GNUBoot/coreboot-kgpe-d16
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
coreboot-kgpe-d16/src/southbridge/intel/bd82x6x/pch.c

431 lines
11 KiB
C
Raw Normal View History

Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
/*
* This file is part of the coreboot project.
*
* Copyright (C) 2008-2009 coresystems GmbH
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
* Copyright (C) 2012 The Chromium OS Authors. All rights reserved.
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License as
* published by the Free Software Foundation; version 2 of
* the License.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
GPLv2 notice: Unify all files to just use one space in »MA 02110-1301« In the file `COPYING` in the coreboot repository and upstream [1] just one space is used. The following command was used to convert all files. $ git grep -l 'MA 02' | xargs sed -i 's/MA 02/MA 02/' [1] http://www.gnu.org/licenses/gpl-2.0.txt Change-Id: Ic956dab2820a9e2ccb7841cab66966ba168f305f Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net> Reviewed-on: http://review.coreboot.org/2490 Tested-by: build bot (Jenkins) Reviewed-by: Anton Kochkov <anton.kochkov@gmail.com>
2013-02-23 18:37:27 +01:00
* Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
*/
#include <console/console.h>
#include <delay.h>
Add bd82x6x PCH functions to SMM Add the PCH function to SMM for follow-on SMM patches that require these functions. Change-Id: I7f3a512c5e98446e835b59934d63a99e8af15280 Signed-off-by: Marc Jones <marc.jones@se-eng.com> Reviewed-on: http://review.coreboot.org/2758 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-11 22:36:35 +01:00
#ifdef __SMM__
#include <arch/io.h>
#include <arch/romcc_io.h>
#include <device/pci_def.h>
#else /* !__SMM__ */
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
#include <device/device.h>
#include <device/pci.h>
Add bd82x6x PCH functions to SMM Add the PCH function to SMM for follow-on SMM patches that require these functions. Change-Id: I7f3a512c5e98446e835b59934d63a99e8af15280 Signed-off-by: Marc Jones <marc.jones@se-eng.com> Reviewed-on: http://review.coreboot.org/2758 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-11 22:36:35 +01:00
#endif
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
#include "pch.h"
static int pch_revision_id = -1;
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
static int pch_type = -1;
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
int pch_silicon_revision(void)
{
Add bd82x6x PCH functions to SMM Add the PCH function to SMM for follow-on SMM patches that require these functions. Change-Id: I7f3a512c5e98446e835b59934d63a99e8af15280 Signed-off-by: Marc Jones <marc.jones@se-eng.com> Reviewed-on: http://review.coreboot.org/2758 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-11 22:36:35 +01:00
device_t dev;
#ifdef __SMM__
dev = PCI_DEV(0, 0x1f, 0);
#else
dev = dev_find_slot(0, PCI_DEVFN(0x1f, 0));
#endif
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
if (pch_revision_id < 0)
Add bd82x6x PCH functions to SMM Add the PCH function to SMM for follow-on SMM patches that require these functions. Change-Id: I7f3a512c5e98446e835b59934d63a99e8af15280 Signed-off-by: Marc Jones <marc.jones@se-eng.com> Reviewed-on: http://review.coreboot.org/2758 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-11 22:36:35 +01:00
pch_revision_id = pci_read_config8(dev, PCI_REVISION_ID);
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
return pch_revision_id;
}
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
int pch_silicon_type(void)
{
