Re: [pci] bug in pci_behaviorial_master.v and system.v?

["Miha Dolenc" <mihad@opencores.org>]

Hi!

This is not the bug, it's just the way things are.
As a matter of fact, pci bus model is always repeating the transaction that
was done last time, until you call do_pause(n) task, it doesn't matter if it
was retried or not.
do_pause(n) call stops pci master from repeating the transaction. What that
means is, that you have to simulate retry of disconnected transaction by
yourself.

Regards,
Miha Dolenc

----- Original Message -----
From: "sumnow" <sumnow@263.sina.com>
To: <pci@opencores.org>
Sent: Thursday, June 13, 2002 11:54 PM
Subject: [pci] bug in pci_behaviorial_master.v and system.v?


> hi all,
>
> I am using the opencores pci bridge and I found that in
pci_target_unit(upstream transaction), we can not initiate a burst transfer
with burst size larger than FIFO depth. If do so, pci bridge will issue
DISCONNECT when FIFO is almost full, but pci_behaviorial_master will take it
as RETRY and only retry the transaction from the beginning, and so it always
fails.
>
> As above, in system.v, the largest size in test_normal_wr_rd is
`PCIW_DEPTH - 2, which is as follows:
>
>         test_normal_wr_rd( `Test_Master_2, Target_Base_Addr_R[image_num] +
32'h20, 32'h7147_a5c3, `Test_Byte_2, image_num,
>                             `PCIW_DEPTH - 2, 1'b1, 1'b1, cache_lsize,
1'b1, 1'b1 );
>
> I wonder if this is a bug in pci_behaviorial_master.v and system.v? If it
is, has someone corrected it?
>
> Best regards.
>
> sumnow
>
>
>
>
> --
> To unsubscribe from pci mailing list please visit
http://www.opencores.org/mailinglists.shtml
>
>


--
To unsubscribe from pci mailing list please visit http://www.opencores.org/mailinglists.shtml