freebsd-dev/sys/dev/pci/pci_host_generic_acpi.c

489 lines
13 KiB
C
Raw Normal View History

/*-
* Copyright (C) 2018 Cavium Inc.
* Copyright (c) 2015 Ruslan Bukin <br@bsdpad.com>
* Copyright (c) 2014 The FreeBSD Foundation
* All rights reserved.
*
* This software was developed by Semihalf under
* the sponsorship of the FreeBSD Foundation.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
/* Generic ECAM PCIe driver */
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_platform.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/malloc.h>
#include <sys/kernel.h>
#include <sys/rman.h>
#include <sys/module.h>
#include <sys/bus.h>
#include <sys/endian.h>
#include <sys/cpuset.h>
#include <sys/rwlock.h>
#include <contrib/dev/acpica/include/acpi.h>
#include <contrib/dev/acpica/include/accommon.h>
#include <dev/acpica/acpivar.h>
#include <dev/acpica/acpi_pcibvar.h>
#include <dev/pci/pcivar.h>
#include <dev/pci/pcireg.h>
#include <dev/pci/pcib_private.h>
#include <dev/pci/pci_host_generic.h>
#include <dev/pci/pci_host_generic_acpi.h>
#include <machine/cpu.h>
#include <machine/bus.h>
#include <machine/intr.h>
#include "pcib_if.h"
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
#include "acpi_bus_if.h"
/* Assembling ECAM Configuration Address */
#define PCIE_BUS_SHIFT 20
#define PCIE_SLOT_SHIFT 15
#define PCIE_FUNC_SHIFT 12
#define PCIE_BUS_MASK 0xFF
#define PCIE_SLOT_MASK 0x1F
#define PCIE_FUNC_MASK 0x07
#define PCIE_REG_MASK 0xFFF
#define PCIE_ADDR_OFFSET(bus, slot, func, reg) \
((((bus) & PCIE_BUS_MASK) << PCIE_BUS_SHIFT) | \
(((slot) & PCIE_SLOT_MASK) << PCIE_SLOT_SHIFT) | \
(((func) & PCIE_FUNC_MASK) << PCIE_FUNC_SHIFT) | \
((reg) & PCIE_REG_MASK))
#define PCI_IO_WINDOW_OFFSET 0x1000
#define SPACE_CODE_SHIFT 24
#define SPACE_CODE_MASK 0x3
#define SPACE_CODE_IO_SPACE 0x1
#define PROPS_CELL_SIZE 1
#define PCI_ADDR_CELL_SIZE 2
/* Forward prototypes */
static int generic_pcie_acpi_probe(device_t dev);
static ACPI_STATUS pci_host_generic_acpi_parse_resource(ACPI_RESOURCE *, void *);
static int generic_pcie_acpi_read_ivar(device_t, device_t, int, uintptr_t *);
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
/*
* generic_pcie_acpi_probe - look for root bridge flag
*/
static int
generic_pcie_acpi_probe(device_t dev)
{
ACPI_DEVICE_INFO *devinfo;
ACPI_HANDLE h;
int root;
if (acpi_disabled("pcib") || (h = acpi_get_handle(dev)) == NULL ||
ACPI_FAILURE(AcpiGetObjectInfo(h, &devinfo)))
return (ENXIO);
root = (devinfo->Flags & ACPI_PCI_ROOT_BRIDGE) != 0;
AcpiOsFree(devinfo);
if (!root)
return (ENXIO);
device_set_desc(dev, "Generic PCI host controller");
return (BUS_PROBE_GENERIC);
}
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
/*
* pci_host_generic_acpi_parse_resource - parse PCI memory, IO and bus spaces
* 'produced' by this bridge
*/
static ACPI_STATUS
pci_host_generic_acpi_parse_resource(ACPI_RESOURCE *res, void *arg)
{
device_t dev = (device_t)arg;
struct generic_pcie_acpi_softc *sc;
struct rman *rm;
rman_res_t min, max, off;
int r;
rm = NULL;
sc = device_get_softc(dev);
r = sc->base.nranges;
switch (res->Type) {
case ACPI_RESOURCE_TYPE_ADDRESS16:
min = res->Data.Address16.Address.Minimum;
max = res->Data.Address16.Address.Maximum;
break;
case ACPI_RESOURCE_TYPE_ADDRESS32:
min = res->Data.Address32.Address.Minimum;
max = res->Data.Address32.Address.Maximum;
off = res->Data.Address32.Address.TranslationOffset;
break;
case ACPI_RESOURCE_TYPE_ADDRESS64:
min = res->Data.Address64.Address.Minimum;
max = res->Data.Address64.Address.Maximum;
off = res->Data.Address64.Address.TranslationOffset;
break;
default:
