I started noticing odd I/O errors on my disk. Example: [398374.597662] sd 0:0:4:0: [sde] UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x00 [398374.597665] sd 0:0:4:0: [sde] CDB: opcode=0x88 88 00 00 00 00 00 c0 52 c3 90 00 00 00 08 00 00 [398374.597667] blk_update_request: I/O error, dev sde, sector 3226649488 It happened to all my disks. This is on a new motherboard with a LSI 2308 SAS controller. I can't seem to notice any problems due to this, but something does not seem right about getting this error. I also checked the offending sectors and it's no problem reading them. So i started to investigate this problem and soon found that the problem occurs when the disk is in standby mode. Ie. when I try to access the disk when it is in standby mode. Somehow the system tries to access the disk without spinning it up first, and hence getting this error. I'm running the 4.0.9-gentoo kernel source (had the same problem with 4.0.5-gentto also). Reproducible: Always Steps to Reproduce: 1.hdparm -y <any device on SAS controller> 2.Access the standby disk 3.Get I/O error in logs, but all seems OK and don't notice anything strange. Actual Results: I get one I/O error on the disk in standby. It's as if it does not notice the disk is in standby and tries to access it anyway. Expected Results: No I/O error in kernel logs.
Hello Mikael. Yes this may be related to the kernel / configuration / hardware, but I think the Gentoo Forums will be a better place where you can get help. I'm closing this for now, if you find any bugs in packages/kernel, don't hesitate to open a new bug report. Good luck.
