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

750 lines
17 KiB
C
Raw Normal View History

src/southbridge: Use SPDX for GPL-2.0-only files Done with sed and God Lines. Only done for C-like code for now. Change-Id: I5b00b3e38edda90f35f0679cd4171a3499288f24 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/40059 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
2020-04-02 23:49:05 +02:00
/* SPDX-License-Identifier: GPL-2.0-only */
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
#include <assert.h>
#include <commonlib/helpers.h>
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
#include <console/console.h>
#include <device/device.h>
#include <device/pci.h>
intel/pci: Utilise pci_def.h for PCI_BRIDGE_CONTROL This is a PCI standard register, no need to alias its definitions under different names. Change-Id: Iea6b198dd70fe1e49b5dc0824dba62628dedc69a Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/35521 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2019-09-21 17:35:37 +02:00
#include <device/pci_def.h>
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
#include <device/pciexp.h>
#include <device/pci_ids.h>
pci: Move inline PCI functions to pci_ops.h Move inline function where they belong to. Fixes compilation on non x86 platforms. Change-Id: Ia05391c43b8d501bd68df5654bcfb587f8786f71 Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com> Reviewed-on: https://review.coreboot.org/25720 Reviewed-by: Aaron Durbin <adurbin@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-04-18 10:11:59 +02:00
#include <device/pci_ops.h>
sb/intel/lynxpoint: Move IOBP API to its own compilation unit Change-Id: Icb6114302cebe19bc3c1971929ea4fc085b454be Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/41946 Reviewed-by: Michael Niewöhner Reviewed-by: Felix Held <felix-coreboot@felixheld.de> Reviewed-by: Jonathan Kollasch <jakllsch@kollasch.net> Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-05-31 00:55:35 +02:00
#include "iobp.h"
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
#include "pch.h"
southbridge/intel/lynxpoint: Use common gpio.c Use shared gpio code from common folder, except for INTEL_LYNXPOINT_LP, which has it's own gpio code. Needs test on real hardware ! Change-Id: Iccc6d254bafb927b6470704cec7c9dd7528e2c68 Signed-off-by: Patrick Rudolph <siro@das-labor.org> Reviewed-on: https://review.coreboot.org/13615 Tested-by: build bot (Jenkins) Reviewed-by: Martin Roth <martinroth@google.com>
2016-02-06 18:07:59 +01:00
#include <southbridge/intel/common/gpio.h>
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
#include <stdint.h>
southbridge/intel: Tidy up preprocessor and headers Change-Id: I52a7b39566acd64ac21a345046675e05649a40f5 Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/34980 Reviewed-by: Furquan Shaikh <furquan@google.com> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2019-08-18 15:33:39 +02:00
#include "chip.h"
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
#define MAX_NUM_ROOT_PORTS 8
struct root_port_config {
/* RPFN is a write-once register so keep a copy until it is written */
u32 orig_rpfn;
u32 new_rpfn;
u32 pin_ownership;
u32 strpfusecfg1;
u32 strpfusecfg2;
u32 strpfusecfg3;
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
u32 b0d28f0_32c;
u32 b0d28f4_32c;
u32 b0d28f5_32c;
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
int coalesce;
int gbe_port;
int num_ports;
sb/intel/lynxpoint: Get rid of device_t Use of device_t has been abandoned in ramstage. Change-Id: I064ff5e76dd95c1770cd24139195b2a5fff2d382 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/26974 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2018-06-08 17:20:38 +02:00
struct device *ports[MAX_NUM_ROOT_PORTS];
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
};
static struct root_port_config rpc;
static inline int max_root_ports(void)
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
{
sb/intel/lynxpoint: Handle H81 only having 6 PCIe root ports The H81 chipset is the only non-LP Lynx Point chipset with 6 PCIe root ports, all others have 8 [1]. The existing PCIe code assumed that all non-LP chipsets had 8 root ports, which meant that port 6 would not be considered the last root port on H81, so `root_port_commit_config()` would not run. Ultimately, while PCIe still worked on H81, all the root ports would remain enabled, even if disabled in the devicetree. Also, remove `PCI_DEVICE_ID_INTEL_LYNXPOINT_MOB_DESK_{MIN,MAX}`, as they are unused, and the MAX constant is incorrect. Interestingly, this fixes an issue where GRUB is unable to halt the system. Tested on an ASRock H81M-HDS. The root ports disabled in the devicetree do indeed end up disabled. [1] Intel® 8 Series/C220 Series Chipset Family Platform Controller Hub (PCH) Datasheet, revision 003, document number 328904. Change-Id: If3ce217e8a4f4ea4e111e4525b03dbbfc63f92b0 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30077 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Patrick Rudolph <siro@das-labor.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:46:58 +01:00
if (pch_is_lp() || pch_silicon_id() == PCI_DEVICE_ID_INTEL_LPT_H81)
return 6;
return 8;
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
}
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
sb/intel/lynxpoint: Get rid of device_t Use of device_t has been abandoned in ramstage. Change-Id: I064ff5e76dd95c1770cd24139195b2a5fff2d382 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/26974 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2018-06-08 17:20:38 +02:00
static inline int root_port_is_first(struct device *dev)
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
{
return PCI_FUNC(dev->path.pci.devfn) == 0;
}
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
sb/intel/lynxpoint: Get rid of device_t Use of device_t has been abandoned in ramstage. Change-Id: I064ff5e76dd95c1770cd24139195b2a5fff2d382 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/26974 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2018-06-08 17:20:38 +02:00
static inline int root_port_is_last(struct device *dev)
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
{
return PCI_FUNC(dev->path.pci.devfn) == (rpc.num_ports - 1);
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
}
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
/* Root ports are numbered 1..N in the documentation. */
sb/intel/lynxpoint: Get rid of device_t Use of device_t has been abandoned in ramstage. Change-Id: I064ff5e76dd95c1770cd24139195b2a5fff2d382 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/26974 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2018-06-08 17:20:38 +02:00
static inline int root_port_number(struct device *dev)
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
{
return PCI_FUNC(dev->path.pci.devfn) + 1;
}
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
static bool is_rp_enabled(int rp)
{
ASSERT(rp > 0 && rp <= ARRAY_SIZE(rpc.ports));
if (rpc.ports[rp - 1] == NULL)
return false;
return rpc.ports[rp - 1]->enabled;
}
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
static void root_port_config_update_gbe_port(void)
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
{
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
/* Is the Gbe Port enabled? */
if (!((rpc.strpfusecfg1 >> 19) & 1))
return;
if (pch_is_lp()) {
switch ((rpc.strpfusecfg1 >> 16) & 0x7) {
case 0:
rpc.gbe_port = 3;
break;
case 1:
rpc.gbe_port = 4;
break;
case 2:
case 3:
case 4:
case 5:
/* Lanes 0-4 of Root Port 5. */
rpc.gbe_port = 5;
break;
default:
printk(BIOS_DEBUG, "Invalid GbE Port Selection.\n");
}
} else {
/* Non-LP has 1:1 mapping with root ports. */
rpc.gbe_port = ((rpc.strpfusecfg1 >> 16) & 0x7) + 1;
}
}
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
sb/intel/lynxpoint: Consider root ports being disabled by strap PCIe RPC (Root Port Configuration) straps will force-disable some root port functions if some root ports have a width greater than x1. In two cases, this affects the last function. The PCIe init code will never finish configuring the root ports if that is the case: it assumes that the last function will eventually run through the code, but it doesn't. If PCIe initialization does not complete, pressing the power button will not power off the board, unless it is held for about five seconds. Also, Windows 10 will show a BSOD about MACHINE CHECK EXCEPTION, and lock up instead of rebooting. Depending on the microcode version, the BSOD may not be visible. This happens even when the root port is not populated. Use the strap fuse configuration value to know which configuration the PCH is strapped to. If needed, update the number of ports accordingly. In addition, print the updated value to ease debugging PCIe init code. Existing code in coreboot disagrees with public documentation about the root port width straps. Assume existing code is correct and document these assumptions in a table, as an explanation for the added code. Tested on Asrock B85M Pro4, PCIe initialization completes successfully. Change-Id: Id6da3a1f45467f00002a5ed41df8650f4a74eeba Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/44155 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-08-04 00:26:45 +02:00
static void update_num_ports(void)
{
/*
* According to existing code in 'root_port_check_disable()', which does
* not agree with the confusing information on the datasheets, the last
* visible function depends on the strapped root port width as follows:
*
* +-----+----+----+----+----+
* | RPC | #5 | #6 | #7 | #8 |
* +-----+----+----+----+----+
* | 0 | x1 | x1 | x1 | x1 |
* | 1 | x2 | | x1 | x1 |
* | 2 | x2 | | x2 | |
* | 3 | x4 | | | |
* +-----+----+----+----+----+
*/
switch ((rpc.strpfusecfg2 >> 14) & 0x3) {
case 0:
case 1:
break;
case 2:
rpc.num_ports = MIN(rpc.num_ports, 7);
break;
case 3:
rpc.num_ports = MIN(rpc.num_ports, 5);
break;
}
printk(BIOS_DEBUG, "Adjusted number of PCIe root ports to %d as per strpfusecfg2\n",
rpc.num_ports);
}
sb/intel/lynxpoint: Get rid of device_t Use of device_t has been abandoned in ramstage. Change-Id: I064ff5e76dd95c1770cd24139195b2a5fff2d382 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/26974 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2018-06-08 17:20:38 +02:00
static void root_port_init_config(struct device *dev)
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
{
int rp;
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
if (root_port_is_first(dev)) {
rpc.orig_rpfn = RCBA32(RPFN);
rpc.new_rpfn = rpc.orig_rpfn;
rpc.num_ports = max_root_ports();
rpc.gbe_port = -1;
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
rpc.pin_ownership = pci_read_config32(dev, 0x410);
root_port_config_update_gbe_port();
if (dev->chip_info != NULL) {
struct southbridge_intel_lynxpoint_config *config;
config = dev->chip_info;
rpc.coalesce = config->pcie_port_coalesce;
}
}
rp = root_port_number(dev);
if (rp > rpc.num_ports) {
printk(BIOS_ERR, "Found Root Port %d, expecting %d\n",
rp, rpc.num_ports);
return;
}
/* Read the fuse configuration and pin ownership. */
switch (rp) {
case 1:
rpc.strpfusecfg1 = pci_read_config32(dev, 0xfc);
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
rpc.b0d28f0_32c = pci_read_config32(dev, 0x32c);
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
break;
case 5:
rpc.strpfusecfg2 = pci_read_config32(dev, 0xfc);
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
rpc.b0d28f4_32c = pci_read_config32(dev, 0x32c);
sb/intel/lynxpoint: Consider root ports being disabled by strap PCIe RPC (Root Port Configuration) straps will force-disable some root port functions if some root ports have a width greater than x1. In two cases, this affects the last function. The PCIe init code will never finish configuring the root ports if that is the case: it assumes that the last function will eventually run through the code, but it doesn't. If PCIe initialization does not complete, pressing the power button will not power off the board, unless it is held for about five seconds. Also, Windows 10 will show a BSOD about MACHINE CHECK EXCEPTION, and lock up instead of rebooting. Depending on the microcode version, the BSOD may not be visible. This happens even when the root port is not populated. Use the strap fuse configuration value to know which configuration the PCH is strapped to. If needed, update the number of ports accordingly. In addition, print the updated value to ease debugging PCIe init code. Existing code in coreboot disagrees with public documentation about the root port width straps. Assume existing code is correct and document these assumptions in a table, as an explanation for the added code. Tested on Asrock B85M Pro4, PCIe initialization completes successfully. Change-Id: Id6da3a1f45467f00002a5ed41df8650f4a74eeba Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/44155 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-08-04 00:26:45 +02:00
if (!pch_is_lp())
update_num_ports();
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
break;
case 6:
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
rpc.b0d28f5_32c = pci_read_config32(dev, 0x32c);
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
rpc.strpfusecfg3 = pci_read_config32(dev, 0xfc);
break;
default:
break;
}
/* Cache pci device. */
rpc.ports[rp - 1] = dev;
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
}
/* Update devicetree with new Root Port function number assignment */
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
static void pch_pcie_device_set_func(int index, int pci_func)
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
{
sb/intel/lynxpoint: Get rid of device_t Use of device_t has been abandoned in ramstage. Change-Id: I064ff5e76dd95c1770cd24139195b2a5fff2d382 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/26974 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2018-06-08 17:20:38 +02:00
struct device *dev;
src/southbridge: change "unsigned" to "unsigned int" Signed-off-by: Martin Roth <martin@coreboot.org> Change-Id: Iee2056a50a1201626fa29194afdbfc1f11094420 Reviewed-on: https://review.coreboot.org/c/coreboot/+/36333 Reviewed-by: Patrick Rudolph <siro@das-labor.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2019-10-24 05:46:03 +02:00
unsigned int new_devfn;
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
dev = rpc.ports[index];
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
/* Set the new PCI function field for this Root Port. */
rpc.new_rpfn &= ~RPFN_FNMASK(index);
rpc.new_rpfn |= RPFN_FNSET(index, pci_func);
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
/* Determine the new devfn for this port */
new_devfn = PCI_DEVFN(PCH_PCIE_DEV_SLOT, pci_func);
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
if (dev && dev->path.pci.devfn != new_devfn) {
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
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));
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
dev->path.pci.devfn = new_devfn;
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
}
}
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
static void pcie_enable_clock_gating(void)
{
int i;
int is_lp;
int enabled_ports;
is_lp = pch_is_lp();
enabled_ports = 0;
for (i = 0; i < rpc.num_ports; i++) {
sb/intel/lynxpoint: Get rid of device_t Use of device_t has been abandoned in ramstage. Change-Id: I064ff5e76dd95c1770cd24139195b2a5fff2d382 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/26974 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2018-06-08 17:20:38 +02:00
struct device *dev;
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
int rp;
dev = rpc.ports[i];
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
if (!dev)
continue;
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
rp = root_port_number(dev);
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
if (!is_rp_enabled(rp)) {
southbridge/intel/lynxpoint: Fix undefined behavior Fix undefined behavior found by clang's -Wshift-sign-overflow, grep, and source inspection. Left shifting an int where the right operand is >= the width of the type is undefined. Add UL suffix since it's safe for unsigned types. Change-Id: I10db2566199200ceb3068721cfb35eadb2be1f68 Signed-off-by: Ryan Salsamendi <rsalsamendi@hotmail.com> Reviewed-on: https://review.coreboot.org/20464 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2017-07-04 22:14:16 +02:00
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
/* Configure shared resource clock gating. */
if (rp == 1 || rp == 5 || (rp == 6 && is_lp))
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config8(dev, 0xe1, 0x3c);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
if (!is_lp) {
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
if (rp == 1 && !is_rp_enabled(2) &&
!is_rp_enabled(3) && !is_rp_enabled(4)) {
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config8(dev, 0xe2, 1);
pci_or_config8(dev, 0xe1, 1 << 7);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
}
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
if (rp == 5 && !is_rp_enabled(6) &&
!is_rp_enabled(7) && !is_rp_enabled(8)) {
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config8(dev, 0xe2, 1);
pci_or_config8(dev, 0xe1, 1 << 7);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
}
continue;
}
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config8(dev, 0xe2, 3 << 4);
pci_or_config32(dev, 0x420, 1 << 31);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
/* Per-Port CLKREQ# handling. */
if (is_lp && gpio_is_native(18 + rp - 1))
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config32(dev, 0x420, 3 << 29);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
/* Enable static clock gating. */
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
if (rp == 1 && !is_rp_enabled(2) &&
!is_rp_enabled(3) && !is_rp_enabled(4)) {
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config8(dev, 0xe2, 1);
pci_or_config8(dev, 0xe1, 1 << 7);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
} else if (rp == 5 || rp == 6) {
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config8(dev, 0xe2, 1);
pci_or_config8(dev, 0xe1, 1 << 7);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
}
continue;
}
enabled_ports++;
/* Enable dynamic clock gating. */
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config8(dev, 0xe1, 0x03);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
if (is_lp) {
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config8(dev, 0xe2, 1 << 6);
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config8(dev, 0xe8, ~(3 << 2), (2 << 2));
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
}
/* Update PECR1 register. */
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config8(dev, 0xe8, 1);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
/* FIXME: Are we supposed to update this register with a constant boolean? */
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config8(dev, 0x324, ~(1 << 5), (1 < 5));
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
/* Per-Port CLKREQ# handling. */
if (is_lp && gpio_is_native(18 + rp - 1))
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config32(dev, 0x420, 3 << 29);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
/* Configure shared resource clock gating. */
if (rp == 1 || rp == 5 || (rp == 6 && is_lp))
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config8(dev, 0xe1, 0x3c);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
}
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
if (!enabled_ports && is_lp && rpc.ports[0])
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config8(rpc.ports[0], 0xe1, 1 << 6);
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
}
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
static void root_port_commit_config(void)
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
{
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
int i;
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
/* If the first root port is disabled the coalesce ports. */
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
if (!is_rp_enabled(1))
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
rpc.coalesce = 1;
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
lynxpoint: enable clock gating Implement the LynxPoint BIOS Spec for clock gating. Change-Id: Iaa84cb447bd29b0d13cdda481a1661ea40499de1 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59590 Reviewed-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/4325 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-21 21:06:11 +02:00
/* Perform clock gating configuration. */
pcie_enable_clock_gating();
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
for (i = 0; i < rpc.num_ports; i++) {
sb/intel/lynxpoint: Get rid of device_t Use of device_t has been abandoned in ramstage. Change-Id: I064ff5e76dd95c1770cd24139195b2a5fff2d382 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/26974 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2018-06-08 17:20:38 +02:00
struct device *dev;
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
dev = rpc.ports[i];
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
if (dev == NULL) {
printk(BIOS_ERR, "Root Port %d device is NULL?\n", i+1);
continue;
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
}
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
if (dev->enabled)
continue;
printk(BIOS_DEBUG, "%s: Disabling device\n", dev_path(dev));
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
/* Ensure memory, io, and bus master are all disabled */
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_and_config16(dev, PCI_COMMAND,
~(PCI_COMMAND_MASTER | PCI_COMMAND_MEMORY | PCI_COMMAND_IO));
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
/* Disable this device if possible */
pch_disable_devfn(dev);
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
}
if (rpc.coalesce) {
int current_func;
/* For all Root Ports N enabled ports get assigned the lower
* PCI function number. The disabled ones get upper PCI
* function numbers. */
current_func = 0;
for (i = 0; i < rpc.num_ports; i++) {
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
if (!is_rp_enabled(i + 1))
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
continue;
pch_pcie_device_set_func(i, current_func);
current_func++;
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
}
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
/* Allocate the disabled devices' PCI function number. */
for (i = 0; i < rpc.num_ports; i++) {
sb/intel/lynxpoint/pcie.c: Add more checks for NULL pointers If PCIe root port `n` is disabled, then `rpc.ports[n - 1]` remains NULL. The existing Lynx Point systems probably don't end up dereferencing NULL pointers this way. However, it might occur on a system using Flexible I/O to remap PCIe root ports to other functions. Tested on an ASRock H81M-HDS and an Acer C720 (Google Peppy). No issues presented themselves. Change-Id: I2c22fa36217766c2c4d6e8046f99989063066b16 Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30079 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2018-12-06 10:47:42 +01:00
if (is_rp_enabled(i + 1))
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
continue;
pch_pcie_device_set_func(i, current_func);
current_func++;
}
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
}
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
printk(BIOS_SPEW, "PCH: RPFN 0x%08x -> 0x%08x\n",
rpc.orig_rpfn, rpc.new_rpfn);
RCBA32(RPFN) = rpc.new_rpfn;
}
sb/intel/lynxpoint: Get rid of device_t Use of device_t has been abandoned in ramstage. Change-Id: I064ff5e76dd95c1770cd24139195b2a5fff2d382 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/26974 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2018-06-08 17:20:38 +02:00
static void root_port_mark_disable(struct device *dev)
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
{
/* Mark device as disabled. */
dev->enabled = 0;
/* Mark device to be hidden. */
rpc.new_rpfn |= RPFN_HIDE(PCI_FUNC(dev->path.pci.devfn));
}
sb/intel/lynxpoint: Get rid of device_t Use of device_t has been abandoned in ramstage. Change-Id: I064ff5e76dd95c1770cd24139195b2a5fff2d382 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/26974 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2018-06-08 17:20:38 +02:00
static void root_port_check_disable(struct device *dev)
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
{
int rp;
int is_lp;
/* Device already disabled. */
if (!dev->enabled) {
root_port_mark_disable(dev);
return;
}
rp = root_port_number(dev);
/* Is the GbE port mapped to this Root Port? */
if (rp == rpc.gbe_port) {
root_port_mark_disable(dev);
return;
}
is_lp = pch_is_lp();
/* Check Root Port Configuration. */
switch (rp) {
case 2:
/* Root Port 2 is disabled for all lane configurations
* but config 00b (4x1 links). */
if ((rpc.strpfusecfg1 >> 14) & 0x3) {
root_port_mark_disable(dev);
return;
}
break;
case 3:
/* Root Port 3 is disabled in config 11b (1x4 links). */
if (((rpc.strpfusecfg1 >> 14) & 0x3) == 0x3) {
root_port_mark_disable(dev);
return;
}
break;
case 4:
/* Root Port 4 is disabled in configs 11b (1x4 links)
* and 10b (2x2 links). */
if ((rpc.strpfusecfg1 >> 14) & 0x2) {
root_port_mark_disable(dev);
return;
}
break;
case 6:
if (is_lp)
break;
/* Root Port 6 is disabled for all lane configurations
* but config 00b (4x1 links). */
if ((rpc.strpfusecfg2 >> 14) & 0x3) {
root_port_mark_disable(dev);
return;
}
break;
case 7:
if (is_lp)
break;
sb/intel/lynxpoint/pcie.c: Fix a mistake in a comment The code annotated by the comment is dealing with root port 7, so update the comment to reflect that. It looks like the comment was copied from the root port 3 case, but not updated. Change-Id: I0e27e4453f4c3b2b1b9dffb0c89b71373c6b303e Signed-off-by: Tristan Corrick <tristan@corrick.kiwi> Reviewed-on: https://review.coreboot.org/c/30078 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Angel Pons <th3fanbus@gmail.com>
2018-12-06 10:47:21 +01:00
/* Root Port 7 is disabled in config 11b (1x4 links). */
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
if (((rpc.strpfusecfg2 >> 14) & 0x3) == 0x3) {
root_port_mark_disable(dev);
return;
}
break;
case 8:
if (is_lp)
break;
/* Root Port 8 is disabled in configs 11b (1x4 links)
* and 10b (2x2 links). */
if ((rpc.strpfusecfg2 >> 14) & 0x2) {
root_port_mark_disable(dev);
return;
}
break;
}
/* Check Pin Ownership. */
if (is_lp) {
switch (rp) {
case 1:
/* Bit 0 is Root Port 1 ownership. */
if ((rpc.pin_ownership & 0x1) == 0) {
root_port_mark_disable(dev);
return;
}
break;
case 2:
/* Bit 2 is Root Port 2 ownership. */
if ((rpc.pin_ownership & 0x4) == 0) {
root_port_mark_disable(dev);
return;
}
break;
case 6:
/* Bits 7:4 are Root Port 6 pin-lane ownership. */
if ((rpc.pin_ownership & 0xf0) == 0) {
root_port_mark_disable(dev);
return;
}
break;
}
} else {
switch (rp) {
case 1:
/* Bits 4 and 0 are Root Port 1 ownership. */
if ((rpc.pin_ownership & 0x11) == 0) {
root_port_mark_disable(dev);
return;
}
break;
case 2:
/* Bits 5 and 2 are Root Port 2 ownership. */
if ((rpc.pin_ownership & 0x24) == 0) {
root_port_mark_disable(dev);
return;
}
break;
}
lynxpoint: move all pcie device handling to pcie.c Some of the pcie logic was located in pch.c as well as pcie.c. Move all pcie logic to the same pcie.c file. This is a straight cut-and-paste (no logic changes) except for a rename from pch_pcie_enable() -> pch_pcie_enable_dev(). Change-Id: I338c53039b95f255ab9ced313c51193a9d34b404 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59277 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4251 Tested-by: build bot (Jenkins) Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
2013-06-19 20:28:04 +02:00
}
}
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
static void pcie_add_0x0202000_iobp(u32 reg)
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
{
u32 reg32;
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
reg32 = pch_iobp_read(reg);
reg32 += (0x2 << 16) | (0x2 << 8);
pch_iobp_write(reg, reg32);
}
static void pch_pcie_early(struct device *dev)
{
int rp;
int do_aspm;
int is_lp;
lynxpoint: Add devicetree config option to force enable ASPM The PCIe root port has ASPM settings/workarounds that are only applied based on the value of an undocumented bit in PCI config register 0x32C. If that bit is not set for some reason then the settings are not applied. This devicetree config option will force the ASPM settings for each port based on the bit map. Change-Id: I40b08ca9a0ef52742609bac72fb821454a373799 Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/65314 Reviewed-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/4453 Tested-by: build bot (Jenkins) Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
2013-08-09 18:06:41 +02:00
struct southbridge_intel_lynxpoint_config *config = dev->chip_info;
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
rp = root_port_number(dev);
do_aspm = 0;
is_lp = pch_is_lp();
if (is_lp) {
switch (rp) {
case 1:
case 2:
case 3:
case 4:
/* Bits 31:28 of b0d28f0 0x32c register correspnd to
* Root Ports 4:1. */
do_aspm = !!(rpc.b0d28f0_32c & (1 << (28 + rp - 1)));
break;
case 5:
/* Bit 28 of b0d28f4 0x32c register correspnd to
* Root Ports 4:1. */
do_aspm = !!(rpc.b0d28f4_32c & (1 << 28));
break;
case 6:
/* Bit 28 of b0d28f5 0x32c register correspnd to
* Root Ports 4:1. */
do_aspm = !!(rpc.b0d28f5_32c & (1 << 28));
break;
}
} else {
switch (rp) {
case 1:
case 2:
case 3:
case 4:
/* Bits 31:28 of b0d28f0 0x32c register correspnd to
* Root Ports 4:1. */
do_aspm = !!(rpc.b0d28f0_32c & (1 << (28 + rp - 1)));
break;
case 5:
case 6:
case 7:
case 8:
/* Bit 31:28 of b0d28f4 0x32c register correspnd to
* Root Ports 8:5. */
do_aspm = !!(rpc.b0d28f4_32c & (1 << (28 + rp - 5)));
break;
}
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
}
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
lynxpoint: Add devicetree config option to force enable ASPM The PCIe root port has ASPM settings/workarounds that are only applied based on the value of an undocumented bit in PCI config register 0x32C. If that bit is not set for some reason then the settings are not applied. This devicetree config option will force the ASPM settings for each port based on the bit map. Change-Id: I40b08ca9a0ef52742609bac72fb821454a373799 Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/65314 Reviewed-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/4453 Tested-by: build bot (Jenkins) Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
2013-08-09 18:06:41 +02:00
/* Allow ASPM to be forced on in devicetree */
if (config && (config->pcie_port_force_aspm & (1 << (rp - 1))))
do_aspm = 1;
printk(BIOS_DEBUG, "PCIe Root Port %d ASPM is %sabled\n",
rp, do_aspm ? "en" : "dis");
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
if (do_aspm) {
/* Set ASPM bits in MPC2 register. */
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config32(dev, 0xd4, ~(0x3 << 2), (1 << 4) | (0x2 << 2));
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
/* Set unique clock exit latency in MPC register. */
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config32(dev, 0xd8, ~(0x7 << 18), (0x7 << 18));
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
/* Set L1 exit latency in LCAP register. */
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config32(dev, 0x4c, ~(0x7 << 15), (0x4 << 15));
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
if (is_lp) {
switch (rp) {
case 1:
pcie_add_0x0202000_iobp(0xe9002440);
break;
case 2:
pcie_add_0x0202000_iobp(0xe9002640);
break;
case 3:
pcie_add_0x0202000_iobp(0xe9000840);
break;
case 4:
pcie_add_0x0202000_iobp(0xe9000a40);
break;
case 5:
pcie_add_0x0202000_iobp(0xe9000c40);
pcie_add_0x0202000_iobp(0xe9000e40);
pcie_add_0x0202000_iobp(0xe9001040);
pcie_add_0x0202000_iobp(0xe9001240);
break;
case 6:
/* Update IOBP based on lane ownership. */
if (rpc.pin_ownership & (1 << 4))
pcie_add_0x0202000_iobp(0xea002040);
if (rpc.pin_ownership & (1 << 5))
pcie_add_0x0202000_iobp(0xea002240);
if (rpc.pin_ownership & (1 << 6))
pcie_add_0x0202000_iobp(0xea002440);
if (rpc.pin_ownership & (1 << 7))
pcie_add_0x0202000_iobp(0xea002640);
break;
}
} else {
switch (rp) {
case 1:
if ((rpc.pin_ownership & 0x3) == 1)
pcie_add_0x0202000_iobp(0xe9002e40);
else
pcie_add_0x0202000_iobp(0xea002040);
break;
case 2:
if ((rpc.pin_ownership & 0xc) == 0x4)
pcie_add_0x0202000_iobp(0xe9002c40);
else
pcie_add_0x0202000_iobp(0xea002240);
break;
case 3:
pcie_add_0x0202000_iobp(0xe9002a40);
break;
case 4:
pcie_add_0x0202000_iobp(0xe9002840);
break;
case 5:
pcie_add_0x0202000_iobp(0xe9002640);
break;
case 6:
pcie_add_0x0202000_iobp(0xe9002440);
break;
case 7:
pcie_add_0x0202000_iobp(0xe9002240);
break;
case 8:
pcie_add_0x0202000_iobp(0xe9002040);
break;
}
}
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_and_config32(dev, 0x338, ~(1 << 26));
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
}
/* Enable LTR in Root Port. */
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config32(dev, 0x64, 1 << 11);
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config32(dev, 0x68, ~(1 << 10), (1 << 10));
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
southbridge/intel/lynxpoint: Fix undefined behavior Fix reports found by undefined behavior sanitizer. Left shifting an int where the right operand is >= the width of the type is undefined. Add UL suffix since it's safe for unsigned types. Change-Id: I755b3c80a8d1b6cb6b6e5f411c6691e5dd17c266 Signed-off-by: Ryan Salsamendi <rsalsamendi@hotmail.com> Reviewed-on: https://review.coreboot.org/20443 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
2017-07-01 02:15:57 +02:00
pci_update_config32(dev, 0x318, ~(0xffffUL << 16), (0x1414UL << 16));
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
/* Set L1 exit latency in LCAP register. */
if (!do_aspm && (pci_read_config8(dev, 0xf5) & 0x1))
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config32(dev, 0x4c, ~(0x7 << 15), (0x4 << 15));
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
else
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config32(dev, 0x4c, ~(0x7 << 15), (0x2 << 15));
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_update_config32(dev, 0x314, 0, 0x743a361b);
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
/* Set Common Clock Exit Latency in MPC register. */
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config32(dev, 0xd8, ~(0x7 << 15), (0x3 << 15));
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config32(dev, 0x33c, ~0x00ffffff, 0x854c74);
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
/* Set Invalid Recieve Range Check Enable in MPC register. */
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config32(dev, 0xd8, 1 << 25);
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_and_config8(dev, 0xf5, 0x3f);
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
if (rp == 1 || rp == 5 || (is_lp && rp == 6))
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_and_config8(dev, 0xf7, ~0x0c);
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
/* Set EOI forwarding disable. */
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config32(dev, 0xd4, 1 << 1);
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
/* Set something involving advanced error reporting. */
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config32(dev, 0x100, ~((1 << 20) - 1), 0x10001);
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
if (is_lp)
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_or_config32(dev, 0x100, 1 << 29);
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
/* Read and write back write-once capability registers. */
PCI ops: Define read-modify-write routines globally Change-Id: I7d64f46bb4ec3229879a60159efc8a8408512acd Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com> Reviewed-on: https://review.coreboot.org/17690 Tested-by: build bot (Jenkins) Reviewed-by: Aaron Durbin <adurbin@chromium.org>
2013-07-26 07:53:59 +02:00
pci_update_config32(dev, 0x34, ~0, 0);
pci_update_config32(dev, 0x40, ~0, 0);
pci_update_config32(dev, 0x80, ~0, 0);
pci_update_config32(dev, 0x90, ~0, 0);
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
}
static void pci_init(struct device *dev)
{
printk(BIOS_DEBUG, "Initializing PCH PCIe bridge.\n");
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
/* Enable SERR */
sb/intel/lynxpoint: Fix 16-bit read/write PCI_COMMAND register Change-Id: I81b740e0cfcf0e1bf096427b45ffba06d357fee6 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/c/coreboot/+/40792 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Patrick Georgi <pgeorgi@google.com>
2020-04-28 10:13:05 +02:00
pci_or_config16(dev, PCI_COMMAND, PCI_COMMAND_SERR);
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
/* Enable Bus Master */
sb/intel/lynxpoint: Fix 16-bit read/write PCI_COMMAND register Change-Id: I81b740e0cfcf0e1bf096427b45ffba06d357fee6 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/c/coreboot/+/40792 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Patrick Georgi <pgeorgi@google.com>
2020-04-28 10:13:05 +02:00
pci_or_config16(dev, PCI_COMMAND, PCI_COMMAND_MASTER);
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
/* Set Cache Line Size to 0x10 */
// This has no effect but the OS might expect it
pci_write_config8(dev, 0x0c, 0x10);
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_and_config16(dev, PCI_BRIDGE_CONTROL, ~PCI_BRIDGE_CTL_PARITY);
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
/* Clear errors in status registers */
sb/intel/lynxpoint: Use PCI bitwise ops Some cases could not be factored out while keeping reproducibility. Also mark some potential bugs with a FIXME comment, since fixing them while also keeping the binary unchanged is pretty much impossible. Tested with BUILD_TIMELESS=1, Asrock B85M Pro4 does not change. Change-Id: I27d6aaa59e12a337f80a6d3387cc9c8ae5949384 Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/42154 Reviewed-by: Arthur Heymans <arthur@aheymans.xyz> Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org> Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
2020-06-08 00:12:43 +02:00
pci_update_config16(dev, 0x06, ~0, 0);
pci_update_config16(dev, 0x1e, ~0, 0);
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
}
sb/intel/lynxpoint: Get rid of device_t Use of device_t has been abandoned in ramstage. Change-Id: I064ff5e76dd95c1770cd24139195b2a5fff2d382 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/26974 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
2018-06-08 17:20:38 +02:00
static void pch_pcie_enable(struct device *dev)
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
{
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
/* Add this device to the root port config structure. */
root_port_init_config(dev);
/* Check to see if this Root Port should be disabled. */
root_port_check_disable(dev);
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
/* Power Management init before enumeration */
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
if (dev->enabled)
lynxpoint: implement additional programming steps The BIOS spec for LynxPoint calls out additional programming steps for the PCIe Root Ports. Implement those steps from the BIOS spec. These steps are completed before deeper PCIe probing. The "late" programming was removed as that was applicable to Cougar/Panther point where this code was originally copied, though there was some overlap. Change-Id: I64f25e4451e035d98ca6b66b0335bd280b70b074 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59558 Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4323 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-12-03 21:13:26 +01:00
pch_pcie_early(dev);
lynxpoint: disable pcie devices based on config PCIe Root Ports should be disabled based on pin ownership and the strapping configuration. Implement this logic for LynxPoint. The chip_ops->enable_dev() path is no longer used. Instead the PCIe driver handles the enabling and disabling of devices. This allows for having an empty or incomplete device tree since those "allocated" devices do not travel through the chip_ops->enable_dev() path. The coalescing was tested to be working properly, however not all configurations were tested. Change-Id: I1e8bfe5e447b72ff8a4b04b650982d8c1ae0823c Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: https://gerrit.chromium.org/gerrit/59424 Reviewed-by: Stefan Reinauer <reinauer@chromium.org> Reviewed-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/4322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-06-20 08:20:30 +02:00
/*
* When processing the last PCIe root port we can now
* update the Root Port Function Number and Hide register.
*/
if (root_port_is_last(dev))
root_port_commit_config();
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
}
static struct device_operations device_ops = {
.read_resources = pci_bus_read_resources,
.set_resources = pci_dev_set_resources,
.enable_resources = pci_bus_enable_resources,
.init = pci_init,
.enable = pch_pcie_enable,
.scan_bus = pciexp_scan_bridge,
src: Use pci_dev_ops_pci where applicable Change-Id: Ie004a94a49fc8f53c370412bee1c3e7eacbf8beb Signed-off-by: Angel Pons <th3fanbus@gmail.com> Reviewed-on: https://review.coreboot.org/c/coreboot/+/41944 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Michael Niewöhner Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net> Reviewed-by: Nico Huber <nico.h@gmx.de>
2020-05-31 00:03:28 +02:00
.ops_pci = &pci_dev_ops_pci,
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
};
lynxpoint: Update device IDs and clock gating setup - Add device IDs for lynxpoint mobile and LP variants. - Update the clock gating setup based on BWG - Update the SATA programming based on BWG - Add a DEVSLP0 mux config register Change-Id: Icf4d7bab7f3df7adef5eb7c5e310a6995227a0e5 Signed-off-by: Duncan Laurie <dlaurie@chromium.org> Reviewed-on: http://review.coreboot.org/2649 Tested-by: build bot (Jenkins) Reviewed-by: Marc Jones <marc.jones@se-eng.com>
2012-12-17 20:31:40 +01:00
static const unsigned short pci_device_ids[] = {
/* Lynxpoint Mobile */
0x8c10, 0x8c12, 0x8c14, 0x8c16, 0x8c18, 0x8c1a, 0x8c1c, 0x8c1e,
/* Lynxpoint Low Power */
0x9c10, 0x9c12, 0x9c14, 0x9c16, 0x9c18, 0x9c1a,
0
};
haswell: Add initial support for Haswell platforms The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore, the southbridge support is included as well. The basis for this code is the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires more attention, but this is a good starting point. This code partially gets up through the romstage just before training memory on a Haswell reference board. Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4 Signed-off-by: Aaron Durbin <adurbin@chromium.org> Reviewed-on: http://review.coreboot.org/2616 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2012-10-30 15:03:43 +01:00
static const struct pci_driver pch_pcie __pci_driver = {
.ops = &device_ops,
.vendor = PCI_VENDOR_ID_INTEL,
.devices = pci_device_ids,
};
Reference in a new issue Copy permalink