* Shaohua Li <shaohua.li@intel.com>,
* Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>,
* Fenghua Yu <fenghua.yu@intel.com>
+ * Joerg Roedel <jroedel@suse.de>
*/
+#define pr_fmt(fmt) "DMAR: " fmt
+
#include <linux/init.h>
#include <linux/bitmap.h>
#include <linux/debugfs.h>
#include <linux/mempool.h>
#include <linux/memory.h>
#include <linux/timer.h>
+#include <linux/io.h>
#include <linux/iova.h>
#include <linux/iommu.h>
#include <linux/intel-iommu.h>
#include <linux/pci-ats.h>
#include <linux/memblock.h>
#include <linux/dma-contiguous.h>
+#include <linux/crash_dump.h>
#include <asm/irq_remapping.h>
#include <asm/cacheflush.h>
#include <asm/iommu.h>
};
#define ROOT_ENTRY_NR (VTD_PAGE_SIZE/sizeof(struct root_entry))
+/*
+ * Take a root_entry and return the Lower Context Table Pointer (LCTP)
+ * if marked present.
+ */
+static phys_addr_t root_entry_lctp(struct root_entry *re)
+{
+ if (!(re->lo & 1))
+ return 0;
+
+ return re->lo & VTD_PAGE_MASK;
+}
+/*
+ * Take a root_entry and return the Upper Context Table Pointer (UCTP)
+ * if marked present.
+ */
+static phys_addr_t root_entry_uctp(struct root_entry *re)
+{
+ if (!(re->hi & 1))
+ return 0;
+
+ return re->hi & VTD_PAGE_MASK;
+}
/*
* low 64 bits:
* 0: present
u64 hi;
};
-static inline bool context_present(struct context_entry *context)
+static inline void context_clear_pasid_enable(struct context_entry *context)
+{
+ context->lo &= ~(1ULL << 11);
+}
+
+static inline bool context_pasid_enabled(struct context_entry *context)
+{
+ return !!(context->lo & (1ULL << 11));
+}
+
+static inline void context_set_copied(struct context_entry *context)
+{
+ context->hi |= (1ull << 3);
+}
+
+static inline bool context_copied(struct context_entry *context)
+{
+ return !!(context->hi & (1ULL << 3));
+}
+
+static inline bool __context_present(struct context_entry *context)
{
return (context->lo & 1);
}
+
+static inline bool context_present(struct context_entry *context)
+{
+ return context_pasid_enabled(context) ?
+ __context_present(context) :
+ __context_present(context) && !context_copied(context);
+}
+
static inline void context_set_present(struct context_entry *context)
{
context->lo |= 1;
context->hi |= (value & ((1 << 16) - 1)) << 8;
}
+static inline int context_domain_id(struct context_entry *c)
+{
+ return((c->hi >> 8) & 0xffff);
+}
+
static inline void context_clear_entry(struct context_entry *context)
{
context->lo = 0;
static struct dmar_domain *si_domain;
static int hw_pass_through = 1;
-/* domain represents a virtual machine, more than one devices
+/*
+ * Domain represents a virtual machine, more than one devices
* across iommus may be owned in one domain, e.g. kvm guest.
*/
#define DOMAIN_FLAG_VIRTUAL_MACHINE (1 << 0)
/* si_domain contains mulitple devices */
#define DOMAIN_FLAG_STATIC_IDENTITY (1 << 1)
+#define for_each_domain_iommu(idx, domain) \
+ for (idx = 0; idx < g_num_of_iommus; idx++) \
+ if (domain->iommu_refcnt[idx])
+
struct dmar_domain {
- int id; /* domain id */
int nid; /* node id */
- DECLARE_BITMAP(iommu_bmp, DMAR_UNITS_SUPPORTED);
- /* bitmap of iommus this domain uses*/
+
+ unsigned iommu_refcnt[DMAR_UNITS_SUPPORTED];
+ /* Refcount of devices per iommu */
+
+
+ u16 iommu_did[DMAR_UNITS_SUPPORTED];
+ /* Domain ids per IOMMU. Use u16 since
+ * domain ids are 16 bit wide according
+ * to VT-d spec, section 9.3 */
struct list_head devices; /* all devices' list */
struct iova_domain iovad; /* iova's that belong to this domain */
int iommu_superpage;/* Level of superpages supported:
0 == 4KiB (no superpages), 1 == 2MiB,
2 == 1GiB, 3 == 512GiB, 4 == 1TiB */
- spinlock_t iommu_lock; /* protect iommu set in domain */
u64 max_addr; /* maximum mapped address */
struct iommu_domain domain; /* generic domain data structure for
struct list_head global; /* link to global list */
u8 bus; /* PCI bus number */
u8 devfn; /* PCI devfn number */
+ u8 pasid_supported:3;
+ u8 pasid_enabled:1;
+ u8 pri_supported:1;
+ u8 pri_enabled:1;
+ u8 ats_supported:1;
+ u8 ats_enabled:1;
+ u8 ats_qdep;
struct device *dev; /* it's NULL for PCIe-to-PCI bridge */
struct intel_iommu *iommu; /* IOMMU used by this device */
struct dmar_domain *domain; /* pointer to domain */
static void domain_exit(struct dmar_domain *domain);
static void domain_remove_dev_info(struct dmar_domain *domain);
-static void domain_remove_one_dev_info(struct dmar_domain *domain,
- struct device *dev);
-static void iommu_detach_dependent_devices(struct intel_iommu *iommu,
- struct device *dev);
+static void dmar_remove_one_dev_info(struct dmar_domain *domain,
+ struct device *dev);
+static void __dmar_remove_one_dev_info(struct device_domain_info *info);
+static void domain_context_clear(struct intel_iommu *iommu,
+ struct device *dev);
static int domain_detach_iommu(struct dmar_domain *domain,
struct intel_iommu *iommu);
static int intel_iommu_strict;
static int intel_iommu_superpage = 1;
static int intel_iommu_ecs = 1;
+static int intel_iommu_pasid28;
+static int iommu_identity_mapping;
-/* We only actually use ECS when PASID support (on the new bit 40)
- * is also advertised. Some early implementations — the ones with
- * PASID support on bit 28 — have issues even when we *only* use
- * extended root/context tables. */
+#define IDENTMAP_ALL 1
+#define IDENTMAP_GFX 2
+#define IDENTMAP_AZALIA 4
+
+/* Broadwell and Skylake have broken ECS support — normal so-called "second
+ * level" translation of DMA requests-without-PASID doesn't actually happen
+ * unless you also set the NESTE bit in an extended context-entry. Which of
+ * course means that SVM doesn't work because it's trying to do nested
+ * translation of the physical addresses it finds in the process page tables,
+ * through the IOVA->phys mapping found in the "second level" page tables.
+ *
+ * The VT-d specification was retroactively changed to change the definition
+ * of the capability bits and pretend that Broadwell/Skylake never happened...
+ * but unfortunately the wrong bit was changed. It's ECS which is broken, but
+ * for some reason it was the PASID capability bit which was redefined (from
+ * bit 28 on BDW/SKL to bit 40 in future).
+ *
+ * So our test for ECS needs to eschew those implementations which set the old
+ * PASID capabiity bit 28, since those are the ones on which ECS is broken.
+ * Unless we are working around the 'pasid28' limitations, that is, by putting
+ * the device into passthrough mode for normal DMA and thus masking the bug.
+ */
#define ecs_enabled(iommu) (intel_iommu_ecs && ecap_ecs(iommu->ecap) && \
- ecap_pasid(iommu->ecap))
+ (intel_iommu_pasid28 || !ecap_broken_pasid(iommu->ecap)))
+/* PASID support is thus enabled if ECS is enabled and *either* of the old
+ * or new capability bits are set. */
+#define pasid_enabled(iommu) (ecs_enabled(iommu) && \
+ (ecap_pasid(iommu->ecap) || ecap_broken_pasid(iommu->ecap)))
int intel_iommu_gfx_mapped;
EXPORT_SYMBOL_GPL(intel_iommu_gfx_mapped);
static const struct iommu_ops intel_iommu_ops;
+static bool translation_pre_enabled(struct intel_iommu *iommu)
+{
+ return (iommu->flags & VTD_FLAG_TRANS_PRE_ENABLED);
+}
+
+static void clear_translation_pre_enabled(struct intel_iommu *iommu)
+{
+ iommu->flags &= ~VTD_FLAG_TRANS_PRE_ENABLED;
+}
+
+static void init_translation_status(struct intel_iommu *iommu)
+{
+ u32 gsts;
+
+ gsts = readl(iommu->reg + DMAR_GSTS_REG);
+ if (gsts & DMA_GSTS_TES)
+ iommu->flags |= VTD_FLAG_TRANS_PRE_ENABLED;
+}
+
/* Convert generic 'struct iommu_domain to private struct dmar_domain */
static struct dmar_domain *to_dmar_domain(struct iommu_domain *dom)
{
while (*str) {
if (!strncmp(str, "on", 2)) {
dmar_disabled = 0;
- printk(KERN_INFO "Intel-IOMMU: enabled\n");
+ pr_info("IOMMU enabled\n");
} else if (!strncmp(str, "off", 3)) {
dmar_disabled = 1;
- printk(KERN_INFO "Intel-IOMMU: disabled\n");
+ pr_info("IOMMU disabled\n");
} else if (!strncmp(str, "igfx_off", 8)) {
dmar_map_gfx = 0;
- printk(KERN_INFO
- "Intel-IOMMU: disable GFX device mapping\n");
+ pr_info("Disable GFX device mapping\n");
} else if (!strncmp(str, "forcedac", 8)) {
- printk(KERN_INFO
- "Intel-IOMMU: Forcing DAC for PCI devices\n");
+ pr_info("Forcing DAC for PCI devices\n");
dmar_forcedac = 1;
} else if (!strncmp(str, "strict", 6)) {
- printk(KERN_INFO
- "Intel-IOMMU: disable batched IOTLB flush\n");
+ pr_info("Disable batched IOTLB flush\n");
intel_iommu_strict = 1;
} else if (!strncmp(str, "sp_off", 6)) {
- printk(KERN_INFO
- "Intel-IOMMU: disable supported super page\n");
+ pr_info("Disable supported super page\n");
intel_iommu_superpage = 0;
} else if (!strncmp(str, "ecs_off", 7)) {
printk(KERN_INFO
"Intel-IOMMU: disable extended context table support\n");
intel_iommu_ecs = 0;
+ } else if (!strncmp(str, "pasid28", 7)) {
+ printk(KERN_INFO
+ "Intel-IOMMU: enable pre-production PASID support\n");
+ intel_iommu_pasid28 = 1;
