freebsd-dev/sys/x86/iommu/intel_intrmap.c

386 lines
10 KiB
C
Raw Normal View History

Use VT-d interrupt remapping block (IR) to perform FSB messages translation. In particular, despite IO-APICs only take 8bit apic id, IR translation structures accept 32bit APIC Id, which allows x2APIC mode to function properly. Extend msi_cpu of struct msi_intrsrc and io_cpu of ioapic_intsrc to full int from one byte. KPI of IR is isolated into the x86/iommu/iommu_intrmap.h, to avoid bringing all dmar headers into interrupt code. The non-PCI(e) devices which generate message interrupts on FSB require special handling. The HPET FSB interrupts are remapped, while DMAR interrupts are not. For each msi and ioapic interrupt source, the iommu cookie is added, which is in fact index of the IRE (interrupt remap entry) in the IR table. Cookie is made at the source allocation time, and then used at the map time to fill both IRE and device registers. The MSI address/data registers and IO-APIC redirection registers are programmed with the special values which are recognized by IR and used to restore the IRE index, to find proper delivery mode and target. Map all MSI interrupts in the block when msi_map() is called. Since an interrupt source setup and dismantle code are done in the non-sleepable context, flushing interrupt entries cache in the IR hardware, which is done async and ideally waits for the interrupt, requires busy-wait for queue to drain. The dmar_qi_wait_for_seq() is modified to take a boolean argument requesting busy-wait for the written sequence number instead of waiting for interrupt. Some interrupts are configured before IR is initialized, e.g. ACPI SCI. Add intr_reprogram() function to reprogram all already configured interrupts, and call it immediately before an IR unit is enabled. There is still a small window after the IO-APIC redirection entry is reprogrammed with cookie but before the unit is enabled, but to fix this properly, IR must be started much earlier. Add workarounds for 5500 and X58 northbridges, some revisions of which have severe flaws in handling IR. Use the same identification methods as employed by Linux. Review: https://reviews.freebsd.org/D1892 Reviewed by: neel Discussed with: jhb Tested by: glebius, pho (previous versions) Sponsored by: The FreeBSD Foundation MFC after: 3 weeks
2015-03-19 13:57:47 +00:00
/*-
* Copyright (c) 2015 The FreeBSD Foundation
* All rights reserved.
*
* This software was developed by Konstantin Belousov <kib@FreeBSD.org>
* under sponsorship from 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.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/bus.h>
#include <sys/kernel.h>
#include <sys/lock.h>
#include <sys/malloc.h>
#include <sys/memdesc.h>
#include <sys/mutex.h>
Use VT-d interrupt remapping block (IR) to perform FSB messages translation. In particular, despite IO-APICs only take 8bit apic id, IR translation structures accept 32bit APIC Id, which allows x2APIC mode to function properly. Extend msi_cpu of struct msi_intrsrc and io_cpu of ioapic_intsrc to full int from one byte. KPI of IR is isolated into the x86/iommu/iommu_intrmap.h, to avoid bringing all dmar headers into interrupt code. The non-PCI(e) devices which generate message interrupts on FSB require special handling. The HPET FSB interrupts are remapped, while DMAR interrupts are not. For each msi and ioapic interrupt source, the iommu cookie is added, which is in fact index of the IRE (interrupt remap entry) in the IR table. Cookie is made at the source allocation time, and then used at the map time to fill both IRE and device registers. The MSI address/data registers and IO-APIC redirection registers are programmed with the special values which are recognized by IR and used to restore the IRE index, to find proper delivery mode and target. Map all MSI interrupts in the block when msi_map() is called. Since an interrupt source setup and dismantle code are done in the non-sleepable context, flushing interrupt entries cache in the IR hardware, which is done async and ideally waits for the interrupt, requires busy-wait for queue to drain. The dmar_qi_wait_for_seq() is modified to take a boolean argument requesting busy-wait for the written sequence number instead of waiting for interrupt. Some interrupts are configured before IR is initialized, e.g. ACPI SCI. Add intr_reprogram() function to reprogram all already configured interrupts, and call it immediately before an IR unit is enabled. There is still a small window after the IO-APIC redirection entry is reprogrammed with cookie but before the unit is enabled, but to fix this properly, IR must be started much earlier. Add workarounds for 5500 and X58 northbridges, some revisions of which have severe flaws in handling IR. Use the same identification methods as employed by Linux. Review: https://reviews.freebsd.org/D1892 Reviewed by: neel Discussed with: jhb Tested by: glebius, pho (previous versions) Sponsored by: The FreeBSD Foundation MFC after: 3 weeks
