https://bugzilla.novell.com/show_bug.cgi?id=308647#c1
--- Comment #1 from Olaf Dabrunz 2007-09-07 08:23:25 MST ---
The reason for using a shift value (the ">>3" above) instead of a percentage
for the scaled crashkernel size is that the shift value will give a size
that is usually a power of two (or the same number of powers of two that the
total size of the memory has: if total_mem == 768 MB == 512 MB + 256 MB,
then crashkernel == 768 MB >> 3 == 96 MB == 64 MB + 32 MB).
One advantage is that less initial fragmentation is introduced.
But this is probably a minor point.
--
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.