Re: Problems with the block-layer timeouts

[Mike Anderson <andmike@linux.vnet.ibm.com>]

Alan Stern <stern@rowland.harvard.edu> wrote:
> I spent most of the day yesterday debugging some tricky problems in the
> new block-layer timeout scheme.  Clearly it is in need of more work.
> 
> A major reason for these problems was that there doesn't seem to be a 
> clear a idea of when the timeout period should begin.  In 
> blk_add_timer() a comment says:

> How should this be fixed?  It would help to call scsi_dev_queue_ready()  
> before elv_next_request(), but that's not sufficient.  
> scsi_times_out() needs to recognize that a timeout for a non-running
> request can be handled by directly returning BLK_EH_HANDLED.  Right?
> 
> 

Tejun described a similar issue here.
http://thread.gmane.org/gmane.linux.ide/35603

And a fix to address the issue here.
http://thread.gmane.org/gmane.linux.ide/35725

Does the patch posted by Tejun address your issue?

> While I'm on the subject, there are a few related items that could be 
> improved.  In my tests, I was generating I/O requests simply by doing
> 
> 	dd if=/dev/sda ...
> 
> I don't know where the timeouts for these requests are determined, but
> they were set to 60 seconds.  That seems much too long.
> 

It is set by a udev rule and the value is historical.
http://thread.gmane.org/gmane.linux.scsi/45631/focus=45646


-andmike
--
Michael Anderson
andmike@linux.vnet.ibm.com