Bug#404143: Fans unreliable under load, permanent memory leak
Some more information.
1) On my machine, reading the temperature using, say, yacpi, causes
one processor to process all the pending ACPI events. On a
uniprocessor machine, the machine would appear to hang for several
seconds; not so on my dual-core machine :)
2) The lare slab usage (1.1 Gb) was in part due to the XFS cache data;
all three of my machine's filesystems are XFS. So the Acpi-State
line in /proc/slabinfo is the really meaningful one.
Here is my complete log so far, with annotations.
2006-06-21T20:06:10: Slab: 30296 kB
2006-17-21T20:17:01: Slab: 37756 kB
2006-17-21T21:17:01: Slab: 48116 kB
2006-17-21T22:17:01: Slab: 55764 kB
2006-17-21T23:17:01: Slab: 69904 kB
-- Reboot with acpi=noirq: only one CPU found --
2006-24-21T23:24:10: Slab: 10444 kB
-- Reboot with pci=noacpi: only one CPU found --
2006-30-21T23:30:26: Slab: 9676 kB
2006-30-21T23:30:26: Acpi-State 0 0 80 48 1 : tunables 120 60 8 : slabdata 0 0 0
-- Reboot with no options: OK, both CPUs found --
2006-34-21T23:34:23: Slab: 10584 kB
2006-34-21T23:34:23: Acpi-State 0 0 80 48 1 : tunables 120 60 8 : slabdata 0 0 0
2006-17-22T00:17:01: Slab: 15424 kB
2006-17-22T00:17:01: Acpi-State 23088 23088 80 48 1 : tunables 120 60 8 : slabdata 481 481 0
2006-17-22T01:17:01: Slab: 29956 kB
2006-17-22T01:17:01: Acpi-State 59136 59136 80 48 1 : tunables 120 60 8 : slabdata 1232 1232 0
2006-17-22T02:17:01: Slab: 37764 kB
2006-17-22T02:17:01: Acpi-State 95088 95088 80 48 1 : tunables 120 60 8 : slabdata 1981 1981 0
2006-17-22T03:17:01: Slab: 45544 kB
2006-17-22T03:17:01: Acpi-State 130992 130992 80 48 1 : tunables 120 60 8 : slabdata 2729 2729 0
2006-17-22T04:17:01: Slab: 53328 kB
2006-17-22T04:17:01: Acpi-State 166944 166944 80 48 1 : tunables 120 60 8 : slabdata 3478 3478 0
2006-17-22T05:17:01: Slab: 61120 kB
2006-17-22T05:17:01: Acpi-State 202896 202896 80 48 1 : tunables 120 60 8 : slabdata 4227 4227 0
2006-17-22T06:17:01: Slab: 68904 kB
2006-17-22T06:17:01: Acpi-State 238800 238800 80 48 1 : tunables 120 60 8 : slabdata 4975 4975 0
2006-17-22T07:17:01: Slab: 1152624 kB
2006-17-22T07:17:01: Acpi-State 274656 274656 80 48 1 : tunables 120 60 8 : slabdata 5722 5722 0
2006-17-22T08:17:01: Slab: 1160376 kB
2006-17-22T08:17:01: Acpi-State 310608 310608 80 48 1 : tunables 120 60 8 : slabdata 6471 6471 0
2006-17-22T09:17:01: Slab: 1168168 kB
2006-17-22T09:17:01: Acpi-State 346464 346464 80 48 1 : tunables 120 60 8 : slabdata 7218 7218 0
2006-17-22T10:17:01: Slab: 1175892 kB
2006-17-22T10:17:01: Acpi-State 382176 382176 80 48 1 : tunables 120 60 8 : slabdata 7962 7962 0
2006-17-22T11:17:01: Slab: 1183660 kB
2006-17-22T11:17:01: Acpi-State 417984 417984 80 48 1 : tunables 120 60 8 : slabdata 8708 8708 0
2006-17-22T12:17:01: Slab: 1191400 kB
2006-17-22T12:17:01: Acpi-State 453744 453744 80 48 1 : tunables 120 60 8 : slabdata 9453 9453 0
2006-17-22T13:17:01: Slab: 1202924 kB
2006-17-22T13:17:01: Acpi-State 489696 489696 80 48 1 : tunables 120 60 8 : slabdata 10202 10202 0
-- Start yacpi, monitoring the temperature every second.