+ iommu_identity_mapping |= IDENTMAP_GFX;
}
str += strcspn(str, ",");
static struct kmem_cache *iommu_domain_cache;
static struct kmem_cache *iommu_devinfo_cache;
+static struct dmar_domain* get_iommu_domain(struct intel_iommu *iommu, u16 did)
+{
+ struct dmar_domain **domains;
+ int idx = did >> 8;
+
+ domains = iommu->domains[idx];
+ if (!domains)
+ return NULL;
+
+ return domains[did & 0xff];
+}
+
+static void set_iommu_domain(struct intel_iommu *iommu, u16 did,
+ struct dmar_domain *domain)
+{
+ struct dmar_domain **domains;
+ int idx = did >> 8;
+
+ if (!iommu->domains[idx]) {
+ size_t size = 256 * sizeof(struct dmar_domain *);
+ iommu->domains[idx] = kzalloc(size, GFP_ATOMIC);
+ }
+
+ domains = iommu->domains[idx];
+ if (WARN_ON(!domains))
+ return;
+ else
+ domains[did & 0xff] = domain;
+}
+
static inline void *alloc_pgtable_page(int node)
{
struct page *page;
return domain->flags & DOMAIN_FLAG_VIRTUAL_MACHINE;
}
+static inline int domain_type_is_si(struct dmar_domain *domain)
+{
+ return domain->flags & DOMAIN_FLAG_STATIC_IDENTITY;
+}
+
static inline int domain_type_is_vm_or_si(struct dmar_domain *domain)
{
return domain->flags & (DOMAIN_FLAG_VIRTUAL_MACHINE |
/* si_domain and vm domain should not get here. */
BUG_ON(domain_type_is_vm_or_si(domain));
- iommu_id = find_first_bit(domain->iommu_bmp, g_num_of_iommus);
+ for_each_domain_iommu(iommu_id, domain)
+ break;
+
if (iommu_id < 0 || iommu_id >= g_num_of_iommus)
return NULL;
domain->iommu_coherency = 1;
- for_each_set_bit(i, domain->iommu_bmp, g_num_of_iommus) {
+ for_each_domain_iommu(i, domain) {
found = true;
if (!ecap_coherent(g_iommus[i]->ecap)) {
domain->iommu_coherency = 0;
/* We can't just free the pages because the IOMMU may still be walking
the page tables, and may have cached the intermediate levels. The
pages can only be freed after the IOTLB flush has been done. */
-struct page *domain_unmap(struct dmar_domain *domain,
- unsigned long start_pfn,
- unsigned long last_pfn)
+static struct page *domain_unmap(struct dmar_domain *domain,
+ unsigned long start_pfn,
+ unsigned long last_pfn)
{
struct page *freelist = NULL;
return freelist;
}
-void dma_free_pagelist(struct page *freelist)
+static void dma_free_pagelist(struct page *freelist)
{
struct page *pg;
root = (struct root_entry *)alloc_pgtable_page(iommu->node);
if (!root) {
- pr_err("IOMMU: allocating root entry for %s failed\n",
+ pr_err("Allocating root entry for %s failed\n",
iommu->name);
return -ENOMEM;
}
/* check IOTLB invalidation granularity */
if (DMA_TLB_IAIG(val) == 0)
- printk(KERN_ERR"IOMMU: flush IOTLB failed\n");
+ pr_err("Flush IOTLB failed\n");
if (DMA_TLB_IAIG(val) != DMA_TLB_IIRG(type))
- pr_debug("IOMMU: tlb flush request %Lx, actual %Lx\n",
+ pr_debug("TLB flush request %Lx, actual %Lx\n",
(unsigned long long)DMA_TLB_IIRG(type),
(unsigned long long)DMA_TLB_IAIG(val));
}
iommu_support_dev_iotlb (struct dmar_domain *domain, struct intel_iommu *iommu,
u8 bus, u8 devfn)
{
- bool found = false;
- unsigned long flags;
struct device_domain_info *info;
- struct pci_dev *pdev;
- if (!ecap_dev_iotlb_support(iommu->ecap))
- return NULL;
+ assert_spin_locked(&device_domain_lock);
if (!iommu->qi)
return NULL;
- spin_lock_irqsave(&device_domain_lock, flags);
list_for_each_entry(info, &domain->devices, link)
if (info->iommu == iommu && info->bus == bus &&
info->devfn == devfn) {
- found = true;
+ if (info->ats_supported && info->dev)
+ return info;
break;
}
- spin_unlock_irqrestore(&device_domain_lock, flags);
-
- if (!found || !info->dev || !dev_is_pci(info->dev))
- return NULL;
-
- pdev = to_pci_dev(info->dev);
- if (!pci_find_ext_capability(pdev, PCI_EXT_CAP_ID_ATS))
- return NULL;
-
- if (!dmar_find_matched_atsr_unit(pdev))
- return NULL;
-
- return info;
+ return NULL;
}
static void iommu_enable_dev_iotlb(struct device_domain_info *info)
{
+ struct pci_dev *pdev;
+
if (!info || !dev_is_pci(info->dev))
return;
- pci_enable_ats(to_pci_dev(info->dev), VTD_PAGE_SHIFT);
+ pdev = to_pci_dev(info->dev);
+
+#ifdef CONFIG_INTEL_IOMMU_SVM
+ /* The PCIe spec, in its wisdom, declares that the behaviour of
+ the device if you enable PASID support after ATS support is
+ undefined. So always enable PASID support on devices which
+ have it, even if we can't yet know if we're ever going to
+ use it. */
+ if (info->pasid_supported && !pci_enable_pasid(pdev, info->pasid_supported & ~1))
+ info->pasid_enabled = 1;
+
+ if (info->pri_supported && !pci_reset_pri(pdev) && !pci_enable_pri(pdev, 32))
+ info->pri_enabled = 1;
+#endif
+ if (info->ats_supported && !pci_enable_ats(pdev, VTD_PAGE_SHIFT)) {
+ info->ats_enabled = 1;
+ info->ats_qdep = pci_ats_queue_depth(pdev);
+ }
}
static void iommu_disable_dev_iotlb(struct device_domain_info *info)
{
- if (!info->dev || !dev_is_pci(info->dev) ||
- !pci_ats_enabled(to_pci_dev(info->dev)))
+ struct pci_dev *pdev;
+
+ if (!dev_is_pci(info->dev))
return;
- pci_disable_ats(to_pci_dev(info->dev));
+ pdev = to_pci_dev(info->dev);
+
+ if (info->ats_enabled) {
+ pci_disable_ats(pdev);
+ info->ats_enabled = 0;
+ }
+#ifdef CONFIG_INTEL_IOMMU_SVM
+ if (info->pri_enabled) {
+ pci_disable_pri(pdev);
+ info->pri_enabled = 0;
+ }
+ if (info->pasid_enabled) {
+ pci_disable_pasid(pdev);
+ info->pasid_enabled = 0;
+ }
+#endif
}
static void iommu_flush_dev_iotlb(struct dmar_domain *domain,
spin_lock_irqsave(&device_domain_lock, flags);
list_for_each_entry(info, &domain->devices, link) {
- struct pci_dev *pdev;
- if (!info->dev || !dev_is_pci(info->dev))
- continue;
-
- pdev = to_pci_dev(info->dev);
- if (!pci_ats_enabled(pdev))
+ if (!info->ats_enabled)
continue;
sid = info->bus << 8 | info->devfn;
- qdep = pci_ats_queue_depth(pdev);
+ qdep = info->ats_qdep;
qi_flush_dev_iotlb(info->iommu, sid, qdep, addr, mask);
}
spin_unlock_irqrestore(&device_domain_lock, flags);
}
-static void iommu_flush_iotlb_psi(struct intel_iommu *iommu, u16 did,
- unsigned long pfn, unsigned int pages, int ih, int map)
+static void iommu_flush_iotlb_psi(struct intel_iommu *iommu,
+ struct dmar_domain *domain,
+ unsigned long pfn, unsigned int pages,
+ int ih, int map)
{
unsigned int mask = ilog2(__roundup_pow_of_two(pages));
uint64_t addr = (uint64_t)pfn << VTD_PAGE_SHIFT;
+ u16 did = domain->iommu_did[iommu->seq_id];
BUG_ON(pages == 0);
* flush. However, device IOTLB doesn't need to be flushed in this case.
*/
if (!cap_caching_mode(iommu->cap) || !map)
- iommu_flush_dev_iotlb(iommu->domains[did], addr, mask);
+ iommu_flush_dev_iotlb(get_iommu_domain(iommu, did),
+ addr, mask);
}
static void iommu_disable_protect_mem_regions(struct intel_iommu *iommu)
static int iommu_init_domains(struct intel_iommu *iommu)
{
- unsigned long ndomains;
- unsigned long nlongs;
+ u32 ndomains, nlongs;
+ size_t size;
ndomains = cap_ndoms(iommu->cap);
- pr_debug("IOMMU%d: Number of Domains supported <%ld>\n",
- iommu->seq_id, ndomains);
+ pr_debug("%s: Number of Domains supported <%d>\n",
+ iommu->name, ndomains);
nlongs = BITS_TO_LONGS(ndomains);
spin_lock_init(&iommu->lock);
- /* TBD: there might be 64K domains,
- * consider other allocation for future chip
- */
iommu->domain_ids = kcalloc(nlongs, sizeof(unsigned long), GFP_KERNEL);
if (!iommu->domain_ids) {
- pr_err("IOMMU%d: allocating domain id array failed\n",
- iommu->seq_id);
+ pr_err("%s: Allocating domain id array failed\n",
+ iommu->name);
return -ENOMEM;
}
- iommu->domains = kcalloc(ndomains, sizeof(struct dmar_domain *),
- GFP_KERNEL);
- if (!iommu->domains) {
- pr_err("IOMMU%d: allocating domain array failed\n",
- iommu->seq_id);
+
+ size = ((ndomains >> 8) + 1) * sizeof(struct dmar_domain **);
+ iommu->domains = kzalloc(size, GFP_KERNEL);
+
+ if (iommu->domains) {
+ size = 256 * sizeof(struct dmar_domain *);
+ iommu->domains[0] = kzalloc(size, GFP_KERNEL);
+ }
+
+ if (!iommu->domains || !iommu->domains[0]) {
+ pr_err("%s: Allocating domain array failed\n",
+ iommu->name);
kfree(iommu->domain_ids);
+ kfree(iommu->domains);
iommu->domain_ids = NULL;
+ iommu->domains = NULL;
return -ENOMEM;
}
+
+
/*
- * if Caching mode is set, then invalid translations are tagged
- * with domainid 0. Hence we need to pre-allocate it.
+ * If Caching mode is set, then invalid translations are tagged
+ * with domain-id 0, hence we need to pre-allocate it. We also
+ * use domain-id 0 as a marker for non-allocated domain-id, so
+ * make sure it is not used for a real domain.
*/
- if (cap_caching_mode(iommu->cap))
- set_bit(0, iommu->domain_ids);
+ set_bit(0, iommu->domain_ids);
+
return 0;
}
static void disable_dmar_iommu(struct intel_iommu *iommu)
{
- struct dmar_domain *domain;
- int i;
+ struct device_domain_info *info, *tmp;
+ unsigned long flags;
- if ((iommu->domains) && (iommu->domain_ids)) {
- for_each_set_bit(i, iommu->domain_ids, cap_ndoms(iommu->cap)) {
- /*
- * Domain id 0 is reserved for invalid translation
- * if hardware supports caching mode.
