> -----Original Message----- > From: Dan Williams [mailto:dan.j.williams@xxxxxxxxx] > Sent: Monday, July 27, 2015 11:25 AM > To: James Bottomley > Cc: Christoph Hellwig; Praveen Murali; linux-scsi; stable@xxxxxxxxxxxxxxx > Subject: Re: [for 4.1 PATCH resend] libsas: fix "sysfs group not found" warnings > at port teardown time > > > I didn't ask you to justify your process, I asked you how important you > > thought the patch was mainly because of the conflicting signals you've > > sent. I get that you think I should treat all your patches as important > > whether you do or not, but the world doesn't quite work like that: patch > > application is a process of triage. Patches, like this, which have > > timing related issues potentially leading to races get looked at by me > > as the last reviewer. The speed of review depends on several factors, > > but one of which is what type of user visible issue is this causing. > > The user visible effects of this are a nasty warning message and nothing > > more, I believe? A useful indicator in this triage is how important the > > submitter thinks the patch is, which was originally why I asked. > > > > That would be a question to Praveen. It wasn't clear to me whether > this sysfs backtrace was a simply a warning or eventually fatal to the > box. As far as I remember, the issue was mostly with the sysfs backtraces. I don’t remember it causing a fatal error; but that could very well be because I did not run it long enough. ��.n��������+%������w��{.n�����������ܨ}���Ơz�j:+v�����w����ޙ��&�)ߡ�a����z�ޗ���ݢj��w�f