On 06/13/2015 01:52 AM, Dave Chinner wrote:
> Yup, that's looking like a toolchain bug. Thread about arm directory
> read corruption:
>
> http://oss.sgi.com/archives/xfs/2013-02/msg00505.html
>
> cross-gcc version results:
>
> http://oss.sgi.com/archives/xfs/2013-02/msg00563.html
>
> "A quick rundown:
> -cross-gcc4.4: OK
> -cross-gcc4.5: OK
> -cross-gcc4.6: BAD
> -cross-gcc4.7: BAD
> -cross-gcc4.8: OK"
>
Just tested the new kernels, they're both good:
GOOD: 3.14.3-std-00094-g9035cb4, gcc version 4.7.3 (Ubuntu/Linaro
4.7.3-12ubuntu1)
GOOD: 3.14.44-std-00095-g0425932, gcc version 4.9.2 (4.9.2-10)
BAD: 3.14.3-00088-g7651c68, gcc version 4.6.3 (Ubuntu/Linaro 4.6.3-1ubuntu5)
On 06/12/2015 03:21 PM, Brian Foster wrote:
> I echo Dave's suggestion down thread with regard to toolchain. This
> kernel was compiled with the following cross-gcc (installed via Fedora
> package):
>
> gcc version 4.9.2 20150212 (Red Hat Cross 4.9.2-5) (GCC)
>
> Are you using something different?
>
> Brian
Indeed, it looks like a compiler bug, thanks a lot for helping me track it down.
I'll see if I can find out more about whats different between the two kernels
compiled by 4.6 and 4.7.
On 06/13/2015 12:41 AM, Karanvir Singh wrote:
>
> Hi Edwin,
>
>
> PFA the newer uimages:
>
> uImage3.14.44.gcc.4.9.2: its a 3.14.44 compiled with gcc 4.9.4
> uImage.gcc.4.7.2-1: its 3.14.3 compiled with gcc 4.7.2
Thanks, both of these images work correctly: I was not able to reproduce the
bug with them (rebooting to the original uImage 3.14.3 with gcc 4.6.3
reproduces bug immediately).
Best regards,
--Edwin
|