@@ -114,10 +114,6 @@ struct hmm_bo_device {
struct kmem_cache *bo_cache;
};
-struct hmm_page_object {
- struct page *page;
-};
-
struct hmm_buffer_object {
struct hmm_bo_device *bdev;
struct list_head list;
@@ -128,7 +124,6 @@ struct hmm_buffer_object {
/* mutex protecting this BO */
struct mutex mutex;
enum hmm_bo_type type;
- struct hmm_page_object *page_obj; /* physical pages */
int mmap_count;
int status;
int mem_type;
@@ -295,7 +295,7 @@ static int load_and_flush_by_kmap(ia_css_ptr virt, void *data,
idx = (virt - bo->start) >> PAGE_SHIFT;
offset = (virt - bo->start) - (idx << PAGE_SHIFT);
- src = (char *)kmap_local_page(bo->page_obj[idx].page) + offset;
+ src = (char *)kmap_local_page(bo->pages[idx]) + offset;
if ((bytes + offset) >= PAGE_SIZE) {
len = PAGE_SIZE - offset;
@@ -427,7 +427,7 @@ int hmm_store(ia_css_ptr virt, const void *data, unsigned int bytes)
idx = (virt - bo->start) >> PAGE_SHIFT;
offset = (virt - bo->start) - (idx << PAGE_SHIFT);
- des = (char *)kmap_local_page(bo->page_obj[idx].page);
+ des = (char *)kmap_local_page(bo->pages[idx]);
if (!des) {
dev_err(atomisp_dev,
@@ -498,7 +498,7 @@ int hmm_set(ia_css_ptr virt, int c, unsigned int bytes)
idx = (virt - bo->start) >> PAGE_SHIFT;
offset = (virt - bo->start) - (idx << PAGE_SHIFT);
- des = (char *)kmap_local_page(bo->page_obj[idx].page) + offset;
+ des = (char *)kmap_local_page(bo->pages[idx]) + offset;
if ((bytes + offset) >= PAGE_SIZE) {
len = PAGE_SIZE - offset;
@@ -537,7 +537,7 @@ phys_addr_t hmm_virt_to_phys(ia_css_ptr virt)
idx = (virt - bo->start) >> PAGE_SHIFT;
offset = (virt - bo->start) - (idx << PAGE_SHIFT);
- return page_to_phys(bo->page_obj[idx].page) + offset;
+ return page_to_phys(bo->pages[idx]) + offset;
}
int hmm_mmap(struct vm_area_struct *vma, ia_css_ptr virt)
@@ -631,7 +631,7 @@ static void free_private_bo_pages(struct hmm_buffer_object *bo,
int i, ret;
for (i = 0; i < free_pgnr; i++) {
- ret = set_pages_wb(bo->page_obj[i].page, 1);
+ ret = set_pages_wb(bo->pages[i], 1);
if (ret)
dev_err(atomisp_dev,
"set page to WB err ...ret = %d\n",
@@ -644,7 +644,7 @@ static void free_private_bo_pages(struct hmm_buffer_object *bo,
address be valid,it maybe memory corruption by lowmemory
*/
if (!ret) {
- __free_pages(bo->page_obj[i].page, 0);
+ __free_pages(bo->pages[i], 0);
}
}
}
@@ -663,11 +663,6 @@ static int alloc_private_pages(struct hmm_buffer_object *bo)
pgnr = bo->pgnr;
- bo->page_obj = kmalloc_array(pgnr, sizeof(struct hmm_page_object),
- GFP_KERNEL);
- if (unlikely(!bo->page_obj))
- return -ENOMEM;
-
i = 0;
alloc_pgnr = 0;
@@ -739,7 +734,7 @@ static int alloc_private_pages(struct hmm_buffer_object *bo)
}
for (j = 0; j < blk_pgnr; j++, i++) {
- bo->page_obj[i].page = pages + j;
+ bo->pages[i] = pages + j;
}
pgnr -= blk_pgnr;
@@ -759,18 +754,9 @@ static int alloc_private_pages(struct hmm_buffer_object *bo)
cleanup:
alloc_pgnr = i;
free_private_bo_pages(bo, alloc_pgnr);
-
- kfree(bo->page_obj);
-
return -ENOMEM;
}
-static void free_private_pages(struct hmm_buffer_object *bo)
-{
- free_private_bo_pages(bo, bo->pgnr);
- kfree(bo->page_obj);
-}
-
static void free_user_pages(struct hmm_buffer_object *bo,
unsigned int page_nr)
{
@@ -782,8 +768,6 @@ static void free_user_pages(struct hmm_buffer_object *bo,
for (i = 0; i < page_nr; i++)
put_page(bo->pages[i]);
}
- kfree(bo->pages);
- kfree(bo->page_obj);
}
/*
@@ -793,20 +777,7 @@ static int alloc_user_pages(struct hmm_buffer_object *bo,
const void __user *userptr)
{
int page_nr;
- int i;
struct vm_area_struct *vma;
- struct page **pages;
-
- pages = kmalloc_array(bo->pgnr, sizeof(struct page *), GFP_KERNEL);
- if (unlikely(!pages))
- return -ENOMEM;
-
- bo->page_obj = kmalloc_array(bo->pgnr, sizeof(struct hmm_page_object),