2015-03-19 13:57:47 +00:00
#include <sys/rman.h>
#include <sys/rwlock.h>
#include <sys/sysctl.h>
Use VT-d interrupt remapping block (IR) to perform FSB messages translation. In particular, despite IO-APICs only take 8bit apic id, IR translation structures accept 32bit APIC Id, which allows x2APIC mode to function properly. Extend msi_cpu of struct msi_intrsrc and io_cpu of ioapic_intsrc to full int from one byte. KPI of IR is isolated into the x86/iommu/iommu_intrmap.h, to avoid bringing all dmar headers into interrupt code. The non-PCI(e) devices which generate message interrupts on FSB require special handling. The HPET FSB interrupts are remapped, while DMAR interrupts are not. For each msi and ioapic interrupt source, the iommu cookie is added, which is in fact index of the IRE (interrupt remap entry) in the IR table. Cookie is made at the source allocation time, and then used at the map time to fill both IRE and device registers. The MSI address/data registers and IO-APIC redirection registers are programmed with the special values which are recognized by IR and used to restore the IRE index, to find proper delivery mode and target. Map all MSI interrupts in the block when msi_map() is called. Since an interrupt source setup and dismantle code are done in the non-sleepable context, flushing interrupt entries cache in the IR hardware, which is done async and ideally waits for the interrupt, requires busy-wait for queue to drain. The dmar_qi_wait_for_seq() is modified to take a boolean argument requesting busy-wait for the written sequence number instead of waiting for interrupt. Some interrupts are configured before IR is initialized, e.g. ACPI SCI. Add intr_reprogram() function to reprogram all already configured interrupts, and call it immediately before an IR unit is enabled. There is still a small window after the IO-APIC redirection entry is reprogrammed with cookie but before the unit is enabled, but to fix this properly, IR must be started much earlier. Add workarounds for 5500 and X58 northbridges, some revisions of which have severe flaws in handling IR. Use the same identification methods as employed by Linux. Review: https://reviews.freebsd.org/D1892 Reviewed by: neel Discussed with: jhb Tested by: glebius, pho (previous versions) Sponsored by: The FreeBSD Foundation MFC after: 3 weeks
2015-03-19 13:57:47 +00:00
#include <sys/taskqueue.h>
#include <sys/tree.h>
#include <sys/vmem.h>
#include <vm/vm.h>
#include <vm/vm_extern.h>
#include <vm/vm_kern.h>
#include <vm/vm_object.h>
#include <vm/vm_page.h>
#include <dev/pci/pcireg.h>
#include <dev/pci/pcivar.h>
#include <machine/bus.h>
#include <machine/intr_machdep.h>
Use VT-d interrupt remapping block (IR) to perform FSB messages translation. In particular, despite IO-APICs only take 8bit apic id, IR translation structures accept 32bit APIC Id, which allows x2APIC mode to function properly. Extend msi_cpu of struct msi_intrsrc and io_cpu of ioapic_intsrc to full int from one byte. KPI of IR is isolated into the x86/iommu/iommu_intrmap.h, to avoid bringing all dmar headers into interrupt code. The non-PCI(e) devices which generate message interrupts on FSB require special handling. The HPET FSB interrupts are remapped, while DMAR interrupts are not. For each msi and ioapic interrupt source, the iommu cookie is added, which is in fact index of the IRE (interrupt remap entry) in the IR table. Cookie is made at the source allocation time, and then used at the map time to fill both IRE and device registers. The MSI address/data registers and IO-APIC redirection registers are programmed with the special values which are recognized by IR and used to restore the IRE index, to find proper delivery mode and target. Map all MSI interrupts in the block when msi_map() is called. Since an interrupt source setup and dismantle code are done in the non-sleepable context, flushing interrupt entries cache in the IR hardware, which is done async and ideally waits for the interrupt, requires busy-wait for queue to drain. The dmar_qi_wait_for_seq() is modified to take a boolean argument requesting busy-wait for the written sequence number instead of waiting for interrupt. Some interrupts are configured before IR is initialized, e.g. ACPI SCI. Add intr_reprogram() function to reprogram all already configured interrupts, and call it immediately before an IR unit is enabled. There is still a small window after the IO-APIC redirection entry is reprogrammed with cookie but before the unit is enabled, but to fix this properly, IR must be started much earlier. Add workarounds for 5500 and X58 northbridges, some revisions of which have severe flaws in handling IR. Use the same identification methods as employed by Linux. Review: https://reviews.freebsd.org/D1892 Reviewed by: neel Discussed with: jhb Tested by: glebius, pho (previous versions) Sponsored by: The FreeBSD Foundation MFC after: 3 weeks
2015-03-19 13:57:47 +00:00
#include <x86/include/apicreg.h>
#include <x86/include/apicvar.h>
#include <x86/include/busdma_impl.h>
#include <dev/iommu/busdma_iommu.h>
#include <x86/iommu/intel_reg.h>
Use VT-d interrupt remapping block (IR) to perform FSB messages translation. In particular, despite IO-APICs only take 8bit apic id, IR translation structures accept 32bit APIC Id, which allows x2APIC mode to function properly. Extend msi_cpu of struct msi_intrsrc and io_cpu of ioapic_intsrc to full int from one byte. KPI of IR is isolated into the x86/iommu/iommu_intrmap.h, to avoid bringing all dmar headers into interrupt code. The non-PCI(e) devices which generate message interrupts on FSB require special handling. The HPET FSB interrupts are remapped, while DMAR interrupts are not. For each msi and ioapic interrupt source, the iommu cookie is added, which is in fact index of the IRE (interrupt remap entry) in the IR table. Cookie is made at the source allocation time, and then used at the map time to fill both IRE and device registers. The MSI address/data registers and IO-APIC redirection registers are programmed with the special values which are recognized by IR and used to restore the IRE index, to find proper delivery mode and target. Map all MSI interrupts in the block when msi_map() is called. Since an interrupt source setup and dismantle code are done in the non-sleepable context, flushing interrupt entries cache in the IR hardware, which is done async and ideally waits for the interrupt, requires busy-wait for queue to drain. The dmar_qi_wait_for_seq() is modified to take a boolean argument requesting busy-wait for the written sequence number instead of waiting for interrupt. Some interrupts are configured before IR is initialized, e.g. ACPI SCI. Add intr_reprogram() function to reprogram all already configured interrupts, and call it immediately before an IR unit is enabled. There is still a small window after the IO-APIC redirection entry is reprogrammed with cookie but before the unit is enabled, but to fix this properly, IR must be started much earlier. Add workarounds for 5500 and X58 northbridges, some revisions of which have severe flaws in handling IR. Use the same identification methods as employed by Linux. Review: https://reviews.freebsd.org/D1892 Reviewed by: neel Discussed with: jhb Tested by: glebius, pho (previous versions) Sponsored by: The FreeBSD Foundation MFC after: 3 weeks
2015-03-19 13:57:47 +00:00
#include <x86/iommu/intel_dmar.h>
#include <x86/iommu/iommu_intrmap.h>
static struct dmar_unit *dmar_ir_find(device_t src, uint16_t *rid,
int *is_dmar);
static void dmar_ir_program_irte(struct dmar_unit *unit, u_int idx,
uint64_t low, uint16_t rid);
static int dmar_ir_free_irte(struct dmar_unit *unit, u_int cookie);
int
iommu_alloc_msi_intr(device_t src, u_int *cookies, u_int count)
{
struct dmar_unit *unit;
vmem_addr_t vmem_res;
u_int idx, i;
int error;
unit = dmar_ir_find(src, NULL, NULL);
if (unit == NULL || !unit->ir_enabled) {
for (i = 0; i < count; i++)
cookies[i] = -1;
return (EOPNOTSUPP);
}
error = vmem_alloc(unit->irtids, count, M_FIRSTFIT | M_NOWAIT,
&vmem_res);
if (error != 0) {
KASSERT(error != EOPNOTSUPP,
("impossible EOPNOTSUPP from vmem"));
return (error);
}
idx = vmem_res;
for (i = 0; i < count; i++)
cookies[i] = idx + i;
return (0);
}
int
iommu_map_msi_intr(device_t src, u_int cpu, u_int vector, u_int cookie,
uint64_t *addr, uint32_t *data)
{
struct dmar_unit *unit;
uint64_t low;
uint16_t rid;
int is_dmar;
unit = dmar_ir_find(src, &rid, &is_dmar);
if (is_dmar) {
KASSERT(unit == NULL, ("DMAR cannot translate itself"));
/*
* See VT-d specification, 5.1.6 Remapping Hardware -
* Interrupt Programming.
