freebsd-skq/sys/powerpc/powermac/cpcht.c

738 lines
18 KiB
C
Raw Normal View History

/*-
* Copyright (C) 2008-2010 Nathan Whitehorn
* All rights reserved.
*
* 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 ``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 TOOLS GMBH 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.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/module.h>
#include <sys/bus.h>
#include <sys/conf.h>
#include <sys/kernel.h>
#include <sys/pciio.h>
#include <sys/rman.h>
#include <dev/ofw/openfirm.h>
#include <dev/ofw/ofw_pci.h>
#include <dev/pci/pcivar.h>
#include <dev/pci/pcireg.h>
#include <machine/bus.h>
#include <machine/intr_machdep.h>
#include <machine/md_var.h>
#include <machine/openpicreg.h>
#include <machine/openpicvar.h>
#include <machine/pio.h>
#include <machine/resource.h>
#include <dev/ofw/ofw_bus.h>
#include <dev/ofw/ofw_bus_subr.h>
#include <dev/ofw/ofwpci.h>
#include <vm/vm.h>
#include <vm/pmap.h>
#include "pcib_if.h"
#include <dev/pci/pcib_private.h>
#include "pic_if.h"
/*
* IBM CPC9X5 Hypertransport Device interface.
*/
static int cpcht_probe(device_t);
static int cpcht_attach(device_t);
static void cpcht_configure_htbridge(device_t, phandle_t);
/*
* pcib interface.
*/
static u_int32_t cpcht_read_config(device_t, u_int, u_int, u_int,
u_int, int);
static void cpcht_write_config(device_t, u_int, u_int, u_int,
u_int, u_int32_t, int);
static int cpcht_route_interrupt(device_t, device_t, int);
static int cpcht_alloc_msi(device_t dev, device_t child,
int count, int maxcount, int *irqs);
static int cpcht_release_msi(device_t dev, device_t child,
int count, int *irqs);
static int cpcht_alloc_msix(device_t dev, device_t child,
int *irq);
static int cpcht_release_msix(device_t dev, device_t child,
int irq);
static int cpcht_map_msi(device_t dev, device_t child,
int irq, uint64_t *addr, uint32_t *data);
/*
* Driver methods.
*/
static device_method_t cpcht_methods[] = {
/* Device interface */
DEVMETHOD(device_probe, cpcht_probe),
DEVMETHOD(device_attach, cpcht_attach),
/* pcib interface */
DEVMETHOD(pcib_read_config, cpcht_read_config),
DEVMETHOD(pcib_write_config, cpcht_write_config),
DEVMETHOD(pcib_route_interrupt, cpcht_route_interrupt),
DEVMETHOD(pcib_alloc_msi, cpcht_alloc_msi),
DEVMETHOD(pcib_release_msi, cpcht_release_msi),
DEVMETHOD(pcib_alloc_msix, cpcht_alloc_msix),
DEVMETHOD(pcib_release_msix, cpcht_release_msix),
DEVMETHOD(pcib_map_msi, cpcht_map_msi),
DEVMETHOD(pcib_request_feature, pcib_request_feature_allow),
DEVMETHOD_END
};
struct cpcht_irq {
enum {
IRQ_NONE, IRQ_HT, IRQ_MSI, IRQ_INTERNAL
} irq_type;
int ht_source;
vm_offset_t ht_base;
vm_offset_t apple_eoi;
uint32_t eoi_data;
int edge;
};
static struct cpcht_irq *cpcht_irqmap = NULL;
uint32_t cpcht_msipic = 0;
struct cpcht_softc {
struct ofw_pci_softc pci_sc;
vm_offset_t sc_data;
uint64_t sc_populated_slots;
struct cpcht_irq htirq_map[128];
struct mtx htirq_mtx;
};
static devclass_t cpcht_devclass;
DEFINE_CLASS_1(pcib, cpcht_driver, cpcht_methods, sizeof(struct cpcht_softc),
ofw_pci_driver);
DRIVER_MODULE(cpcht, ofwbus, cpcht_driver, cpcht_devclass, 0, 0);
#define CPCHT_IOPORT_BASE 0xf4000000UL /* Hardwired */
#define CPCHT_IOPORT_SIZE 0x00400000UL
#define HTAPIC_REQUEST_EOI 0x20
#define HTAPIC_TRIGGER_LEVEL 0x02
#define HTAPIC_MASK 0x01
static int
cpcht_probe(device_t dev)
{
const char *type, *compatible;
type = ofw_bus_get_type(dev);
compatible = ofw_bus_get_compat(dev);
if (type == NULL || compatible == NULL)
return (ENXIO);
if (strcmp(type, "ht") != 0)
return (ENXIO);
if (strcmp(compatible, "u3-ht") != 0)
return (ENXIO);
device_set_desc(dev, "IBM CPC9X5 HyperTransport Tunnel");
return (0);
}
static int
cpcht_attach(device_t dev)
{
struct cpcht_softc *sc;
phandle_t node, child;
u_int32_t reg[3];
int i;
node = ofw_bus_get_node(dev);
sc = device_get_softc(dev);
if (OF_getencprop(node, "reg", reg, sizeof(reg)) < 12)
return (ENXIO);
2012-02-04 23:25:18 +00:00
if (OF_getproplen(node, "ranges") <= 0)
sc->pci_sc.sc_quirks = OFW_PCI_QUIRK_RANGES_ON_CHILDREN;
sc->sc_populated_slots = 0;
sc->sc_data = (vm_offset_t)pmap_mapdev(reg[1], reg[2]);
/*
* Set up the resource manager and the HT->MPIC mapping. For cpcht,
* the ranges are properties of the child bridges, and this is also
* where we get the HT interrupts properties.
*/
#if 0
/* I/O port mappings are usually not in the device tree */
rman_manage_region(&sc->pci_sc.sc_io_rman, 0, CPCHT_IOPORT_SIZE - 1);
#endif
bzero(sc->htirq_map, sizeof(sc->htirq_map));
mtx_init(&sc->htirq_mtx, "cpcht irq", NULL, MTX_DEF);
for (i = 0; i < 8; i++)
sc->htirq_map[i].irq_type = IRQ_INTERNAL;
for (child = OF_child(node); child != 0; child = OF_peer(child))
cpcht_configure_htbridge(dev, child);
/* Now make the mapping table available to the MPIC */
cpcht_irqmap = sc->htirq_map;
return (ofw_pci_attach(dev));
}
static void
cpcht_configure_htbridge(device_t dev, phandle_t child)
{
struct cpcht_softc *sc;
struct ofw_pci_register pcir;
int ptr, nextptr;
uint32_t vend, val;
int i, nirq, irq;
2012-02-04 23:25:18 +00:00
u_int b, f, s;
sc = device_get_softc(dev);
if (OF_getencprop(child, "reg", (pcell_t *)&pcir, sizeof(pcir)) == -1)
return;
2012-02-04 23:25:18 +00:00
b = OFW_PCI_PHYS_HI_BUS(pcir.phys_hi);
s = OFW_PCI_PHYS_HI_DEVICE(pcir.phys_hi);
f = OFW_PCI_PHYS_HI_FUNCTION(pcir.phys_hi);
/*
* Mark this slot is populated. The remote south bridge does
* not like us talking to unpopulated slots on the root bus.
*/
sc->sc_populated_slots |= (1 << s);
/*
* Next build up any HT->MPIC mappings for this sub-bus. One would
* naively hope that enabling, disabling, and EOIing interrupts would
* cause the appropriate HT bus transactions to that effect. This is
* not the case.
*
* Instead, we have to muck about on the HT peer's root PCI bridges,
* figure out what interrupts they send, enable them, and cache
* the location of their WaitForEOI registers so that we can
* send EOIs later.