- */
- if (cap_caching_mode(iommu->cap) && i == 0)
- continue;
+ if (!iommu->domains || !iommu->domain_ids)
+ return;
- domain = iommu->domains[i];
- clear_bit(i, iommu->domain_ids);
- if (domain_detach_iommu(domain, iommu) == 0 &&
- !domain_type_is_vm(domain))
- domain_exit(domain);
- }
+ spin_lock_irqsave(&device_domain_lock, flags);
+ list_for_each_entry_safe(info, tmp, &device_domain_list, global) {
+ struct dmar_domain *domain;
+
+ if (info->iommu != iommu)
+ continue;
+
+ if (!info->dev || !info->domain)
+ continue;
+
+ domain = info->domain;
+
+ dmar_remove_one_dev_info(domain, info->dev);
+
+ if (!domain_type_is_vm_or_si(domain))
+ domain_exit(domain);
}
+ spin_unlock_irqrestore(&device_domain_lock, flags);
if (iommu->gcmd & DMA_GCMD_TE)
iommu_disable_translation(iommu);
static void free_dmar_iommu(struct intel_iommu *iommu)
{
if ((iommu->domains) && (iommu->domain_ids)) {
+ int elems = (cap_ndoms(iommu->cap) >> 8) + 1;
+ int i;
+
+ for (i = 0; i < elems; i++)
+ kfree(iommu->domains[i]);
kfree(iommu->domains);
kfree(iommu->domain_ids);
iommu->domains = NULL;
/* free context mapping */
free_context_table(iommu);
+
+#ifdef CONFIG_INTEL_IOMMU_SVM
+ if (pasid_enabled(iommu)) {
+ if (ecap_prs(iommu->ecap))
+ intel_svm_finish_prq(iommu);
+ intel_svm_free_pasid_tables(iommu);
+ }
+#endif
}
static struct dmar_domain *alloc_domain(int flags)
{
- /* domain id for virtual machine, it won't be set in context */
- static atomic_t vm_domid = ATOMIC_INIT(0);
struct dmar_domain *domain;
domain = alloc_domain_mem();
memset(domain, 0, sizeof(*domain));
domain->nid = -1;
domain->flags = flags;
- spin_lock_init(&domain->iommu_lock);
INIT_LIST_HEAD(&domain->devices);
- if (flags & DOMAIN_FLAG_VIRTUAL_MACHINE)
- domain->id = atomic_inc_return(&vm_domid);
return domain;
}
-static int __iommu_attach_domain(struct dmar_domain *domain,
- struct intel_iommu *iommu)
-{
- int num;
- unsigned long ndomains;
-
- ndomains = cap_ndoms(iommu->cap);
- num = find_first_zero_bit(iommu->domain_ids, ndomains);
- if (num < ndomains) {
- set_bit(num, iommu->domain_ids);
- iommu->domains[num] = domain;
- } else {
- num = -ENOSPC;
- }
-
- return num;
-}
-
-static int iommu_attach_domain(struct dmar_domain *domain,
+/* Must be called with iommu->lock */
+static int domain_attach_iommu(struct dmar_domain *domain,
struct intel_iommu *iommu)
{
- int num;
- unsigned long flags;
-
- spin_lock_irqsave(&iommu->lock, flags);
- num = __iommu_attach_domain(domain, iommu);
- spin_unlock_irqrestore(&iommu->lock, flags);
- if (num < 0)
- pr_err("IOMMU: no free domain ids\n");
-
- return num;
-}
-
-static int iommu_attach_vm_domain(struct dmar_domain *domain,
- struct intel_iommu *iommu)
-{
- int num;
unsigned long ndomains;
+ int num;
- ndomains = cap_ndoms(iommu->cap);
- for_each_set_bit(num, iommu->domain_ids, ndomains)
- if (iommu->domains[num] == domain)
- return num;
-
- return __iommu_attach_domain(domain, iommu);
-}
-
-static void iommu_detach_domain(struct dmar_domain *domain,
- struct intel_iommu *iommu)
-{
- unsigned long flags;
- int num, ndomains;
+ assert_spin_locked(&device_domain_lock);
+ assert_spin_locked(&iommu->lock);
- spin_lock_irqsave(&iommu->lock, flags);
- if (domain_type_is_vm_or_si(domain)) {
+ domain->iommu_refcnt[iommu->seq_id] += 1;
+ domain->iommu_count += 1;
+ if (domain->iommu_refcnt[iommu->seq_id] == 1) {
ndomains = cap_ndoms(iommu->cap);
- for_each_set_bit(num, iommu->domain_ids, ndomains) {
- if (iommu->domains[num] == domain) {
- clear_bit(num, iommu->domain_ids);
- iommu->domains[num] = NULL;
- break;
- }
+ num = find_first_zero_bit(iommu->domain_ids, ndomains);
+
+ if (num >= ndomains) {
+ pr_err("%s: No free domain ids\n", iommu->name);
+ domain->iommu_refcnt[iommu->seq_id] -= 1;
+ domain->iommu_count -= 1;
+ return -ENOSPC;
}
- } else {
- clear_bit(domain->id, iommu->domain_ids);
- iommu->domains[domain->id] = NULL;
- }
- spin_unlock_irqrestore(&iommu->lock, flags);
-}
-static void domain_attach_iommu(struct dmar_domain *domain,
- struct intel_iommu *iommu)
-{
- unsigned long flags;
+ set_bit(num, iommu->domain_ids);
+ set_iommu_domain(iommu, num, domain);
+
+ domain->iommu_did[iommu->seq_id] = num;
+ domain->nid = iommu->node;
- spin_lock_irqsave(&domain->iommu_lock, flags);
- if (!test_and_set_bit(iommu->seq_id, domain->iommu_bmp)) {
- domain->iommu_count++;
- if (domain->iommu_count == 1)
- domain->nid = iommu->node;
domain_update_iommu_cap(domain);
}
- spin_unlock_irqrestore(&domain->iommu_lock, flags);
+
+ return 0;
}
static int domain_detach_iommu(struct dmar_domain *domain,
struct intel_iommu *iommu)
{
- unsigned long flags;
- int count = INT_MAX;
+ int num, count = INT_MAX;
+
+ assert_spin_locked(&device_domain_lock);
+ assert_spin_locked(&iommu->lock);
+
+ domain->iommu_refcnt[iommu->seq_id] -= 1;
+ count = --domain->iommu_count;
+ if (domain->iommu_refcnt[iommu->seq_id] == 0) {
+ num = domain->iommu_did[iommu->seq_id];
+ clear_bit(num, iommu->domain_ids);
+ set_iommu_domain(iommu, num, NULL);
- spin_lock_irqsave(&domain->iommu_lock, flags);
- if (test_and_clear_bit(iommu->seq_id, domain->iommu_bmp)) {
- count = --domain->iommu_count;
domain_update_iommu_cap(domain);
+ domain->iommu_did[iommu->seq_id] = 0;
}
- spin_unlock_irqrestore(&domain->iommu_lock, flags);
return count;
}
iova = reserve_iova(&reserved_iova_list, IOVA_PFN(IOAPIC_RANGE_START),
IOVA_PFN(IOAPIC_RANGE_END));
if (!iova) {
- printk(KERN_ERR "Reserve IOAPIC range failed\n");
+ pr_err("Reserve IOAPIC range failed\n");
return -ENODEV;
}
IOVA_PFN(r->start),
IOVA_PFN(r->end));
if (!iova) {
- printk(KERN_ERR "Reserve iova failed\n");
+ pr_err("Reserve iova failed\n");
return -ENODEV;
}
}
return agaw;
}
-static int domain_init(struct dmar_domain *domain, int guest_width)
+static int domain_init(struct dmar_domain *domain, struct intel_iommu *iommu,
+ int guest_width)
{
- struct intel_iommu *iommu;
int adjust_width, agaw;
unsigned long sagaw;
domain_reserve_special_ranges(domain);
/* calculate AGAW */
- iommu = domain_get_iommu(domain);
if (guest_width > cap_mgaw(iommu->cap))
guest_width = cap_mgaw(iommu->cap);
domain->gaw = guest_width;
sagaw = cap_sagaw(iommu->cap);
if (!test_bit(agaw, &sagaw)) {
/* hardware doesn't support it, choose a bigger one */
- pr_debug("IOMMU: hardware doesn't support agaw %d\n", agaw);
+ pr_debug("Hardware doesn't support agaw %d\n", agaw);
agaw = find_next_bit(&sagaw, 5, agaw);
if (agaw >= 5)
return -ENODEV;
static void domain_exit(struct dmar_domain *domain)
{
- struct dmar_drhd_unit *drhd;
- struct intel_iommu *iommu;
struct page *freelist = NULL;
/* Domain 0 is reserved, so dont process it */
if (!intel_iommu_strict)
flush_unmaps_timeout(0);
- /* remove associated devices */
+ /* Remove associated devices and clear attached or cached domains */
+ rcu_read_lock();
domain_remove_dev_info(domain);
+ rcu_read_unlock();
/* destroy iovas */
put_iova_domain(&domain->iovad);
freelist = domain_unmap(domain, 0, DOMAIN_MAX_PFN(domain->gaw));
- /* clear attached or cached domains */
- rcu_read_lock();
- for_each_active_iommu(iommu, drhd)
- if (domain_type_is_vm(domain) ||
- test_bit(iommu->seq_id, domain->iommu_bmp))
- iommu_detach_domain(domain, iommu);
- rcu_read_unlock();
-
dma_free_pagelist(freelist);
free_domain_mem(domain);
static int domain_context_mapping_one(struct dmar_domain *domain,
struct intel_iommu *iommu,
- u8 bus, u8 devfn, int translation)
+ u8 bus, u8 devfn)
{
+ u16 did = domain->iommu_did[iommu->seq_id];
+ int translation = CONTEXT_TT_MULTI_LEVEL;
+ struct device_domain_info *info = NULL;
struct context_entry *context;
unsigned long flags;
struct dma_pte *pgd;
- int id;
- int agaw;
- struct device_domain_info *info = NULL;
+ int ret, agaw;
+
+ WARN_ON(did == 0);
+
+ if (hw_pass_through && domain_type_is_si(domain))
+ translation = CONTEXT_TT_PASS_THROUGH;
pr_debug("Set context mapping for %02x:%02x.%d\n",
bus, PCI_SLOT(devfn), PCI_FUNC(devfn));
BUG_ON(!domain->pgd);
- BUG_ON(translation != CONTEXT_TT_PASS_THROUGH &&
- translation != CONTEXT_TT_MULTI_LEVEL);
- spin_lock_irqsave(&iommu->lock, flags);
+ spin_lock_irqsave(&device_domain_lock, flags);
+ spin_lock(&iommu->lock);
+
+ ret = -ENOMEM;
context = iommu_context_addr(iommu, bus, devfn, 1);
- spin_unlock_irqrestore(&iommu->lock, flags);
if (!context)
- return -ENOMEM;
- spin_lock_irqsave(&iommu->lock, flags);
- if (context_present(context)) {
- spin_unlock_irqrestore(&iommu->lock, flags);
- return 0;
- }
+ goto out_unlock;
+
+ ret = 0;
+ if (context_present(context))
+ goto out_unlock;
- id = domain->id;
pgd = domain->pgd;
- if (domain_type_is_vm_or_si(domain)) {
- if (domain_type_is_vm(domain)) {
- id = iommu_attach_vm_domain(domain, iommu);
- if (id < 0) {
- spin_unlock_irqrestore(&iommu->lock, flags);
- pr_err("IOMMU: no free domain ids\n");
- return -EFAULT;
- }
- }
+ context_clear_entry(context);
+ context_set_domain_id(context, did);
- /* Skip top levels of page tables for
- * iommu which has less agaw than default.
- * Unnecessary for PT mode.
- */
- if (translation != CONTEXT_TT_PASS_THROUGH) {
- for (agaw = domain->agaw; agaw != iommu->agaw; agaw--) {
- pgd = phys_to_virt(dma_pte_addr(pgd));
- if (!dma_pte_present(pgd)) {
- spin_unlock_irqrestore(&iommu->lock, flags);
- return -ENOMEM;
- }
- }
+ /*
+ * Skip top levels of page tables for iommu which has less agaw
+ * than default. Unnecessary for PT mode.