*/
*data = vector;
*addr = MSI_INTEL_ADDR_BASE | ((cpu & 0xff) << 12);
if (x2apic_mode)
*addr |= ((uint64_t)cpu & 0xffffff00) << 32;
else
KASSERT(cpu <= 0xff, ("cpu id too big %d", cpu));
return (0);
}
if (unit == NULL || !unit->ir_enabled || cookie == -1)
return (EOPNOTSUPP);
low = (DMAR_X2APIC(unit) ? DMAR_IRTE1_DST_x2APIC(cpu) :
DMAR_IRTE1_DST_xAPIC(cpu)) | DMAR_IRTE1_V(vector) |
DMAR_IRTE1_DLM_FM | DMAR_IRTE1_TM_EDGE | DMAR_IRTE1_RH_DIRECT |
DMAR_IRTE1_DM_PHYSICAL | DMAR_IRTE1_P;
dmar_ir_program_irte(unit, cookie, low, rid);
if (addr != NULL) {
/*
* See VT-d specification, 5.1.5.2 MSI and MSI-X
* Register Programming.
*/
*addr = MSI_INTEL_ADDR_BASE | ((cookie & 0x7fff) << 5) |
((cookie & 0x8000) << 2) | 0x18;
*data = 0;
}
return (0);
}
int
iommu_unmap_msi_intr(device_t src, u_int cookie)
{
struct dmar_unit *unit;
if (cookie == -1)
return (0);
unit = dmar_ir_find(src, NULL, NULL);
return (dmar_ir_free_irte(unit, cookie));
}
int
iommu_map_ioapic_intr(u_int ioapic_id, u_int cpu, u_int vector, bool edge,
bool activehi, int irq, u_int *cookie, uint32_t *hi, uint32_t *lo)
{
struct dmar_unit *unit;
vmem_addr_t vmem_res;
uint64_t low, iorte;
u_int idx;
int error;
uint16_t rid;
unit = dmar_find_ioapic(ioapic_id, &rid);
if (unit == NULL || !unit->ir_enabled) {
*cookie = -1;
return (EOPNOTSUPP);
}
error = vmem_alloc(unit->irtids, 1, M_FIRSTFIT | M_NOWAIT, &vmem_res);
if (error != 0) {
KASSERT(error != EOPNOTSUPP,
("impossible EOPNOTSUPP from vmem"));
return (error);
}
idx = vmem_res;
low = 0;
switch (irq) {
case IRQ_EXTINT:
low |= DMAR_IRTE1_DLM_ExtINT;
break;
case IRQ_NMI:
low |= DMAR_IRTE1_DLM_NMI;
break;
case IRQ_SMI:
low |= DMAR_IRTE1_DLM_SMI;
break;
default:
KASSERT(vector != 0, ("No vector for IRQ %u", irq));
low |= DMAR_IRTE1_DLM_FM | DMAR_IRTE1_V(vector);
break;
}
low |= (DMAR_X2APIC(unit) ? DMAR_IRTE1_DST_x2APIC(cpu) :
DMAR_IRTE1_DST_xAPIC(cpu)) |
(edge ? DMAR_IRTE1_TM_EDGE : DMAR_IRTE1_TM_LEVEL) |
DMAR_IRTE1_RH_DIRECT | DMAR_IRTE1_DM_PHYSICAL | DMAR_IRTE1_P;
dmar_ir_program_irte(unit, idx, low, rid);
if (hi != NULL) {
/*
* See VT-d specification, 5.1.5.1 I/OxAPIC
* Programming.
*/
iorte = (1ULL << 48) | ((uint64_t)(idx & 0x7fff) << 49) |
((idx & 0x8000) != 0 ? (1 << 11) : 0) |
(edge ? IOART_TRGREDG : IOART_TRGRLVL) |
(activehi ? IOART_INTAHI : IOART_INTALO) |
IOART_DELFIXED | vector;
*hi = iorte >> 32;
*lo = iorte;
}
*cookie = idx;
return (0);
}
int
iommu_unmap_ioapic_intr(u_int ioapic_id, u_int *cookie)
{
struct dmar_unit *unit;
u_int idx;
idx = *cookie;
if (idx == -1)
return (0);
*cookie = -1;
unit = dmar_find_ioapic(ioapic_id, NULL);
KASSERT(unit != NULL && unit->ir_enabled,
("unmap: cookie %d unit %p", idx, unit));
return (dmar_ir_free_irte(unit, idx));
}
static struct dmar_unit *
dmar_ir_find(device_t src, uint16_t *rid, int *is_dmar)
{
devclass_t src_class;
struct dmar_unit *unit;
/*
* We need to determine if the interrupt source generates FSB
* interrupts. If yes, it is either DMAR, in which case
* interrupts are not remapped. Or it is HPET, and interrupts
* are remapped. For HPET, source id is reported by HPET
* record in DMAR ACPI table.