*/
/* All the devices we are interested in have caps */
2012-02-04 23:25:18 +00:00
if (!(PCIB_READ_CONFIG(dev, b, s, f, PCIR_STATUS, 2)
& PCIM_STATUS_CAPPRESENT))
return;
2012-02-04 23:25:18 +00:00
nextptr = PCIB_READ_CONFIG(dev, b, s, f, PCIR_CAP_PTR, 1);
while (nextptr != 0) {
ptr = nextptr;
2012-02-04 23:25:18 +00:00
nextptr = PCIB_READ_CONFIG(dev, b, s, f,
ptr + PCICAP_NEXTPTR, 1);
/* Find the HT IRQ capabilities */
2012-02-04 23:25:18 +00:00
if (PCIB_READ_CONFIG(dev, b, s, f,
ptr + PCICAP_ID, 1) != PCIY_HT)
continue;
2012-02-04 23:25:18 +00:00
val = PCIB_READ_CONFIG(dev, b, s, f, ptr + PCIR_HT_COMMAND, 2);
if ((val & PCIM_HTCMD_CAP_MASK) != PCIM_HTCAP_INTERRUPT)
continue;
/* Ask for the IRQ count */
2012-02-04 23:25:18 +00:00
PCIB_WRITE_CONFIG(dev, b, s, f, ptr + PCIR_HT_COMMAND, 0x1, 1);
nirq = PCIB_READ_CONFIG(dev, b, s, f, ptr + 4, 4);
nirq = ((nirq >> 16) & 0xff) + 1;
device_printf(dev, "%d HT IRQs on device %d.%d\n", nirq, s, f);
for (i = 0; i < nirq; i++) {
2012-02-04 23:25:18 +00:00
PCIB_WRITE_CONFIG(dev, b, s, f,
ptr + PCIR_HT_COMMAND, 0x10 + (i << 1), 1);
2012-02-04 23:25:18 +00:00
irq = PCIB_READ_CONFIG(dev, b, s, f, ptr + 4, 4);
/*
* Mask this interrupt for now.
*/
2012-02-04 23:25:18 +00:00
PCIB_WRITE_CONFIG(dev, b, s, f, ptr + 4,
irq | HTAPIC_MASK, 4);
irq = (irq >> 16) & 0xff;
sc->htirq_map[irq].irq_type = IRQ_HT;
sc->htirq_map[irq].ht_source = i;
sc->htirq_map[irq].ht_base = sc->sc_data +
(((((s & 0x1f) << 3) | (f & 0x07)) << 8) | (ptr));
2012-02-04 23:25:18 +00:00
PCIB_WRITE_CONFIG(dev, b, s, f,
ptr + PCIR_HT_COMMAND, 0x11 + (i << 1), 1);
sc->htirq_map[irq].eoi_data =
2012-02-04 23:25:18 +00:00
PCIB_READ_CONFIG(dev, b, s, f, ptr + 4, 4) |
0x80000000;
/*
* Apple uses a non-compliant IO/APIC that differs
* in how we signal EOIs. Check if this device was
* made by Apple, and act accordingly.