+ */
+ if (translation != CONTEXT_TT_PASS_THROUGH) {
+ for (agaw = domain->agaw; agaw != iommu->agaw; agaw--) {
+ ret = -ENOMEM;
+ pgd = phys_to_virt(dma_pte_addr(pgd));
+ if (!dma_pte_present(pgd))
+ goto out_unlock;
}
- }
- context_set_domain_id(context, id);
-
- if (translation != CONTEXT_TT_PASS_THROUGH) {
info = iommu_support_dev_iotlb(domain, iommu, bus, devfn);
- translation = info ? CONTEXT_TT_DEV_IOTLB :
- CONTEXT_TT_MULTI_LEVEL;
- }
- /*
- * In pass through mode, AW must be programmed to indicate the largest
- * AGAW value supported by hardware. And ASR is ignored by hardware.
- */
- if (unlikely(translation == CONTEXT_TT_PASS_THROUGH))
- context_set_address_width(context, iommu->msagaw);
- else {
+ if (info && info->ats_supported)
+ translation = CONTEXT_TT_DEV_IOTLB;
+ else
+ translation = CONTEXT_TT_MULTI_LEVEL;
+
context_set_address_root(context, virt_to_phys(pgd));
context_set_address_width(context, iommu->agaw);
+ } else {
+ /*
+ * In pass through mode, AW must be programmed to
+ * indicate the largest AGAW value supported by
+ * hardware. And ASR is ignored by hardware.
+ */
+ context_set_address_width(context, iommu->msagaw);
}
context_set_translation_type(context, translation);
(((u16)bus) << 8) | devfn,
DMA_CCMD_MASK_NOBIT,
DMA_CCMD_DEVICE_INVL);
- iommu->flush.flush_iotlb(iommu, id, 0, 0, DMA_TLB_DSI_FLUSH);
+ iommu->flush.flush_iotlb(iommu, did, 0, 0, DMA_TLB_DSI_FLUSH);
} else {
iommu_flush_write_buffer(iommu);
}
iommu_enable_dev_iotlb(info);
- spin_unlock_irqrestore(&iommu->lock, flags);
- domain_attach_iommu(domain, iommu);
+ ret = 0;
+
+out_unlock:
+ spin_unlock(&iommu->lock);
+ spin_unlock_irqrestore(&device_domain_lock, flags);
return 0;
}
struct domain_context_mapping_data {
struct dmar_domain *domain;
struct intel_iommu *iommu;
- int translation;
};
static int domain_context_mapping_cb(struct pci_dev *pdev,
struct domain_context_mapping_data *data = opaque;
return domain_context_mapping_one(data->domain, data->iommu,
- PCI_BUS_NUM(alias), alias & 0xff,
- data->translation);
+ PCI_BUS_NUM(alias), alias & 0xff);
}
static int
-domain_context_mapping(struct dmar_domain *domain, struct device *dev,
- int translation)
+domain_context_mapping(struct dmar_domain *domain, struct device *dev)
{
struct intel_iommu *iommu;
u8 bus, devfn;
return -ENODEV;
if (!dev_is_pci(dev))
- return domain_context_mapping_one(domain, iommu, bus, devfn,
- translation);
+ return domain_context_mapping_one(domain, iommu, bus, devfn);
data.domain = domain;
data.iommu = iommu;
- data.translation = translation;
return pci_for_each_dma_alias(to_pci_dev(dev),
&domain_context_mapping_cb, &data);
return -ENOMEM;
/* It is large page*/
if (largepage_lvl > 1) {
+ unsigned long nr_superpages, end_pfn;
+
pteval |= DMA_PTE_LARGE_PAGE;
lvl_pages = lvl_to_nr_pages(largepage_lvl);
+
+ nr_superpages = sg_res / lvl_pages;
+ end_pfn = iov_pfn + nr_superpages * lvl_pages - 1;
+
/*
* Ensure that old small page tables are
- * removed to make room for superpage,
- * if they exist.
+ * removed to make room for superpage(s).
*/
- dma_pte_free_pagetable(domain, iov_pfn,
- iov_pfn + lvl_pages - 1);
+ dma_pte_free_pagetable(domain, iov_pfn, end_pfn);
} else {
pteval &= ~(uint64_t)DMA_PTE_LARGE_PAGE;
}
tmp = cmpxchg64_local(&pte->val, 0ULL, pteval);
if (tmp) {
static int dumps = 5;
- printk(KERN_CRIT "ERROR: DMA PTE for vPFN 0x%lx already set (to %llx not %llx)\n",
- iov_pfn, tmp, (unsigned long long)pteval);
+ pr_crit("ERROR: DMA PTE for vPFN 0x%lx already set (to %llx not %llx)\n",
+ iov_pfn, tmp, (unsigned long long)pteval);
if (dumps) {
dumps--;
debug_dma_dump_mappings(NULL);
return __domain_mapping(domain, iov_pfn, NULL, phys_pfn, nr_pages, prot);
}
-static void iommu_detach_dev(struct intel_iommu *iommu, u8 bus, u8 devfn)
+static void domain_context_clear_one(struct intel_iommu *iommu, u8 bus, u8 devfn)
{
if (!iommu)
return;
unsigned long flags;
spin_lock_irqsave(&device_domain_lock, flags);
- list_for_each_entry_safe(info, tmp, &domain->devices, link) {
- unlink_domain_info(info);
- spin_unlock_irqrestore(&device_domain_lock, flags);
-
- iommu_disable_dev_iotlb(info);
- iommu_detach_dev(info->iommu, info->bus, info->devfn);
-
- if (domain_type_is_vm(domain)) {
- iommu_detach_dependent_devices(info->iommu, info->dev);
- domain_detach_iommu(domain, info->iommu);
- }
-
- free_devinfo_mem(info);
- spin_lock_irqsave(&device_domain_lock, flags);
- }
+ list_for_each_entry_safe(info, tmp, &domain->devices, link)
+ __dmar_remove_one_dev_info(info);
spin_unlock_irqrestore(&device_domain_lock, flags);
}
return NULL;
}
-static struct dmar_domain *dmar_insert_dev_info(struct intel_iommu *iommu,
- int bus, int devfn,
- struct device *dev,
- struct dmar_domain *domain)
+static struct dmar_domain *dmar_insert_one_dev_info(struct intel_iommu *iommu,
+ int bus, int devfn,
+ struct device *dev,
+ struct dmar_domain *domain)
{
struct dmar_domain *found = NULL;
struct device_domain_info *info;
unsigned long flags;
+ int ret;
info = alloc_devinfo_mem();
if (!info)
info->bus = bus;
info->devfn = devfn;
+ info->ats_supported = info->pasid_supported = info->pri_supported = 0;
+ info->ats_enabled = info->pasid_enabled = info->pri_enabled = 0;
+ info->ats_qdep = 0;
info->dev = dev;
info->domain = domain;
info->iommu = iommu;
+ if (dev && dev_is_pci(dev)) {
+ struct pci_dev *pdev = to_pci_dev(info->dev);
+
+ if (ecap_dev_iotlb_support(iommu->ecap) &&
+ pci_find_ext_capability(pdev, PCI_EXT_CAP_ID_ATS) &&
+ dmar_find_matched_atsr_unit(pdev))
+ info->ats_supported = 1;
+
+ if (ecs_enabled(iommu)) {
+ if (pasid_enabled(iommu)) {
+ int features = pci_pasid_features(pdev);
+ if (features >= 0)
+ info->pasid_supported = features | 1;
+ }
+
+ if (info->ats_supported && ecap_prs(iommu->ecap) &&
+ pci_find_ext_capability(pdev, PCI_EXT_CAP_ID_PRI))
+ info->pri_supported = 1;
+ }
+ }
+
spin_lock_irqsave(&device_domain_lock, flags);
if (dev)
found = find_domain(dev);
- else {
+
+ if (!found) {
struct device_domain_info *info2;
info2 = dmar_search_domain_by_dev_info(iommu->segment, bus, devfn);
- if (info2)
- found = info2->domain;
+ if (info2) {
+ found = info2->domain;
+ info2->dev = dev;
+ }
}
+
if (found) {
spin_unlock_irqrestore(&device_domain_lock, flags);
free_devinfo_mem(info);
return found;
}
+ spin_lock(&iommu->lock);
+ ret = domain_attach_iommu(domain, iommu);
+ spin_unlock(&iommu->lock);
+
+ if (ret) {
+ spin_unlock_irqrestore(&device_domain_lock, flags);
+ free_devinfo_mem(info);
+ return NULL;
+ }
+
list_add(&info->link, &domain->devices);
list_add(&info->global, &device_domain_list);
if (dev)
dev->archdata.iommu = info;
spin_unlock_irqrestore(&device_domain_lock, flags);
+ if (dev && domain_context_mapping(domain, dev)) {
+ pr_err("Domain context map for %s failed\n", dev_name(dev));
+ dmar_remove_one_dev_info(domain, dev);
+ return NULL;
+ }
+
return domain;
}
/* domain is initialized */
static struct dmar_domain *get_domain_for_dev(struct device *dev, int gaw)
{
+ struct device_domain_info *info = NULL;
struct dmar_domain *domain, *tmp;
struct intel_iommu *iommu;
- struct device_domain_info *info;
- u16 dma_alias;
+ u16 req_id, dma_alias;
unsigned long flags;
u8 bus, devfn;
if (!iommu)
return NULL;
+ req_id = ((u16)bus << 8) | devfn;
+
if (dev_is_pci(dev)) {
struct pci_dev *pdev = to_pci_dev(dev);
domain = alloc_domain(0);
if (!domain)
return NULL;
- domain->id = iommu_attach_domain(domain, iommu);
- if (domain->id < 0) {
- free_domain_mem(domain);
- return NULL;
- }
- domain_attach_iommu(domain, iommu);
- if (domain_init(domain, gaw)) {
+ if (domain_init(domain, iommu, gaw)) {
domain_exit(domain);
return NULL;
}
/* register PCI DMA alias device */
- if (dev_is_pci(dev)) {
- tmp = dmar_insert_dev_info(iommu, PCI_BUS_NUM(dma_alias),
- dma_alias & 0xff, NULL, domain);
+ if (req_id != dma_alias && dev_is_pci(dev)) {
+ tmp = dmar_insert_one_dev_info(iommu, PCI_BUS_NUM(dma_alias),
+ dma_alias & 0xff, NULL, domain);
if (!tmp || tmp != domain) {
domain_exit(domain);
}
found_domain:
- tmp = dmar_insert_dev_info(iommu, bus, devfn, dev, domain);
+ tmp = dmar_insert_one_dev_info(iommu, bus, devfn, dev, domain);
if (!tmp || tmp != domain) {
domain_exit(domain);
return domain;
}
-static int iommu_identity_mapping;
-#define IDENTMAP_ALL 1
-#define IDENTMAP_GFX 2
-#define IDENTMAP_AZALIA 4
-
static int iommu_domain_identity_map(struct dmar_domain *domain,
unsigned long long start,
unsigned long long end)
if (!reserve_iova(&domain->iovad, dma_to_mm_pfn(first_vpfn),
dma_to_mm_pfn(last_vpfn))) {
- printk(KERN_ERR "IOMMU: reserve iova failed\n");
+ pr_err("Reserving iova failed\n");
return -ENOMEM;
}
