Re: port-sparc64/46461: Sun E450 drops to DDB on boot of 6.0_BETA

[Michael <macallan%netbsd.org@localhost>]

The following reply was made to PR port-sparc64/46461; it has been noted by 
GNATS.

From: Michael <macallan%netbsd.org@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: 
Subject: Re: port-sparc64/46461: Sun E450 drops to DDB on boot of 6.0_BETA
Date: Sat, 19 May 2012 10:09:58 -0400

 -----BEGIN PGP SIGNED MESSAGE-----
 Hash: SHA1
 
 > The machine's not explicitly set to use serial console, so it
 > complains on each reboot:
 >
 >  Can't open input device.
 >  Keyboard not present.  Using ttya for input and output.
 >
 > And indeed there is no keyboard connected, but the console is on
 > the serial port.
 
 Out of curiosity - what happens if you explicitly set the console to =20
 ttya?
 
 > f009e590: /pci/TSI,gfxp@4
 >
 > address                 fe74e000 ........ ........ ........   .t..
 > assigned-addresses      82002010 00000000 00020000 =20
 > 00000000   .. .............
 >            0010:       00020000 82002014 00000000 =20
 > 00800000   ...... .........
 >            0020:       00000000 00800000 82002018 =20
 > 00000000   .......... .....
 >            0030:       01000000 00000000 00800000 =20
 > 82002030   .............. 0
 >            0040:       00000000 01800000 00000000 =20
 > 00010000   ................
 > character-set           49534f38 3835392d 3100.... ........   =20
 > ISO8859-1.
 > class-code              00038000 ........ ........ ........   ....
 > depth                   00000020 ........ ........ ........   ...
 > device-id               00000009 ........ ........ ........   ....
 > device_type             64697370 6c617900 ........ ........   =20
 > "display"
 > devsel-speed            00000001 ........ ........ ........   ....
 > dotclockm               00000042 ........ ........ ........   ...B
 > dotclockn               00000005 ........ ........ ........   ....
 > dotclockp               00000001 ........ ........ ........   ....
 > fast-back-to-back
 > fcode-rom-offset        00000000 ........ ........ ........   ....
 > height                  00000384 ........ ........ ........   ....
 > interrupts              00000001 ........ ........ ........   ....
 > latency-timer           00000040 ........ ........ ........   ...@
 > linebytes               00000480 ........ ........ ........   ....
 > max-latency             000000c0 ........ ........ ........   ....
 > min-grant               000000c0 ........ ........ ........   ....
 > model                   47465850 00...... ........ ........   "GFXP"
 > name                    5453492c 67667870 00...... ........   =20
 > "TSI,gfxp"
 > prom-revision           312e3131 2039392f 30372f30 36205453   1.11 =20
 > 99/07/06 TS
 >            0010:       4900.... ........ ........ ........   I.
 > reg                     00002000 00000000 00000000 =20
 > 00000000   .. .............
 >            0010:       00000000 02002010 00000000 =20
 > 00000000   ...... .........
 >            0020:       00000000 00020000 02002014 =20
 > 00000000   .......... .....
 >            0030:       00000000 00000000 00800000 =20
 > 02002018   .............. .
 >            0040:       00000000 00000000 00000000 =20
 > 00800000   ................
 >            0050:       02002030 00000000 00000000 00000000   .. =20
 > 0............
 >            0060:       00010000 ........ ........ ........   ....
 > revision-id             00000001 ........ ........ ........   ....
 > subsystem-id            00000006 ........ ........ ........   ....
 > subsystem-vendor-id     00001227 ........ ........ ........   ...'
 > timing-numbers          31313532 2c203337 362c2034 302c2031   1152, =20=
 
 > 376, 40, 1
 >            0010:       32382c20 20393030 2c203337 2c202032   28,  =20
 > 900, 37,  2
 >            0020:       2c20342c 20203934 35303030 30302c20   , 4,  =20
 > 94500000,
 >            0030:       32303536 00...... ........ ........   2056.
 > vendor-id               00003d3d ........ ........ ........   ..=3D=3D
 > width                   00000480 ........ ........ ........   ....
 
 This looks sane and should work, console or not - address, width, =20
 height, linebytes etc. are present ( and pm2fb doesn't even use =20
 address - unlike genfb it knows which BARs to map )
 Could you sprinkle printf()s over pm2fb_attach() to figure out where =20
 exactly we crash?
 
 > Why it said "no width property" in 5.99.38 and then suddenly | it =20
 > finds it in 6.0_BETA will excercise the latter parts of | =20
 > pm2fb_attach().
 
 I dimly remember that some (all?) versions of the Raptor firmware have =20=
 
 a bug which results in some property names having trailing spaces =20
 which messes with of_getprop() and friends. At some point I added code =20=
 
 to deal with that. Nothing to do with pm2fb though, that code is =20
 somewhere in arch/sparc64/
 
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1.4.7 (Darwin)
 Comment: GPGTools - http://gpgtools.org
 
 iQEVAwUBT7eptspnzkX8Yg2nAQIR1ggAhtO+Bz7Y083XMO9GIjN54WRkXWwwVvsg
 l+M5fx/BFEMrJYynUyHhsEtIksLMVkz1eX0TIBMnC677l+Uq/vPRIlKClqkOOYNc
 r0TEwDhSC+KgIjJfFppBMmn8CJ/Vy+WGsObOGz2yRMRBoFgfzrZqpJqwRYSySswQ
 bD5KWkZoxt6CG0qOqQTQ8R86bwUqyEt7rItHYwnndTFemABC2+TJRgw62kDU/I9i
 3ZV5RRJ21MI5IkmnBq3m3VQgnBmP6ld/wfGpXMI/23R/NBfF8JXvRTbv9p+2yJvV
 0NyORl1HuSsgtDA/yfCiEb0KyvitlqgMiA6aOaQPKlHMR5OyyegUaQ=3D=3D
 =3DiTuq
 -----END PGP SIGNATURE-----