Re: kern/56686: wd(4) device timeouts

[Michael van Elst <mlelstv%serpens.de@localhost>]

The following reply was made to PR kern/56686; it has been noted by GNATS.

From: Michael van Elst <mlelstv%serpens.de@localhost>
To: Andreas Gustafsson <gson%gson.org@localhost>
Cc: gnats-bugs%netbsd.org@localhost
Subject: Re: kern/56686: wd(4) device timeouts
Date: Fri, 4 Feb 2022 15:58:51 +0100

 On Fri, Feb 04, 2022 at 03:45:00PM +0200, Andreas Gustafsson wrote:
 > Michael van Elst wrote:
 > >  > [     5.921953] wd3: <ST5000LM000-2AN170>
 > >  
 > >  ST5000LM000 is a SMR disk, the timeouts might be real and
 > >  the driver might need to wait longer for such hardware.
 > 
 > Agreed.  Where is the timeout defined, and does it take the queue
 > length into account?
 
 There are several kinds of timeouts, but this here should be a command
 timeout. It's a callout started when the command is issued to the
 controller, the timeout period is a constant that depends on the
 particular command. Regular I/O commands have 10s, that's the
 ATA_DELAY constant in
 
 sys/dev/ic/ahcisata_core.c
 sys/dev/ata/ata_wdc.c
 
 Other controllers may have something else...
 
 There are a few commands that have their own timeouts, like flushing
 the drive cache (used by WAPBL if you let it). They should not run
 concurrently with I/O commands, but I'm not sure.
 
 
 > >  The 'uncorrectable data error' comes from the drive, so
 > >  that's probably real too.
 > 
 > Probably, but I don't see it in the SMART error log:
 
 
 could be all hidden in that number:
 
 >   1 Raw_Read_Error_Rate     0x000f   072   062   006    Pre-fail  Always       -       74443112
 
 
 
 > SMART Self-test log structure revision number 1
 > No self-tests have been logged.  [To run self-tests, use: smartctl -t]
 
 Maybe a long self-test will log something.
 
 
 Greetings,
 -- 
                                 Michael van Elst
 Internet: mlelstv%serpens.de@localhost
                                 "A potential Snark may lurk in every tree."