*/
2012-02-04 23:25:18 +00:00
vend = PCIB_READ_CONFIG(dev, b, s, f,
PCIR_DEVVENDOR, 4);
if ((vend & 0xffff) == 0x106b)
sc->htirq_map[irq].apple_eoi =
(sc->htirq_map[irq].ht_base - ptr) + 0x60;
}
}
}
static u_int32_t
cpcht_read_config(device_t dev, u_int bus, u_int slot, u_int func, u_int reg,
int width)
{
struct cpcht_softc *sc;
vm_offset_t caoff;
sc = device_get_softc(dev);
caoff = sc->sc_data +
(((((slot & 0x1f) << 3) | (func & 0x07)) << 8) | reg);
if (bus == 0 && (!(sc->sc_populated_slots & (1 << slot)) || func > 0))
return (0xffffffff);
if (bus > 0)
caoff += 0x01000000UL + (bus << 16);
switch (width) {
case 1:
return (in8rb(caoff));
break;
case 2:
return (in16rb(caoff));
break;
case 4:
return (in32rb(caoff));
break;
}
return (0xffffffff);
}
static void
cpcht_write_config(device_t dev, u_int bus, u_int slot, u_int func,
u_int reg, u_int32_t val, int width)
{
struct cpcht_softc *sc;
vm_offset_t caoff;
sc = device_get_softc(dev);
caoff = sc->sc_data +
(((((slot & 0x1f) << 3) | (func & 0x07)) << 8) | reg);
if (bus == 0 && (!(sc->sc_populated_slots & (1 << slot)) || func > 0))
return;
if (bus > 0)
caoff += 0x01000000UL + (bus << 16);
switch (width) {
case 1:
out8rb(caoff, val);
break;
case 2:
out16rb(caoff, val);
break;
case 4:
out32rb(caoff, val);
break;
}
}
static int
cpcht_route_interrupt(device_t bus, device_t dev, int pin)
{
return (pin);
}
static int
cpcht_alloc_msi(device_t dev, device_t child, int count, int maxcount,
int *irqs)
{
struct cpcht_softc *sc;
int i, j;
sc = device_get_softc(dev);
j = 0;
/* Bail if no MSI PIC yet */
if (cpcht_msipic == 0)
return (ENXIO);
mtx_lock(&sc->htirq_mtx);
for (i = 8; i < 124 - count; i++) {
for (j = 0; j < count; j++) {
if (sc->htirq_map[i+j].irq_type != IRQ_NONE)
break;
}
if (j == count)
break;
i += j; /* We know there isn't a large enough run */
}
if (j != count) {
mtx_unlock(&sc->htirq_mtx);
return (ENXIO);
}
for (j = 0; j < count; j++) {
irqs[j] = MAP_IRQ(cpcht_msipic, i+j);
sc->htirq_map[i+j].irq_type = IRQ_MSI;
}
mtx_unlock(&sc->htirq_mtx);
return (0);
}
static int
cpcht_release_msi(device_t dev, device_t child, int count, int *irqs)
{
struct cpcht_softc *sc;
int i;
sc = device_get_softc(dev);
mtx_lock(&sc->htirq_mtx);
for (i = 0; i < count; i++)
sc->htirq_map[irqs[i] & 0xff].irq_type = IRQ_NONE;
mtx_unlock(&sc->htirq_mtx);
return (0);
}
static int
cpcht_alloc_msix(device_t dev, device_t child, int *irq)
{
struct cpcht_softc *sc;
int i;
sc = device_get_softc(dev);
/* Bail if no MSI PIC yet */
if (cpcht_msipic == 0)
return (ENXIO);
mtx_lock(&sc->htirq_mtx);
for (i = 8; i < 124; i++) {
if (sc->htirq_map[i].irq_type == IRQ_NONE) {
sc->htirq_map[i].irq_type = IRQ_MSI;
*irq = MAP_IRQ(cpcht_msipic, i);
mtx_unlock(&sc->htirq_mtx);
return (0);
}
}
mtx_unlock(&sc->htirq_mtx);
return (ENXIO);
}
static int
cpcht_release_msix(device_t dev, device_t child, int irq)
{
struct cpcht_softc *sc;
sc = device_get_softc(dev);
mtx_lock(&sc->htirq_mtx);
sc->htirq_map[irq & 0xff].irq_type = IRQ_NONE;
mtx_unlock(&sc->htirq_mtx);
return (0);
}
static int
cpcht_map_msi(device_t dev, device_t child, int irq, uint64_t *addr,
uint32_t *data)
{
device_t pcib;
struct pci_devinfo *dinfo;
struct pcicfg_ht *ht = NULL;
for (pcib = child; pcib != dev; pcib =
device_get_parent(device_get_parent(pcib))) {
dinfo = device_get_ivars(pcib);
ht = &dinfo->cfg.ht;
if (ht == NULL)
continue;
}
if (ht == NULL)
return (ENXIO);
*addr = ht->ht_msiaddr;
*data = irq & 0xff;
return (0);
}
/*
* Driver for the integrated MPIC on U3/U4 (CPC925/CPC945)
*/
static int openpic_cpcht_probe(device_t);
static int openpic_cpcht_attach(device_t);
static void openpic_cpcht_config(device_t, u_int irq,
enum intr_trigger trig, enum intr_polarity pol);
static void openpic_cpcht_enable(device_t, u_int irq, u_int vector);
