>-----Original Message----- >From: linux-rdma-owner@xxxxxxxxxxxxxxx [mailto:linux-rdma- >owner@xxxxxxxxxxxxxxx] On Behalf Of Jason Gunthorpe >Sent: Friday, February 8, 2019 10:57 AM >To: Ruhl, Michael J <michael.j.ruhl@xxxxxxxxx> >Cc: Dalessandro, Dennis <dennis.dalessandro@xxxxxxxxx>; >dledford@xxxxxxxxxx; linux-rdma@xxxxxxxxxxxxxxx; Marciniszyn, Mike ><mike.marciniszyn@xxxxxxxxx> >Subject: Re: [PATCH for-rc 1/7] IB/{hfi1, qib, rdmavt}: Do not depend on IB >Verbs name for driver logging > >On Fri, Feb 08, 2019 at 03:17:08PM +0000, Ruhl, Michael J wrote: >> I have an alternate patch that uses: >> >> 1) a device name private to the rdmavt instance (for rdmavt logging) >> (adds the name as a char) > >And how does rename work? > >> 2) doesn't use the kobject from HFI > >Delete it :( > >> 3) doesn't use the char in QIB > >> 4) uses hardwired text rather than a string in the logging >> function.: i.e : "hfi1_%d", dd->unit. > >Would rather see this loggin as a core function like netdev does. > >Why don't you fix the core code to allow early name assignment? > >In fact my series here: > >https://github.com/jgunthorpe/linux/commits/device_locking_cleanup > >Does pretty much just that already, it needs two more lines and some >EXPORT_SYMBOLs, but otherwise does the job. Ok, I will take a look at this path. Mike >Jason
