[U-Boot-Users] common/cmd_mii.c bug and patch
The mii info command (the !CONFIG_TERSE_MII version) has a bug in current cvs. If the phy address is passed to the command the code assigns the variable addr to start, instead of the values of addrlo which is extracted from the command line. At that point in the code addr is always 0.
to demonstrate find a board with a phy at a non-zero address, compare the results of 'mii info' and mii info 1' (substitute the phy address on the test board for 1).
Also in this file on lines 55 and 418 there are two places where a call to mii_init() is conditionally compiled in, but one condition includes defined(CONFIG_MCF52x2) and the other doesn't. Is this intentional?
In message c166aa9f04122820342e4a5c55@mail.gmail.com you wrote:
The mii info command (the !CONFIG_TERSE_MII version) has a bug in current cvs. If the phy address is passed to the command the code assigns the variable addr to start, instead of the values of addrlo which is extracted from the command line. At that point in the code addr is always 0.
Added, thanks. But please don;t forget to include a CHANGELOG entry!
Also in this file on lines 55 and 418 there are two places where a call to mii_init() is conditionally compiled in, but one condition includes defined(CONFIG_MCF52x2) and the other doesn't. Is this intentional?
Probably not, but I don't have any MCF52x2 systems to test.
Best regards,
Wolfgang Denk
participants (2)
-
Andrew Dyer -
Wolfgang Denk