Merge commit 'v3.6-rc5' into pci/gavin-window-alignment

* commit 'v3.6-rc5': (1098 commits)
  Linux 3.6-rc5
  HID: tpkbd: work even if the new Lenovo Keyboard driver is not configured
  Remove user-triggerable BUG from mpol_to_str
  xen/pciback: Fix proper FLR steps.
  uml: fix compile error in deliver_alarm()
  dj: memory scribble in logi_dj
  Fix order of arguments to compat_put_time[spec|val]
  xen: Use correct masking in xen_swiotlb_alloc_coherent.
  xen: fix logical error in tlb flushing
  xen/p2m: Fix one-off error in checking the P2M tree directory.
  powerpc: Don't use __put_user() in patch_instruction
  powerpc: Make sure IPI handlers see data written by IPI senders
  powerpc: Restore correct DSCR in context switch
  powerpc: Fix DSCR inheritance in copy_thread()
  powerpc: Keep thread.dscr and thread.dscr_inherit in sync
  powerpc: Update DSCR on all CPUs when writing sysfs dscr_default
  powerpc/powernv: Always go into nap mode when CPU is offline
  powerpc: Give hypervisor decrementer interrupts their own handler
  powerpc/vphn: Fix arch_update_cpu_topology() return value
  ARM: gemini: fix the gemini build
  ...

Conflicts:
	drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c
	drivers/rapidio/devices/tsi721.c

1 files changed, 110 insertions, 16 deletions

diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c
index c2cdf6500f75..051204fc4617 100644
--- a/arch/arm/mm/dma-mapping.c
+++ b/arch/arm/mm/dma-mapping.c
@@ -267,17 +267,19 @@ static void __dma_free_remap(void *cpu_addr, size_t size)
         vunmap(cpu_addr);
 }
 
+#define DEFAULT_DMA_COHERENT_POOL_SIZE  SZ_256K
+
 struct dma_pool {
         size_t size;
         spinlock_t lock;
         unsigned long *bitmap;
         unsigned long nr_pages;
         void *vaddr;
-        struct page *page;
+        struct page **pages;
 };
 
 static struct dma_pool atomic_pool = {
-        .size = SZ_256K,
+        .size = DEFAULT_DMA_COHERENT_POOL_SIZE,
 };
 
 static int __init early_coherent_pool(char *p)
@@ -287,6 +289,21 @@ static int __init early_coherent_pool(char *p)
 }
 early_param("coherent_pool", early_coherent_pool);
 
+void __init init_dma_coherent_pool_size(unsigned long size)
+{
+        /*
+         * Catch any attempt to set the pool size too late.
+         */
+        BUG_ON(atomic_pool.vaddr);
+
+        /*
+         * Set architecture specific coherent pool size only if
+         * it has not been changed by kernel command line parameter.
+         */
+        if (atomic_pool.size == DEFAULT_DMA_COHERENT_POOL_SIZE)
+                atomic_pool.size = size;
+}
+
 /*
  * Initialise the coherent pool for atomic allocations.
  */
@@ -297,6 +314,7 @@ static int __init atomic_pool_init(void)
         unsigned long nr_pages = pool->size >> PAGE_SHIFT;
         unsigned long *bitmap;
         struct page *page;
+        struct page **pages;
         void *ptr;
         int bitmap_size = BITS_TO_LONGS(nr_pages) * sizeof(long);
 
@@ -304,21 +322,31 @@ static int __init atomic_pool_init(void)
         if (!bitmap)
                 goto no_bitmap;
 
+        pages = kzalloc(nr_pages * sizeof(struct page *), GFP_KERNEL);
+        if (!pages)
+                goto no_pages;
+
         if (IS_ENABLED(CONFIG_CMA))
                 ptr = __alloc_from_contiguous(NULL, pool->size, prot, &page);
         else
                 ptr = __alloc_remap_buffer(NULL, pool->size, GFP_KERNEL, prot,
                                            &page, NULL);
         if (ptr) {
+                int i;
+
+                for (i = 0; i < nr_pages; i++)
+                        pages[i] = page + i;
+
                 spin_lock_init(&pool->lock);
                 pool->vaddr = ptr;
-                pool->page = page;
+                pool->pages = pages;
                 pool->bitmap = bitmap;
                 pool->nr_pages = nr_pages;
                 pr_info("DMA: preallocated %u KiB pool for atomic coherent allocations\n",
                        (unsigned)pool->size / 1024);
                 return 0;
         }
+no_pages:
         kfree(bitmap);
 no_bitmap:
         pr_err("DMA: failed to allocate %u KiB pool for atomic coherent allocation\n",
@@ -358,7 +386,7 @@ void __init dma_contiguous_remap(void)
                 if (end > arm_lowmem_limit)
                         end = arm_lowmem_limit;
                 if (start >= end)
-                        return;
+                        continue;
 
                 map.pfn = __phys_to_pfn(start);
                 map.virtual = __phys_to_virt(start);
@@ -423,7 +451,7 @@ static void *__alloc_from_pool(size_t size, struct page **ret_page)
         unsigned int pageno;
         unsigned long flags;
         void *ptr = NULL;
-        size_t align;
+        unsigned long align_mask;
 
         if (!pool->vaddr) {
                 WARN(1, "coherent pool not initialised!\n");
@@ -435,35 +463,53 @@ static void *__alloc_from_pool(size_t size, struct page **ret_page)
          * small, so align them to their order in pages, minimum is a page
          * size. This helps reduce fragmentation of the DMA space.
          */
-        align = PAGE_SIZE << get_order(size);
+        align_mask = (1 << get_order(size)) - 1;
 
