Skip to contentSkip to conversation

Reimplement old functionality of vm_munmap to vm_munmap_mm

PATCH In review James Custer 3 participants 3 messages Nov 19, 2013

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 Reimplement old functionality of vm_munmap to vm_munmap_mm

Commit bfce281c287a427d0841fadf5d59242757b4e620 killed the mm parameter to vm_munmap. Although the mm parameter was not used in any in-tree kernel modules, it is used by some out-of-tree modules.

We create a new function vm_munmap_mm that has the same functionality as vm_munmap, whereas vm_munmap uses current->mm, vm_munmap_mm takes the mm as a paramter.

Since this is a newly exported symbol it is marked EXPORT_SYMBOL_GPL.

---
 include/linux/mm.h | 1 +
 mm/mmap.c          | 9 +++++++--
 mm/nommu.c         | 9 +++++++--
 3 files changed, 15 insertions(+), 4 deletions(-)
include/linux/mm.h +1 -0
index 0548eb2..6e9917e 100644
@@ -1724,6 +1724,7 @@ static inline void mm_populate(unsigned long addr, unsigned long len) {}
 /* These take the mm semaphore themselves */
 extern unsigned long vm_brk(unsigned long, unsigned long);
 extern int vm_munmap(unsigned long, size_t);
+extern int vm_munmap_mm(struct mm_struct *, unsigned long, size_t);
 extern unsigned long vm_mmap(struct file *, unsigned long,
         unsigned long, unsigned long,
         unsigned long, unsigned long);
mm/mmap.c +7 -2
index 834b2d7..63eb96e 100644
@@ -2539,16 +2539,21 @@ int do_munmap(struct mm_struct *mm, unsigned long start, size_t len)
 	return 0;
 }
 
-int vm_munmap(unsigned long start, size_t len)
+int vm_munmap_mm(struct mm_struct *mm, unsigned long start, size_t len)
 {
 	int ret;
-	struct mm_struct *mm = current->mm;
 
 	down_write(&mm->mmap_sem);
 	ret = do_munmap(mm, start, len);
 	up_write(&mm->mmap_sem);
 	return ret;
 }
+EXPORT_SYMBOL_GPL(vm_munmap_mm);
+
+int vm_munmap(unsigned long start, size_t len)
+{
+	return vm_munmap_mm(current->mm, start, len);
+}
 EXPORT_SYMBOL(vm_munmap);
 
 SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len)
mm/nommu.c +7 -2
index fec093a..5cf8677 100644
@@ -1717,9 +1717,8 @@ erase_whole_vma:
 }
 EXPORT_SYMBOL(do_munmap);
 
-int vm_munmap(unsigned long addr, size_t len)
+int vm_munmap_mm(struct mm_struct *mm, unsigned long addr, size_t len)
 {
-	struct mm_struct *mm = current->mm;
 	int ret;
 
 	down_write(&mm->mmap_sem);
@@ -1727,6 +1726,12 @@ int vm_munmap(unsigned long addr, size_t len)
 	up_write(&mm->mmap_sem);
 	return ret;
 }
+EXPORT_SYMBOL_GPL(vm_munmap_mm);
+
+int vm_munmap(unsigned long addr, size_t len)
+{
+	return vm_munmap_mm(current->mm, addr, len);
+}
 EXPORT_SYMBOL(vm_munmap);
 
 SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len)
Signature · 1 line
1.8.2.1
#1 lore raw
Quoted 3 lines · James Custer

Commit bfce281c287a427d0841fadf5d59242757b4e620 killed the mm parameter to vm_munmap. Although the mm parameter was not used in any in-tree kernel modules, it is used by some out-of-tree modules.

Which doesn't matter at all.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>

#2 lore raw
In reply to James Custer
Quoted 9 lines · James Custer

Commit bfce281c287a427d0841fadf5d59242757b4e620 killed the mm parameter to vm_munmap. Although the mm parameter was not used in any in-tree kernel modules, it is used by some out-of-tree modules.

We create a new function vm_munmap_mm that has the same functionality as vm_munmap, whereas vm_munmap uses current->mm, vm_munmap_mm takes the mm as a paramter.

Since this is a newly exported symbol it is marked EXPORT_SYMBOL_GPL.

Which modules and what are they doing with it? More to the point, what prevents races with e.g. dumping core? And that's not an idle question - for example, fs/aio.c used to contain very unpleasant races of that kind exactly because it was playing games with modifying ->mm other than current->mm.

In other words, NAK.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>