Allocate dma-remapping structures using numa locality info. On platforms having remapping hardware units span different nodes, this enables optimized dma-remapping transalation structures access by remapping hardware. Signed-off-by: Suresh Siddha --- drivers/pci/intel-iommu.c | 27 +++++++++++++++++++-------- 1 file changed, 19 insertions(+), 8 deletions(-) Index: tip/drivers/pci/intel-iommu.c =================================================================== --- tip.orig/drivers/pci/intel-iommu.c +++ tip/drivers/pci/intel-iommu.c @@ -267,6 +267,7 @@ struct dmar_domain *si_domain; struct dmar_domain { int id; /* domain id */ + int nid; /* node id */ unsigned long iommu_bmp; /* bitmap of iommus this domain uses*/ struct list_head devices; /* all devices' list */ @@ -390,15 +391,18 @@ static inline void *iommu_kmem_cache_all } -static inline void *alloc_pgtable_page(void) +static inline void *alloc_pgtable_page(int node) { unsigned int flags; - void *vaddr; + struct page *page; + void *vaddr = NULL; /* trying to avoid low memory issues */ flags = current->flags & PF_MEMALLOC; current->flags |= PF_MEMALLOC; - vaddr = (void *)get_zeroed_page(GFP_ATOMIC); + page = alloc_pages_node(node, GFP_ATOMIC | __GFP_ZERO, 0); + if (page) + vaddr = page_address(page); current->flags &= (~PF_MEMALLOC | flags); return vaddr; } @@ -579,7 +583,8 @@ static struct context_entry * device_to_ root = &iommu->root_entry[bus]; context = get_context_addr_from_root(root); if (!context) { - context = (struct context_entry *)alloc_pgtable_page(); + context = (struct context_entry *) + alloc_pgtable_page(iommu->node); if (!context) { spin_unlock_irqrestore(&iommu->lock, flags); return NULL; @@ -722,7 +727,7 @@ static struct dma_pte *pfn_to_dma_pte(st if (!dma_pte_present(pte)) { uint64_t pteval; - tmp_page = alloc_pgtable_page(); + tmp_page = alloc_pgtable_page(domain->nid); if (!tmp_page) return NULL; @@ -855,7 +860,7 @@ static int iommu_alloc_root_entry(struct struct root_entry *root; unsigned long flags; - root = (struct root_entry *)alloc_pgtable_page(); + root = (struct root_entry *)alloc_pgtable_page(iommu->node); if (!root) return -ENOMEM; @@ -1249,6 +1254,7 @@ static struct dmar_domain *alloc_domain( if (!domain) return NULL; + domain->nid = -1; memset(&domain->iommu_bmp, 0, sizeof(unsigned long)); domain->flags = 0; @@ -1409,9 +1415,10 @@ static int domain_init(struct dmar_domai domain->iommu_snooping = 0; domain->iommu_count = 1; + domain->nid = iommu->node; /* always allocate the top pgd */ - domain->pgd = (struct dma_pte *)alloc_pgtable_page(); + domain->pgd = (struct dma_pte *)alloc_pgtable_page(domain->nid); if (!domain->pgd) return -ENOMEM; __iommu_flush_cache(iommu, domain->pgd, PAGE_SIZE); @@ -1566,6 +1573,8 @@ static int domain_context_mapping_one(st 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); @@ -3397,6 +3406,7 @@ static struct dmar_domain *iommu_alloc_v return NULL; domain->id = vm_domid++; + domain->nid = -1; memset(&domain->iommu_bmp, 0, sizeof(unsigned long)); domain->flags = DOMAIN_FLAG_VIRTUAL_MACHINE; @@ -3423,9 +3433,10 @@ static int md_domain_init(struct dmar_do domain->iommu_coherency = 0; domain->iommu_snooping = 0; domain->max_addr = 0; + domain->nid = -1; /* always allocate the top pgd */ - domain->pgd = (struct dma_pte *)alloc_pgtable_page(); + domain->pgd = (struct dma_pte *)alloc_pgtable_page(domain->nid); if (!domain->pgd) return -ENOMEM; domain_flush_cache(domain, domain->pgd, PAGE_SIZE); -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/