Re: [PATCH] I/O stall finally solved

2009-08-18 Thread Mike Christie
On 08/17/2009 10:00 AM, Hannes Reinecke wrote: > Mike Christie wrote: >> On 08/14/2009 07:01 AM, Hannes Reinecke wrote: >>> Hi Mike, >>> >>> finally I've found the mysterious I/O stall I've been chasing >>> for the last two months. >>> >>> Problem is iscsi_conn_queue_work(); we're just calling >>>

Re: [PATCH] I/O stall finally solved

2009-08-18 Thread Mike Christie
On 08/17/2009 10:00 AM, Hannes Reinecke wrote: > However, I still have the feeling there still is potential > for an I/O stall; this iscsi_check_cmdsn_window_closed() check > simply doesn't feel safe. > Do you still have that test in iscsi_data_xmit or iscsi_xmit_task? If so there are the other

Re: [PATCH] I/O stall finally solved

2009-08-17 Thread Hannes Reinecke
Mike Christie wrote: > On 08/14/2009 07:01 AM, Hannes Reinecke wrote: >> Hi Mike, >> >> finally I've found the mysterious I/O stall I've been chasing >> for the last two months. >> >> Problem is iscsi_conn_queue_work(); we're just calling >> queue_work() without checking the return value. >> >> Ho

Re: [PATCH] I/O stall finally solved

2009-08-15 Thread Mike Christie
On 08/14/2009 07:01 AM, Hannes Reinecke wrote: > Hi Mike, > > finally I've found the mysterious I/O stall I've been chasing > for the last two months. > > Problem is iscsi_conn_queue_work(); we're just calling > queue_work() without checking the return value. > > However, queue_work() will be effe