pci: Fix expansion ROM programming for multi-function devices
PATCH 1/1 pci: Fix expansion ROM programming for multi-function devices
PCI_HEADER_TYPE register (offset 0x0e) bit 7 is an indicator for multi-function devices. We should mask it off before using it as the header type.
---
drivers/pci/pci_auto.c | 1 + 1 file changed, 1 insertion(+)
drivers/pci/pci_auto.c +1 -0
@@ -185,6 +185,7 @@ void pciauto_setup_device(struct pci_controller *hose,#ifndef CONFIG_PCI_ENUM_ONLY /* Configure the expansion ROM address */ pci_hose_read_config_byte (hose, dev,PCI_HEADER_TYPE , &header_type); + header_type &=0x7f ;if (header_type !=PCI_HEADER_TYPE_CARDBUS ) { rom_addr = (header_type ==PCI_HEADER_TYPE_NORMAL ) ?PCI_ROM_ADDRESS :PCI_ROM_ADDRESS1 ;
Signature · 1 line
1.8.2.1
PATCH 1/1 pci: Fix expansion ROM programming for multi-function devices
Hi Tom,
Quoted 15 lines · Bin Meng
PCI_HEADER_TYPEregister (offset0x0e) bit 7 is an indicator for multi-function devices. We should mask it off before using it as the header type.Signed-off-by Bin Meng <[email protected]>---
drivers/pci/pci_auto.c | 1 + 1 file changed, 1 insertion(+)drivers/pci/pci_auto.c +0
-0
drivers/pci/pci_auto.c void pciauto_setup_device(struct pci_controller *hose,
Quoted 8 lines
drivers/pci/pci_auto.c +1
-1@@ -185,6 +185,7 @@ void pciauto_setup_device(struct pci_controller *hose,#ifndef CONFIG_PCI_ENUM_ONLY /* Configure the expansion ROM address */ pci_hose_read_config_byte (hose, dev,PCI_HEADER_TYPE , &header_type); + header_type &=0x7f ;if (header_type !=PCI_HEADER_TYPE_CARDBUS ) { rom_addr = (header_type ==PCI_HEADER_TYPE_NORMAL ) ?PCI_ROM_ADDRESS :PCI_ROM_ADDRESS1 ; --
I think we should get this patch in v2015.10 as it's a bug fix otherwise multi-function devices won't get ROM address assigned.
Regards,
Bin
PATCH 1/1 pci: Fix expansion ROM programming for multi-function devices
In reply to Bin MengQuoted 9 lines · Bin Meng
PCI_HEADER_TYPEregister (offset0x0e) bit 7 is an indicator for multi-function devices. We should mask it off before using it as the header type.Signed-off-by Bin Meng <[email protected]>---
drivers/pci/pci_auto.c | 1 + 1 file changed, 1 insertion(+)
PATCH 1/1 pci: Fix expansion ROM programming for multi-function devices
Hi Bin,
Quoted 5 lines · Simon Glass
Quoted 4 lines · Bin Meng
PCI_HEADER_TYPEregister (offset0x0e) bit 7 is an indicator for multi-function devices. We should mask it off before using it as the header type.
Quoted 7 lines
Quoted 5 lines
Signed-off-by Bin Meng <[email protected]>---
drivers/pci/pci_auto.c | 1 + 1 file changed, 1 insertion(+)Acked-by Simon Glass <[email protected]>
I'm not sure if this fix actually fixes anything for existing boards, but I'll pick it up.
Is there anything else I have missed?
Regards,
Simon
PATCH 1/1 pci: Fix expansion ROM programming for multi-function devices
Hi Simon,
Quoted 11 lines · Simon Glass
Hi Bin,
Quoted 8 lines · Simon Glass
Quoted 7 lines · Bin Meng
PCI_HEADER_TYPEregister (offset0x0e) bit 7 is an indicator for multi-function devices. We should mask it off before using it as the header type.Signed-off-by Bin Meng <[email protected]>---
Quoted 8 lines
Quoted 4 lines
Quoted 2 lines
drivers/pci/pci_auto.c | 1 + 1 file changed, 1 insertion(+)Acked-by Simon Glass <[email protected]>I'm not sure if this fix actually fixes anything for existing boards, but I'll pick it up.
Yes, it fixes the Crown Bay board where Topcliff PCI device D0:F0 is a multi-function device and has a ROM.
Quoted 2 lines
Is there anything else I have missed?
So far I believe everything for this release is in. Thanks!
Regards,
Bin
PATCH 1/1 pci: Fix expansion ROM programming for multi-function devices
Quoted 26 lines · Bin Meng
Hi Simon,
Quoted 11 lines · Simon Glass
Hi Bin,
Quoted 8 lines · Simon Glass
Quoted 7 lines · Bin Meng
PCI_HEADER_TYPEregister (offset0x0e) bit 7 is an indicator for multi-function devices. We should mask it off before using it as the header type.Signed-off-by Bin Meng <[email protected]>---
Quoted 8 lines
Quoted 4 lines
Quoted 2 lines
drivers/pci/pci_auto.c | 1 + 1 file changed, 1 insertion(+)Acked-by Simon Glass <[email protected]>I'm not sure if this fix actually fixes anything for existing boards, but I'll pick it up.
Yes, it fixes the Crown Bay board where Topcliff PCI device D0:F0 is a multi-function device and has a ROM.
Quoted 7 lines
Quoted 2 lines
Is there anything else I have missed?
So far I believe everything for this release is in. Thanks!
Regards,
Bin
Applied to u-boot-x86, and now in mainline.