large number of minor faults 2.6.10-rc1

large number of minor faults 2.6.10-rc1

[Richard Hubbell]

I've noticed with the X process and with seti@home that they are
generating a large
number of minor faults. 

I fiddled around with these thinking it was all glibc's fault.  And
that reduced the number
(for seti, but didn't try with X)

setenv MALLOC_MMAP_MAX_ 0
setenv MALLOC_TRIM_THRESHOLD_ -1

I do see improved efficiencies after changing those glibc settings so
I am betting
that to reduce the number of minor faults even further will improve
efficiencies further.

I don't have any history of data to look at for previous kernels, so
would like to know
if anyone might like to try an educated guess as to what's going on.

I'll have to find some time to try with a previous kernel and see if it changes.
Or maybe it's purely an application problem

Thanks,
Richard