- pr_debug("Mapping reserved region %llx-%llx for domain %d\n",
- start, end, domain->id);
+ pr_debug("Mapping reserved region %llx-%llx\n", start, end);
/*
* RMRR range might have overlap with physical memory range,
* clear it first
DMA_PTE_READ|DMA_PTE_WRITE);
}
-static int iommu_prepare_identity_map(struct device *dev,
- unsigned long long start,
- unsigned long long end)
+static int domain_prepare_identity_map(struct device *dev,
+ struct dmar_domain *domain,
+ unsigned long long start,
+ unsigned long long end)
{
- struct dmar_domain *domain;
- int ret;
-
- domain = get_domain_for_dev(dev, DEFAULT_DOMAIN_ADDRESS_WIDTH);
- if (!domain)
- return -ENOMEM;
-
/* For _hardware_ passthrough, don't bother. But for software
passthrough, we do it anyway -- it may indicate a memory
range which is reserved in E820, so which didn't get set
up to start with in si_domain */
if (domain == si_domain && hw_pass_through) {
- printk("Ignoring identity map for HW passthrough device %s [0x%Lx - 0x%Lx]\n",
- dev_name(dev), start, end);
+ pr_warn("Ignoring identity map for HW passthrough device %s [0x%Lx - 0x%Lx]\n",
+ dev_name(dev), start, end);
return 0;
}
- printk(KERN_INFO
- "IOMMU: Setting identity map for device %s [0x%Lx - 0x%Lx]\n",
- dev_name(dev), start, end);
-
+ pr_info("Setting identity map for device %s [0x%Lx - 0x%Lx]\n",
+ dev_name(dev), start, end);
+
if (end < start) {
WARN(1, "Your BIOS is broken; RMRR ends before it starts!\n"
"BIOS vendor: %s; Ver: %s; Product Version: %s\n",
dmi_get_system_info(DMI_BIOS_VENDOR),
dmi_get_system_info(DMI_BIOS_VERSION),
dmi_get_system_info(DMI_PRODUCT_VERSION));
- ret = -EIO;
- goto error;
+ return -EIO;
}
if (end >> agaw_to_width(domain->agaw)) {
dmi_get_system_info(DMI_BIOS_VENDOR),
dmi_get_system_info(DMI_BIOS_VERSION),
dmi_get_system_info(DMI_PRODUCT_VERSION));
- ret = -EIO;
- goto error;
+ return -EIO;
}
- ret = iommu_domain_identity_map(domain, start, end);
- if (ret)
- goto error;
+ return iommu_domain_identity_map(domain, start, end);
+}
+
+static int iommu_prepare_identity_map(struct device *dev,
+ unsigned long long start,
+ unsigned long long end)
+{
+ struct dmar_domain *domain;
+ int ret;
+
+ domain = get_domain_for_dev(dev, DEFAULT_DOMAIN_ADDRESS_WIDTH);
+ if (!domain)
+ return -ENOMEM;
- /* context entry init */
- ret = domain_context_mapping(domain, dev, CONTEXT_TT_MULTI_LEVEL);
+ ret = domain_prepare_identity_map(dev, domain, start, end);
if (ret)
- goto error;
-
- return 0;
+ domain_exit(domain);
- error:
- domain_exit(domain);
return ret;
}
if (!pdev)
return;
- printk(KERN_INFO "IOMMU: Prepare 0-16MiB unity mapping for LPC\n");
+ pr_info("Prepare 0-16MiB unity mapping for LPC\n");
ret = iommu_prepare_identity_map(&pdev->dev, 0, 16*1024*1024 - 1);
if (ret)
- printk(KERN_ERR "IOMMU: Failed to create 0-16MiB identity map; "
- "floppy might not work\n");
+ pr_err("Failed to create 0-16MiB identity map - floppy might not work\n");
pci_dev_put(pdev);
}
static int __init si_domain_init(int hw)
{
- struct dmar_drhd_unit *drhd;
- struct intel_iommu *iommu;
int nid, ret = 0;
- bool first = true;
si_domain = alloc_domain(DOMAIN_FLAG_STATIC_IDENTITY);
if (!si_domain)
return -EFAULT;
- for_each_active_iommu(iommu, drhd) {
- ret = iommu_attach_domain(si_domain, iommu);
- if (ret < 0) {
- domain_exit(si_domain);
- return -EFAULT;
- } else if (first) {
- si_domain->id = ret;
- first = false;
- } else if (si_domain->id != ret) {
- domain_exit(si_domain);
- return -EFAULT;
- }
- domain_attach_iommu(si_domain, iommu);
- }
-
if (md_domain_init(si_domain, DEFAULT_DOMAIN_ADDRESS_WIDTH)) {
domain_exit(si_domain);
return -EFAULT;
}
- pr_debug("IOMMU: identity mapping domain is domain %d\n",
- si_domain->id);
+ pr_debug("Identity mapping domain allocated\n");
if (hw)
return 0;
return 0;
}
-static int domain_add_dev_info(struct dmar_domain *domain,
- struct device *dev, int translation)
+static int domain_add_dev_info(struct dmar_domain *domain, struct device *dev)
{
struct dmar_domain *ndomain;
struct intel_iommu *iommu;
u8 bus, devfn;
- int ret;
iommu = device_to_iommu(dev, &bus, &devfn);
if (!iommu)
return -ENODEV;
- ndomain = dmar_insert_dev_info(iommu, bus, devfn, dev, domain);
+ ndomain = dmar_insert_one_dev_info(iommu, bus, devfn, dev, domain);
if (ndomain != domain)
return -EBUSY;
- ret = domain_context_mapping(domain, dev, translation);
- if (ret) {
- domain_remove_one_dev_info(domain, dev);
- return ret;
- }
-
return 0;
}
if (!iommu_should_identity_map(dev, 1))
return 0;
- ret = domain_add_dev_info(si_domain, dev,
- hw ? CONTEXT_TT_PASS_THROUGH :
- CONTEXT_TT_MULTI_LEVEL);
+ ret = domain_add_dev_info(si_domain, dev);
if (!ret)
- pr_info("IOMMU: %s identity mapping for device %s\n",
- hw ? "hardware" : "software", dev_name(dev));
+ pr_info("%s identity mapping for device %s\n",
+ hw ? "Hardware" : "Software", dev_name(dev));
else if (ret == -ENODEV)
/* device not associated with an iommu */
ret = 0;
int i;
int ret = 0;
- ret = si_domain_init(hw);
- if (ret)
- return -EFAULT;
-
for_each_pci_dev(pdev) {
ret = dev_prepare_static_identity_mapping(&pdev->dev, hw);
if (ret)
if (dev->bus != &acpi_bus_type)
continue;
-
+
adev= to_acpi_device(dev);
mutex_lock(&adev->physical_node_lock);
list_for_each_entry(pn, &adev->physical_node_list, node) {
*/
iommu->flush.flush_context = __iommu_flush_context;
iommu->flush.flush_iotlb = __iommu_flush_iotlb;
- pr_info("IOMMU: %s using Register based invalidation\n",
+ pr_info("%s: Using Register based invalidation\n",
iommu->name);
} else {
iommu->flush.flush_context = qi_flush_context;
iommu->flush.flush_iotlb = qi_flush_iotlb;
- pr_info("IOMMU: %s using Queued invalidation\n", iommu->name);
+ pr_info("%s: Using Queued invalidation\n", iommu->name);
+ }
+}
+
+static int copy_context_table(struct intel_iommu *iommu,
+ struct root_entry *old_re,
+ struct context_entry **tbl,
+ int bus, bool ext)
+{
+ int tbl_idx, pos = 0, idx, devfn, ret = 0, did;
+ struct context_entry *new_ce = NULL, ce;
+ struct context_entry *old_ce = NULL;
+ struct root_entry re;
+ phys_addr_t old_ce_phys;
+
+ tbl_idx = ext ? bus * 2 : bus;
+ memcpy(&re, old_re, sizeof(re));
+
+ for (devfn = 0; devfn < 256; devfn++) {
+ /* First calculate the correct index */
+ idx = (ext ? devfn * 2 : devfn) % 256;
+
+ if (idx == 0) {
+ /* First save what we may have and clean up */
+ if (new_ce) {
+ tbl[tbl_idx] = new_ce;
+ __iommu_flush_cache(iommu, new_ce,
+ VTD_PAGE_SIZE);
+ pos = 1;
+ }
+
+ if (old_ce)
+ iounmap(old_ce);
+
+ ret = 0;
+ if (devfn < 0x80)
+ old_ce_phys = root_entry_lctp(&re);
+ else
+ old_ce_phys = root_entry_uctp(&re);
+
+ if (!old_ce_phys) {
+ if (ext && devfn == 0) {
+ /* No LCTP, try UCTP */
+ devfn = 0x7f;
+ continue;
+ } else {
+ goto out;
+ }
+ }
+
+ ret = -ENOMEM;
+ old_ce = memremap(old_ce_phys, PAGE_SIZE,
+ MEMREMAP_WB);
+ if (!old_ce)
+ goto out;
+
+ new_ce = alloc_pgtable_page(iommu->node);
+ if (!new_ce)
+ goto out_unmap;
+
+ ret = 0;
+ }
+
+ /* Now copy the context entry */
+ memcpy(&ce, old_ce + idx, sizeof(ce));
+
+ if (!__context_present(&ce))
+ continue;
+
+ did = context_domain_id(&ce);
+ if (did >= 0 && did < cap_ndoms(iommu->cap))
+ set_bit(did, iommu->domain_ids);
+
+ /*
+ * We need a marker for copied context entries. This
+ * marker needs to work for the old format as well as
+ * for extended context entries.
+ *
+ * Bit 67 of the context entry is used. In the old
+ * format this bit is available to software, in the
+ * extended format it is the PGE bit, but PGE is ignored
+ * by HW if PASIDs are disabled (and thus still
+ * available).
+ *
+ * So disable PASIDs first and then mark the entry
+ * copied. This means that we don't copy PASID
+ * translations from the old kernel, but this is fine as
+ * faults there are not fatal.
+ */
+ context_clear_pasid_enable(&ce);
+ context_set_copied(&ce);
+
+ new_ce[idx] = ce;
+ }
+
+ tbl[tbl_idx + pos] = new_ce;
+
+ __iommu_flush_cache(iommu, new_ce, VTD_PAGE_SIZE);
+
+out_unmap:
+ memunmap(old_ce);
+
+out:
+ return ret;
+}
+
+static int copy_translation_tables(struct intel_iommu *iommu)
+{
+ struct context_entry **ctxt_tbls;
+ struct root_entry *old_rt;
+ phys_addr_t old_rt_phys;
+ int ctxt_table_entries;
+ unsigned long flags;
+ u64 rtaddr_reg;
+ int bus, ret;
+ bool new_ext, ext;
+
+ rtaddr_reg = dmar_readq(iommu->reg + DMAR_RTADDR_REG);
+ ext = !!(rtaddr_reg & DMA_RTADDR_RTT);
+ new_ext = !!ecap_ecs(iommu->ecap);
+
+ /*
+ * The RTT bit can only be changed when translation is disabled,
+ * but disabling translation means to open a window for data
+ * corruption. So bail out and don't copy anything if we would
+ * have to change the bit.
