ARM FCSE - random programs segfaults

Martix martix.cz at gmail.com
Fri Mar 26 19:38:09 CET 2010


> What processor is your target running (namely, what core)?
The processor is Samsung S3C2442B [1] with ARM920T (ARMv4t) core.

> The way I test the FCSE patch is by running LTP with and without FCSE.
> And on the boards I tested it, there was no problem (LTP returned the
> same results with and without FCSE best effort mode).
>
> However, some processors may not be suitable for the FCSE patch.
>
> Have you a
> simple program which, when run, allow to reproduce these segmentation
> faults on a vanilla kernel?
Unfortunately, I don't have any simpler program than SHR-settings [2]
and it's not always reproducible, but perhaps in 7 from 10 cases. I
could provide more debug info, if you have any suggestions.

[1] http://wiki.openmoko.org/wiki/Samsung_S3C2442B_B54
[2] http://git.shr-project.org/git/?p=shr-settings.git;a=summary



More information about the openmoko-kernel mailing list