return (AE_OK);
}
/* Save detected ranges */
if (res->Data.Address.ResourceType == ACPI_MEMORY_RANGE ||
res->Data.Address.ResourceType == ACPI_IO_RANGE) {
sc->base.ranges[r].pci_base = min;
sc->base.ranges[r].phys_base = min + off;
sc->base.ranges[r].size = max - min + 1;
if (res->Data.Address.ResourceType == ACPI_MEMORY_RANGE)
sc->base.ranges[r].flags |= FLAG_MEM;
else if (res->Data.Address.ResourceType == ACPI_IO_RANGE)
sc->base.ranges[r].flags |= FLAG_IO;
sc->base.nranges++;
} else if (res->Data.Address.ResourceType == ACPI_BUS_NUMBER_RANGE) {
sc->base.bus_start = min;
sc->base.bus_end = max;
}
return (AE_OK);
}
static int
pci_host_acpi_get_ecam_resource(device_t dev)
{
struct generic_pcie_acpi_softc *sc;
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
struct acpi_device *ad;
struct resource_list *rl;
ACPI_TABLE_HEADER *hdr;
ACPI_MCFG_ALLOCATION *mcfg_entry, *mcfg_end;
ACPI_HANDLE handle;
ACPI_STATUS status;
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
rman_res_t base, start, end;
int found, val;
sc = device_get_softc(dev);
handle = acpi_get_handle(dev);
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
/* Try MCFG first */
status = AcpiGetTable(ACPI_SIG_MCFG, 1, &hdr);
if (ACPI_SUCCESS(status)) {
found = FALSE;
mcfg_end = (ACPI_MCFG_ALLOCATION *)((char *)hdr + hdr->Length);
mcfg_entry = (ACPI_MCFG_ALLOCATION *)((ACPI_TABLE_MCFG *)hdr + 1);
while (mcfg_entry < mcfg_end && !found) {
if (mcfg_entry->PciSegment == sc->base.ecam &&
mcfg_entry->StartBusNumber <= sc->base.bus_start &&
mcfg_entry->EndBusNumber >= sc->base.bus_start)
found = TRUE;
else
mcfg_entry++;
}
if (found) {
sc->base.bus_end = mcfg_entry->EndBusNumber;
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
base = mcfg_entry->Address;
} else {
device_printf(dev, "MCFG exists, but does not have bus %d-%d\n",
sc->base.bus_start, sc->base.bus_end);
return (ENXIO);
}
} else {
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
status = acpi_GetInteger(handle, "_CBA", &val);
if (ACPI_SUCCESS(status)) {
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
base = val;
sc->base.bus_end = 255;
} else
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
return (ENXIO);
}
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
/* add as MEM rid 0 */
ad = device_get_ivars(dev);
rl = &ad->ad_rl;
start = base + (sc->base.bus_start << PCIE_BUS_SHIFT);
end = base + ((sc->base.bus_end + 1) << PCIE_BUS_SHIFT) - 1;
resource_list_add(rl, SYS_RES_MEMORY, 0, start, end, end - start + 1);
if (bootverbose)
device_printf(dev, "ECAM for bus %d-%d at mem %jx-%jx\n",
sc->base.bus_start, sc->base.bus_end, start, end);
return (0);
}
int
pci_host_generic_acpi_init(device_t dev)
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
{
struct generic_pcie_acpi_softc *sc;
ACPI_HANDLE handle;
uint64_t phys_base;
uint64_t pci_base;
uint64_t size;
ACPI_STATUS status;
int error;
int tuple;
sc = device_get_softc(dev);
handle = acpi_get_handle(dev);
/* Get Start bus number for the PCI host bus is from _BBN method */
status = acpi_GetInteger(handle, "_BBN", &sc->base.bus_start);
if (ACPI_FAILURE(status)) {
device_printf(dev, "No _BBN, using start bus 0\n");
sc->base.bus_start = 0;
}
/* Get PCI Segment (domain) needed for MCFG lookup */
status = acpi_GetInteger(handle, "_SEG", &sc->base.ecam);
if (ACPI_FAILURE(status)) {
device_printf(dev, "No _SEG for PCI Bus, using segment 0\n");
sc->base.ecam = 0;
}
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
/* Bus decode ranges */
status = AcpiWalkResources(handle, "_CRS",
pci_host_generic_acpi_parse_resource, (void *)dev);