+ */
+ if (new_ext != ext)
+ return -EINVAL;
+
+ old_rt_phys = rtaddr_reg & VTD_PAGE_MASK;
+ if (!old_rt_phys)
+ return -EINVAL;
+
+ old_rt = memremap(old_rt_phys, PAGE_SIZE, MEMREMAP_WB);
+ if (!old_rt)
+ return -ENOMEM;
+
+ /* This is too big for the stack - allocate it from slab */
+ ctxt_table_entries = ext ? 512 : 256;
+ ret = -ENOMEM;
+ ctxt_tbls = kzalloc(ctxt_table_entries * sizeof(void *), GFP_KERNEL);
+ if (!ctxt_tbls)
+ goto out_unmap;
+
+ for (bus = 0; bus < 256; bus++) {
+ ret = copy_context_table(iommu, &old_rt[bus],
+ ctxt_tbls, bus, ext);
+ if (ret) {
+ pr_err("%s: Failed to copy context table for bus %d\n",
+ iommu->name, bus);
+ continue;
+ }
+ }
+
+ spin_lock_irqsave(&iommu->lock, flags);
+
+ /* Context tables are copied, now write them to the root_entry table */
+ for (bus = 0; bus < 256; bus++) {
+ int idx = ext ? bus * 2 : bus;
+ u64 val;
+
+ if (ctxt_tbls[idx]) {
+ val = virt_to_phys(ctxt_tbls[idx]) | 1;
+ iommu->root_entry[bus].lo = val;
+ }
+
+ if (!ext || !ctxt_tbls[idx + 1])
+ continue;
+
+ val = virt_to_phys(ctxt_tbls[idx + 1]) | 1;
+ iommu->root_entry[bus].hi = val;
}
+
+ spin_unlock_irqrestore(&iommu->lock, flags);
+
+ kfree(ctxt_tbls);
+
+ __iommu_flush_cache(iommu, iommu->root_entry, PAGE_SIZE);
+
+ ret = 0;
+
+out_unmap:
+ memunmap(old_rt);
+
+ return ret;
}
static int __init init_dmars(void)
{
struct dmar_drhd_unit *drhd;
struct dmar_rmrr_unit *rmrr;
+ bool copied_tables = false;
struct device *dev;
struct intel_iommu *iommu;
int i, ret;
g_num_of_iommus++;
continue;
}
- printk_once(KERN_ERR "intel-iommu: exceeded %d IOMMUs\n",
- DMAR_UNITS_SUPPORTED);
+ pr_err_once("Exceeded %d IOMMUs\n", DMAR_UNITS_SUPPORTED);
}
/* Preallocate enough resources for IOMMU hot-addition */
g_iommus = kcalloc(g_num_of_iommus, sizeof(struct intel_iommu *),
GFP_KERNEL);
if (!g_iommus) {
- printk(KERN_ERR "Allocating global iommu array failed\n");
+ pr_err("Allocating global iommu array failed\n");
ret = -ENOMEM;
goto error;
}
for_each_active_iommu(iommu, drhd) {
g_iommus[iommu->seq_id] = iommu;
+ intel_iommu_init_qi(iommu);
+
ret = iommu_init_domains(iommu);
if (ret)
goto free_iommu;
+ init_translation_status(iommu);
+
+ if (translation_pre_enabled(iommu) && !is_kdump_kernel()) {
+ iommu_disable_translation(iommu);
+ clear_translation_pre_enabled(iommu);
+ pr_warn("Translation was enabled for %s but we are not in kdump mode\n",
+ iommu->name);
+ }
+
/*
* TBD:
* we could share the same root & context tables
ret = iommu_alloc_root_entry(iommu);
if (ret)
goto free_iommu;
+
+ if (translation_pre_enabled(iommu)) {
+ pr_info("Translation already enabled - trying to copy translation structures\n");
+
+ ret = copy_translation_tables(iommu);
+ if (ret) {
+ /*
+ * We found the IOMMU with translation
+ * enabled - but failed to copy over the
+ * old root-entry table. Try to proceed
+ * by disabling translation now and
+ * allocating a clean root-entry table.
+ * This might cause DMAR faults, but
+ * probably the dump will still succeed.
+ */
+ pr_err("Failed to copy translation tables from previous kernel for %s\n",
+ iommu->name);
+ iommu_disable_translation(iommu);
+ clear_translation_pre_enabled(iommu);
+ } else {
+ pr_info("Copied translation tables from previous kernel for %s\n",
+ iommu->name);
+ copied_tables = true;
+ }
+ }
+
+ iommu_flush_write_buffer(iommu);
+ iommu_set_root_entry(iommu);
+ iommu->flush.flush_context(iommu, 0, 0, 0, DMA_CCMD_GLOBAL_INVL);
+ iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_GLOBAL_FLUSH);
+
if (!ecap_pass_through(iommu->ecap))
hw_pass_through = 0;
+#ifdef CONFIG_INTEL_IOMMU_SVM
+ if (pasid_enabled(iommu))
+ intel_svm_alloc_pasid_tables(iommu);
+#endif
}
- for_each_active_iommu(iommu, drhd)
- intel_iommu_init_qi(iommu);
-
if (iommu_pass_through)
iommu_identity_mapping |= IDENTMAP_ALL;
iommu_identity_mapping |= IDENTMAP_GFX;
#endif
+ if (iommu_identity_mapping) {
+ ret = si_domain_init(hw_pass_through);
+ if (ret)
+ goto free_iommu;
+ }
+
check_tylersburg_isoch();
+ /*
+ * If we copied translations from a previous kernel in the kdump
+ * case, we can not assign the devices to domains now, as that
+ * would eliminate the old mappings. So skip this part and defer
+ * the assignment to device driver initialization time.
+ */
+ if (copied_tables)
+ goto domains_done;
+
/*
* If pass through is not set or not enabled, setup context entries for
* identity mappings for rmrr, gfx, and isa and may fall back to static
if (iommu_identity_mapping) {
ret = iommu_prepare_static_identity_mapping(hw_pass_through);
if (ret) {
- printk(KERN_CRIT "Failed to setup IOMMU pass-through\n");
+ pr_crit("Failed to setup IOMMU pass-through\n");
goto free_iommu;
}
}
* endfor
* endfor
*/
- printk(KERN_INFO "IOMMU: Setting RMRR:\n");
+ pr_info("Setting RMRR:\n");
for_each_rmrr_units(rmrr) {
/* some BIOS lists non-exist devices in DMAR table. */
for_each_active_dev_scope(rmrr->devices, rmrr->devices_cnt,
i, dev) {
ret = iommu_prepare_rmrr_dev(rmrr, dev);
if (ret)
- printk(KERN_ERR
- "IOMMU: mapping reserved region failed\n");
+ pr_err("Mapping reserved region failed\n");
}
}
iommu_prepare_isa();
+domains_done:
+
/*
* for each drhd
* enable fault log
iommu_flush_write_buffer(iommu);
+#ifdef CONFIG_INTEL_IOMMU_SVM
+ if (pasid_enabled(iommu) && ecap_prs(iommu->ecap)) {
+ ret = intel_svm_enable_prq(iommu);
+ if (ret)
+ goto free_iommu;
+ }
+#endif
ret = dmar_set_interrupt(iommu);
if (ret)
goto free_iommu;
- iommu_set_root_entry(iommu);
+ if (!translation_pre_enabled(iommu))
+ iommu_enable_translation(iommu);
- iommu->flush.flush_context(iommu, 0, 0, 0, DMA_CCMD_GLOBAL_INVL);
- iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_GLOBAL_FLUSH);
- iommu_enable_translation(iommu);
iommu_disable_protect_mem_regions(iommu);
}
/* Restrict dma_mask to the width that the iommu can handle */
dma_mask = min_t(uint64_t, DOMAIN_MAX_ADDR(domain->gaw), dma_mask);
+ /* Ensure we reserve the whole size-aligned region */
+ nrpages = __roundup_pow_of_two(nrpages);
if (!dmar_forcedac && dma_mask > DMA_BIT_MASK(32)) {
/*
}
iova = alloc_iova(&domain->iovad, nrpages, IOVA_PFN(dma_mask), 1);
if (unlikely(!iova)) {
- printk(KERN_ERR "Allocating %ld-page iova for %s failed",
+ pr_err("Allocating %ld-page iova for %s failed",
nrpages, dev_name(dev));
return NULL;
}
static struct dmar_domain *__get_valid_domain_for_dev(struct device *dev)
{
+ struct dmar_rmrr_unit *rmrr;
struct dmar_domain *domain;
- int ret;
+ struct device *i_dev;
+ int i, ret;
domain = get_domain_for_dev(dev, DEFAULT_DOMAIN_ADDRESS_WIDTH);
if (!domain) {
- printk(KERN_ERR "Allocating domain for %s failed",
+ pr_err("Allocating domain for %s failed\n",
dev_name(dev));
return NULL;
}
- /* make sure context mapping is ok */
- if (unlikely(!domain_context_mapped(dev))) {
- ret = domain_context_mapping(domain, dev, CONTEXT_TT_MULTI_LEVEL);
- if (ret) {
- printk(KERN_ERR "Domain context map for %s failed",
- dev_name(dev));
- return NULL;
+ /* We have a new domain - setup possible RMRRs for the device */
+ rcu_read_lock();
+ for_each_rmrr_units(rmrr) {
+ for_each_active_dev_scope(rmrr->devices, rmrr->devices_cnt,
+ i, i_dev) {
+ if (i_dev != dev)
+ continue;
+
+ ret = domain_prepare_identity_map(dev, domain,
+ rmrr->base_address,
+ rmrr->end_address);
+ if (ret)
+ dev_err(dev, "Mapping reserved region failed\n");
}
}
+ rcu_read_unlock();
return domain;
}
* 32 bit DMA is removed from si_domain and fall back
* to non-identity mapping.
*/
- domain_remove_one_dev_info(si_domain, dev);
- printk(KERN_INFO "32bit %s uses non-identity mapping\n",
- dev_name(dev));
+ dmar_remove_one_dev_info(si_domain, dev);
+ pr_info("32bit %s uses non-identity mapping\n",
+ dev_name(dev));
return 0;
}
} else {
*/
if (iommu_should_identity_map(dev, 0)) {
int ret;
- ret = domain_add_dev_info(si_domain, dev,
- hw_pass_through ?
