On Wed, 19 Nov 2008 10:09:38 +0100 Heiko Schocher hs@denx.de wrote:
Hello Heiko,
- rebased against current HEAD commit e0b0ec843085f96f4fe0738424835ee90e58bc00
drivers/qe/uec_phy.c | 59 ++++++++++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 59 insertions(+), 0 deletions(-)
sorry for the late reply, this looks like it's subject prefix should be "net:" and this patch go through Ben Warren (net maintainer), since there's really nothing 83xx specific in here.
it looks all good except for the two comments I have below.
+static struct phy_info phy_info_smsclan8700 = {
- .phy_id = 0x0007c0c0,
- .phy_id_mask = 0xfffffff0,
- .name = "SMSC LAN",
"SMSC LAN8700" (because there are other SMSC LAN PHYs with different phy_id's for that mask).
@@ -670,6 +728,7 @@ static struct phy_info *phy_info[] = { &phy_info_marvell, &phy_info_bcm5481, &phy_info_fixedphy,
- &phy_info_smsclan8700, &phy_info_genmii, NULL
if I'm not mistaken, I believe the smsclan8700 entries here should precede the fixedphy implementation (both in the phy_info list here, and for each of the chunks above, to maintain a consistent intra-file order). The reason is that if you have a board with both a smsclan8700 phy and a fixedphy (or with a use for fixedphy), uec_get_phy_info will fail to match on the smsclan8700, since the phy_info list is processed in order and will stop before the smsclan8700 because it matched the fixedphy entry first. I believe that's the case, although I could be wrong - I've never set up and run a fixed phy.
Kim