiki.org/wiki/Tp_smapi
  http://www.nabble.com/tp_smapi-0.34-and-hdaps-td14712633.html

Shem, by the way, I've got the same EC values as were reported on the thread, and patched the call in hdaps.c to use:

  if (!((data.val[0x1]!=0x00 || data.val[0x2]!=0x60 ||data.val[0x3]!=0x00 || data.val[0xF]!=0x00) ||
      (data.val[0x1]!=0x01 || data.val[0x2]!=0x00 ||data.val[0x3]!=0x00 || data.val[0xF]!=0x00))) {
          printk(KERN_WARNING
   "bad hdaps_check_ec reply 0x%02x,0x%02x,0x%02x,0x%02x",
   data.val[0x1], data.val[0x2],
   data.val[0x3], data.val[0xF]);

   return -EIO;
 }

I didn't have to make any changes to the kernel, I just needed to build and load tp_smapi and the replacement hdaps module, so this could be put out as an alternative package, so long as it was kept in sync with the kernel releases, if there is some reason for keeping the mainline kernel module.