- CONTEXT_TT_PASS_THROUGH :
- CONTEXT_TT_MULTI_LEVEL);
+ ret = domain_add_dev_info(si_domain, dev);
if (!ret) {
- printk(KERN_INFO "64bit %s uses identity mapping\n",
- dev_name(dev));
+ pr_info("64bit %s uses identity mapping\n",
+ dev_name(dev));
return 1;
}
}
/* it's a non-present to present mapping. Only flush if caching mode */
if (cap_caching_mode(iommu->cap))
- iommu_flush_iotlb_psi(iommu, domain->id, mm_to_dma_pfn(iova->pfn_lo), size, 0, 1);
+ iommu_flush_iotlb_psi(iommu, domain,
+ mm_to_dma_pfn(iova->pfn_lo),
+ size, 0, 1);
else
iommu_flush_write_buffer(iommu);
error:
if (iova)
__free_iova(&domain->iovad, iova);
- printk(KERN_ERR"Device %s request: %zx@%llx dir %d --- failed\n",
+ pr_err("Device %s request: %zx@%llx dir %d --- failed\n",
dev_name(dev), size, (unsigned long long)paddr, dir);
return 0;
}
/* On real hardware multiple invalidations are expensive */
if (cap_caching_mode(iommu->cap))
- iommu_flush_iotlb_psi(iommu, domain->id,
+ iommu_flush_iotlb_psi(iommu, domain,
iova->pfn_lo, iova_size(iova),
!deferred_flush[i].freelist[j], 0);
else {
freelist = domain_unmap(domain, start_pfn, last_pfn);
if (intel_iommu_strict) {
- iommu_flush_iotlb_psi(iommu, domain->id, start_pfn,
+ iommu_flush_iotlb_psi(iommu, domain, start_pfn,
last_pfn - start_pfn + 1, !freelist, 0);
/* free iova */
__free_iova(&domain->iovad, iova);
flags |= GFP_DMA32;
}
- if (flags & __GFP_WAIT) {
+ if (gfpflags_allow_blocking(flags)) {
unsigned int count = size >> PAGE_SHIFT;
page = dma_alloc_from_contiguous(dev, count, order);
/* it's a non-present to present mapping. Only flush if caching mode */
if (cap_caching_mode(iommu->cap))
- iommu_flush_iotlb_psi(iommu, domain->id, start_vpfn, size, 0, 1);
+ iommu_flush_iotlb_psi(iommu, domain, start_vpfn, size, 0, 1);
else
iommu_flush_write_buffer(iommu);
NULL);
if (!iommu_domain_cache) {
- printk(KERN_ERR "Couldn't create iommu_domain cache\n");
+ pr_err("Couldn't create iommu_domain cache\n");
ret = -ENOMEM;
}
SLAB_HWCACHE_ALIGN,
NULL);
if (!iommu_devinfo_cache) {
- printk(KERN_ERR "Couldn't create devinfo cache\n");
+ pr_err("Couldn't create devinfo cache\n");
ret = -ENOMEM;
}
static int __init iommu_init_mempool(void)
{
int ret;
- ret = iommu_iova_cache_init();
+ ret = iova_cache_get();
if (ret)
return ret;
kmem_cache_destroy(iommu_domain_cache);
domain_error:
- iommu_iova_cache_destroy();
+ iova_cache_put();
return -ENOMEM;
}
{
kmem_cache_destroy(iommu_devinfo_cache);
kmem_cache_destroy(iommu_domain_cache);
- iommu_iova_cache_destroy();
+ iova_cache_put();
}
static void quirk_ioat_snb_local_iommu(struct pci_dev *pdev)
return 0;
if (hw_pass_through && !ecap_pass_through(iommu->ecap)) {
- pr_warn("IOMMU: %s doesn't support hardware pass through.\n",
+ pr_warn("%s: Doesn't support hardware pass through.\n",
iommu->name);
return -ENXIO;
}
if (!ecap_sc_support(iommu->ecap) &&
domain_update_iommu_snooping(iommu)) {
- pr_warn("IOMMU: %s doesn't support snooping.\n",
+ pr_warn("%s: Doesn't support snooping.\n",
iommu->name);
return -ENXIO;
}
sp = domain_update_iommu_superpage(iommu) - 1;
if (sp >= 0 && !(cap_super_page_val(iommu->cap) & (1 << sp))) {
- pr_warn("IOMMU: %s doesn't support large page.\n",
+ pr_warn("%s: Doesn't support large page.\n",
iommu->name);
return -ENXIO;
}
if (ret)
goto out;
+#ifdef CONFIG_INTEL_IOMMU_SVM
+ if (pasid_enabled(iommu))
+ intel_svm_alloc_pasid_tables(iommu);
+#endif
+
if (dmaru->ignored) {
/*
* we always have to disable PMRs or DMA may fail on this device
intel_iommu_init_qi(iommu);
iommu_flush_write_buffer(iommu);
+
+#ifdef CONFIG_INTEL_IOMMU_SVM
+ if (pasid_enabled(iommu) && ecap_prs(iommu->ecap)) {
+ ret = intel_svm_enable_prq(iommu);
+ if (ret)
+ goto disable_iommu;
+ }
+#endif
ret = dmar_set_interrupt(iommu);
if (ret)
goto disable_iommu;
iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_GLOBAL_FLUSH);
iommu_enable_translation(iommu);
- if (si_domain) {
- ret = iommu_attach_domain(si_domain, iommu);
- if (ret < 0 || si_domain->id != ret)
- goto disable_iommu;
- domain_attach_iommu(si_domain, iommu);
- }
-
iommu_disable_protect_mem_regions(iommu);
return 0;
dev = pci_physfn(dev);
for (bus = dev->bus; bus; bus = bus->parent) {
bridge = bus->self;
- if (!bridge || !pci_is_pcie(bridge) ||
+ /* If it's an integrated device, allow ATS */
+ if (!bridge)
+ return 1;
+ /* Connected via non-PCIe: no ATS */
+ if (!pci_is_pcie(bridge) ||
pci_pcie_type(bridge) == PCI_EXP_TYPE_PCI_BRIDGE)
return 0;
+ /* If we found the root port, look it up in the ATSR */
if (pci_pcie_type(bridge) == PCI_EXP_TYPE_ROOT_PORT)
break;
}
- if (!bridge)
- return 0;
rcu_read_lock();
list_for_each_entry_rcu(atsru, &dmar_atsr_units, list) {
rmrru->devices_cnt);
if(ret < 0)
return ret;
- } else if (info->event == BUS_NOTIFY_DEL_DEVICE) {
+ } else if (info->event == BUS_NOTIFY_REMOVED_DEVICE) {
dmar_remove_dev_scope(info, rmrr->segment,
rmrru->devices, rmrru->devices_cnt);
}
break;
else if(ret < 0)
return ret;
- } else if (info->event == BUS_NOTIFY_DEL_DEVICE) {
+ } else if (info->event == BUS_NOTIFY_REMOVED_DEVICE) {
if (dmar_remove_dev_scope(info, atsr->segment,
atsru->devices, atsru->devices_cnt))
break;
if (!domain)
return 0;
- down_read(&dmar_global_lock);
- domain_remove_one_dev_info(domain, dev);
+ dmar_remove_one_dev_info(domain, dev);
if (!domain_type_is_vm_or_si(domain) && list_empty(&domain->devices))
domain_exit(domain);
- up_read(&dmar_global_lock);
return 0;
}
start = mhp->start_pfn << PAGE_SHIFT;
end = ((mhp->start_pfn + mhp->nr_pages) << PAGE_SHIFT) - 1;
if (iommu_domain_identity_map(si_domain, start, end)) {
- pr_warn("dmar: failed to build identity map for [%llx-%llx]\n",
+ pr_warn("Failed to build identity map for [%llx-%llx]\n",
start, end);
return NOTIFY_BAD;
}
iova = find_iova(&si_domain->iovad, start_vpfn);
if (iova == NULL) {
- pr_debug("dmar: failed get IOVA for PFN %lx\n",
+ pr_debug("Failed get IOVA for PFN %lx\n",
start_vpfn);
break;
}
iova = split_and_remove_iova(&si_domain->iovad, iova,
start_vpfn, last_vpfn);
if (iova == NULL) {
- pr_warn("dmar: failed to split IOVA PFN [%lx-%lx]\n",
+ pr_warn("Failed to split IOVA PFN [%lx-%lx]\n",
start_vpfn, last_vpfn);
return NOTIFY_BAD;
}
rcu_read_lock();
for_each_active_iommu(iommu, drhd)
- iommu_flush_iotlb_psi(iommu, si_domain->id,
+ iommu_flush_iotlb_psi(iommu, si_domain,
iova->pfn_lo, iova_size(iova),
!freelist, 0);
rcu_read_unlock();
}
static DEVICE_ATTR(ecap, S_IRUGO, intel_iommu_show_ecap, NULL);
+static ssize_t intel_iommu_show_ndoms(struct device *dev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct intel_iommu *iommu = dev_get_drvdata(dev);
+ return sprintf(buf, "%ld\n", cap_ndoms(iommu->cap));
+}
+static DEVICE_ATTR(domains_supported, S_IRUGO, intel_iommu_show_ndoms, NULL);
+
+static ssize_t intel_iommu_show_ndoms_used(struct device *dev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct intel_iommu *iommu = dev_get_drvdata(dev);
+ return sprintf(buf, "%d\n", bitmap_weight(iommu->domain_ids,
+ cap_ndoms(iommu->cap)));
+}
+static DEVICE_ATTR(domains_used, S_IRUGO, intel_iommu_show_ndoms_used, NULL);
+
static struct attribute *intel_iommu_attrs[] = {
&dev_attr_version.attr,
&dev_attr_address.attr,
&dev_attr_cap.attr,
&dev_attr_ecap.attr,
+ &dev_attr_domains_supported.attr,
+ &dev_attr_domains_used.attr,
NULL,
};
goto out_free_dmar;
}
- /*
- * Disable translation if already enabled prior to OS handover.
- */
- for_each_active_iommu(iommu, drhd)
- if (iommu->gcmd & DMA_GCMD_TE)
- iommu_disable_translation(iommu);
-
if (dmar_dev_scope_init() < 0) {
if (force_on)
panic("tboot: Failed to initialize DMAR device scope\n");
goto out_free_dmar;
if (list_empty(&dmar_rmrr_units))
- printk(KERN_INFO "DMAR: No RMRR found\n");
+ pr_info("No RMRR found\n");
if (list_empty(&dmar_atsr_units))
- printk(KERN_INFO "DMAR: No ATSR found\n");
+ pr_info("No ATSR found\n");
if (dmar_init_reserved_ranges()) {
if (force_on)
if (ret) {
if (force_on)
panic("tboot: Failed to initialize DMARs\n");
- printk(KERN_ERR "IOMMU: dmar init failed\n");
+ pr_err("Initialization failed\n");
goto out_free_reserved_range;
}
up_write(&dmar_global_lock);
- printk(KERN_INFO
- "PCI-DMA: Intel(R) Virtualization Technology for Directed I/O\n");
+ pr_info("Intel(R) Virtualization Technology for Directed I/O\n");
init_timer(&unmap_timer);
#ifdef CONFIG_SWIOTLB
for_each_active_iommu(iommu, drhd)
iommu->iommu_dev = iommu_device_create(NULL, iommu,
intel_iommu_groups,
- iommu->name);
+ "%s", iommu->name);
bus_set_iommu(&pci_bus_type, &intel_iommu_ops);
bus_register_notifier(&pci_bus_type, &device_nb);
return ret;
}
-static int iommu_detach_dev_cb(struct pci_dev *pdev, u16 alias, void *opaque)
+static int domain_context_clear_one_cb(struct pci_dev *pdev, u16 alias, void *opaque)
{
struct intel_iommu *iommu = opaque;
- iommu_detach_dev(iommu, PCI_BUS_NUM(alias), alias & 0xff);
+ domain_context_clear_one(iommu, PCI_BUS_NUM(alias), alias & 0xff);
return 0;
}
* devices, unbinding the driver from any one of them will possibly leave
* the others unable to operate.