-- Note how the slab allocation drops by ~100M and then stays constant.
2006-17-22T14:17:01: Slab: 1097584 kB
2006-17-22T14:17:01: Acpi-State 109 144 80 48 1 : tunables 120 60 8 : slabdata 3 3 0
2006-17-22T15:17:01: Slab: 1097532 kB
2006-17-22T15:17:01: Acpi-State 45 96 80 48 1 : tunables 120 60 8 : slabdata 2 2 0
2006-17-22T16:17:01: Slab: 1097536 kB
2006-17-22T16:17:01: Acpi-State 75 144 80 48 1 : tunables 120 60 8 : slabdata 3 3 0
2006-17-22T17:17:01: Slab: 1097668 kB
2006-17-22T17:17:01: Acpi-State 141 144 80 48 1 : tunables 120 60 8 : slabdata 3 3 0
-- Stop the yacpi monitoring.
2006-17-22T18:17:01: Slab: 1098904 kB
2006-17-22T18:17:01: Acpi-State 5808 5808 80 48 1 : tunables 120 60 8 : slabdata 121 121 0
-- At this point the Acpi-State has started increasing again, but is still
-- small. Most of the slab allocations are in the XFS caches (all three
-- filesystems on this computer are XFS).
-- To make sure the memory can be released, start a fairly large compilation
-- using both CPUs and 2x370 M of RAM. Just before compilation:
2006-48-22T18:48:56: Slab: 1103244 kB
2006-48-22T18:48:56: Acpi-State 24528 24528 80 48 1 : tunables 120 60 8 : slabdata 511 511 0
-- A couple of minutes into the compilation, the fans have still not turned on
-- and the CPU is getting so hot it burns my hand. Restart yacpi, monitoring
-- temperature every second. The temp is 85°C (dangerous!!) One CPU starts
-- processing the backlog of ACPI events, the other continues the compilation.
-- Fans start. Temperature drops to 71°C and stays there.
2006-00-22T19:00:44: Slab: 861828 kB
2006-00-22T19:00:44: Acpi-State 74 96 80 48 1 : tunables 120 60 8 : slabdata 2 2 0
-- End of compilation. During the final packaging stages, the temperature has
-- dropped to 57°C as the CPUs were less used. Stop the yacpi monitoring.
2006-07-22T19:07:13: Slab: 865660 kB
2006-07-22T19:07:13: Acpi-State 73 96 80 48 1 : tunables 120 60 8 : slabdata 2 2 0
2006-17-22T19:17:01: Slab: 865028 kB
2006-17-22T19:17:01: Acpi-State 71 144 80 48 1 : tunables 120 60 8 : slabdata 3 3 0
2006-17-22T20:17:01: Slab: 871224 kB
2006-17-22T20:17:01: Acpi-State 34704 34704 80 48 1 : tunables 120 60 8 : slabdata 723 723 0
2006-17-22T21:17:01: Slab: 879112 kB
2006-17-22T21:17:01: Acpi-State 69552 69552 80 48 1 : tunables 120 60 8 : slabdata 1449 1449 0
2006-17-22T22:17:01: Slab: 887908 kB
2006-17-22T22:17:01: Acpi-State 104784 104784 80 48 1 : tunables 120 60 8 : slabdata 2183 2183 0
2006-17-22T23:17:01: Slab: 896024 kB
2006-17-22T23:17:01: Acpi-State 139920 139968 80 48 1 : tunables 120 60 8 : slabdata 2915 2916 0
Reply to: