[PATCH 39/64] arch/m68k: use mm locking wrappers
From: Davidlohr Bueso
Date: Sun Feb 04 2018 - 20:38:22 EST
From: Davidlohr Bueso <dave@xxxxxxxxxxxx>
This becomes quite straightforward with the mmrange in place.
Signed-off-by: Davidlohr Bueso <dbueso@xxxxxxx>
---
arch/m68k/kernel/sys_m68k.c | 18 +++++++++++-------
arch/m68k/mm/fault.c | 8 ++++----
2 files changed, 15 insertions(+), 11 deletions(-)
diff --git a/arch/m68k/kernel/sys_m68k.c b/arch/m68k/kernel/sys_m68k.c
index 27e10af5153a..d151bd19385c 100644
--- a/arch/m68k/kernel/sys_m68k.c
+++ b/arch/m68k/kernel/sys_m68k.c
@@ -378,6 +378,7 @@ asmlinkage int
sys_cacheflush (unsigned long addr, int scope, int cache, unsigned long len)
{
int ret = -EINVAL;
+ DEFINE_RANGE_LOCK_FULL(mmrange);
if (scope < FLUSH_SCOPE_LINE || scope > FLUSH_SCOPE_ALL ||
cache & ~FLUSH_CACHE_BOTH)
@@ -399,7 +400,7 @@ sys_cacheflush (unsigned long addr, int scope, int cache, unsigned long len)
* Verify that the specified address region actually belongs
* to this process.
*/
- down_read(¤t->mm->mmap_sem);
+ mm_read_lock(current->mm, &mmrange);
vma = find_vma(current->mm, addr);
if (!vma || addr < vma->vm_start || addr + len > vma->vm_end)
goto out_unlock;
@@ -450,7 +451,7 @@ sys_cacheflush (unsigned long addr, int scope, int cache, unsigned long len)
}
}
out_unlock:
- up_read(¤t->mm->mmap_sem);
+ mm_read_unlock(current->mm, &mmrange);
out:
return ret;
}
@@ -461,6 +462,8 @@ asmlinkage int
sys_atomic_cmpxchg_32(unsigned long newval, int oldval, int d3, int d4, int d5,
unsigned long __user * mem)
{
+ DEFINE_RANGE_LOCK_FULL(mmrange);
+
/* This was borrowed from ARM's implementation. */
for (;;) {
struct mm_struct *mm = current->mm;
@@ -470,7 +473,7 @@ sys_atomic_cmpxchg_32(unsigned long newval, int oldval, int d3, int d4, int d5,
spinlock_t *ptl;
unsigned long mem_value;
- down_read(&mm->mmap_sem);
+ mm_read_lock(mm, &mmrange);
pgd = pgd_offset(mm, (unsigned long)mem);
if (!pgd_present(*pgd))
goto bad_access;
@@ -493,11 +496,11 @@ sys_atomic_cmpxchg_32(unsigned long newval, int oldval, int d3, int d4, int d5,
__put_user(newval, mem);
pte_unmap_unlock(pte, ptl);
- up_read(&mm->mmap_sem);
+ mm_read_unlock(mm, &mmrange);
return mem_value;
bad_access:
- up_read(&mm->mmap_sem);
+ mm_read_unlock(mm, &mmrange);
/* This is not necessarily a bad access, we can get here if
a memory we're trying to write to should be copied-on-write.
Make the kernel do the necessary page stuff, then re-iterate.
@@ -536,14 +539,15 @@ sys_atomic_cmpxchg_32(unsigned long newval, int oldval, int d3, int d4, int d5,
{
struct mm_struct *mm = current->mm;
unsigned long mem_value;
+ DEFINE_RANGE_LOCK_FULL(mmrange);
- down_read(&mm->mmap_sem);
+ mm_read_lock(mm, &mmrange);
mem_value = *mem;
if (mem_value == oldval)
*mem = newval;
- up_read(&mm->mmap_sem);
+ mm_read_unlock(mm, &mmrange);
return mem_value;
}
diff --git a/arch/m68k/mm/fault.c b/arch/m68k/mm/fault.c
index ec32a193726f..426d22924852 100644
--- a/arch/m68k/mm/fault.c
+++ b/arch/m68k/mm/fault.c
@@ -90,7 +90,7 @@ int do_page_fault(struct pt_regs *regs, unsigned long address,
if (user_mode(regs))
flags |= FAULT_FLAG_USER;
retry:
- down_read(&mm->mmap_sem);
+ mm_read_lock(mm, &mmrange);
vma = find_vma(mm, address);
if (!vma)
@@ -181,7 +181,7 @@ int do_page_fault(struct pt_regs *regs, unsigned long address,
}
}
- up_read(&mm->mmap_sem);
+ mm_read_unlock(mm, &mmrange);
return 0;
/*
@@ -189,7 +189,7 @@ int do_page_fault(struct pt_regs *regs, unsigned long address,
* us unable to handle the page fault gracefully.
*/
out_of_memory:
- up_read(&mm->mmap_sem);
+ mm_read_unlock(mm, &mmrange);
if (!user_mode(regs))
goto no_context;
pagefault_out_of_memory();
@@ -218,6 +218,6 @@ int do_page_fault(struct pt_regs *regs, unsigned long address,
current->thread.faddr = address;
send_sig:
- up_read(&mm->mmap_sem);
+ mm_read_unlock(mm, &mmrange);
return send_fault_sig(regs);
}
--
2.13.6