On Thu, 27 Jun 2013 19:22:15 +0200 Stefan Tauner stefan.tauner@student.tuwien.ac.at wrote:
On Wed, 26 Jun 2013 15:59:54 -0700 Aaron Durbin adurbin@chromium.org wrote:
On Wed, Jun 26, 2013 at 3:58 PM, Stefan Tauner stefan.tauner@student.tuwien.ac.at wrote:
One thing he did not mention is that cbmem seems to not work on his device. I don't know anything about cbmem but I looked into it a bit. mapping the first address (0) works fine, but cbmem does not find anything there and the mapping of the second address (0xf0000) fails then with: Looking for coreboot table at f0000 Mapping 1MB of physical memory at 0xf0000.
any ideas why this could happen?
Could someone post the full serial log from the boot? I could take a look then.
Guess what we tried to gather from cbmem ;) I have not followed the discussion in detail, but IIRC he has no docking station and hence no serial port... and most probably no USB debugging device either.
There are some possibilities: -> cbmem -c only seem to work with native graphics init or the VGA option rom on X60/T60 -> it doesn't work after resuming from suspend -> it may depend on some compilation options of the kernel(some options filter /dev/mem etc...) -> some cmdline options of the kernel can relax some checks about /dev/mem -> you need to be root to run ./cbmem -c (included for completeness, since you already know that).
I've attached my cbmem -c on the T60.
Denis.