*/
if (is_dmar != NULL)
*is_dmar = FALSE;
src_class = device_get_devclass(src);
if (src_class == devclass_find("dmar")) {
unit = NULL;
if (is_dmar != NULL)
*is_dmar = TRUE;
} else if (src_class == devclass_find("hpet")) {
unit = dmar_find_hpet(src, rid);
} else {
unit = dmar_find(src, bootverbose);
Use VT-d interrupt remapping block (IR) to perform FSB messages translation. In particular, despite IO-APICs only take 8bit apic id, IR translation structures accept 32bit APIC Id, which allows x2APIC mode to function properly. Extend msi_cpu of struct msi_intrsrc and io_cpu of ioapic_intsrc to full int from one byte. KPI of IR is isolated into the x86/iommu/iommu_intrmap.h, to avoid bringing all dmar headers into interrupt code. The non-PCI(e) devices which generate message interrupts on FSB require special handling. The HPET FSB interrupts are remapped, while DMAR interrupts are not. For each msi and ioapic interrupt source, the iommu cookie is added, which is in fact index of the IRE (interrupt remap entry) in the IR table. Cookie is made at the source allocation time, and then used at the map time to fill both IRE and device registers. The MSI address/data registers and IO-APIC redirection registers are programmed with the special values which are recognized by IR and used to restore the IRE index, to find proper delivery mode and target. Map all MSI interrupts in the block when msi_map() is called. Since an interrupt source setup and dismantle code are done in the non-sleepable context, flushing interrupt entries cache in the IR hardware, which is done async and ideally waits for the interrupt, requires busy-wait for queue to drain. The dmar_qi_wait_for_seq() is modified to take a boolean argument requesting busy-wait for the written sequence number instead of waiting for interrupt. Some interrupts are configured before IR is initialized, e.g. ACPI SCI. Add intr_reprogram() function to reprogram all already configured interrupts, and call it immediately before an IR unit is enabled. There is still a small window after the IO-APIC redirection entry is reprogrammed with cookie but before the unit is enabled, but to fix this properly, IR must be started much earlier. Add workarounds for 5500 and X58 northbridges, some revisions of which have severe flaws in handling IR. Use the same identification methods as employed by Linux. Review: https://reviews.freebsd.org/D1892 Reviewed by: neel Discussed with: jhb Tested by: glebius, pho (previous versions) Sponsored by: The FreeBSD Foundation MFC after: 3 weeks
2015-03-19 13:57:47 +00:00
if (unit != NULL && rid != NULL)
iommu_get_requester(src, rid);
Use VT-d interrupt remapping block (IR) to perform FSB messages translation. In particular, despite IO-APICs only take 8bit apic id, IR translation structures accept 32bit APIC Id, which allows x2APIC mode to function properly. Extend msi_cpu of struct msi_intrsrc and io_cpu of ioapic_intsrc to full int from one byte. KPI of IR is isolated into the x86/iommu/iommu_intrmap.h, to avoid bringing all dmar headers into interrupt code. The non-PCI(e) devices which generate message interrupts on FSB require special handling. The HPET FSB interrupts are remapped, while DMAR interrupts are not. For each msi and ioapic interrupt source, the iommu cookie is added, which is in fact index of the IRE (interrupt remap entry) in the IR table. Cookie is made at the source allocation time, and then used at the map time to fill both IRE and device registers. The MSI address/data registers and IO-APIC redirection registers are programmed with the special values which are recognized by IR and used to restore the IRE index, to find proper delivery mode and target. Map all MSI interrupts in the block when msi_map() is called. Since an interrupt source setup and dismantle code are done in the non-sleepable context, flushing interrupt entries cache in the IR hardware, which is done async and ideally waits for the