static void openpic_cpcht_unmask(device_t, u_int irq);
static void openpic_cpcht_eoi(device_t, u_int irq);
static device_method_t openpic_cpcht_methods[] = {
/* Device interface */
DEVMETHOD(device_probe, openpic_cpcht_probe),
DEVMETHOD(device_attach, openpic_cpcht_attach),
/* PIC interface */
DEVMETHOD(pic_bind, openpic_bind),
DEVMETHOD(pic_config, openpic_cpcht_config),
DEVMETHOD(pic_dispatch, openpic_dispatch),
DEVMETHOD(pic_enable, openpic_cpcht_enable),
DEVMETHOD(pic_eoi, openpic_cpcht_eoi),
DEVMETHOD(pic_ipi, openpic_ipi),
DEVMETHOD(pic_mask, openpic_mask),
DEVMETHOD(pic_unmask, openpic_cpcht_unmask),
{ 0, 0 },
};
struct openpic_cpcht_softc {
struct openpic_softc sc_openpic;
struct mtx sc_ht_mtx;
};
static driver_t openpic_cpcht_driver = {
"htpic",
openpic_cpcht_methods,
2010-05-19 01:37:47 +00:00
sizeof(struct openpic_cpcht_softc),
};
DRIVER_MODULE(openpic, unin, openpic_cpcht_driver, openpic_devclass, 0, 0);
static int
openpic_cpcht_probe(device_t dev)
{
const char *type = ofw_bus_get_type(dev);
if (strcmp(type, "open-pic") != 0)
return (ENXIO);
device_set_desc(dev, OPENPIC_DEVSTR);
return (0);
}
static int
openpic_cpcht_attach(device_t dev)
{
struct openpic_cpcht_softc *sc;
Fix the interrupt code, broken 7 months ago. The interrupt framework already supported nested PICs, but was limited to having a nested AT-PIC only. With G5 support the need for nested OpenPIC controllers needed to be added. This was done the wrong way and broke the MPC8555 eval system in the process. OFW, as well as FDT, describe the interrupt routing in terms of a controller and an interrupt pin on it. This needs to be mapped to a flat and global resource: the IRQ. The IRQ is the same as the PCI intline and as such needs to be representable in 8 bits. Secondly, ISA support pretty much dictates that IRQ 0-15 should be reserved for ISA interrupts, because of the internal workins of south bridges. Both were broken. This change reverts revision 209298 for a big part and re-implements it simpler. In particular: o The id() method of the PIC I/F is removed again. It's not needed. o The openpic_attach() function has been changed to take the OFW or FDT phandle of the controller as a second argument. All bus attachments that previously used openpic_attach() as the attach method of the device I/F now implement as bus-specific method and pass the phandle_t to the renamed openpic_attach(). o Change powerpc_register_pic() to take a few more arguments. In particular: - Pass the number of IPIs specificly. The number of IRQs carved out for a PIC is the sum of the number of int. pins and IPIs. - Pass a flag indicating whether the PIC is an AT-PIC or not. This tells the interrupt framework whether to assign IRQ 0-15 or some other range. o Until we implement proper multi-pass bus enumeration, we have to handle the case where we need to map from PIC+pin to IRQ *before* the PIC gets registered. This is done in a similar way as before, but rather than carving out 256 IRQs per PIC, we carve out 128 IRQs (124 pins + 4 IPIs). This is supposed to handle the G5 case, but should really be fixed properly using multiple passes. o Have the interrupt framework set root_pic in most cases and not put that burden in PIC drivers (for the most part). o Remove powerpc_ign_lookup() and replace it with powerpc_get_irq(). Remove IGN_SHIFT, INTR_INTLINE and INTR_IGN. Related to the above, fix the Freescale PCI controller driver, broken by the FDT code. Besides not attaching properly, bus numbers were assigned improperly and enumeration was broken in general. This prevented the AT PIC from being discovered and interrupt routing to work properly. Consequently, the ata(4) controller stopped functioning. Fix the driver, and FDT PCI support, enough to get the MPC8555CDS going again. The FDT PCI code needs a whole lot more work. No breakages are expected, but lackiong G5 hardware, it's possible that there are unpleasant side-effects. At least MPC85xx support is back to where it was 7 months ago -- it's amazing how badly support can be broken in just 7 months... Sponsored by: Juniper Networks
