| commit | e17ba8f51f69782abc4575dd30848ba5eafaa797 | |
| author | David S. Miller <davem@davemloft.net> | |
| Sun, 24 Sep 2006 01:31:42 +0000 (23 18:31 -0700) | ||
| committer | David S. Miller <davem@sunset.davemloft.net> | |
| Sun, 24 Sep 2006 01:34:06 +0000 (23 18:34 -0700) | ||
| tree | 534d829913806ce56e32f3d2e0404f3ab236bddf | treesnapshot (tar.gz zip) |
| parent | 1ab9dd0902df4f4ff56fbf672220549090ab28ba | commitdiff |
[SPARC64]: Fix sched_clock() wrapping every ~17 seconds.
Unfortunately, sparc64 doesn't have an easy way to do a "64 X 64 -->
128" bit multiply like PowerPC and IA64 do. We were doing a
"64 X 64 --> 64" bit multiple which causes overflow very quickly with
a 30-bit quotient shift.
So use a quotientshift count of 10 instead of 30, just like x86 and
ARM do.
This also fixes the wrapping of printk timestamp values every ~17
seconds.
Signed-off-by: David S. Miller <davem@davemloft.net>
Unfortunately, sparc64 doesn't have an easy way to do a "64 X 64 -->
128" bit multiply like PowerPC and IA64 do. We were doing a
"64 X 64 --> 64" bit multiple which causes overflow very quickly with
a 30-bit quotient shift.
So use a quotientshift count of 10 instead of 30, just like x86 and
ARM do.
This also fixes the wrapping of printk timestamp values every ~17
seconds.
Signed-off-by: David S. Miller <davem@davemloft.net>
| arch/sparc64/kernel/time.c | diffblobblamehistory |