hv: add the support of vector remapping for pre-launched VMs

For pre-launched VMs MSI/MSI-x configuration writes are not intercepted by ACRN.
It is pass-thru and interrupts land in ACRN and the guest vector is injected into
the VM's vLAPIC. With this patch, ACRN intercepts MSI/MSI-x config writes and take
the code path to remap interrupt vector/APIC ID as it does for SOS/UOS.

Tracked-On: #2879
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Reviewed-by: Eddie Dong <eddie.dong@intel.com>
This commit is contained in:
Sainath Grandhi
2019-04-02 22:51:32 -07:00
committed by Eddie Dong
parent c4ec7ac358
commit 5b795a3312
7 changed files with 235 additions and 160 deletions

View File

@@ -31,7 +31,6 @@
#include <vm.h>
#include <errno.h>
#include <vtd.h>
#include <ept.h>
#include <mmu.h>
#include <logmsg.h>
@@ -42,81 +41,6 @@ static inline uint32_t pci_bar_base(uint32_t bar)
return bar & PCIM_BAR_MEM_BASE;
}
#if defined(HV_DEBUG)
/**
* @pre vdev != NULL
*/
static int32_t validate(const struct pci_vdev *vdev)
{
uint32_t idx;
int32_t ret = 0;
for (idx = 0U; idx < PCI_BAR_COUNT; idx++) {
if ((vdev->bar[idx].base != 0x0UL)
|| ((vdev->bar[idx].size & 0xFFFUL) != 0x0UL)
|| ((vdev->bar[idx].type != PCIBAR_MEM32)
&& (vdev->bar[idx].type != PCIBAR_NONE))) {
ret = -EINVAL;
break;
}
}
return ret;
}
#endif
/**
* @pre vdev != NULL
* @pre vdev->vpci != NULL
* @pre vdev->vpci->vm != NULL
*/
void vdev_pt_init(const struct pci_vdev *vdev)
{
int32_t ret;
struct acrn_vm *vm = vdev->vpci->vm;
uint16_t pci_command;
ASSERT(validate(vdev) == 0, "Error, invalid bar defined");
/* Create an iommu domain for target VM if not created */
if (vm->iommu == NULL) {
if (vm->arch_vm.nworld_eptp == 0UL) {
vm->arch_vm.nworld_eptp = vm->arch_vm.ept_mem_ops.get_pml4_page(vm->arch_vm.ept_mem_ops.info);
sanitize_pte((uint64_t *)vm->arch_vm.nworld_eptp);
}
vm->iommu = create_iommu_domain(vm->vm_id,
hva2hpa(vm->arch_vm.nworld_eptp), 48U);
}
ret = assign_iommu_device(vm->iommu, (uint8_t)vdev->pdev->bdf.bits.b,
(uint8_t)(vdev->pdev->bdf.value & 0xFFU));
if (ret != 0) {
panic("failed to assign iommu device!");
}
pci_command = (uint16_t)pci_pdev_read_cfg(vdev->pdev->bdf, PCIR_COMMAND, 2U);
/* Disable INTX */
pci_command |= 0x400U;
pci_pdev_write_cfg(vdev->pdev->bdf, PCIR_COMMAND, 2U, pci_command);
}
/**
* @pre vdev != NULL
* @pre vdev->vpci != NULL
* @pre vdev->vpci->vm != NULL
*/
void vdev_pt_deinit(const struct pci_vdev *vdev)
{
int32_t ret;
struct acrn_vm *vm = vdev->vpci->vm;
ret = unassign_iommu_device(vm->iommu, (uint8_t)vdev->pdev->bdf.bits.b,
(uint8_t)(vdev->pdev->bdf.value & 0xFFU));
if (ret != 0) {
panic("failed to unassign iommu device!");
}
}
/**
* @pre vdev != NULL
*/
@@ -135,11 +59,123 @@ int32_t vdev_pt_cfgread(const struct pci_vdev *vdev, uint32_t offset,
}
/**
* @pre vdev != NULL
* @pre vdev->pdev != NULL
* @pre vdev->pdev->msix.table_bar < (PCI_BAR_COUNT - 1U)
*/
void vdev_pt_remap_msix_table_bar(struct pci_vdev *vdev)
{
uint32_t i;
uint64_t addr_hi, addr_lo;
struct pci_msix *msix = &vdev->msix;
struct pci_pdev *pdev = vdev->pdev;
struct pci_bar *bar;
struct acrn_vm *vm = vdev->vpci->vm;
struct acrn_vm_config *vm_config;
vm_config = get_vm_config(vm->vm_id);
ASSERT(vdev->pdev->msix.table_bar < (PCI_BAR_COUNT - 1U), "msix->table_bar out of range");
/* Mask all table entries */
for (i = 0U; i < msix->table_count; i++) {
msix->tables[i].vector_control = PCIM_MSIX_VCTRL_MASK;
msix->tables[i].addr = 0U;
msix->tables[i].data = 0U;
}
bar = &pdev->bar[msix->table_bar];
if (bar != NULL) {
vdev->msix.mmio_hpa = bar->base;
if (vm_config->type == PRE_LAUNCHED_VM) {
vdev->msix.mmio_gpa = vdev->bar[msix->table_bar].base;
} else {
vdev->msix.mmio_gpa = sos_vm_hpa2gpa(bar->base);
}
vdev->msix.mmio_size = bar->size;
}
/*
* For SOS:
* --------
* MSI-X Table BAR Contains:
* Other Info + Tables + PBA Ohter info already mapped into EPT (since SOS)
* Tables are handled by HV MMIO handler (4k adjusted up and down)