2011-01-29 20:58:38 +00:00
phandle_t node;
int err, irq;
Fix the interrupt code, broken 7 months ago. The interrupt framework already supported nested PICs, but was limited to having a nested AT-PIC only. With G5 support the need for nested OpenPIC controllers needed to be added. This was done the wrong way and broke the MPC8555 eval system in the process. OFW, as well as FDT, describe the interrupt routing in terms of a controller and an interrupt pin on it. This needs to be mapped to a flat and global resource: the IRQ. The IRQ is the same as the PCI intline and as such needs to be representable in 8 bits. Secondly, ISA support pretty much dictates that IRQ 0-15 should be reserved for ISA interrupts, because of the internal workins of south bridges. Both were broken. This change reverts revision 209298 for a big part and re-implements it simpler. In particular: o The id() method of the PIC I/F is removed again. It's not needed. o The openpic_attach() function has been changed to take the OFW or FDT phandle of the controller as a second argument. All bus attachments that previously used openpic_attach() as the attach method of the device I/F now implement as bus-specific method and pass the phandle_t to the renamed openpic_attach(). o Change powerpc_register_pic() to take a few more arguments. In particular: - Pass the number of IPIs specificly. The number of IRQs carved out for a PIC is the sum of the number of int. pins and IPIs. - Pass a flag indicating whether the PIC is an AT-PIC or not. This tells the interrupt framework whether to assign IRQ 0-15 or some other range. o Until we implement proper multi-pass bus enumeration, we have to handle the case where we need to map from PIC+pin to IRQ *before* the PIC gets registered. This is done in a similar way as before, but rather than carving out 256 IRQs per PIC, we carve out 128 IRQs (124 pins + 4 IPIs). This is supposed to handle the G5 case, but should really be fixed properly using multiple passes. o Have the interrupt framework set root_pic in most cases and not put that burden in PIC drivers (for the most part). o Remove powerpc_ign_lookup() and replace it with powerpc_get_irq(). Remove IGN_SHIFT, INTR_INTLINE and INTR_IGN. Related to the above, fix the Freescale PCI controller driver, broken by the FDT code. Besides not attaching properly, bus numbers were assigned improperly and enumeration was broken in general. This prevented the AT PIC from being discovered and interrupt routing to work properly. Consequently, the ata(4) controller stopped functioning. Fix the driver, and FDT PCI support, enough to get the MPC8555CDS going again. The FDT PCI code needs a whole lot more work. No breakages are expected, but lackiong G5 hardware, it's possible that there are unpleasant side-effects. At least MPC85xx support is back to where it was 7 months ago -- it's amazing how badly support can be broken in just 7 months... Sponsored by: Juniper Networks
2011-01-29 20:58:38 +00:00
node = ofw_bus_get_node(dev);
err = openpic_common_attach(dev, node);
if (err != 0)
return (err);
/*
* The HT APIC stuff is not thread-safe, so we need a mutex to
* protect it.
*/
sc = device_get_softc(dev);
mtx_init(&sc->sc_ht_mtx, "htpic", NULL, MTX_SPIN);
/*
* Interrupts 0-3 are internally sourced and are level triggered
* active low. Interrupts 4-123 are connected to a pulse generator
* and should be programmed as edge triggered low-to-high.
*
* IBM CPC945 Manual, Section 9.3.
