Re: [PATCH v4] kmemleak: survive in a low-memory situation

[Qian Cai]

On Thu, 2019-03-28 at 15:05 +0000, Catalin Marinas wrote:
> > It takes 2 runs of LTP oom01 tests to disable kmemleak.
> 
> What configuration are you using (number of CPUs, RAM)? I tried this on
> an arm64 guest under kvm with 4 CPUs and 512MB of RAM, together with
> fault injection on kmemleak_object cache and running oom01 several times
> without any failures.

Apparently, the CPUs are so fast and the disk is so slow (swapping). It ends up
taking a long time for OOM to kick in.

# lscpu
Architecture:ÂÂÂÂÂÂÂÂx86_64
CPU op-mode(s):ÂÂÂÂÂÂ32-bit, 64-bit
Byte Order:ÂÂÂÂÂÂÂÂÂÂLittle Endian
CPU(s):ÂÂÂÂÂÂÂÂÂÂÂÂÂÂ48
On-line CPU(s) list: 0-47
Thread(s) per core:ÂÂ2
Core(s) per socket:ÂÂ12
Socket(s):ÂÂÂÂÂÂÂÂÂÂÂ2
NUMA node(s):ÂÂÂÂÂÂÂÂ2
Vendor ID:ÂÂÂÂÂÂÂÂÂÂÂGenuineIntel
CPU family:ÂÂÂÂÂÂÂÂÂÂ6
Model:ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ85
Model name:ÂÂÂÂÂÂÂÂÂÂIntel(R) Xeon(R) Gold 6126T CPU @ 2.60GHz
Stepping:ÂÂÂÂÂÂÂÂÂÂÂÂ4
CPU MHz:ÂÂÂÂÂÂÂÂÂÂÂÂÂ3300.002
BogoMIPS:ÂÂÂÂÂÂÂÂÂÂÂÂ5200.00
Virtualization:ÂÂÂÂÂÂVT-x
L1d cache:ÂÂÂÂÂÂÂÂÂÂÂ32K
L1i cache:ÂÂÂÂÂÂÂÂÂÂÂ32K
L2 cache:ÂÂÂÂÂÂÂÂÂÂÂÂ1024K
L3 cache:ÂÂÂÂÂÂÂÂÂÂÂÂ19712K
NUMA node0 CPU(s):ÂÂÂ0-11,24-35
NUMA node1 CPU(s):ÂÂÂ12-23,36-47

# free -m
ÂÂÂÂÂÂÂÂÂÂÂÂÂÂtotalÂÂÂÂÂÂÂÂusedÂÂÂÂÂÂÂÂfreeÂÂÂÂÂÂsharedÂÂbuff/cacheÂÂÂavailable
Mem:ÂÂÂÂÂÂÂÂÂ166206ÂÂÂÂÂÂÂ31737ÂÂÂÂÂÂ134063ÂÂÂÂÂÂÂÂÂÂ33ÂÂÂÂÂÂÂÂÂ406ÂÂÂÂÂÂ133584
Swap:ÂÂÂÂÂÂÂÂÂÂ4095ÂÂÂÂÂÂÂÂÂÂÂ0ÂÂÂÂÂÂÂÂ4095

# lspci | grep -i sata
00:11.5 SATA controller: Intel Corporation C620 Series Chipset Family SSATA
Controller [AHCI mode] (rev 08)
00:17.0 SATA controller: Intel Corporation C620 Series Chipset Family SATA
Controller [AHCI mode] (rev 08)