Re: Ultra 45 onboard bge

(2013/10/29 16:31), Julian Coleman wrote:
> Hi,
>
>> Could you test the following patch?
>>
>> 1) Add extra delay for BCM570[0-4]. This is not documented but
>>     Linux tg3 does.
>> 2) Insert a completion barrier at the end of bge_blockinit()
>
> Unfortunately, that doesn't make things work - the same symptoms as before -
> packets are transmitted with the correct mac address, but none are received
> by the driver.
>
> Thanks,
>
> J

  I added another fix into -current.

> Module Name:	src
> Committed By:	msaitoh
> Date:		Thu Oct 31 06:01:39 UTC 2013
>
> Modified Files:
> 	src/sys/dev/pci: if_bge.c if_bgereg.h
>
> Log Message:
>  Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD (part of r199670).
> This workaround is for 5714/5715 controllers and is not actually a MAC
> controller bug but an issue with the embedded PCIe to PCI-X bridge in the
> device.
>  This change uses bus_dmatag_subregion(), so this workaround won't work on some
> archs which doesn't support bus_dmatag_subregion().
>
>
> To generate a diff of this commit:
> cvs rdiff -u -r1.261 -r1.262 src/sys/dev/pci/if_bge.c
> cvs rdiff -u -r1.85 -r1.86 src/sys/dev/pci/if_bgereg.h

Ultra 45 has BCM5715. I don't know if this bug is related
to the problems we have.

  For sparc64, bus_dmatag_subregion() is not supported yet,
so this change won't fix any problem.

-- 
-----------------------------------------------
                 SAITOH Masanobu ([email protected]
                                  [email protected])