Re: beaglebone black: is mem=... broken?

2015-01-19 Thread Paolo Pisati
On Sun, Jan 18, 2015 at 06:38:46PM +0100, Geert Uytterhoeven wrote: > > The boot loader copied the DT to the end of real RAM, not to the end of > the 256 MiB block? Hence the kernel accesses unmapped memory > when checking the FDT header? actually everything is below 256M since physical memory st

Re: beaglebone black: is mem=... broken?

2015-01-18 Thread Geert Uytterhoeven
On Sun, Jan 18, 2015 at 3:34 PM, Paolo Pisati wrote: > Boot hangs when passing mem=256M to a 3.16 kernel (but i was able to reproduce > it with multi_v7_defconfig on a 3.19rcX kernel too): > > 80e6d694: 0024 65746e49 6c616e72..$.Internal > 80e6d6a4: 72726520 203a726f 73706f4f

beaglebone black: is mem=... broken?

2015-01-18 Thread Paolo Pisati
Boot hangs when passing mem=256M to a 3.16 kernel (but i was able to reproduce it with multi_v7_defconfig on a 3.19rcX kernel too): 80e6d694: 0024 65746e49 6c616e72..$.Internal 80e6d6a4: 72726520 203a726f 73706f4f 2035203a error: Oops: 5 80e6d6b4: 5d31235b 504d5320 4d5241