Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 2 Jan 2016 20:23:26 -0600
From:      Jason Browning <jbrowning@uh.edu>
To:        freebsd-mips@freebsd.org
Subject:   [RT5350] Boot from flash --> TLB miss (2)
Message-ID:  <CAK-cGiW%2BiXe%2BON5sryG_ex=1O5-9TWT-e0P0ErRQwd0_=YZRog@mail.gmail.com>

next in thread | raw e-mail | index | archive | help
Hi!
I have an RT5350-based board called a =E2=80=98VoCore=E2=80=99
http://vocore.io
It=E2=80=99s small - 32M RAM, 16M SPI flash, WiFi, and GPIOs aplenty.

I have a software-dev background and am new to embedded-dev, but
the FreeBSD documentation and searching this list have brought me
a long way.  I am now at my limit.

I=E2=80=99ve built kernels for (release)RT305X and (current)RT5350,
and both work as expected when uploaded as an ELF to RAM and
kicked-off with U-Boot=E2=80=99s =E2=80=98go=E2=80=99 command.  They both f=
ail with the
following output when booted from on-board SPI flash:

Booting image at bc050000 .
   Image Name:
   Created:      2016-01-01   7:14:33 UTC
   Image Type:   MIPS NetBSD Kernel Image (gzip compressed)
   Data Size:    1580815 Bytes =3D  1.5 MB
   Load Address: 80000000
   Entry Point:  80000100
   Load Kernel:  .........................

   Verifying Checksum ... OK
   Uncompressing Kernel Image ... OK

U-Boot args (from 0 args):
None
Environment:
memsize=3D32
initrd_start=3D0x00000000
initrd_size=3D0x0
flash_start=3D0x00000000
flash_size=3D0x400000
entry: mips_init()
Cache info:
  picache_stride    =3D 4096
  picache_loopcount =3D 8
  pdcache_stride    =3D 4096
  pdcache_loopcount =3D 4
cpu0: MIPS Technologies processor v76.150
  MMU: Standard TLB, 32 entries (4K 16K 64K 256K 1M 16M 64M 256M pg sizes)
  L1 i-cache: 4 ways of 256 sets, 32 bytes per line
  L1 d-cache: 4 ways of 128 sets, 32 bytes per line
  L2 cache: disabled
  Config1=3D0xbea3319e<PerfCount,WatchRegs,MIPS16,EJTAG>
  Config2=3D0x80000000
  Config3=3D0x420
Physical memory chunk(s):
0x3e0000 - 0x1ffffff, 29491200 bytes (7200 pages)
Maxmem is 0x2000000
GDB: debug ports: uart
GDB: current port: uart
KDB: debugger backends: ddb gdb
KDB: current backend: ddb
[ thread pid 0 tid 0 ]
Stopped at      0x801f43b4
db> bt
Tracing pid 0 tid 0 td 0x803646c0
8027896c+30 (?,?,?,?) ra 1880364268 sp 0 sz 0
8000f1b0+114 (0,?,ffffffff,?) ra 2080364280 sp 1 sz 1
8000e4f4+388 (?,?,?,?) ra a8803642a0 sp 0 sz 0
8000e990+70 (?,?,?,?) ra 1880364348 sp 0 sz 0
800110c8+f4 (?,?,?,?) ra 1a880364360 sp 0 sz 0
800f4590+110 (?,?,?,?) ra 3080364508 sp 0 sz 0
trap+d28 (?,?,?,?) ra c080364538 sp 0 sz 0
MipsKernGenException+134 (803f1e98,80355f1c,aba9500,0) ra c8803645f8
sp 100000001 sz 1
801f432c+88 (?,?,?,?) ra 20803646c0 sp 0 sz 0
pid 0
db> reset
Trap cause =3D 2 (TLB miss (load or instr. fetch) - kernel mode)
panic: trap
Uptime: 1s

=3D=3D
More info: This board came with U-Boot+OpenWRT installed; per the
U-Boot config, the FreeBSD kernels were compiled with makeoption
KERNLOADADDR=3D0x80100000
when initiated with =E2=80=98go 80100100=E2=80=99 from RAM, and with makeop=
tion
KERNLOADADDR=3D0x80000000
when initiated with =E2=80=98bootm bc050000=E2=80=99 from SPI flash.

I=E2=80=99m out of my depth here, but I *think* things go wrong from within
locore.S ([srcRoot]/sys/mips/mips/locore.S) at ~line:177 where:

PTR_L   a0, TD_PCB(sp)

I say this because the problem-scenario seems to make it out of
=E2=80=98platform_start=E2=80=99
([srcRoot]/sys/mips/rt305x/rt305x_machdep.c),
and also seems never to enter =E2=80=98mi_startup=E2=80=99
([srcRoot]/sys/mips/kern/init_main.c).
Additionally, I do not know why the flash_size is reported to be only
4M, it=E2=80=99s 16M - verified.
I=E2=80=99ve not delved into the intricacies of U-Boot yet.

Am I on the wrong track?  I=E2=80=99m at my wit=E2=80=99s end with this pro=
blem.

Jason



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAK-cGiW%2BiXe%2BON5sryG_ex=1O5-9TWT-e0P0ErRQwd0_=YZRog>