         spin_lock_irqsave(&pool->lock, flags);
         pageno = bitmap_find_next_zero_area(pool->bitmap, pool->nr_pages,
-                                            0, count, (1 << align) - 1);
+                                            0, count, align_mask);
         if (pageno < pool->nr_pages) {
                 bitmap_set(pool->bitmap, pageno, count);
                 ptr = pool->vaddr + PAGE_SIZE * pageno;
-                *ret_page = pool->page + pageno;
+                *ret_page = pool->pages[pageno];
+        } else {
+                pr_err_once("ERROR: %u KiB atomic DMA coherent pool is too small!\n"
+                            "Please increase it with coherent_pool= kernel parameter!\n",
+                            (unsigned)pool->size / 1024);
         }
         spin_unlock_irqrestore(&pool->lock, flags);
 
         return ptr;
 }
 
+static bool __in_atomic_pool(void *start, size_t size)
+{
+        struct dma_pool *pool = &atomic_pool;
+        void *end = start + size;
+        void *pool_start = pool->vaddr;
+        void *pool_end = pool->vaddr + pool->size;
+
+        if (start < pool_start || start > pool_end)
+                return false;
+
+        if (end <= pool_end)
+                return true;
+
+        WARN(1, "Wrong coherent size(%p-%p) from atomic pool(%p-%p)\n",
+             start, end - 1, pool_start, pool_end - 1);
+
+        return false;
+}
+
 static int __free_from_pool(void *start, size_t size)
 {
         struct dma_pool *pool = &atomic_pool;
         unsigned long pageno, count;
         unsigned long flags;
 
-        if (start < pool->vaddr || start > pool->vaddr + pool->size)
+        if (!__in_atomic_pool(start, size))
                 return 0;
 
-        if (start + size > pool->vaddr + pool->size) {
-                WARN(1, "freeing wrong coherent size from pool\n");
-                return 0;
-        }
-
         pageno = (start - pool->vaddr) >> PAGE_SHIFT;
         count = size >> PAGE_SHIFT;
 
@@ -648,12 +694,12 @@ void arm_dma_free(struct device *dev, size_t size, void *cpu_addr,
 
         if (arch_is_coherent() || nommu()) {
                 __dma_free_buffer(page, size);
+        } else if (__free_from_pool(cpu_addr, size)) {
+                return;
         } else if (!IS_ENABLED(CONFIG_CMA)) {
                 __dma_free_remap(cpu_addr, size);
                 __dma_free_buffer(page, size);
         } else {
-                if (__free_from_pool(cpu_addr, size))
-                        return;
                 /*
                  * Non-atomic allocations cannot be freed with IRQs disabled
                  */
@@ -1090,10 +1136,22 @@ static int __iommu_remove_mapping(struct device *dev, dma_addr_t iova, size_t si
         return 0;
 }
 
+static struct page **__atomic_get_pages(void *addr)
+{
+        struct dma_pool *pool = &atomic_pool;
+        struct page **pages = pool->pages;
+        int offs = (addr - pool->vaddr) >> PAGE_SHIFT;
+
+        return pages + offs;
+}
+
 static struct page **__iommu_get_pages(void *cpu_addr, struct dma_attrs *attrs)
 {
         struct vm_struct *area;
 
+        if (__in_atomic_pool(cpu_addr, PAGE_SIZE))
+                return __atomic_get_pages(cpu_addr);
+
         if (dma_get_attr(DMA_ATTR_NO_KERNEL_MAPPING, attrs))
                 return cpu_addr;
 
@@ -1103,6 +1161,34 @@ static struct page **__iommu_get_pages(void *cpu_addr, struct dma_attrs *attrs)
         return NULL;
 }
 
+static void *__iommu_alloc_atomic(struct device *dev, size_t size,
+                                  dma_addr_t *handle)
+{
+        struct page *page;
+        void *addr;
+
+        addr = __alloc_from_pool(size, &page);
+        if (!addr)
+                return NULL;
+
+        *handle = __iommu_create_mapping(dev, &page, size);
+        if (*handle == DMA_ERROR_CODE)
+                goto err_mapping;
+
+        return addr;
+
+err_mapping:
+        __free_from_pool(addr, size);
+        return NULL;
+}
+
+static void __iommu_free_atomic(struct device *dev, struct page **pages,
+                                dma_addr_t handle, size_t size)
+{
+        __iommu_remove_mapping(dev, handle, size);
+        __free_from_pool(page_address(pages[0]), size);
+}
+
 static void *arm_iommu_alloc_attrs(struct device *dev, size_t size,
             dma_addr_t *handle, gfp_t gfp, struct dma_attrs *attrs)
 {
@@ -1113,6 +1199,9 @@ static void *arm_iommu_alloc_attrs(struct device *dev, size_t size,
         *handle = DMA_ERROR_CODE;
         size = PAGE_ALIGN(size);
 
+        if (gfp & GFP_ATOMIC)
+                return __iommu_alloc_atomic(dev, size, handle);
+
         pages = __iommu_alloc_buffer(dev, size, gfp);
         if (!pages)
                 return NULL;
@@ -1179,6 +1268,11 @@ void arm_iommu_free_attrs(struct device *dev, size_t size, void *cpu_addr,
                 return;
         }
 
+        if (__in_atomic_pool(cpu_addr, size)) {
+                __iommu_free_atomic(dev, pages, handle, size);
+                return;
+        }
+
         if (!dma_get_attr(DMA_ATTR_NO_KERNEL_MAPPING, attrs)) {
                 unmap_kernel_range((unsigned long)cpu_addr, size);
                 vunmap(cpu_addr);