* Remove KTR_CPUMASK, it is not correct to set this

* Explain why 32768 entries is usually not enough

* Increase the scaling ratio to 10 to deal with 32-bit overflows that
  can occur in calculating the canvas offsets
This commit is contained in:
kris 2008-01-13 22:46:24 +00:00
parent fa806e1d17
commit 4c2abb46e1

View File

@ -38,10 +38,11 @@
# options KTR_ENTRIES=32768 # options KTR_ENTRIES=32768
# options KTR_COMPILE=(KTR_SCHED) # options KTR_COMPILE=(KTR_SCHED)
# options KTR_MASK=(KTR_SCHED) # options KTR_MASK=(KTR_SCHED)
# options KTR_CPUMASK=0x3
# - It is encouraged to increase KTR_ENTRIES size to gather enough # - It is encouraged to increase KTR_ENTRIES size to gather enough
# information for analysis; e.g. # information for analysis; e.g.
# options KTR_ENTRIES=262144 # options KTR_ENTRIES=262144
# as 32768 entries may only correspond to a second or two of profiling
# data depending on your workload.
# - Rebuild kernel with proper changes to KERNCONF and boot new kernel. # - Rebuild kernel with proper changes to KERNCONF and boot new kernel.
# - Run your workload to be profiled. # - Run your workload to be profiled.
# - While the workload is continuing (i.e. before it finishes), disable # - While the workload is continuing (i.e. before it finishes), disable
@ -1225,7 +1226,7 @@ def fixup(self):
class SchedDisplay(Canvas): class SchedDisplay(Canvas):
def __init__(self, master): def __init__(self, master):
self.ratio = 1 self.ratio = 10
self.ktrfile = None self.ktrfile = None
self.sources = None self.sources = None
self.bdheight = 10 self.bdheight = 10