-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Somebody in the thread at some point said:
|> The best way I heard to explain it yet is just a mismatch between the |> worst case slowness of that card and the biggest timeout the Glamo can |> express at the faster clock rates. If the timeout register had another |> nybble in its counter we would just have set it to FFFF and forgotten |> about it. | Ah, OK. Could this also be related to the length of the transfer? What | if I reduce the number of blocks that can be read/written at once? | Something with req_size or seg_size? Maybe... but I seem to recall they are all quite modest, 4096 bytes or something (dim memory though). - -Andy -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iEYEARECAAYFAkgnS2IACgkQOjLpvpq7dMrSawCfaEVNeP75T4Ok0r2BggtNYNYP kO0An0HU3N72H2gnpAo5Gj8lHSL1UOeJ =Uugd -----END PGP SIGNATURE-----
