Date: Sun, 29 Jul 2018 08:40:33 -0700 From: John Kennedy <warlock@phouka.net> To: Per olof Ljungmark <peo@nethead.se> Cc: freebsd-arm <freebsd-arm@freebsd.org> Subject: Re: rpi3 and Adafruit GPS hat continued Message-ID: <20180729154033.GI75644@phouka1.phouka.net> In-Reply-To: <2a0c3063-d1d3-00c5-ee59-0a78ca0aaad4@nethead.se> References: <d1341718-f8a3-31f9-9ac4-24b7a2de51d9@nethead.se> <CANCZdfrDXytg%2B3a5O%2BHch5fMhm85Uf5GGKKVP=AfbPydO-c22Q@mail.gmail.com> <ba6d3be7-73ed-46b7-df87-0cf98b035397@nethead.se> <2f55aaea-6832-7441-1011-b8288628a4a9@nethead.se> <CF8A1708-2D3E-414C-92BA-442999DDBB37@ralphsmith.org> <2aee4fd4-c221-23a6-00e4-1db2694bfef6@nethead.se> <6DCECF08-1366-4289-9929-94A967284291@ralphsmith.org> <eb0c71e3-35ba-1e38-afa1-4957d1100db0@nethead.se> <20180729134340.GH75644@phouka1.phouka.net> <2a0c3063-d1d3-00c5-ee59-0a78ca0aaad4@nethead.se>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, Jul 29, 2018 at 04:28:53PM +0200, Per olof Ljungmark wrote: > Nice to hear! The flakiness I have seen is mainly with the USB > subsystem, for example every other time the RPI3B here loads u-boot it > detects 3 devices, and the next 4 devices, this is from cold start > without any soft- or hardware changes. Hmm. So for a while, experimenting with ZFS, I was trying to boot off of USB. In my case, that was a DYNEX DX-C112 SD/microSD memory card reader with the SD card I might have otherwise plugged into the front inserted. I was juggling multiple variables at the time (not that I knew them all at that instant), but it never failed to boot past u-boot (to be clear in my mind, into the software loaded into the msdosfs partition). However, I think I only every had one bootable "disk" device in there on startup, so probably much simpler than your setup. My usual device list looks like this with nothing plugged in except for the wireless USB-keyboard receiver: # usbconfig list ugen0.1: <DWCOTG OTG Root HUB> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=SAVE (0mA) ugen0.2: <vendor 0x0424 product 0x2514> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=SAVE (2mA) ugen0.3: <vendor 0x0424 product 0x2514> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=SAVE (2mA) ugen0.4: <vendor 0x0424 product 0x7800> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=ON (2mA) ugen0.5: <Logitech USB Receiver> at usbus0, cfg=0 md=HOST spd=FULL (12Mbps) pwr=ON (98mA) > Another issue I've seen is file system corruption but of course this is > current and I do not expect everything to be top notch. I have not seen corruption. I've got two SanDisk Extreme PLUS 64GiB SD cards (V30 rated) that were new (now only ~1 month old) that I flip-flop between as I experiment. With ZFS, I'd expect that to show up fairly easily as the OS detected it. I was doing a lot of UFS until I got ZFS running. I wasn't experiencing uncontrolled crashes or reboots and I spent most of my time up until recently running off of the RE images (I'm not sure what extra QA value those have). I've been rebuilding FreeBSD as my stress-test. Mostly with /usr/src and /usr/obj NFS-mounted on there, but with ethernet on the USB bus I'm not sure that it matters as much for beating up the I/O subsystem. I've certainly got the gut feeling that slow USB/SD I/O can back up the OS in unexpected ways that fast I/O won't (which manifests as hangs and timeouts, and if you reboot then maybe corruption). I'm also more concerned about write-I/O wear on this tiny SD (or USB), since it's so tiny and really isn't bragging about wear- leveling. In other words, I'm wondering when I'll start to experience media failure corruption that has nothing to do with the OS. > Flaky or not, would be nice to try another arm or arm64 platform just to > learn, preferably one of the bigger ones, already got an Orange Pi R1 to > delv into now. If I wasn't specifically interested in the RPI, I'd be in the same situation. I'm totally spoiled by the speed of my amd64 system though, plus the graphics I can attach to it.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20180729154033.GI75644>