interrupt, requires busy-wait for queue to drain. The dmar_qi_wait_for_seq() is modified to take a boolean argument requesting busy-wait for the written sequence number instead of waiting for interrupt. Some interrupts are configured before IR is initialized, e.g. ACPI SCI. Add intr_reprogram() function to reprogram all already configured interrupts, and call it immediately before an IR unit is enabled. There is still a small window after the IO-APIC redirection entry is reprogrammed with cookie but before the unit is enabled, but to fix this properly, IR must be started much earlier. Add workarounds for 5500 and X58 northbridges, some revisions of which have severe flaws in handling IR. Use the same identification methods as employed by Linux. Review: https://reviews.freebsd.org/D1892 Reviewed by: neel Discussed with: jhb Tested by: glebius, pho (previous versions) Sponsored by: The FreeBSD Foundation MFC after: 3 weeks
2015-03-19 13:57:47 +00:00
}
return (unit);
}
static void
dmar_ir_program_irte(struct dmar_unit *unit, u_int idx, uint64_t low,
uint16_t rid)
{
dmar_irte_t *irte;
uint64_t high;
KASSERT(idx < unit->irte_cnt,
("bad cookie %d %d", idx, unit->irte_cnt));
irte = &(unit->irt[idx]);
high = DMAR_IRTE2_SVT_RID | DMAR_IRTE2_SQ_RID |
DMAR_IRTE2_SID_RID(rid);
if (bootverbose) {
device_printf(unit->dev,
"programming irte[%d] rid %#x high %#jx low %#jx\n",
idx, rid, (uintmax_t)high, (uintmax_t)low);
}
Use VT-d interrupt remapping block (IR) to perform FSB messages translation. In particular, despite IO-APICs only take 8bit apic id, IR translation structures accept 32bit APIC Id, which allows x2APIC mode to function properly. Extend msi_cpu of struct msi_intrsrc and io_cpu of ioapic_intsrc to full int from one byte. KPI of IR is isolated into the x86/iommu/iommu_intrmap.h, to avoid bringing all dmar headers into interrupt code. The non-PCI(e) devices which generate message interrupts on FSB require special handling. The HPET FSB interrupts are remapped, while DMAR interrupts are not. For each msi and ioapic interrupt source, the iommu cookie is added, which is in fact index of the IRE (interrupt remap entry) in the IR table. Cookie is made at the source allocation time, and then used at the map time to fill both IRE and device registers. The MSI address/data registers and IO-APIC redirection registers are programmed with the special values which are recognized by IR and used to restore the IRE index, to find proper delivery mode and target. Map all MSI interrupts in the block when msi_map() is called. Since an interrupt source setup and dismantle code are done in the non-sleepable context, flushing interrupt entries cache in the IR hardware, which is done async and ideally waits for the interrupt, requires busy-wait for queue to drain. The dmar_qi_wait_for_seq() is modified to take a boolean argument requesting busy-wait for the written sequence number instead of waiting for interrupt. Some interrupts are configured before IR is initialized, e.g. ACPI SCI. Add intr_reprogram() function to reprogram all already configured interrupts, and call it immediately before an IR unit is enabled. There is still a small window after the IO-APIC redirection entry is reprogrammed with cookie but before the unit is enabled, but to fix this properly, IR must be started much earlier. Add workarounds for 5500 and X58 northbridges, some revisions of which have severe flaws in handling IR. Use the same identification methods as employed by Linux. Review: https://reviews.freebsd.org/D1892 Reviewed by: neel Discussed with: jhb Tested by: glebius, pho (previous versions) Sponsored by: The FreeBSD Foundation MFC after: 3 weeks
2015-03-19 13:57:47 +00:00
DMAR_LOCK(unit);
if ((irte->irte1 & DMAR_IRTE1_P) != 0) {
/*
* The rte is already valid. Assume that the request
* is to remap the interrupt for balancing. Only low
* word of rte needs to be changed. Assert that the
* high word contains expected value.
