@@ -42,6 +42,7 @@ struct vm_struct {
unsigned long size;
unsigned long flags;
struct page **pages;
+ unsigned int page_shift;
unsigned int nr_pages;
phys_addr_t phys_addr;
const void *caller;
@@ -216,32 +216,34 @@ static int vmap_p4d_range(pgd_t *pgd, unsigned long addr,
* Ie. pte at addr+N*PAGE_SIZE shall point to pfn corresponding to pages[N]
*/
static int vmap_page_range_noflush(unsigned long start, unsigned long end,
- pgprot_t prot, struct page **pages)
+ pgprot_t prot, struct page **pages,
+ unsigned int page_shift)
{
- pgd_t *pgd;
- unsigned long next;
unsigned long addr = start;
- int err = 0;
- int nr = 0;
+ unsigned int i, nr = (end - start) >> (PAGE_SHIFT + page_shift);
- BUG_ON(addr >= end);
- pgd = pgd_offset_k(addr);
- do {
- next = pgd_addr_end(addr, end);
- err = vmap_p4d_range(pgd, addr, next, prot, pages, &nr);
+ for (i = 0; i < nr; i++) {
+ int err;
+
+ err = ioremap_page_range(addr,
+ addr + (PAGE_SIZE << page_shift),
+ __pa(page_address(pages[i])), prot);
if (err)
return err;
- } while (pgd++, addr = next, addr != end);
+
+ addr += PAGE_SIZE << page_shift;
+ }
return nr;
}
static int vmap_page_range(unsigned long start, unsigned long end,
- pgprot_t prot, struct page **pages)
+ pgprot_t prot, struct page **pages,
+ unsigned int page_shift)
{
int ret;
- ret = vmap_page_range_noflush(start, end, prot, pages);
+ ret = vmap_page_range_noflush(start, end, prot, pages, page_shift);
flush_cache_vmap(start, end);
return ret;
}
@@ -1189,7 +1191,7 @@ void *vm_map_ram(struct page **pages, unsigned int count, int node, pgprot_t pro
addr = va->va_start;
mem = (void *)addr;
}
- if (vmap_page_range(addr, addr + size, prot, pages) < 0) {
+ if (vmap_page_range(addr, addr + size, prot, pages, 0) < 0) {
vm_unmap_ram(mem, count);
return NULL;
}
@@ -1305,7 +1307,7 @@ void __init vmalloc_init(void)
int map_kernel_range_noflush(unsigned long addr, unsigned long size,
pgprot_t prot, struct page **pages)
{
- return vmap_page_range_noflush(addr, addr + size, prot, pages);
+ return vmap_page_range_noflush(addr, addr + size, prot, pages, 0);
}
/**
@@ -1352,7 +1354,7 @@ int map_vm_area(struct vm_struct *area, pgprot_t prot, struct page **pages)
unsigned long end = addr + get_vm_area_size(area);
int err;
- err = vmap_page_range(addr, end, prot, pages);
+ err = vmap_page_range(addr, end, prot, pages, 0);
return err > 0 ? 0 : err;
}
@@ -1395,8 +1397,9 @@ static struct vm_struct *__get_vm_area_node(unsigned long size,
return NULL;
if (flags & VM_IOREMAP)
- align = 1ul << clamp_t(int, get_count_order_long(size),
- PAGE_SHIFT, IOREMAP_MAX_ORDER);
+ align = max(align,
+ 1ul << clamp_t(int, get_count_order_long(size),
+ PAGE_SHIFT, IOREMAP_MAX_ORDER));
area = kzalloc_node(sizeof(*area), gfp_mask & GFP_RECLAIM_MASK, node);
if (unlikely(!area))
@@ -1608,7 +1611,7 @@ static void __vunmap(const void *addr, int deallocate_pages)
struct page *page = area->pages[i];
BUG_ON(!page);
- __free_pages(page, 0);
+ __free_pages(page, area->page_shift);
}
kvfree(area->pages);
@@ -1751,14 +1754,17 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
pgprot_t prot, int node)
{
struct page **pages;
+ unsigned long addr = (unsigned long)area->addr;
+ unsigned long size = get_vm_area_size(area);
+ unsigned int page_shift = area->page_shift;
+ unsigned int shift = page_shift + PAGE_SHIFT;
unsigned int nr_pages, array_size, i;
const gfp_t nested_gfp = (gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO;
const gfp_t alloc_mask = gfp_mask | __GFP_NOWARN;
const gfp_t highmem_mask = (gfp_mask & (GFP_DMA | GFP_DMA32)) ?