if (ACPI_FAILURE(status))
return (ENXIO);
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
/* Coherency attribute */
if (ACPI_FAILURE(acpi_GetInteger(handle, "_CCA", &sc->base.coherent)))
sc->base.coherent = 0;
if (bootverbose)
device_printf(dev, "Bus is%s cache-coherent\n",
sc->base.coherent ? "" : " not");
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
/* add config space resource */
pci_host_acpi_get_ecam_resource(dev);
acpi_pcib_fetch_prt(dev, &sc->ap_prt);
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
error = pci_host_generic_core_attach(dev);
if (error != 0)
return (error);
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
for (tuple = 0; tuple < MAX_RANGES_TUPLES; tuple++) {
phys_base = sc->base.ranges[tuple].phys_base;
pci_base = sc->base.ranges[tuple].pci_base;
size = sc->base.ranges[tuple].size;
if (phys_base == 0 || size == 0)
continue; /* empty range element */
if (sc->base.ranges[tuple].flags & FLAG_MEM) {
error = rman_manage_region(&sc->base.mem_rman,
pci_base, pci_base + size - 1);
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
} else if (sc->base.ranges[tuple].flags & FLAG_IO) {
error = rman_manage_region(&sc->base.io_rman,
pci_base + PCI_IO_WINDOW_OFFSET,
pci_base + PCI_IO_WINDOW_OFFSET + size - 1);
} else
continue;
if (error) {
device_printf(dev, "rman_manage_region() failed."
"error = %d\n", error);
rman_fini(&sc->base.mem_rman);
return (error);
}
}
return (0);
}
static int
pci_host_generic_acpi_attach(device_t dev)
{
int error;
error = pci_host_generic_acpi_init(dev);
if (error != 0)
return (error);
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
device_add_child(dev, "pci", -1);
return (bus_generic_attach(dev));
}
static int
generic_pcie_acpi_read_ivar(device_t dev, device_t child, int index,
uintptr_t *result)
{
struct generic_pcie_acpi_softc *sc;
sc = device_get_softc(dev);
if (index == PCIB_IVAR_BUS) {
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
*result = sc->base.bus_start;
return (0);
}
if (index == PCIB_IVAR_DOMAIN) {
*result = sc->base.ecam;
return (0);
}
if (bootverbose)
device_printf(dev, "ERROR: Unknown index %d.\n", index);
return (ENOENT);
}
static int
generic_pcie_acpi_route_interrupt(device_t bus, device_t dev, int pin)
{
struct generic_pcie_acpi_softc *sc;
sc = device_get_softc(bus);
return (acpi_pcib_route_interrupt(bus, dev, pin, &sc->ap_prt));
}
static u_int
generic_pcie_get_xref(device_t pci, device_t child)
{
struct generic_pcie_acpi_softc *sc;
uintptr_t rid;
u_int xref, devid;
int err;
sc = device_get_softc(pci);
err = pcib_get_id(pci, child, PCI_ID_RID, &rid);
if (err != 0)
return (ACPI_MSI_XREF);
err = acpi_iort_map_pci_msi(sc->base.ecam, rid, &xref, &devid);
if (err != 0)
return (ACPI_MSI_XREF);
return (xref);
}
static u_int
generic_pcie_map_id(device_t pci, device_t child, uintptr_t *id)
{
struct generic_pcie_acpi_softc *sc;
uintptr_t rid;
u_int xref, devid;
int err;
sc = device_get_softc(pci);
err = pcib_get_id(pci, child, PCI_ID_RID, &rid);
if (err != 0)
return (err);
err = acpi_iort_map_pci_msi(sc->base.ecam, rid, &xref, &devid);
if (err == 0)
*id = devid;
else
*id = rid; /* RID not in IORT, likely FW bug, ignore */
return (0);
}
static int
generic_pcie_acpi_alloc_msi(device_t pci, device_t child, int count,
int maxcount, int *irqs)
{
#if defined(INTRNG)
return (intr_alloc_msi(pci, child, generic_pcie_get_xref(pci, child),
count, maxcount, irqs));
#else
return (ENXIO);
#endif
}
static int
generic_pcie_acpi_release_msi(device_t pci, device_t child, int count,
int *irqs)
{
#if defined(INTRNG)
return (intr_release_msi(pci, child, generic_pcie_get_xref(pci, child),