Add bd82x6x PCH functions to SMM Add the PCH function to SMM for follow-on SMM patches that require these functions. Change-Id: I7f3a512c5e98446e835b59934d63a99e8af15280 Signed-off-by: Marc Jones <marc.jones@se-eng.com> Reviewed-on: http://review.coreboot.org/2758 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-11 22:36:35 +01:00
device_t dev;
#ifdef __SMM__
dev = PCI_DEV(0, 0x1f, 0);
#else
dev = dev_find_slot(0, PCI_DEVFN(0x1f, 0));
#endif
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
if (pch_type < 0)
Add bd82x6x PCH functions to SMM Add the PCH function to SMM for follow-on SMM patches that require these functions. Change-Id: I7f3a512c5e98446e835b59934d63a99e8af15280 Signed-off-by: Marc Jones <marc.jones@se-eng.com> Reviewed-on: http://review.coreboot.org/2758 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-11 22:36:35 +01:00
pch_type = pci_read_config8(dev, PCI_DEVICE_ID + 1);
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
return pch_type;
}
int pch_silicon_supported(int type, int rev)
{
int cur_type = pch_silicon_type();
int cur_rev = pch_silicon_revision();
switch (type) {
case PCH_TYPE_CPT:
/* CougarPoint minimum revision */
if (cur_type == PCH_TYPE_CPT && cur_rev >= rev)
return 1;
/* PantherPoint any revision */
if (cur_type == PCH_TYPE_PPT)
return 1;
break;
case PCH_TYPE_PPT:
/* PantherPoint minimum revision */
if (cur_type == PCH_TYPE_PPT && cur_rev >= rev)
return 1;
break;
}
return 0;
}
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
#define IOBP_RETRY 1000
static inline int iobp_poll(void)
{
unsigned try = IOBP_RETRY;
u32 data;
while (try--) {
data = RCBA32(IOBPS);
if ((data & 1) == 0)
return 1;
udelay(10);
}
printk(BIOS_ERR, "IOBP timeout\n");
return 0;
}
void pch_iobp_update(u32 address, u32 andvalue, u32 orvalue)
{
u32 data;
/* Set the address */
RCBA32(IOBPIRI) = address;
/* READ OPCODE */
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
if (pch_silicon_supported(PCH_TYPE_CPT, PCH_STEP_B0))
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
RCBA32(IOBPS) = IOBPS_RW_BX;
else
RCBA32(IOBPS) = IOBPS_READ_AX;
if (!iobp_poll())
return;
/* Read IOBP data */
data = RCBA32(IOBPD);
if (!iobp_poll())
return;
/* Check for successful transaction */
if ((RCBA32(IOBPS) & 0x6) != 0) {
printk(BIOS_ERR, "IOBP read 0x%08x failed\n", address);
return;
}
/* Update the data */
data &= andvalue;
data |= orvalue;
/* WRITE OPCODE */
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
if (pch_silicon_supported(PCH_TYPE_CPT, PCH_STEP_B0))
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
RCBA32(IOBPS) = IOBPS_RW_BX;
else
RCBA32(IOBPS) = IOBPS_WRITE_AX;
if (!iobp_poll())
return;
/* Write IOBP data */
RCBA32(IOBPD) = data;
if (!iobp_poll())
return;
}
Add bd82x6x PCH functions to SMM Add the PCH function to SMM for follow-on SMM patches that require these functions. Change-Id: I7f3a512c5e98446e835b59934d63a99e8af15280 Signed-off-by: Marc Jones <marc.jones@se-eng.com> Reviewed-on: http://review.coreboot.org/2758 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-11 22:36:35 +01:00
#ifndef __SMM__
/* Set bit in Function Disble register to hide this device */
static void pch_hide_devfn(unsigned devfn)
{
switch (devfn) {
case PCI_DEVFN(22, 0): /* MEI #1 */
RCBA32_OR(FD2, PCH_DISABLE_MEI1);
break;
case PCI_DEVFN(22, 1): /* MEI #2 */
RCBA32_OR(FD2, PCH_DISABLE_MEI2);
break;
case PCI_DEVFN(22, 2): /* IDE-R */
RCBA32_OR(FD2, PCH_DISABLE_IDER);
break;
case PCI_DEVFN(22, 3): /* KT */
RCBA32_OR(FD2, PCH_DISABLE_KT);
break;
case PCI_DEVFN(25, 0): /* Gigabit Ethernet */
RCBA32_OR(BUC, PCH_DISABLE_GBE);
break;
case PCI_DEVFN(26, 0): /* EHCI #2 */
RCBA32_OR(FD, PCH_DISABLE_EHCI2);
break;
case PCI_DEVFN(27, 0): /* HD Audio Controller */
RCBA32_OR(FD, PCH_DISABLE_HD_AUDIO);
break;
case PCI_DEVFN(28, 0): /* PCI Express Root Port 1 */
