Hi,
在 2023/10/07 23:16, Oleg Nesterov 写道:
On 10/07, Yu Kuai wrote:
probably need to remove the mul_u64_u64_div_u64 and check for
overflow/potential overflow ourselves?
probably yes...
How about this?
diff --git a/block/blk-throttle.c b/block/blk-throttle.c
index 1101fb6f6cc8..5482c316a103 100644
--- a/block/blk-throttle.c
+++ b/block/blk-throttle.c
@@ -723,6 +723,10 @@ static unsigned int calculate_io_allowed(u32
iops_limit,
static u64 calculate_bytes_allowed(u64 bps_limit, unsigned long
jiffy_elapsed)
{
+ if (jiffy_elapsed > HZ &&
+ bps_limit > mul_u64_u64_div_u64(U64_MAX, (u64)HZ,
(u64)jiffy_elapsed);
+ return U64_MAX;
+
I can't suggest anything better...
but I do not know if it is possible that HZ > jiffy_elapsed. If yes, then
mul_u64_u64_div_u64() above is not safe too.
Well, 'jiffy_elapsed > HZ' is judged before mul_u64_u64_div_u64().
Overflow can only happen if the above 2 conditions passed, and U64_MAX
is returned.
Thanks,
Kuai
Oleg.
.