*/
-static void iommu_detach_dependent_devices(struct intel_iommu *iommu,
- struct device *dev)
+static void domain_context_clear(struct intel_iommu *iommu, struct device *dev)
{
if (!iommu || !dev || !dev_is_pci(dev))
return;
- pci_for_each_dma_alias(to_pci_dev(dev), &iommu_detach_dev_cb, iommu);
+ pci_for_each_dma_alias(to_pci_dev(dev), &domain_context_clear_one_cb, iommu);
}
-static void domain_remove_one_dev_info(struct dmar_domain *domain,
- struct device *dev)
+static void __dmar_remove_one_dev_info(struct device_domain_info *info)
{
- struct device_domain_info *info, *tmp;
struct intel_iommu *iommu;
unsigned long flags;
- bool found = false;
- u8 bus, devfn;
- iommu = device_to_iommu(dev, &bus, &devfn);
- if (!iommu)
+ assert_spin_locked(&device_domain_lock);
+
+ if (WARN_ON(!info))
return;
- spin_lock_irqsave(&device_domain_lock, flags);
- list_for_each_entry_safe(info, tmp, &domain->devices, link) {
- if (info->iommu == iommu && info->bus == bus &&
- info->devfn == devfn) {
- unlink_domain_info(info);
- spin_unlock_irqrestore(&device_domain_lock, flags);
+ iommu = info->iommu;
- iommu_disable_dev_iotlb(info);
- iommu_detach_dev(iommu, info->bus, info->devfn);
- iommu_detach_dependent_devices(iommu, dev);
- free_devinfo_mem(info);
+ if (info->dev) {
+ iommu_disable_dev_iotlb(info);
+ domain_context_clear(iommu, info->dev);
+ }
- spin_lock_irqsave(&device_domain_lock, flags);
+ unlink_domain_info(info);
- if (found)
- break;
- else
- continue;
- }
+ spin_lock_irqsave(&iommu->lock, flags);
+ domain_detach_iommu(info->domain, iommu);
+ spin_unlock_irqrestore(&iommu->lock, flags);
- /* if there is no other devices under the same iommu
- * owned by this domain, clear this iommu in iommu_bmp
- * update iommu count and coherency
- */
- if (info->iommu == iommu)
- found = true;
- }
+ free_devinfo_mem(info);
+}
- spin_unlock_irqrestore(&device_domain_lock, flags);
+static void dmar_remove_one_dev_info(struct dmar_domain *domain,
+ struct device *dev)
+{
+ struct device_domain_info *info;
+ unsigned long flags;
- if (found == 0) {
- domain_detach_iommu(domain, iommu);
- if (!domain_type_is_vm_or_si(domain))
- iommu_detach_domain(domain, iommu);
- }
+ spin_lock_irqsave(&device_domain_lock, flags);
+ info = dev->archdata.iommu;
+ __dmar_remove_one_dev_info(info);
+ spin_unlock_irqrestore(&device_domain_lock, flags);
}
static int md_domain_init(struct dmar_domain *domain, int guest_width)
dmar_domain = alloc_domain(DOMAIN_FLAG_VIRTUAL_MACHINE);
if (!dmar_domain) {
- printk(KERN_ERR
- "intel_iommu_domain_init: dmar_domain == NULL\n");
+ pr_err("Can't allocate dmar_domain\n");
return NULL;
}
if (md_domain_init(dmar_domain, DEFAULT_DOMAIN_ADDRESS_WIDTH)) {
- printk(KERN_ERR
- "intel_iommu_domain_init() failed\n");
+ pr_err("Domain initialization failed\n");
domain_exit(dmar_domain);
return NULL;
}
old_domain = find_domain(dev);
if (old_domain) {
- if (domain_type_is_vm_or_si(dmar_domain))
- domain_remove_one_dev_info(old_domain, dev);
- else
- domain_remove_dev_info(old_domain);
+ rcu_read_lock();
+ dmar_remove_one_dev_info(old_domain, dev);
+ rcu_read_unlock();
if (!domain_type_is_vm_or_si(old_domain) &&
list_empty(&old_domain->devices))
addr_width = cap_mgaw(iommu->cap);
if (dmar_domain->max_addr > (1LL << addr_width)) {
- printk(KERN_ERR "%s: iommu width (%d) is not "
+ pr_err("%s: iommu width (%d) is not "
"sufficient for the mapped address (%llx)\n",
__func__, addr_width, dmar_domain->max_addr);
return -EFAULT;
dmar_domain->agaw--;
}
- return domain_add_dev_info(dmar_domain, dev, CONTEXT_TT_MULTI_LEVEL);
+ return domain_add_dev_info(dmar_domain, dev);
}
static void intel_iommu_detach_device(struct iommu_domain *domain,
struct device *dev)
{
- domain_remove_one_dev_info(to_dmar_domain(domain), dev);
+ dmar_remove_one_dev_info(to_dmar_domain(domain), dev);
}
static int intel_iommu_map(struct iommu_domain *domain,
/* check if minimum agaw is sufficient for mapped address */
end = __DOMAIN_MAX_ADDR(dmar_domain->gaw) + 1;
if (end < max_addr) {
- printk(KERN_ERR "%s: iommu width (%d) is not "
+ pr_err("%s: iommu width (%d) is not "
"sufficient for the mapped address (%llx)\n",
__func__, dmar_domain->gaw, max_addr);
return -EFAULT;
struct intel_iommu *iommu;
unsigned long start_pfn, last_pfn;
unsigned int npages;
- int iommu_id, num, ndomains, level = 0;
+ int iommu_id, level = 0;
/* Cope with horrid API which requires us to unmap more than the
size argument if it happens to be a large-page mapping. */
- if (!pfn_to_dma_pte(dmar_domain, iova >> VTD_PAGE_SHIFT, &level))
- BUG();
+ BUG_ON(!pfn_to_dma_pte(dmar_domain, iova >> VTD_PAGE_SHIFT, &level));
if (size < VTD_PAGE_SIZE << level_to_offset_bits(level))
size = VTD_PAGE_SIZE << level_to_offset_bits(level);
npages = last_pfn - start_pfn + 1;
- for_each_set_bit(iommu_id, dmar_domain->iommu_bmp, g_num_of_iommus) {
- iommu = g_iommus[iommu_id];
-
- /*
- * find bit position of dmar_domain
- */
- ndomains = cap_ndoms(iommu->cap);
- for_each_set_bit(num, iommu->domain_ids, ndomains) {
- if (iommu->domains[num] == dmar_domain)
- iommu_flush_iotlb_psi(iommu, num, start_pfn,
- npages, !freelist, 0);
- }
+ for_each_domain_iommu(iommu_id, dmar_domain) {
+ iommu = g_iommus[iommu_id];
+ iommu_flush_iotlb_psi(g_iommus[iommu_id], dmar_domain,
+ start_pfn, npages, !freelist, 0);
}
dma_free_pagelist(freelist);
iommu_device_unlink(iommu->iommu_dev, dev);
}
+#ifdef CONFIG_INTEL_IOMMU_SVM
+int intel_iommu_enable_pasid(struct intel_iommu *iommu, struct intel_svm_dev *sdev)
+{
+ struct device_domain_info *info;
+ struct context_entry *context;
+ struct dmar_domain *domain;
+ unsigned long flags;
+ u64 ctx_lo;
+ int ret;
+
+ domain = get_valid_domain_for_dev(sdev->dev);
+ if (!domain)
+ return -EINVAL;
+
+ spin_lock_irqsave(&device_domain_lock, flags);
+ spin_lock(&iommu->lock);
+
+ ret = -EINVAL;
+ info = sdev->dev->archdata.iommu;
+ if (!info || !info->pasid_supported)
+ goto out;
+
+ context = iommu_context_addr(iommu, info->bus, info->devfn, 0);
+ if (WARN_ON(!context))
+ goto out;
+
+ ctx_lo = context[0].lo;
+
+ sdev->did = domain->iommu_did[iommu->seq_id];
+ sdev->sid = PCI_DEVID(info->bus, info->devfn);
+
+ if (!(ctx_lo & CONTEXT_PASIDE)) {
+ context[1].hi = (u64)virt_to_phys(iommu->pasid_state_table);
+ context[1].lo = (u64)virt_to_phys(iommu->pasid_table) | ecap_pss(iommu->ecap);
+ wmb();
+ /* CONTEXT_TT_MULTI_LEVEL and CONTEXT_TT_DEV_IOTLB are both
+ * extended to permit requests-with-PASID if the PASIDE bit
+ * is set. which makes sense. For CONTEXT_TT_PASS_THROUGH,
+ * however, the PASIDE bit is ignored and requests-with-PASID
+ * are unconditionally blocked. Which makes less sense.
+ * So convert from CONTEXT_TT_PASS_THROUGH to one of the new
+ * "guest mode" translation types depending on whether ATS
+ * is available or not. Annoyingly, we can't use the new
+ * modes *unless* PASIDE is set. */
+ if ((ctx_lo & CONTEXT_TT_MASK) == (CONTEXT_TT_PASS_THROUGH << 2)) {
+ ctx_lo &= ~CONTEXT_TT_MASK;
+ if (info->ats_supported)
+ ctx_lo |= CONTEXT_TT_PT_PASID_DEV_IOTLB << 2;
+ else
+ ctx_lo |= CONTEXT_TT_PT_PASID << 2;
+ }
+ ctx_lo |= CONTEXT_PASIDE;
+ if (iommu->pasid_state_table)
+ ctx_lo |= CONTEXT_DINVE;
+ if (info->pri_supported)
+ ctx_lo |= CONTEXT_PRS;
+ context[0].lo = ctx_lo;
+ wmb();
+ iommu->flush.flush_context(iommu, sdev->did, sdev->sid,
+ DMA_CCMD_MASK_NOBIT,
+ DMA_CCMD_DEVICE_INVL);
+ }
+
+ /* Enable PASID support in the device, if it wasn't already */
+ if (!info->pasid_enabled)
+ iommu_enable_dev_iotlb(info);
+
+ if (info->ats_enabled) {
+ sdev->dev_iotlb = 1;
+ sdev->qdep = info->ats_qdep;
+ if (sdev->qdep >= QI_DEV_EIOTLB_MAX_INVS)
+ sdev->qdep = 0;
+ }
+ ret = 0;
+
+ out:
+ spin_unlock(&iommu->lock);
+ spin_unlock_irqrestore(&device_domain_lock, flags);
+
+ return ret;
+}
+
+struct intel_iommu *intel_svm_device_to_iommu(struct device *dev)
+{
+ struct intel_iommu *iommu;
+ u8 bus, devfn;
+
+ if (iommu_dummy(dev)) {
+ dev_warn(dev,
+ "No IOMMU translation for device; cannot enable SVM\n");
+ return NULL;
+ }
+
+ iommu = device_to_iommu(dev, &bus, &devfn);
+ if ((!iommu)) {
+ dev_err(dev, "No IOMMU for device; cannot enable SVM\n");
+ return NULL;
+ }
+
+ if (!iommu->pasid_table) {
+ dev_err(dev, "PASID not enabled on IOMMU; cannot enable SVM\n");
+ return NULL;
+ }
+
+ return iommu;
+}
+#endif /* CONFIG_INTEL_IOMMU_SVM */
+
static const struct iommu_ops intel_iommu_ops = {
.capable = intel_iommu_capable,
.domain_alloc = intel_iommu_domain_alloc,
.iova_to_phys = intel_iommu_iova_to_phys,
.add_device = intel_iommu_add_device,
.remove_device = intel_iommu_remove_device,
+ .device_group = pci_device_group,
.pgsize_bitmap = INTEL_IOMMU_PGSIZES,
};
static void quirk_iommu_g4x_gfx(struct pci_dev *dev)
{
/* G4x/GM45 integrated gfx dmar support is totally busted. */
- printk(KERN_INFO "DMAR: Disabling IOMMU for graphics on this chipset\n");
+ pr_info("Disabling IOMMU for graphics on this chipset\n");
dmar_map_gfx = 0;
}
* Mobile 4 Series Chipset neglects to set RWBF capability,
* but needs it. Same seems to hold for the desktop versions.
*/
- printk(KERN_INFO "DMAR: Forcing write-buffer flush capability\n");
+ pr_info("Forcing write-buffer flush capability\n");
rwbf_quirk = 1;
}
return;
if (!(ggc & GGC_MEMORY_VT_ENABLED)) {
- printk(KERN_INFO "DMAR: BIOS has allocated no shadow GTT; disabling IOMMU for graphics\n");
+ pr_info("BIOS has allocated no shadow GTT; disabling IOMMU for graphics\n");
dmar_map_gfx = 0;
} else if (dmar_map_gfx) {
/* we have to ensure the gfx device is idle before we flush */
- printk(KERN_INFO "DMAR: Disabling batched IOTLB flush on Ironlake\n");
+ pr_info("Disabling batched IOTLB flush on Ironlake\n");
intel_iommu_strict = 1;
}
}
iommu_identity_mapping |= IDENTMAP_AZALIA;
return;
}
-
- printk(KERN_WARNING "DMAR: Recommended TLB entries for ISOCH unit is 16; your BIOS set %d\n",
+
+ pr_warn("Recommended TLB entries for ISOCH unit is 16; your BIOS set %d\n",
vtisochctrl);
}
Code Review / kvmfornfv.git / blobdiff