- GFP_KERNEL);
- if (unlikely(!bo->page_obj)) {
- kfree(pages);
- return -ENOMEM;
- }
mutex_unlock(&bo->mutex);
mmap_read_lock(current->mm);
@@ -814,8 +785,6 @@ static int alloc_user_pages(struct hmm_buffer_object *bo,
mmap_read_unlock(current->mm);
if (!vma) {
dev_err(atomisp_dev, "find_vma failed\n");
- kfree(bo->page_obj);
- kfree(pages);
mutex_lock(&bo->mutex);
return -EFAULT;
}
@@ -827,18 +796,16 @@ static int alloc_user_pages(struct hmm_buffer_object *bo,
userptr = untagged_addr(userptr);
- bo->pages = pages;
-
if (vma->vm_flags & (VM_IO | VM_PFNMAP)) {
page_nr = pin_user_pages((unsigned long)userptr, bo->pgnr,
FOLL_LONGTERM | FOLL_WRITE,
- pages, NULL);
+ bo->pages, NULL);
bo->mem_type = HMM_BO_MEM_TYPE_PFN;
} else {
/*Handle frame buffer allocated in user space*/
mutex_unlock(&bo->mutex);
page_nr = get_user_pages_fast((unsigned long)userptr,
- (int)(bo->pgnr), 1, pages);
+ (int)(bo->pgnr), 1, bo->pages);
mutex_lock(&bo->mutex);
bo->mem_type = HMM_BO_MEM_TYPE_USER;
}
@@ -858,10 +825,6 @@ static int alloc_user_pages(struct hmm_buffer_object *bo,
goto out_of_mem;
}
- for (i = 0; i < bo->pgnr; i++) {
- bo->page_obj[i].page = pages[i];
- }
-
return 0;
out_of_mem:
@@ -891,6 +854,12 @@ int hmm_bo_alloc_pages(struct hmm_buffer_object *bo,
mutex_lock(&bo->mutex);
check_bo_status_no_goto(bo, HMM_BO_PAGE_ALLOCED, status_err);
+ bo->pages = kmalloc_array(bo->pgnr, sizeof(struct page *), GFP_KERNEL);
+ if (unlikely(!bo->pages)) {
+ ret = -ENOMEM;
+ goto alloc_err;
+ }
+
/*
* TO DO:
* add HMM_BO_USER type
@@ -915,6 +884,7 @@ int hmm_bo_alloc_pages(struct hmm_buffer_object *bo,
return 0;
alloc_err:
+ kfree(bo->pages);
mutex_unlock(&bo->mutex);
dev_err(atomisp_dev, "alloc pages err...\n");
return ret;
@@ -940,11 +910,13 @@ void hmm_bo_free_pages(struct hmm_buffer_object *bo)
bo->status &= (~HMM_BO_PAGE_ALLOCED);
if (bo->type == HMM_BO_PRIVATE)
- free_private_pages(bo);
+ free_private_bo_pages(bo, bo->pgnr);
else if (bo->type == HMM_BO_USER)
free_user_pages(bo, bo->pgnr);
else
dev_err(atomisp_dev, "invalid buffer type.\n");
+
+ kfree(bo->pages);
mutex_unlock(&bo->mutex);
return;
@@ -989,7 +961,7 @@ int hmm_bo_bind(struct hmm_buffer_object *bo)
for (i = 0; i < bo->pgnr; i++) {
ret =
isp_mmu_map(&bdev->mmu, virt,
- page_to_phys(bo->page_obj[i].page), 1);
+ page_to_phys(bo->pages[i]), 1);
if (ret)
goto map_err;
virt += (1 << PAGE_SHIFT);
@@ -1103,9 +1075,6 @@ int hmm_bo_binded(struct hmm_buffer_object *bo)
void *hmm_bo_vmap(struct hmm_buffer_object *bo, bool cached)
{
- struct page **pages;
- int i;
-
check_bo_null_return(bo, NULL);
mutex_lock(&bo->mutex);
@@ -1122,27 +1091,15 @@ void *hmm_bo_vmap(struct hmm_buffer_object *bo, bool cached)
bo->status &= ~(HMM_BO_VMAPED | HMM_BO_VMAPED_CACHED);
}
- pages = kmalloc_array(bo->pgnr, sizeof(*pages), GFP_KERNEL);
- if (unlikely(!pages)) {
- mutex_unlock(&bo->mutex);
- return NULL;
- }
-
- for (i = 0; i < bo->pgnr; i++)
- pages[i] = bo->page_obj[i].page;
-
- bo->vmap_addr = vmap(pages, bo->pgnr, VM_MAP,
+ bo->vmap_addr = vmap(bo->pages, bo->pgnr, VM_MAP,
cached ? PAGE_KERNEL : PAGE_KERNEL_NOCACHE);
if (unlikely(!bo->vmap_addr)) {
- kfree(pages);
mutex_unlock(&bo->mutex);
dev_err(atomisp_dev, "vmap failed...\n");
return NULL;
}
bo->status |= (cached ? HMM_BO_VMAPED_CACHED : HMM_BO_VMAPED);
- kfree(pages);
-
mutex_unlock(&bo->mutex);
return bo->vmap_addr;
}
@@ -1272,7 +1229,7 @@ int hmm_bo_mmap(struct vm_area_struct *vma, struct hmm_buffer_object *bo)
virt = vma->vm_start;
for (i = 0; i < pgnr; i++) {
- pfn = page_to_pfn(bo->page_obj[i].page);
+ pfn = page_to_pfn(bo->pages[i]);
if (remap_pfn_range(vma, virt, pfn, PAGE_SIZE, PAGE_SHARED)) {
dev_warn(atomisp_dev,
"remap_pfn_range failed: virt = 0x%x, pfn = 0x%x, mapped_pgnr = %d\n",