[Top] [All Lists]

Re: Building the kernel for a Broadcom SB1

To: Andy Isaacson <>
Subject: Re: Building the kernel for a Broadcom SB1
From: Jonathan Day <>
Date: Thu, 22 Sep 2005 10:22:50 -0700 (PDT)
Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024;; h=Message-ID:Received:Date:From:Subject:To:Cc:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=3oZa1rsmUxwe57iSfuAFRUcLZBYKOaW70f9flSQKnvUFV5ugkrLCBBeVPfmf7hksW8e3MFS8wVtiFQyYUC6UB4Z1s+VADe7GO2GJNHR01n+SkOeu4XcQnKlok+rq0X4yUk4aKzE2J4q2upUvLmxWQNd9n3ic+56LHCjXy8Yb0p8= ;
In-reply-to: <>
Original-recipient: rfc822;
Oops, my bad! Ok, well here is the entire log for the
boot sequence, including non-kernel stuff. I've also
attached the kernel configuration I'm using.

Incidently, I'm now not so sure it's a bug in the
kernel that's the factor - I've done some experiments
and discovered that there may be a bug in the ZBus.
Code that runs quickly will hard-crash the processor,
but adding anything that just inserts wait states or
otherwise slows down activity will cause the code to
run just fine.

To see if this could be related to the kernel issues,
I recompiled kernel sources I -know- absolutely works
with the board. Where it had debugging code enabled
for the Infiniband and a few other places, for the
working build, I disabled the debugs. everything else
was kept exactly the same. The whole system locked
absolutely solid.

I'll try the suggested patches & config - with and
without the debugs enabled - to see what happens. I'm
also going to scour around for diagnostic tools to see
if I can get a clearer picture of what is happening
and why.


Initializing Arena.
Initializing PCI. [normal]
PCI bus 0 slot 1/0: Warning: SipReady already set
PCI bus 0 slot 1/0: HT init: InitDone not set
PCI bus 0 slot 1/0:   Link Cmd = 0x20010008, Link Ctrl
= 0x00000000
Initializing Devices.
SENTOSA board revision 2
Config switch: 2
CPU: BCM1250 rev 0x11
L2 Cache Status: Wafer ID:   0x92CEE019  [Lot 9395,
Wafer 23]
Manuf Test: Bin A [2CPU_FI_FD_F2 (OK)]
SysCfg: 0000000024C20800 [PLL_DIV: 16, IOB0_DIV:
CPU type 0x1040102: 800MHz
Total memory: 0x10000000 bytes (256MB)
Total memory used by CFE:  0x8FE99CE0 - 0x90000000
Initialized Data:          0x8FE99CE0 - 0x8FEA3B60
BSS Area:                  0x8FEA3B60 - 0x8FEA41A0
Local Heap:                0x8FEA41A0 - 0x8FFA41A0
Stack Area:                0x8FFA41A0 - 0x8FFA61A0
Text (code) segment:       0x8FFA61A0 - 0x8FFFFFBC
Boot area (physical):      0x0FE58000 - 0x0FE98000
Relocation Factor:         I:F03A61A0 - D:0FE98CE0
eth0: Link speed: 100BaseT FDX
Device eth0:  hwaddr 00-02-4C-FD-0D-3C, ipaddr, mask
        gateway, nameserver, domain
Loader:elf Filesys:tftp Dev:eth0
Loading: 0xffffffff80100000/2903912
0xffffffff803c4f68/305896 Entry at
Closing network.
Starting program at 0xffffffff80396000
Broadcom SiByte BCM1250 B2 @ 800 MHz (SB1 rev 2)
Board type: SiByte BCM91250E (Sentosa)

--- Andy Isaacson <> wrote:

> On Wed, Sep 21, 2005 at 10:24:35AM -0700, Jonathan
> Day wrote:
> > So far, so good. Tried compiling the 2.6.14-rc1
> > kernel, forward-porting all the changes that
> hadn't
> > been merged in. the kernel locks in the initial
> boot
> > sequence, after a couple of printf's.
> We'd be able to help more if you simply pasted in
> the console output
> rather than being vague.

Yahoo! Mail - PC Magazine Editors' Choice 2005

Attachment: config
Description: 3341867309-config

<Prev in Thread] Current Thread [Next in Thread>