mm/vmalloc.c: fix kernel BUG at mm/vmalloc.c:512!

[ Upstream commit afd0738 ]

One of the vmalloc stress test case triggers the kernel BUG():

  <snip>
  [60.562151] ------------[ cut here ]------------
  [60.562154] kernel BUG at mm/vmalloc.c:512!
  [60.562206] invalid opcode: 0000 [MIPS#1] PREEMPT SMP PTI
  [60.562247] CPU: 0 PID: 430 Comm: vmalloc_test/0 Not tainted 4.20.0+ torvalds#161
  [60.562293] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/01/2014
  [60.562351] RIP: 0010:alloc_vmap_area+0x36f/0x390
  <snip>

it can happen due to big align request resulting in overflowing of
calculated address, i.e.  it becomes 0 after ALIGN()'s fixup.

Fix it by checking if calculated address is within vstart/vend range.

Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Uladzislau Rezki (Sony) <[email protected]>
Reviewed-by: Andrew Morton <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Joel Fernandes <[email protected]>
Cc: Matthew Wilcox <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Oleksiy Avramchenko <[email protected]>
Cc: Steven Rostedt <[email protected]>
Cc: Tejun Heo <[email protected]>
Cc: Thomas Garnier <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Signed-off-by: Sasha Levin <[email protected]>

mm/vmalloc.c

@@ -439,7 +439,11 @@ static struct vmap_area *alloc_vmap_area(unsigned long size,
 	}
 
 found:
-	if (addr + size > vend)
+	/*
+	 * Check also calculated address against the vstart,
+	 * because it can be 0 because of big align request.
+	 */
+	if (addr + size > vend || addr < vstart)
 		goto overflow;
 
 	va->va_start = addr;