[Scst-devel] [ofa-general] WinOF_2_0_5/SRP initiator: slow reads and eventually hangs

Chris Worley worleys at gmail.com
Mon Sep 28 05:35:46 PDT 2009


Bart,

I really appreciate all your work here.  It looks like you've really
bullet-proofed scst, and well defined where ib_srp has issues.

Thanks,

Chris
On Mon, Sep 28, 2009 at 12:27 AM, Bart Van Assche
<bart.vanassche at gmail.com> wrote:
> On Mon, Sep 21, 2009 at 8:06 PM, Bart Van Assche
> <bart.vanassche at gmail.com> wrote:
>> My hypothesis is that in your setup running ib_srpt with thread=0
>> resulted in SRPT's completion queue handler (srpt_completion()), which
>> keeps running as long as more completion queue elements can be
>> processed, took up too much time, and that this finally resulted in
>> remote SRP initiator disconnects.
>
> An update regarding the SRPT and SRP issues reported in this e-mail thread:
> * At least on my setup, the latest SRPT revision (r1153 or later) runs
> fine  -- all issues reported in this (really long) e-mail thread
> should have been resolved in that SRPT revision.
> * Unfortunately this doesn't mean that the SRP initiator lockup is
> gone. I'm still able to trigger this issue, both with the SRP
> initiator included in the mainstream Linux kernel and with the SRP
> initiator included in OFED. More details can be found here:
> http://bugzilla.kernel.org/show_bug.cgi?id=14235 and
> https://bugs.openfabrics.org/show_bug.cgi?id=1745.
>
> Bart.
>



More information about the general mailing list