BTX halted trying to boot X4100 from disk

[Dave]

On Nov 27, 2006, at 4:11 AM, Remy Nonnenmacher wrote:

> Michiel Boland wrote:
>> Ok here is a patch for /sys/boot/i386/btx/btx/btx.S that will get  
>> X4100s
>> to boot.
>> It extends the v86 monitor in the BTX so that it catches INT3 and  
>> treats
>> it like any other interrupt. Not sure if it's a great idea, but it's
>> better than nothing at this point.
>> You will need to update the bootblock and /boot/loader since they  
>> both
>> appear to be linking in their own copy of BTX.
>> --- btx.S.orig	Thu Oct  5 17:30:51 2006
>> +++ btx.S	Sun Nov 26 16:26:12 2006
>> @@ -510,6 +510,8 @@
>>  		je v86popf			# Yes
>>  		cmpb $0xcd,%al			# INT imm8?
>>  		je v86intn			# Yes
>> +		cmpb $0xcc,%al			# INT3?
>> +		je v86int3			# Yes
>>  		cmpb $0xcf,%al			# IRET/IRETD?
>>  		je v86iret			# Yes
>>  		popl %ebx			# Restore
>> @@ -651,6 +653,11 @@
>>  reboot:		orb $0x1,btx_hdr+0x7		# Set the reboot flag
>>  		jmp exit			# Terminate BTX and reboot
>> +/*
>> + * Emulate INT3
>> + */
>> +v86int3:	movb $3,%al
>> +		jmp v86intn.1
>>  /*
>>   * Emulate INT imm8... also make sure to check if it's int 15/87
>>   */
>> _______________________________________________
>
> I confirm this patch fixes the same 'BTX Halted' problem on LSI1068  
> SAS/IME (mpt) on Fujitsu/Siemens RX200-S3.
>
> This will probably helps the BTX syndrome on recent LSI bioses on a  
> few machines.

Thanks for the pointer.  We where able to get an older version of the  
LSI MPT BIOS from Sun.  If this is something that may be committed  
and needs testing, let me know soon and I can throw the problem BIOS  
on the machine and give the patch a try.

--
Dave