On 10/17/2010 09:17 PM, Russell Senior wrote: >>>>>> "Jochen" == Jochen Friedrich <[email protected]> writes: > > Jochen> Hi Russell, >>> By judicious application of printk's (note that in order to get the >>> console back for the WGT634U to see them, I had to revert r22663) I >>> managed to figure out what is visible, and now have at least >>> broadcom-diag detecting the board again with this patch: [...] > > Jochen> Thanks a lot. I'm going to test this on my WGT634U. > > Curiously, the WGT634U does not seem to trigger get_router() in > target/linux/brcm47xx/files-2.6.34/drivers/mtd/maps/bcm47xx-flash.c, > at least I didn't get any printk's out of it. > > The get_router() function is called from init_mtd_partitions(). It > appears as the normal nvram_get function is just matching to the > section of flash that it finds with the FLSH header magic. The WGT > has one of those sections, at offset 0x018000 of mtdblock4, but the > only value is the sdram_ncdl and the meat of the board information is > located at offset 0x01e000 of mtdblock4 and has no FLSH magic. I have > no idea how uniquely you can identify the WGT based on those values > though. > > The vanilla linux kernel guesses at identifying the WGT just based on > Netgear allocated macaddrs. Seems like there ought to be a more > robust way. > > Hi WGT634U users, ;-)
I made a patch which could fix the problem: https://dev.openwrt.org/raw-attachment/ticket/7874/999-zz-wgt634u.patch Please please try it and report the results. I do not have this device so I can not try it. It should be placed at target/linux/brcm47xx/patches-2.6.36/999-zz-wgt634u.patch and an image for kernel 2.6.36 must be build. I need some reports done with this patch to fix the problem. Please report if it is working, if behavior changed or if nothing changed. Hauke _______________________________________________ openwrt-devel mailing list [email protected] https://lists.openwrt.org/mailman/listinfo/openwrt-devel
