From mboxrd@z Thu Jan  1 00:00:00 1970
From: Xiaoshan Zuo <xzuo@vinesystech.com>
Date: Tue, 14 Oct 2003 16:49:59 -0700
Subject: [U-Boot-Users] walnut 405gp bridge setting
Message-ID: <3F8C8BA7.9020204@vinesystech.com>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de

I have a customized walnut 405 gp board. While using u-boot the bridge 
was not set correclty. The problem was that the bridge didn't respond to 
requst. The problem was traced down to PCIC0_CMD[MA]=0 on the bridge. 
The bit is set correclty in pci_405gp_init,
later it is got changed in pci_hose_scan_bus. It is not supposed to be 
changed, as suggested by the comment below in the code. But this comment 
contradict with function pci_hose_scan_bus which intentionly doesn't 
skip the bridge.  Removing #ifndef in pci_hose_scan_bus makes the bridge 
work. My question is: is this a typo?

Where can I search for u-boot mailing list archive? linuxppc is very 
easy to serach useful information.  Sorry for the newbie question.

Thanks,

Xiaoshan

/*
 * drivers/pci.c skips every host bridge but the 405GP since it could
 * be set as an Adapter.
 *
 * I (Andrew May) don't know what we should do here, but I don't want
 * the auto setup of a PCI device disabling what is done pci_405gp_init
 * as has happened before.
 */

void pci_405gp_setup_bridge(struct pci_controller *hose, pci_dev_t dev,
                            struct pci_config_table *entry)
{
#ifdef DEBUG
        printf("405gp_setup_bridge\n");
#endif
}

In drivers/pci.c:

int pci_hose_scan_bus(struct pci_controller *hose, int bus)
{
#ifndef CONFIG_405GP /* don't skip host bridge on ppc405gp */
                /* Skip our host bridge */
                if ( dev == PCI_BDF(hose->first_busno,0,0) )
                        continue;
#endif