Hi experts,
stumbling across some kernel panics due to a faulty ethernet-driver, i recently got to use the mii command set for some analysis. Doing so, I noticed some inconvienient behaviour in the parameter choices of the mii-commands. the command in general takes 5 arguments: the 'sub'-command to execute ('read','write','dev','info' or 'dump') and four number/ranges, specifying the mii-addresses and corresponding registers. the command is intelligent somehow and remembers the last used parameters (and reuses them if none are specified), so > mii dump 1-3 0-5 > mii dump gives two exactly same outputs. this is pretty neat. what is irritating to me is: 1. why not check for all valid mii-adresses if none is specified? (i only got a PHY on mii 6, if i call 'mii dump', i want that one) 2. the initialisation of the last used parameters is unclear to me. i wasn't able to find it in the source code, and if i execute > mii dump 6 as first command after reboot, the command only dumps register 0, why not dump 0-5, since those are standards anyway? 3. when calling 'mii dump' on an invalid address, the dump doesn't deliver an error, it returns '0xffff' for the register. why? :) (this is for both, valid but unpresent addresses (e.g. 5), and invalid addresses (e.g. 100))
another example: > mii dump 6 0-5 > mii dump 100 gives 6 fine-dumped registers at first and 6 0xffff-dumped registers for address 100. I really don't think this is critical, but i guess it's not wanted...
btw: i stumbled upon a 'typo' in the mii-cmd-set:
diff --git a/common/cmd_mii.c b/common/cmd_mii.c index d3dd6b1..b82a7ce 100644 --- a/common/cmd_mii.c +++ b/common/cmd_mii.c @@ -78,9 +78,9 @@ static const MII_field_desc_t reg_3_desc_tbl[] = {
static const MII_field_desc_t reg_4_desc_tbl[] = { { 15, 15, 0x01, "next page able" }, - { 14, 14, 0x01, "reserved" }, + { 14, 14, 0x01, "(reserved)" }, { 13, 13, 0x01, "remote fault" }, - { 12, 12, 0x01, "reserved" }, + { 12, 12, 0x01, "(reserved)" }, { 11, 11, 0x01, "asymmetric pause" }, { 10, 10, 0x01, "pause enable" }, { 9, 9, 0x01, "100BASE-T4 able" },
(not critical either, but hey :))
regards, Stephan