On 24 August 2016 at 22:13, Bjorn Andersson [off-list ref] wrote:

On Wed 24 Aug 03:22 PDT 2016, Paolo Pisati wrote:

quoted

On Wed, Aug 17, 2016 at 02:33:40PM -0500, Andy Gross wrote:

quoted

On Wed, Aug 17, 2016 at 10:48:43AM +0530, Rajendra Nayak wrote:

quoted

Hey Andy,

This is a respin of v2 with some minor fixes pointed out by Rob.
Please pull these in for 4.9

Thanks,
Rajendra

I pulled these in.

Did you try to read the content of the qfprom from userspace?

$ uname -a
Linux dragon410c 4.8.0-rc1+ #6 SMP PREEMPT Wed Aug 24 11:11:02 CEST 2016 aarch64
aarch64 aarch64 GNU/Linux
$ lsmod
Module                  Size  Used by
nvmem_qfprom           16384  0
nvmem_core             24576  1 nvmem_qfprom
$ ls -la /sys/bus/nvmem/devices/
total 0
drwxr-xr-x 2 root root 0 Aug 24 10:17 .
drwxr-xr-x 4 root root 0 Aug 24 10:15 ..
lrwxrwxrwx 1 root root 0 Aug 24 10:17 qfprom0 ->
../../../devices/platform/soc/5c000.qfprom/qfprom0
$ cat /sys/bus/nvmem/devices/qfprom0/nvmem

[spontaneous reboot]

This using agross's for-next tree as of today ("54ba896 Merge branch
'arm64-defconfig-for-4.9' into all-for-4.8") and defconfig.

This was reported in some other forum as well, after some investigation
we concluded that it looks like one of the entries are locked down -
probably from some security reason.

I'm not aware of any way to query this configuration. But the main use
case for the qfprom is in-kernel access to certain elements and that we
do get from the driver as is...

Yeah this appears to be the case.  So I guess the response is 'don't
do that'.  At least don't cat or od the file.  You need to seek and
read.

Andy