On 28/04/16 17:42, Bojan Prtvar wrote: > Hello everyone, > > I need to test all RAM cells on a linux ARM embedded system. My use case > is very similar to the one described in [1] expect the fact I also have > strong requirements on minimizing the boot time impact. > Instead of doing that from the bootloader, I decided to evaluate the > linux memtest feature introduced with [2]. > > My questions are: > > 1) > Does the early_memtest() as called in [3] really covers *all* RAM cells? No. > > 2) > As memtest happens very early in boot stage, what primitives I can use > to measure duration of early_memtest()? Are there any known heuristics? > I need to test ~2GB of RAM. > This is my major concern. Stopwatch? > > 3) > It seems reasonable to expose the number of detected bad cells to user > space. I was thinking about sysfs. Are the patches welcomed? > $ dmesg | grep "bad mem" Cheers Vladimir > [1] > http://www.linuxforums.org/forum/newbie/173847-how-do-memory-ram-test-when-linux-running.html > > [2] > http://lkml.iu.edu/hypermail/linux/kernel/1503.1/00566.html > [3] > http://lxr.free-electrons.com/source/arch/arm/mm/init.c#L291 > > Thanks, > Bojan > > -- > To unsubscribe, send a message with 'unsubscribe linux-mm' in > the body to majordomo@xxxxxxxxx. For more info on Linux MM, > see: http://www.linux-mm.org/ . > Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a> > > -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>