The (start, size) tuple represents a range [start, start + size - 1],
which means "start" and "start + size - 1" should be compared to see
whether the range overflows.
For example, a range with (start, size)
(0xffffffff fffffff0, 0x00000000 00000010)
represents
[0xffffffff fffffff0, 0xffffffff ffffffff]
would be judged overflow in the original code, while actually it is not.
This patch fixes this and makes sure it still works when size is zero.
Signed-off-by: Wei Yang <[email protected]>
---
arch/x86/kernel/e820.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c
index 621b501..871f186 100644
--- a/arch/x86/kernel/e820.c
+++ b/arch/x86/kernel/e820.c
@@ -388,11 +388,11 @@ static int __init __append_e820_map(struct e820entry
*biosmap, int nr_map)
while (nr_map) {
u64 start = biosmap->addr;
u64 size = biosmap->size;
- u64 end = start + size;
+ u64 end = start + size - 1;
u32 type = biosmap->type;
/* Overflow in 64 bits? Ignore the memory map. */
- if (start > end)
+ if (start > end && likely(size))
return -1;
e820_add_region(start, size, type);
--
1.7.9.5