*/
KASSERT(irte->irte2 == high,
("irte2 mismatch, %jx %jx", (uintmax_t)irte->irte2,
(uintmax_t)high));
dmar_pte_update(&irte->irte1, low);
} else {
dmar_pte_store(&irte->irte2, high);
dmar_pte_store(&irte->irte1, low);
}
dmar_qi_invalidate_iec(unit, idx, 1);
DMAR_UNLOCK(unit);
}
static int
dmar_ir_free_irte(struct dmar_unit *unit, u_int cookie)
{
dmar_irte_t *irte;
KASSERT(unit != NULL && unit->ir_enabled,
("unmap: cookie %d unit %p", cookie, unit));
KASSERT(cookie < unit->irte_cnt,
("bad cookie %u %u", cookie, unit->irte_cnt));
irte = &(unit->irt[cookie]);
dmar_pte_clear(&irte->irte1);
dmar_pte_clear(&irte->irte2);
DMAR_LOCK(unit);
dmar_qi_invalidate_iec(unit, cookie, 1);
DMAR_UNLOCK(unit);
vmem_free(unit->irtids, cookie, 1);
return (0);
}
static u_int
clp2(u_int v)
{
return (powerof2(v) ? v : 1 << fls(v));
}
int
dmar_init_irt(struct dmar_unit *unit)
{
if ((unit->hw_ecap & DMAR_ECAP_IR) == 0)
return (0);
unit->ir_enabled = 1;
TUNABLE_INT_FETCH("hw.dmar.ir", &unit->ir_enabled);
if (!unit->ir_enabled)
return (0);
if (!unit->qi_enabled) {
unit->ir_enabled = 0;
if (bootverbose)
device_printf(unit->dev,
"QI disabled, disabling interrupt remapping\n");
return (0);
}
Dynamically allocate IRQ ranges on x86. Previously, x86 used static ranges of IRQ values for different types of I/O interrupts. Interrupt pins on I/O APICs and 8259A PICs used IRQ values from 0 to 254. MSI interrupts used a compile-time-defined range starting at 256, and Xen event channels used a compile-time-defined range after MSI. Some recent systems have more than 255 I/O APIC interrupt pins which resulted in those IRQ values overflowing into the MSI range triggering an assertion failure. Replace statically assigned ranges with dynamic ranges. Do a single pass computing the sizes of the IRQ ranges (PICs, MSI, Xen) to determine the total number of IRQs required. Allocate the interrupt source and interrupt count arrays dynamically once this pass has completed. To minimize runtime complexity these arrays are only sized once during bootup. The PIC range is determined by the PICs present in the system. The MSI and Xen ranges continue to use a fixed size, though this does make it possible to turn the MSI range size into a tunable in the future. As a result, various places are updated to use dynamic limits instead of constants. In addition, the vmstat(8) utility has been taught to understand that some kernels may treat 'intrcnt' and 'intrnames' as pointers rather than arrays when extracting interrupt stats from a crashdump. This is determined by the presence (vs absence) of a global 'nintrcnt' symbol. This change reverts r189404 which worked around a buggy BIOS which enumerated an I/O APIC twice (using the same memory mapped address for both entries but using an IRQ base of 256 for one entry and a valid IRQ base for the second entry). Making the "base" of MSI IRQ values dynamic avoids the panic that r189404 worked around, and there may now be valid I/O APICs with an IRQ base above 256 which this workaround would incorrectly skip. If in the future the issue reported in PR 130483 reoccurs, we will have to add a pass over the I/O APIC entries in the MADT to detect duplicates using the memory mapped address and use some strategy to choose the "correct" one. While here, reserve room in intrcnts for the Hyper-V counters. PR: 229429, 130483 Reviewed by: kib, royger, cem Tested by: royger (Xen), kib (DMAR) Approved by: re (gjb) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D16861
2018-08-28 21:09:19 +00:00
unit->irte_cnt = clp2(num_io_irqs);
unit->irt = (dmar_irte_t *)(uintptr_t)kmem_alloc_contig(
Use VT-d interrupt remapping block (IR) to perform FSB messages translation. In particular, despite IO-APICs only take 8bit apic id, IR translation structures accept 32bit APIC Id, which allows x2APIC mode to function properly. Extend msi_cpu of struct msi_intrsrc and io_cpu of ioapic_intsrc to full int from one byte. KPI of IR is isolated into the x86/iommu/iommu_intrmap.h, to avoid bringing all dmar headers into interrupt code. The non-PCI(e) devices which generate message interrupts on FSB require special handling. The HPET FSB interrupts are remapped, while DMAR interrupts are not. For each msi and ioapic interrupt source, the iommu cookie is added, which is in fact index of the IRE (interrupt remap entry) in the IR table. Cookie is made at the source allocation time, and then used at the map time to fill both IRE and device registers. The MSI address/data registers and IO-APIC redirection registers are programmed with the special values which are recognized by IR and used to restore the IRE index, to find proper delivery mode and target. Map all MSI interrupts in the block when msi_map() is called. Since an interrupt source setup and dismantle code are done in the non-sleepable context, flushing interrupt entries cache in the IR hardware, which is done async and ideally waits for the interrupt, requires busy-wait for queue to drain. The dmar_qi_wait_for_seq() is modified to take a boolean argument requesting busy-wait for the written sequence number instead of waiting for interrupt. Some interrupts are configured before IR is initialized, e.g. ACPI SCI. Add intr_reprogram() function to reprogram all already configured interrupts, and call it immediately before an IR unit is enabled. There is still a small window after the IO-APIC redirection entry is reprogrammed with cookie but before the unit is enabled, but to fix this properly, IR must be started much earlier. Add workarounds for 5500 and X58 northbridges, some revisions of which have severe flaws in handling IR. Use the same identification methods as employed by Linux. Review: https://reviews.freebsd.org/D1892 Reviewed by: neel Discussed with: jhb Tested by: glebius, pho (previous versions) Sponsored by: The FreeBSD Foundation MFC after: 3 weeks
2015-03-19 13:57:47 +00:00
unit->irte_cnt * sizeof(dmar_irte_t), M_ZERO | M_WAITOK, 0,
dmar_high, PAGE_SIZE, 0, DMAR_IS_COHERENT(unit) ?
