On 12/04/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
> [...] > > btw, it's sd-0.39, isn't? No! it is not. I have updated it to 2.6.21-rc5-sd-0.39.patch and did the same tests.
It's nice-dependent. At least, that's something one would expect from this design..
PRI values s constantly changing while the git-gc is running. Is this expected behaviour?
Yep. Each nice level has a set of dynamic priorities which can be be used by a task. After exhausting a time quota on a current level, the task is moved to another (if there are no more -> "expired" array). So the "PRI" field shows the current dynamic level. The exceptions are SCHED_BATCH and Real-Time (FIFO and RR) tasks which have different priority scheme. /* * This contains a bitmap for each dynamic priority level with empty slots * for the valid priorities each different nice level can have. It allows * us to stagger the slots where differing priorities run in a way that * keeps latency differences between different nice levels at a minimum. * ie, where 0 means a slot for that priority, priority running from left to * right: * nice -20 0000000000000000000000000000000000000000 * nice -10 1001000100100010001001000100010010001000 * nice 0 0101010101010101010101010101010101010101 * nice 5 1101011010110101101011010110101101011011 * nice 10 0110111011011101110110111011101101110111 * nice 15 0111110111111011111101111101111110111111 * nice 19 1111111111111111111011111111111111111111 As you can see there is only a single possible level for nice 19 tasks, that's why your first results looked strange _but_ it might have been due to the old version which lacks the prio_matrix (or it was broken.. PRI == 0 for NICE 19 is very strabge indeed). So it's good you have upgraded to the recent one. I have made a small debugging extension that allows to monitor all the priority levels (and how many times) being used by a given task (configured via proc).. will test it later (don't have sd-enabled kernel here). Just want to be sure that the "rules" are never broken. Most likely Con did it already but as he is off-line we can't know.
Did another test with hackbench and git-gc both at same priorities and captured the top output, in here all of them seem to have same priority..
Yep, this one looks good.
thanks. surya.
-- Best regards, Dmitry Adamushko - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