count, irqs));
#else
return (ENXIO);
#endif
}
static int
generic_pcie_acpi_map_msi(device_t pci, device_t child, int irq, uint64_t *addr,
uint32_t *data)
{
#if defined(INTRNG)
return (intr_map_msi(pci, child, generic_pcie_get_xref(pci, child), irq,
addr, data));
#else
return (ENXIO);
#endif
}
static int
generic_pcie_acpi_alloc_msix(device_t pci, device_t child, int *irq)
{
#if defined(INTRNG)
return (intr_alloc_msix(pci, child, generic_pcie_get_xref(pci, child),
irq));
#else
return (ENXIO);
#endif
}
static int
generic_pcie_acpi_release_msix(device_t pci, device_t child, int irq)
{
#if defined(INTRNG)
return (intr_release_msix(pci, child, generic_pcie_get_xref(pci, child),
irq));
#else
return (ENXIO);
#endif
}
static int
generic_pcie_acpi_get_id(device_t pci, device_t child, enum pci_id_type type,
uintptr_t *id)
{
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
if (type == PCI_ID_MSI)
return (generic_pcie_map_id(pci, child, id));
acpica, pci_host_generic_acpi: redo pci_host_generic_acpi.c This is a major update for pci_host_generic_acpi.c, the current implementation has some gaps that are better fixed up in one go. The changes are to: * Follow x86 method of not adding PCI resources to PCI host bridge in ACPI code. This has been moved to pci_host_generic_acpi.c, where we walk thru its resources of the host bridge and add them. * Fixup code in pci_host_generic_acpi.c to read all decoded ranges and update the 'ranges' property. This allows us to share most of the code with generic implementation (and the FDT one). * Parse and setup IO ranges and bus ranges when walking the resources above. Drop most of the changes related to this from acpica code. * Add the ECAM memory area as mem resource 0. Implement the logic to get the ECAM area from MCFG (using bus range which we now decode), or from _CBA (using _BBN/bus range). Drop aarch64 ifdefs from acpica code which did part of this. * Switch resource activation to similar code as FDT implementation, this can be moved into generic implementation in a later pass. * Drop the mechanism of using the 7th bit of bus number as the domain, this is not correct and will work only in very specific cases. Use _SEG as PCI domain and use the bus ranges of the host bridge to provide start bus number. This commit should not make any functional change to dev/acpica/acpi.c for other architectures, almost all the changes there are to revert earlier additions in this file done for aarch64. Reviewed by: andrew Differential Revision: https://reviews.freebsd.org/D17791
2018-11-19 03:16:16 +00:00
else
return (pcib_get_id(pci, child, type, id));
}
static device_method_t generic_pcie_acpi_methods[] = {
DEVMETHOD(device_probe, generic_pcie_acpi_probe),
DEVMETHOD(device_attach, pci_host_generic_acpi_attach),
DEVMETHOD(bus_read_ivar, generic_pcie_acpi_read_ivar),
/* pcib interface */
DEVMETHOD(pcib_route_interrupt, generic_pcie_acpi_route_interrupt),
DEVMETHOD(pcib_alloc_msi, generic_pcie_acpi_alloc_msi),
DEVMETHOD(pcib_release_msi, generic_pcie_acpi_release_msi),
DEVMETHOD(pcib_alloc_msix, generic_pcie_acpi_alloc_msix),
DEVMETHOD(pcib_release_msix, generic_pcie_acpi_release_msix),
DEVMETHOD(pcib_map_msi, generic_pcie_acpi_map_msi),
DEVMETHOD(pcib_get_id, generic_pcie_acpi_get_id),
DEVMETHOD_END
};
DEFINE_CLASS_1(pcib, generic_pcie_acpi_driver, generic_pcie_acpi_methods,
sizeof(struct generic_pcie_acpi_softc), generic_pcie_core_driver);
static devclass_t generic_pcie_acpi_devclass;
DRIVER_MODULE(pcib, acpi, generic_pcie_acpi_driver, generic_pcie_acpi_devclass,
0, 0);