Re: [pci] Can read config, cannot write config (yet)

[kclin3@yahoo.com.tw]

Dear all:
     I have the same problem in FPGA implementation. Can read the 
config space but can't write it. Why? could you tell me?
Thanks.    
              K.C

----- Original Message ----- 
From: "Charles Krinke" <ckrinke@p... > 
To: <pci@o... > 
Date: Wed, 19 Jun 2002 14:57:39 -0700 
Subject: [pci] Can read config, cannot write config (yet) 

> 
> 
> Hello all, I have gotten my project where I can synthesize the PCI 
> bridge 
> into a VirtexE and read configuration space reliably. I can both 
> see the 
> address/data phases with the correct data on my logic analyzer and 
> I can see 
> them from the host side of the PCI interface. Thats for reading 
> configuration space. The bridge is implemented as a GUEST. Now 
> comes the 
> curious part. I can see that when I issue a configuration write, 
> that the 
> logic analyzer shows what looks like the proper command for writing 
> 0x1011. 
> And I can see the 32 bit data. The write completes successfully 
> with no 
> master/target abort. The unfortunate thing is that when I read a 
> few seconds 
> later, what I see on the bus and what the host receives on its 
> command line 
> is the same contents that were in the command register before. So, 
> I have 
> yet to figure out how to write to configutation space with the 
> PCI_BRIDGE. 
> In reading the specification, it seems to me that in GUEST mode, 
> the master 
> talking to the PCI_BRIDGE should be able to write both the command 
> register 
> (to enable memory space and bus mastering) and to write BAR’s, but 
> alas, I 
> dont quite get it yet. 
> 
> As usual, any suggestions or criticisms are greatly appreciated. At 
> least 
> progress is being made. 
> 
> Charles Krinke 
> 
--
To unsubscribe from pci mailing list please visit http://www.opencores.org/mailinglists.shtml