> 2022年6月29日 15:56，Nikhil Kshirsagar <nkshirsagar@xxxxxxxxx> 写道： > > 60gb slow hdd, 15gb fast nvme cache. shall i file upstream bug for this issue? If you may trigger this on upstream kernel or stable kernel, please do it. Coly Li > > On Wed, 29 Jun 2022 at 13:20, Coly Li <colyli@xxxxxxx> wrote: >> >> >> >>> 2022年6月29日 15:02，Nikhil Kshirsagar <nkshirsagar@xxxxxxxxx> 写道： >>> >>> Hi Coly, >>> >>> I see the same bug on upstream kernel when I tried with >>> https://kernel.ubuntu.com/~kernel-ppa/mainline/v5.15.50/ ( >>> cod/mainline/v5.15.50 (767db4b286c3e101ac220b813c873f492d9e4ee8) >> >> >> Can you help to try the linux-stable tree, or the latest upstream Linus tree? Then I can have a clean code base. >> >>> >>> Reads seem to trigger it, not writes. So this test triggered it - >>> >>> fio --name=read_iops --directory=/home/ubuntu/bcache_mount --size=12G >>> --ioengine=libaio --direct=1 --verify=0 --bs=4K --iodepth=128 >>> --rw=randread --group_reporting=1 >>> >>> https://pastebin.com/KyVSfnik has all the details. >> >> >> OK, I will try to reproduce with above operation. What is the preferred cache and backing device sizes to reproduce the soft lockup? >> >> Thanks. >> >> Coly Li >> >>