*/
for (irq = 0; irq < 4; irq++)
openpic_config(dev, irq, INTR_TRIGGER_LEVEL, INTR_POLARITY_LOW);
for (irq = 4; irq < 124; irq++)
openpic_config(dev, irq, INTR_TRIGGER_EDGE, INTR_POLARITY_LOW);
/*
* Use this PIC for MSI only if it is the root PIC. This may not
* be necessary, but Linux does it, and I cannot find any U3 machines
* with MSI devices to test.
*/
if (dev == root_pic)
Fix the interrupt code, broken 7 months ago. The interrupt framework already supported nested PICs, but was limited to having a nested AT-PIC only. With G5 support the need for nested OpenPIC controllers needed to be added. This was done the wrong way and broke the MPC8555 eval system in the process. OFW, as well as FDT, describe the interrupt routing in terms of a controller and an interrupt pin on it. This needs to be mapped to a flat and global resource: the IRQ. The IRQ is the same as the PCI intline and as such needs to be representable in 8 bits. Secondly, ISA support pretty much dictates that IRQ 0-15 should be reserved for ISA interrupts, because of the internal workins of south bridges. Both were broken. This change reverts revision 209298 for a big part and re-implements it simpler. In particular: o The id() method of the PIC I/F is removed again. It's not needed. o The openpic_attach() function has been changed to take the OFW or FDT phandle of the controller as a second argument. All bus attachments that previously used openpic_attach() as the attach method of the device I/F now implement as bus-specific method and pass the phandle_t to the renamed openpic_attach(). o Change powerpc_register_pic() to take a few more arguments. In particular: - Pass the number of IPIs specificly. The number of IRQs carved out for a PIC is the sum of the number of int. pins and IPIs. - Pass a flag indicating whether the PIC is an AT-PIC or not. This tells the interrupt framework whether to assign IRQ 0-15 or some other range. o Until we implement proper multi-pass bus enumeration, we have to handle the case where we need to map from PIC+pin to IRQ *before* the PIC gets registered. This is done in a similar way as before, but rather than carving out 256 IRQs per PIC, we carve out 128 IRQs (124 pins + 4 IPIs). This is supposed to handle the G5 case, but should really be fixed properly using multiple passes. o Have the interrupt framework set root_pic in most cases and not put that burden in PIC drivers (for the most part). o Remove powerpc_ign_lookup() and replace it with powerpc_get_irq(). Remove IGN_SHIFT, INTR_INTLINE and INTR_IGN. Related to the above, fix the Freescale PCI controller driver, broken by the FDT code. Besides not attaching properly, bus numbers were assigned improperly and enumeration was broken in general. This prevented the AT PIC from being discovered and interrupt routing to work properly. Consequently, the ata(4) controller stopped functioning. Fix the driver, and FDT PCI support, enough to get the MPC8555CDS going again. The FDT PCI code needs a whole lot more work. No breakages are expected, but lackiong G5 hardware, it's possible that there are unpleasant side-effects. At least MPC85xx support is back to where it was 7 months ago -- it's amazing how badly support can be broken in just 7 months... Sponsored by: Juniper Networks
2011-01-29 20:58:38 +00:00
cpcht_msipic = node;
return (0);
}
static void
openpic_cpcht_config(device_t dev, u_int irq, enum intr_trigger trig,
enum intr_polarity pol)
{
struct openpic_cpcht_softc *sc;
uint32_t ht_irq;
/*
* The interrupt settings for the MPIC are completely determined
* by the internal wiring in the northbridge. Real changes to these
* settings need to be negotiated with the remote IO-APIC on the HT
* link.
