Re: NetBSD/hpcmips userland crashes (Re: IBM Workpad z50 dmesg collection)
David Brownlee <[email protected]>
| Newsgroups | gmane.os.netbsd.ports.mips.devel |
|---|---|
| Message-ID | <CAGN_6paYDAvj1C1CL=Fb45Ycv1AXDPdLgD=K0obyLtNfGA59kw@mail.gmail.com> |
On Mon, 20 Mar 2023 at 17:22, Alexander Bochmann <[email protected]> wrote: > > ...on 2023-03-19 21:37:51, David Brownlee wrote: > > > Sorry - could you confirm if 2016-07-01 works or hangs? > > 2016-07-01 is the last good one. > > > It feels like progress - if can can identify the last working/first > > failing, we can run down the intervening commits and either build another > > kernel ot two, or point at a candidate change. > > Great, thanks :) > > I still have the pcmcia problem on my list, but that's certainly > much later on in the timeline. Will come back to that :-p OK, taking a look at the commit log between 2016-07-01 and 08-01 https://sync.absd.org/hpcmips/2016-07-01-to-2016-08-01.log (this is from a git copy as its just much easier to work with), shows.... - a binutils update (boo) - a whole host of small mips changes - MIPS changing to use the common pmap code My money would be on the last one being the cause, with the common pmap code maybe not handling some edge-case in the VR4121. But... progress! Sounds like I need to fire up my compiler to build another couple of test kernels to narrow it down (watch this space :) David