- 0 :
- __GFP_HIGHMEM;
+ 0 : __GFP_HIGHMEM;
- nr_pages = get_vm_area_size(area) >> PAGE_SHIFT;
+ nr_pages = size >> shift;
array_size = (nr_pages * sizeof(struct page *));
area->nr_pages = nr_pages;
@@ -1779,10 +1785,8 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
for (i = 0; i < area->nr_pages; i++) {
struct page *page;
- if (node == NUMA_NO_NODE)
- page = alloc_page(alloc_mask|highmem_mask);
- else
- page = alloc_pages_node(node, alloc_mask|highmem_mask, 0);
+ page = alloc_pages_node(node,
+ alloc_mask|highmem_mask, page_shift);
if (unlikely(!page)) {
/* Successfully allocated i pages, free them in __vunmap() */
@@ -1794,8 +1798,9 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
cond_resched();
}
- if (map_vm_area(area, prot, pages))
+ if (vmap_page_range(addr, addr + size, prot, pages, page_shift) < 0)
goto fail;
+
return area->addr;
fail:
@@ -1832,19 +1837,35 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
struct vm_struct *area;
void *addr;
unsigned long real_size = size;
+ unsigned long real_align = align;
+ unsigned long size_per_node;
+ unsigned int shift;
size = PAGE_ALIGN(size);
if (!size || (size >> PAGE_SHIFT) > totalram_pages())
goto fail;
+ size_per_node = size;
+ if (node == NUMA_NO_NODE)
+ size_per_node /= num_online_nodes();
+ if (size_per_node >= PMD_SIZE)
+ shift = PMD_SHIFT;
+ else
+ shift = PAGE_SHIFT;
+again:
+ align = max(real_align, 1UL << shift);
+ size = ALIGN(real_size, align);
+
area = __get_vm_area_node(size, align, VM_ALLOC | VM_UNINITIALIZED |
vm_flags, start, end, node, gfp_mask, caller);
if (!area)
goto fail;
+ area->page_shift = shift - PAGE_SHIFT;
+
addr = __vmalloc_area_node(area, gfp_mask, prot, node);
if (!addr)
- return NULL;
+ goto fail;
/*
* In this function, newly allocated vm_struct has VM_UNINITIALIZED
@@ -1858,8 +1879,16 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
return addr;
fail:
- warn_alloc(gfp_mask, NULL,
+ if (shift == PMD_SHIFT) {
+ shift = PAGE_SHIFT;
+ goto again;
+ }
+
+ if (!area) {
+ /* Warn for area allocation, page allocations already warn */
+ warn_alloc(gfp_mask, NULL,
"vmalloc: allocation failure: %lu bytes", real_size);
+ }
return NULL;
}
This appears to help cached git diff performance by about 5% on a POWER9 (with 32MB dentry cache hash). Profiling git diff dTLB misses with a vanilla kernel: 81.75% git [kernel.vmlinux] [k] __d_lookup_rcu 7.21% git [kernel.vmlinux] [k] strncpy_from_user 1.77% git [kernel.vmlinux] [k] find_get_entry 1.59% git [kernel.vmlinux] [k] kmem_cache_free 40,168 dTLB-miss 0.100342754 seconds time elapsed After this patch (and the subsequent powerpc HUGE_VMAP patches), the dentry cache hash gets mapped with 2MB pages: 2,987 dTLB-miss 0.095933138 seconds time elapsed elapsed time improvement isn't too scientific but seems consistent, TLB misses certainly improves an order of magnitude. My laptop takes a lot of misses here too, so x86 would be interesting to test, I think it should just work there. --- include/linux/vmalloc.h | 1 + mm/vmalloc.c | 87 +++++++++++++++++++++++++++-------------- 2 files changed, 59 insertions(+), 29 deletions(-)