* and remaps interrupts
* PBA already mapped into EPT (since SOS)
*
* Other Info + Tables Other info already mapped into EPT (since SOS)
* Tables are handled by HV MMIO handler (4k adjusted up and down)
* and remaps interrupts
*
* Tables Tables are handled by HV MMIO handler (4k adjusted up and down)
* and remaps interrupts
*
* For UOS (launched by DM):
* -------------------------
* MSI-X Table BAR Contains:
* Other Info + Tables + PBA Other info mapped into EPT (4k adjusted) by DM
* Tables are handled by DM MMIO handler (4k adjusted up and down) and SOS writes to tables,
* intercepted by HV MMIO handler and HV remaps interrupts
* PBA already mapped into EPT by DM
*
* Other Info + Tables Other info mapped into EPT by DM
* Tables are handled by DM MMIO handler (4k adjusted up and down) and SOS writes to tables,
* intercepted by HV MMIO handler and HV remaps interrupts.
*
* Tables Tables are handled by DM MMIO handler (4k adjusted up and down) and SOS writes to tables,
* intercepted by HV MMIO handler and HV remaps interrupts.
*
* For Pre-launched VMs (no SOS/DM):
* --------------------------------
* MSI-X Table BAR Contains:
* All 3 cases: Writes to MMIO region in MSI-X Table BAR handled by HV MMIO handler
* If the offset falls within the MSI-X table [offset, offset+tables_size), HV remaps
* interrupts.
* Else, HV writes/reads to/from the corresponding HPA
*/
if (msix->mmio_gpa != 0U) {
if (vm_config->type == PRE_LAUNCHED_VM) {
addr_hi = vdev->msix.mmio_gpa + vdev->msix.mmio_size;
addr_lo = vdev->msix.mmio_gpa;
} else {
/*
* PCI Spec: a BAR may also map other usable address space that is not associated
* with MSI-X structures, but it must not share any naturally aligned 4 KB
* address range with one where either MSI-X structure resides.
* The MSI-X Table and MSI-X PBA are permitted to co-reside within a naturally
* aligned 4 KB address range.
*
* If PBA or others reside in the same BAR with MSI-X Table, devicemodel could
* emulate them and maps these memory range at the 4KB boundary. Here, we should
* make sure only intercept the minimum number of 4K pages needed for MSI-X table.
*/
/* The higher boundary of the 4KB aligned address range for MSI-X table */
addr_hi = msix->mmio_gpa + msix->table_offset + (msix->table_count * MSIX_TABLE_ENTRY_SIZE);
addr_hi = round_page_up(addr_hi);
/* The lower boundary of the 4KB aligned address range for MSI-X table */
addr_lo = round_page_down(msix->mmio_gpa + msix->table_offset);
}
(void)register_mmio_emulation_handler(vdev->vpci->vm, vmsix_table_mmio_access_handler,
addr_lo, addr_hi, vdev);
}
}
/**
* @brief Remaps guest BARs other than MSI-x Table BAR
* This API is invoked upon guest re-programming PCI BAR with MMIO region
* @pre vdev != NULL
* @pre vdev->vpci != NULL
* @pre vdev->vpci->vm != NULL
*/
static void vdev_pt_remap_bar(const struct pci_vdev *vdev, uint32_t idx,
static void vdev_pt_remap_generic_bar(const struct pci_vdev *vdev, uint32_t idx,
uint32_t new_base)
{
struct acrn_vm *vm = vdev->vpci->vm;
@@ -169,6 +205,7 @@ static void vdev_pt_cfgwrite_bar(struct pci_vdev *vdev, uint32_t offset,
uint32_t idx;
uint32_t new_bar, mask;
bool bar_update_normal;
bool is_msix_table_bar;
if ((bytes != 4U) || ((offset & 0x3U) != 0U)) {
return;
@@ -185,12 +222,18 @@ static void vdev_pt_cfgwrite_bar(struct pci_vdev *vdev, uint32_t offset,
case PCIBAR_MEM32:
bar_update_normal = (new_bar_uos != (uint32_t)~0U);
is_msix_table_bar = (has_msix_cap(vdev) && (idx == vdev->msix.table_bar));
new_bar = new_bar_uos & mask;
if (bar_update_normal) {
vdev_pt_remap_bar(vdev, idx,
pci_bar_base(new_bar));
if (is_msix_table_bar) {
vdev->bar[idx].base = pci_bar_base(new_bar);
vdev_pt_remap_msix_table_bar(vdev);
} else {
vdev_pt_remap_generic_bar(vdev, idx,
pci_bar_base(new_bar));
vdev->bar[idx].base = pci_bar_base(new_bar);
vdev->bar[idx].base = pci_bar_base(new_bar);
}
}
break;
@@ -218,4 +261,3 @@ int32_t vdev_pt_cfgwrite(struct pci_vdev *vdev, uint32_t offset,
return ret;
}