Date: Sun, 5 May 2019 15:25:12 -0700 From: Mark Millard <marklmi@yahoo.com> To: Justin Hibbits <chmeeedalf@gmail.com>, FreeBSD PowerPC ML <freebsd-ppc@freebsd.org> Subject: Remote Firewire dcons use shows my G5 "hang-up point" reports have been wrong: shows lines not shown on the problem context's display Message-ID: <E4E3E09D-4945-45E8-91C3-0C89653F06D6@yahoo.com>
next in thread | raw e-mail | index | archive | help
I got remote FireWire dcons use going (partially), = https://www.freebsd.org/doc/en_US.ISO8859-1/books/developers-handbook/kern= eldebug-dcons.html # more /mnt/boot/loader.conf hw.firewire.phydma_enable=3D1 hw.firewire.dcons_crom.force_console=3D1 #dcons_gdb=3D1 #boot_multicons=3D"YES" [Trying to turn on boot_multi_cons gets stuck at db> prompts *very* early and the display on the target ends up with mixed material from two streams of output. So: no use of dcons for low level console activity.] The above combination does show output on the remote, observing PowerMac, during part of the boot sequence. And . . . It shows more lines of output just before the overall hang-up than the target machine's display does. It appears that for most or all the hang-ups that I've reported, my report indicated a point earlier then the actual hang-up: more of the messages were actually generated and the code got farther than I knew. So there might be a separate issue of the screen display stopping its updates before the old PowerMac G5 has actually hung-up (overall). Or it my just be timing relationships do not get a chance to allow the "next screen update" or some such. It appears that tracking down hang-up points via print messages shown on the normal display can not generally be done on the G5s. FireWire dcons use may be required (unless there is some alternative). I'll note that I was not able to break into ddb via the dcons connection. It appeared that I had no ability to input to the target if I had no such on the directly-connected usb keyboard at the same time. In other words: I gained no control but got better observational data. FYI: I used (replacing ??'s with appropriate hexadecimal digits): dconschat -e \# -br -TC 5555 -t ??-??-??-??-??-??-??-?? & telnet localhost 5555 on the observing PowerMac. (I happened to use a G4 as the observer of the target G5.) =3D=3D=3D Mark Millard marklmi at yahoo.com ( dsl-only.net went away in early 2018-Mar)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?E4E3E09D-4945-45E8-91C3-0C89653F06D6>