Re: aic7xxx: aic7892(B): BUG: soft lockup detected on CPU#0!

[thomas schorpp <t.schorpp@gmx.de>]

thomas schorpp wrote:
> James Bottomley wrote:
>> On Fri, 2007-03-23 at 02:26 +0100, thomas schorpp wrote:
>>> ok, overriding the first while(ahc_is_paused) that blocked before (i 
>>> see no sense for doing this in a pci mmap test function, cause proper 
>>> resource setup is required *before* using such I/O functions, 
>>> otherwise the adapter had entered SEQ paused status)
>>> i got the kernel to boot at least at pio mode.
>>>
>>> this is surely not the correct resource and looks like a datatype 
>>> boundary overflow, the upper 0x0f is missing:
>>> [   49.278810] Trying to free nonexistent resource
>>> <00000000fffff000-00000000fff
>>> fffff>
>>
>> That's because ahc->platform_data->mem_busaddr is u32
>>
>>
>>> [   54.513224] scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, 
>>> Rev 7.0
>>> [   54.513226]         <Adaptec 19160B Ultra160 SCSI adapter>
>>> [   54.513227]         aic7892: Ultra160 Wide Channel A, SCSI Id=7, 
>>> 32/253 SCBs
>>
>> The driver code suggests that the 7892 can't do the AHC_LARGE_SCBS
>> features ... which means the card itself cannot address more than 32
>> bits of memory, so it would be unable to decode a BAR that's beyond the
>> 32 bit range.  So this looks like some type of error in the PCI config
>> system (or possibly in the BIOS).  I think this card needs its BARs to
>> be in the lower 32 bits to function.
>>
>> James
>>
>>
> 
> i agree for this to be a 32bit dma busmaster chip,
> since pci_resource_flags and lspci say 64bit mem resource type
> 
> aic7xxx: pci_resource_start fffff000 *maddr 20000 mem64 4
> 
> we've a bug in the x86_64 linux pci config, BIOS is ok, the hardware 
> worked fine in a winxp_x64 test setup a few months ago.
> 
> will ask LKML.
> 
> y
> tom

sorry, wrong according to http://download.adaptec.com/pdfs/aic7892.pdf.

"66 MHz, 64-bit, PCI interface that
supports zero wait-state memory;
also operates on 33 MHz, 32-bit
PCI busses"

this chip is capable of 64bit addressing, as pci_resource_nnnn (checking this) on x86_64 platform 
and lspci on x86_64 *and* AMDK7 configured kernels reports, even on PCI/32, right?
or is it impossible to do multiplexed 64bit mem addressing on PCI/32?

Why are the driver structure address members 32bits wide types if therere PCI/64 card 
models with this chips as listet in aic7xxx.txt kernel doc and stated in aic7892.pdf?
I'll adapt the respective driver structures and function args now to 64bit and see what happens...

can adaptec.inc pls comment? since the aha19160 card is still in production state, 
i assume they want to have a linux x86_64 dma capable driver. so far it is not, 
or can other users having this card pls confirm my pci system broken?

y
tom