case PCI_DEVFN(28, 1): /* PCI Express Root Port 2 */
case PCI_DEVFN(28, 2): /* PCI Express Root Port 3 */
case PCI_DEVFN(28, 3): /* PCI Express Root Port 4 */
case PCI_DEVFN(28, 4): /* PCI Express Root Port 5 */
case PCI_DEVFN(28, 5): /* PCI Express Root Port 6 */
case PCI_DEVFN(28, 6): /* PCI Express Root Port 7 */
case PCI_DEVFN(28, 7): /* PCI Express Root Port 8 */
RCBA32_OR(FD, PCH_DISABLE_PCIE(PCI_FUNC(devfn)));
break;
case PCI_DEVFN(29, 0): /* EHCI #1 */
RCBA32_OR(FD, PCH_DISABLE_EHCI1);
break;
case PCI_DEVFN(30, 0): /* PCI-to-PCI Bridge */
RCBA32_OR(FD, PCH_DISABLE_P2P);
break;
case PCI_DEVFN(31, 0): /* LPC */
RCBA32_OR(FD, PCH_DISABLE_LPC);
break;
case PCI_DEVFN(31, 2): /* SATA #1 */
RCBA32_OR(FD, PCH_DISABLE_SATA1);
break;
case PCI_DEVFN(31, 3): /* SMBUS */
RCBA32_OR(FD, PCH_DISABLE_SMBUS);
break;
case PCI_DEVFN(31, 5): /* SATA #22 */
RCBA32_OR(FD, PCH_DISABLE_SATA2);
break;
case PCI_DEVFN(31, 6): /* Thermal Subsystem */
RCBA32_OR(FD, PCH_DISABLE_THERMAL);
break;
}
}
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
/* Check if any port in set X to X+3 is enabled */
static int pch_pcie_check_set_enabled(device_t dev)
{
device_t port;
int port_func;
int dev_func = PCI_FUNC(dev->path.pci.devfn);
printk(BIOS_DEBUG, "%s: check set enabled\n", dev_path(dev));
/* Go through static device tree list of devices
* because enumeration is still in progress */
for (port = all_devices; port; port = port->next) {
/* Only care about PCIe root ports */
if (PCI_SLOT(port->path.pci.devfn) !=
PCI_SLOT(dev->path.pci.devfn))
continue;
/* Check if port is in range and enabled */
port_func = PCI_FUNC(port->path.pci.devfn);
if (port_func >= dev_func &&
port_func < (dev_func + 4) &&
port->enabled)
return 1;
}
/* None of the ports in this set are enabled */
return 0;
}
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
/* RPFN is a write-once register so keep a copy until it is written */
static u32 new_rpfn;
/* Swap function numbers assigned to two PCIe Root Ports */
static void pch_pcie_function_swap(u8 old_fn, u8 new_fn)
{
u32 old_rpfn = new_rpfn;
printk(BIOS_DEBUG, "PCH: Remap PCIe function %d to %d\n",
old_fn, new_fn);
new_rpfn &= ~(RPFN_FNMASK(old_fn) | RPFN_FNMASK(new_fn));
/* Old function set to new function and disabled */
new_rpfn |= RPFN_FNSET(old_fn, RPFN_FNGET(old_rpfn, new_fn));
new_rpfn |= RPFN_FNSET(new_fn, RPFN_FNGET(old_rpfn, old_fn));
}
/* Update devicetree with new Root Port function number assignment */
static void pch_pcie_devicetree_update(void)
{
device_t dev;
/* Update the function numbers in the static devicetree */
for (dev = all_devices; dev; dev = dev->next) {
u8 new_devfn;
/* Only care about PCH PCIe root ports */
if (PCI_SLOT(dev->path.pci.devfn) !=
PCH_PCIE_DEV_SLOT)
continue;
/* Determine the new devfn for this port */
new_devfn = PCI_DEVFN(PCH_PCIE_DEV_SLOT,
RPFN_FNGET(new_rpfn,
PCI_FUNC(dev->path.pci.devfn)));
if (dev->path.pci.devfn != new_devfn) {
printk(BIOS_DEBUG,
"PCH: PCIe map %02x.%1x -> %02x.%1x\n",
PCI_SLOT(dev->path.pci.devfn),
PCI_FUNC(dev->path.pci.devfn),
PCI_SLOT(new_devfn), PCI_FUNC(new_devfn));
dev->path.pci.devfn = new_devfn;
}
}
}
/* Special handling for PCIe Root Port devices */
static void pch_pcie_enable(device_t dev)
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
{
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
struct southbridge_intel_bd82x6x_config *config = dev->chip_info;
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
u32 reg32;
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
/*
* Save a copy of the Root Port Function Number map when
* starting to walk the list of PCIe Root Ports so it can
* be updated locally and written out when the last port
* has been processed.