*/
sc = device_get_softc(dev);
if (cpcht_irqmap != NULL && irq < 128 &&
cpcht_irqmap[irq].ht_base > 0 && !cpcht_irqmap[irq].edge) {
mtx_lock_spin(&sc->sc_ht_mtx);
/* Program the data port */
out8rb(cpcht_irqmap[irq].ht_base + PCIR_HT_COMMAND,
0x10 + (cpcht_irqmap[irq].ht_source << 1));
/* Grab the IRQ config register */
ht_irq = in32rb(cpcht_irqmap[irq].ht_base + 4);
/* Mask the IRQ while we fiddle settings */
out32rb(cpcht_irqmap[irq].ht_base + 4, ht_irq | HTAPIC_MASK);
/* Program the interrupt sense */
ht_irq &= ~(HTAPIC_TRIGGER_LEVEL | HTAPIC_REQUEST_EOI);
if (trig == INTR_TRIGGER_EDGE) {
cpcht_irqmap[irq].edge = 1;
} else {
cpcht_irqmap[irq].edge = 0;
ht_irq |= HTAPIC_TRIGGER_LEVEL | HTAPIC_REQUEST_EOI;
}
out32rb(cpcht_irqmap[irq].ht_base + 4, ht_irq);
mtx_unlock_spin(&sc->sc_ht_mtx);
}
}
static void
openpic_cpcht_enable(device_t dev, u_int irq, u_int vec)
{
struct openpic_cpcht_softc *sc;
uint32_t ht_irq;
openpic_enable(dev, irq, vec);
sc = device_get_softc(dev);
if (cpcht_irqmap != NULL && irq < 128 &&
cpcht_irqmap[irq].ht_base > 0) {
mtx_lock_spin(&sc->sc_ht_mtx);
/* Program the data port */
out8rb(cpcht_irqmap[irq].ht_base + PCIR_HT_COMMAND,
0x10 + (cpcht_irqmap[irq].ht_source << 1));
/* Unmask the interrupt */
ht_irq = in32rb(cpcht_irqmap[irq].ht_base + 4);
ht_irq &= ~HTAPIC_MASK;
out32rb(cpcht_irqmap[irq].ht_base + 4, ht_irq);
mtx_unlock_spin(&sc->sc_ht_mtx);
}
openpic_cpcht_eoi(dev, irq);
}
static void
openpic_cpcht_unmask(device_t dev, u_int irq)
{
struct openpic_cpcht_softc *sc;
uint32_t ht_irq;
openpic_unmask(dev, irq);
sc = device_get_softc(dev);
if (cpcht_irqmap != NULL && irq < 128 &&
cpcht_irqmap[irq].ht_base > 0) {
mtx_lock_spin(&sc->sc_ht_mtx);
/* Program the data port */
out8rb(cpcht_irqmap[irq].ht_base + PCIR_HT_COMMAND,
0x10 + (cpcht_irqmap[irq].ht_source << 1));
/* Unmask the interrupt */
ht_irq = in32rb(cpcht_irqmap[irq].ht_base + 4);
ht_irq &= ~HTAPIC_MASK;
out32rb(cpcht_irqmap[irq].ht_base + 4, ht_irq);
mtx_unlock_spin(&sc->sc_ht_mtx);
}
openpic_cpcht_eoi(dev, irq);
}
static void
openpic_cpcht_eoi(device_t dev, u_int irq)
{
struct openpic_cpcht_softc *sc;
uint32_t off, mask;
if (irq == 255)
return;
sc = device_get_softc(dev);
if (cpcht_irqmap != NULL && irq < 128 &&
cpcht_irqmap[irq].ht_base > 0 && !cpcht_irqmap[irq].edge) {
/* If this is an HT IRQ, acknowledge it at the remote APIC */
if (cpcht_irqmap[irq].apple_eoi) {
off = (cpcht_irqmap[irq].ht_source >> 3) & ~3;
mask = 1 << (cpcht_irqmap[irq].ht_source & 0x1f);
out32rb(cpcht_irqmap[irq].apple_eoi + off, mask);
} else {
mtx_lock_spin(&sc->sc_ht_mtx);
out8rb(cpcht_irqmap[irq].ht_base + PCIR_HT_COMMAND,
0x11 + (cpcht_irqmap[irq].ht_source << 1));
out32rb(cpcht_irqmap[irq].ht_base + 4,
cpcht_irqmap[irq].eoi_data);
mtx_unlock_spin(&sc->sc_ht_mtx);
}
}
openpic_eoi(dev, irq);
}