[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <eb7dede4-9668-9252-b95a-e20e313c5890@huawei.com>
Date: Mon, 15 Apr 2019 23:20:31 +0800
From: "chengjian (D)" <cj.chengjian@...wei.com>
To: Peter Zijlstra <peterz@...radead.org>
CC: <huawei.libin@...wei.com>, <xiexiuqi@...wei.com>,
<yangyingliang@...wei.com>, <mingo@...hat.com>,
<linux-kernel@...r.kernel.org>,
"chengjian (D)" <cj.chengjian@...wei.com>,
Yang Yingliang <yangyingliang@...wei.com>
Subject: Re: [PATCH] sched/fair: Use 'unsigned long' for
group_shares,group_runnable
Hi, Peter
On 2019/4/15 20:46, Peter Zijlstra wrote:
I write a demo about this, which I described it as overflow.
#cat test.c
//test.c
#include <stdio.h>
#include <stdlib.h>
#include <limits.h>
#include <assert.h>
int main(void)
{
long a = 1048576 * 9144968455305; /* shares = tg_shares * load */
unsigned long b = a;
//unsigned long e = 1048576 * 9144968455305;
printf("LONG_MAX = %ld, 0x%lx\n", LONG_MAX, LONG_MAX);
printf("ULONG_MAX = %lu, 0x%lx\n", ULONG_MAX, ULONG_MAX);
if (b > LONG_MAX)
printf("==overflow!!!==\n"); // OVERFLOW
printf("a = %20ld,0x%016lx\n", a, a);
printf("b = %20lu,0x%016lx\n", b, b);
/* shares /= tg_weight */
printf("a/3 = 0x%016lx\n", (a / 3)); // WRONG
printf("a/3 = 0x%016lx\n", (((unsigned long)a) / 3)); // unsigned
printf("a/3 = 0x%016lx\n", (a / (unsigned long)3)); // unsigned
printf("b/3 = 0x%016lx\n", b / 3); // unsigned
return EXIT_SUCCESS;
}
#./test
LONG_MAX = 9223372036854775807, 0x7fffffffffffffff
ULONG_MAX = 18446744073709551615, 0xffffffffffffffff
==overflow!!!==
a = -8857549630719655936,0x8513a98a48900000
b = 9589194442989895680,0x8513a98a48900000
a/3 = 0xd7068dd8c2daaaab // WRONG
a/3 = 0x2c5be32e18300000
a/3 = 0x2c5be32e18300000
b/3 = 0x2c5be32e18300000
> On Sat, Apr 13, 2019 at 03:32:34AM +0000, Cheng Jian wrote:
>> group_share and group_runnable are tracked as 'unsigned long',
>> however some functions using them as 'long' which is ultimately
>> assigned back to 'unsigned long' variables in reweight_entity.
>>
>> Since there is not scope on using a different and signed type,
>> this change improves code consistency and avoids further type
>> conversions. More important, to prevent undefined behavior
>> caused by overflow.
> There is no undefined behaviour due to overflow.UBSAN is broken,
> upgrade to GCC8 or later.
>
> .
So, function calc_group_shares will return the wrong value.
```cpp
static long calc_group_shares(struct cfs_rq *cfs_rq)
{
// ......
shares = (tg_shares * load);
if (tg_weight)
shares /= tg_weight;
}
```
The same to calc_group_runnable.
Thanks.
CHENG Jian.
Powered by blists - more mailing lists