*/
if (PCI_FUNC(dev->path.pci.devfn) == 0) {
new_rpfn = RCBA32(RPFN);
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
/*
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
* Enable Root Port coalescing if the first port is disabled
* or the other devices will not be enumerated by the OS.
*/
if (!dev->enabled)
config->pcie_port_coalesce = 1;
if (config->pcie_port_coalesce)
printk(BIOS_INFO,
"PCH: PCIe Root Port coalescing is enabled\n");
}
if (!dev->enabled) {
Add PCIe port disable debug message The PCIe device enable function prints when it disables a device. The PCIe ports(bridges) use a different routine that didn't print the message. Add it to be consistent and to provide better debug output. Change-Id: I8462c48e7f4930db68703f0bfb710c01c9643a98 Signed-off-by: Marc Jones <marc.jones@se-eng.com> Reviewed-on: http://review.coreboot.org/1326 Reviewed-by: Patrick Georgi <patrick@georgi-clan.de> Tested-by: build bot (Jenkins)
2012-06-16 07:03:15 +02:00
printk(BIOS_DEBUG, "%s: Disabling device\n", dev_path(dev));
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
/*
* PCIE Power Savings for PantherPoint and CougarPoint/B1+
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
*
* If PCIe 0-3 disabled set Function 0 0xE2[0] = 1
* If PCIe 4-7 disabled set Function 4 0xE2[0] = 1
*
* This check is done here instead of pcie driver
* because the pcie driver enable() handler is not
* called unless the device is enabled.
*/
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
if ((PCI_FUNC(dev->path.pci.devfn) == 0 ||
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
PCI_FUNC(dev->path.pci.devfn) == 4)) {
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
/* Handle workaround for PPT and CPT/B1+ */
if (pch_silicon_supported(PCH_TYPE_CPT, PCH_STEP_B1) &&
!pch_pcie_check_set_enabled(dev)) {
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
u8 reg8 = pci_read_config8(dev, 0xe2);
reg8 |= 1;
pci_write_config8(dev, 0xe2, reg8);
}
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
/*
* Enable Clock Gating for shared PCIe resources
* before disabling this particular port.
*/
pci_write_config8(dev, 0xe1, 0x3c);
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
}
/* Ensure memory, io, and bus master are all disabled */
reg32 = pci_read_config32(dev, PCI_COMMAND);
reg32 &= ~(PCI_COMMAND_MASTER |
Add an option to enable PCIe root port coalescing Background: The PCI spec (3.0-3.2.2.3.4) requires that PCI devices implement function 0. The Linux Kernel therefore will not enumerate a PCI device if it does not present a valid config space at function 0. If a board does not have anything connected to root port 0 and it is desired to disable the unused ports in order to save power then this will cause the other downstream PCIe devices to go missing as they will not be enumerated. Intel chipsets provide a way to map root port numbers to different PCI function numbers, thereby avoiding this issue and allowing root port 0 to be turned off. This change adds a new chip config option 'pcie_port_coalesce' that will collapse the enabled root ports into a linear map starting at zero. This option defaults to disabled as it can have a confusing effect on the system as the declared static devicetree may not match what is seen at runtime. This option is also forced on if the static devicetree disables port 0. When each root port is processed in the early enable stage it looks for a lower numbered root port that has been disabled and then swaps the two assigned function numbers. However the mapping register is write-once so it has to keep track of the proposed mapping changes until all ports have been processed before writing out the final map value. At this point it also updates the function numbers in the static device tree so they are consistent with the new layout. There are a few other closely related fixes in this change: 1) There is a power savings opportunity if an entire bank of ports (0-3 or 4-7) are disabled. This was checking the chipset revision to look for CougarPoint B1+ stepping and that was not passing on PantherPoint where this should always be applied. To fix this I added a function to determine the chipset type based on comparing the upper byte of the device ID. 2) Apply the same chipset type check fix to the IOBP programming. 3) There is another power savings opportunity to enable dynamic clock gating on shared PCIe resources which only applies to ports 0 and 4. However if 0 or 4 is disabled then the later check to enable this would fail as that device is already hidden. LUMPY current: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.3 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B LUMPY with PCIe port coalesce enabled: 00:1c.0 PCI bridge: Intel Corporation Device 1c10 (rev b5) 00:1c.1 PCI bridge: Intel Corporation Device 1c16 (rev b5) 01:00.0 Network controller: Atheros Communications Inc. Device 0030 (rev 01) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B Change-Id: I828aa407fdc9c156c1c42eda8e2d893c0aa66eef Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/979 Tested-by: build bot (Jenkins)
2012-04-27 19:30:51 +02:00
PCI_COMMAND_MEMORY | PCI_COMMAND_IO);
pci_write_config32(dev, PCI_COMMAND, reg32);
/* Do not claim downstream transactions for PCIe ports */
new_rpfn |= RPFN_HIDE(PCI_FUNC(dev->path.pci.devfn));
/* Hide this device if possible */
pch_hide_devfn(dev->path.pci.devfn);
} else {
int fn;
/*
* Check if there is a lower disabled port to swap with this
* port in order to maintain linear order starting at zero.
