csky: Fixup wrong update_mmu_cache implementation

In our stress test, we found some crash problem caused by:

if (!(vma->vm_flags & VM_EXEC))
	return;

in update_mmu_cache().

Seems current update_mmu_cache implementation is wrong and we retread
to the conservative implementation.

Also the usage of kmap_atomic in update_mmu_cache is risky, page-virtual
may be scheduled out and changed, so we must use preempt_disable &
pagefault_disable which is called by kmap_atomic().

Signed-off-by: Guo Ren <ren_guo@c-sky.com>
Cc: Arnd Bergmann <arnd@arndb.de>
This commit is contained in:
Guo Ren 2019-04-10 10:55:07 +08:00
parent f62e31623d
commit 981bbf274b

View file

@ -34,10 +34,6 @@ void update_mmu_cache(struct vm_area_struct *vma, unsigned long address,
{
unsigned long addr, pfn;
struct page *page;
void *va;
if (!(vma->vm_flags & VM_EXEC))
return;
pfn = pte_pfn(*pte);
if (unlikely(!pfn_valid(pfn)))
@ -47,14 +43,9 @@ void update_mmu_cache(struct vm_area_struct *vma, unsigned long address,
if (page == ZERO_PAGE(0))
return;
va = page_address(page);
addr = (unsigned long) va;
if (va == NULL && PageHighMem(page))
addr = (unsigned long) kmap_atomic(page);
addr = (unsigned long) kmap_atomic(page);
cache_wbinv_range(addr, addr + PAGE_SIZE);
if (va == NULL && PageHighMem(page))
kunmap_atomic((void *) addr);
kunmap_atomic((void *) addr);
}