> -----Original Message----- > From: Xiao Ni <xni@xxxxxxxxxx> > Sent: Thursday, February 4, 2021 1:57 PM > To: songliubraving@xxxxxx > Cc: linux-raid@xxxxxxxxxxxxxxx; matthew.ruffell@xxxxxxxxxxxxx; > colyli@xxxxxxx; guoqing.jiang@xxxxxxxxxxxxxxx; ncroxon@xxxxxxxxxx; > hch@xxxxxxxxxxxxx > Subject: [PATCH V2 0/5] md/raid10: Improve handling raid10 discard request > > Xiao Ni (5): > md: add md_submit_discard_bio() for submitting discard bio > md/raid10: extend r10bio devs to raid disks > md/raid10: pull the code that wait for blocked dev into one function > md/raid10: improve raid10 discard request > md/raid10: improve discard request for far layout Hi Xiao Ni, Thanks for this series. I also reproduced this issue when creating a RAID10 disk via Intel VROC. The xfs formatting process was not finished on 5.4.0-66 and 5.12.0-rc2 (waiting for one hour), and there were lots of IO timeouts from dmesg. With this series (on top of 5.12.0-rc2), the xfs formatting process only took 1 second. And, I did not see any IO timeouts from dmesg. The test detail [0] is shown as follows. So, feel free to add my tested-by. [0] https://gist.githubusercontent.com/AdrianHuang/56daafe1b4dbd8b5744d02c5a473e5cd/raw/82f33862698be2567af48b7662f08ccd8e8d27fd/raid10-issue-test-detail.log -- Adrian