*/
if (config->pcie_port_coalesce) {
for (fn=0; fn < PCI_FUNC(dev->path.pci.devfn); fn++) {
if (!(new_rpfn & RPFN_HIDE(fn)))
continue;
/* Swap places with this function */
pch_pcie_function_swap(
PCI_FUNC(dev->path.pci.devfn), fn);
break;
}
}
/* Enable SERR */
reg32 = pci_read_config32(dev, PCI_COMMAND);
reg32 |= PCI_COMMAND_SERR;
pci_write_config32(dev, PCI_COMMAND, reg32);
}
/*
* When processing the last PCIe root port we can now
* update the Root Port Function Number and Hide register.
*/
if (PCI_FUNC(dev->path.pci.devfn) == 7) {
printk(BIOS_SPEW, "PCH: RPFN 0x%08x -> 0x%08x\n",
RCBA32(RPFN), new_rpfn);
RCBA32(RPFN) = new_rpfn;
/* Update static devictree with new function numbers */
if (config->pcie_port_coalesce)
pch_pcie_devicetree_update();
}
}
void pch_enable(device_t dev)
{
u32 reg32;
/* PCH PCIe Root Ports get special handling */
if (PCI_SLOT(dev->path.pci.devfn) == PCH_PCIE_DEV_SLOT)
return pch_pcie_enable(dev);
if (!dev->enabled) {
printk(BIOS_DEBUG, "%s: Disabling device\n", dev_path(dev));
/* Ensure memory, io, and bus master are all disabled */
reg32 = pci_read_config32(dev, PCI_COMMAND);
reg32 &= ~(PCI_COMMAND_MASTER |
PCI_COMMAND_MEMORY | PCI_COMMAND_IO);
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
pci_write_config32(dev, PCI_COMMAND, reg32);
/* Hide this device if possible */
pch_hide_devfn(dev->path.pci.devfn);
} else {
/* Enable SERR */
reg32 = pci_read_config32(dev, PCI_COMMAND);
reg32 |= PCI_COMMAND_SERR;
pci_write_config32(dev, PCI_COMMAND, reg32);
}
}
struct chip_operations southbridge_intel_bd82x6x_ops = {
Sandy/Ivy Bridge and Cougar/Panther Point: Fix names The names were set at various times during development, but the way the code works, you might end up with the wrong name being displayed in the logs. Instead of doing magic, just display both names for each component Change-Id: I1f8ce44d156442f5f7d717e1a2b47ed1218d4527 Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/1413 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-07-26 01:10:36 +02:00
CHIP_NAME("Intel Series 6/7 (Cougar Point/Panther Point) Southbridge")
Add support for Intel Panther Point PCH Change-Id: Iac3cd25b36493bb203e849674320e113cc5fce32 Signed-off-by: Duncan Laurie <dlaurie@google.com> Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/853 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-04-04 00:07:22 +02:00
.enable_dev = pch_enable,
};
Add bd82x6x PCH functions to SMM Add the PCH function to SMM for follow-on SMM patches that require these functions. Change-Id: I7f3a512c5e98446e835b59934d63a99e8af15280 Signed-off-by: Marc Jones <marc.jones@se-eng.com> Reviewed-on: http://review.coreboot.org/2758 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-11 22:36:35 +01:00
#endif
Reference in a new issue Copy permalink