VM_MEMATTR_DEFAULT : VM_MEMATTR_UNCACHEABLE);
if (unit->irt == NULL)
return (ENOMEM);
unit->irt_phys = pmap_kextract((vm_offset_t)unit->irt);
unit->irtids = vmem_create("dmarirt", 0, unit->irte_cnt, 1, 0,
M_FIRSTFIT | M_NOWAIT);
DMAR_LOCK(unit);
dmar_load_irt_ptr(unit);
dmar_qi_invalidate_iec_glob(unit);
DMAR_UNLOCK(unit);
/*
* Initialize mappings for already configured interrupt pins.
* Required, because otherwise the interrupts fault without
* irtes.
*/
intr_reprogram();
DMAR_LOCK(unit);
dmar_enable_ir(unit);
DMAR_UNLOCK(unit);
return (0);
}
void
dmar_fini_irt(struct dmar_unit *unit)
{
unit->ir_enabled = 0;
if (unit->irt != NULL) {
dmar_disable_ir(unit);
dmar_qi_invalidate_iec_glob(unit);
vmem_destroy(unit->irtids);
kmem_free((vm_offset_t)unit->irt, unit->irte_cnt *
sizeof(dmar_irte_t));
Use VT-d interrupt remapping block (IR) to perform FSB messages translation. In particular, despite IO-APICs only take 8bit apic id, IR translation structures accept 32bit APIC Id, which allows x2APIC mode to function properly. Extend msi_cpu of struct msi_intrsrc and io_cpu of ioapic_intsrc to full int from one byte. KPI of IR is isolated into the x86/iommu/iommu_intrmap.h, to avoid bringing all dmar headers into interrupt code. The non-PCI(e) devices which generate message interrupts on FSB require special handling. The HPET FSB interrupts are remapped, while DMAR interrupts are not. For each msi and ioapic interrupt source, the iommu cookie is added, which is in fact index of the IRE (interrupt remap entry) in the IR table. Cookie is made at the source allocation time, and then used at the map time to fill both IRE and device registers. The MSI address/data registers and IO-APIC redirection registers are programmed with the special values which are recognized by IR and used to restore the IRE index, to find proper delivery mode and target. Map all MSI interrupts in the block when msi_map() is called. Since an interrupt source setup and dismantle code are done in the non-sleepable context, flushing interrupt entries cache in the IR hardware, which is done async and ideally waits for the interrupt, requires busy-wait for queue to drain. The dmar_qi_wait_for_seq() is modified to take a boolean argument requesting busy-wait for the written sequence number instead of waiting for interrupt. Some interrupts are configured before IR is initialized, e.g. ACPI SCI. Add intr_reprogram() function to reprogram all already configured interrupts, and call it immediately before an IR unit is enabled. There is still a small window after the IO-APIC redirection entry is reprogrammed with cookie but before the unit is enabled, but to fix this properly, IR must be started much earlier. Add workarounds for 5500 and X58 northbridges, some revisions of which have severe flaws in handling IR. Use the same identification methods as employed by Linux. Review: https://reviews.freebsd.org/D1892 Reviewed by: neel Discussed with: jhb Tested by: glebius, pho (previous versions) Sponsored by: The FreeBSD Foundation MFC after: 3 weeks
2015-03-19 13:57:47 +00:00
}
}