kcsan
data_race
这里为什么需要
void folio_undo_large_rmappable(struct folio *folio)
{
struct deferred_split *ds_queue;
unsigned long flags;
if (folio_order(folio) <= 1)
return;
/*
* At this point, there is no one trying to add the folio to
* deferred_list. If folio is not in deferred_list, it's safe
* to check without acquiring the split_queue_lock.
*/
if (data_race(list_empty(&folio->_deferred_list)))
return;
ds_queue = get_deferred_split_queue(folio);
spin_lock_irqsave(&ds_queue->split_queue_lock, flags);
if (!list_empty(&folio->_deferred_list)) {
ds_queue->split_queue_len--;
list_del_init(&folio->_deferred_list);
}
spin_unlock_irqrestore(&ds_queue->split_queue_lock, flags);
}
else if (data_race(sis->flags & SWP_SYNCHRONOUS_IO))
History: #0
Commit: 7b7aca6d7c0f9b2d9400bfc57cb2b23cfbd5134d
Author: Pei Li <peili.dev@gmail.com>
Committer: Andrew Morton <akpm@linux-foundation.org>
Author Date: Fri 12 Jul 2024 12:32:30 AM CST
Committer Date: Thu 18 Jul 2024 12:05:18 PM CST
如何理解 do_swap_page 中的这个
if (data_race(si->flags & SWP_SYNCHRONOUS_IO) &&
__swap_count(entry) == 1) {
是不是全局变量的访问，要么添加 READ_ONCE / WRITE_ONCE ， 要么添加 date_race 才可以?
调试
只有一打开，就有很多类似的错误 想要解析这个东西并不容易
==================================================================
BUG: KCSAN: data-race in virtqueue_get_buf_ctx_split+0x63/0x220
race at unknown origin, with read to 0xffff88810cab3242 of 2 bytes by interrupt on cpu 2:
virtqueue_get_buf_ctx_split+0x63/0x220
virtqueue_get_buf_ctx+0x41/0x50
virtnet_rq_get_buf+0x5c/0xa0 [virtio_net]
virtnet_poll+0xd22/0xf40 [virtio_net]
__napi_poll+0x5f/0x280
net_rx_action+0x311/0x670
handle_softirqs+0xe3/0x2a0
irq_exit_rcu+0x9a/0xc0
common_interrupt+0x85/0xa0
asm_common_interrupt+0x26/0x40
pv_native_safe_halt+0xf/0x20
default_idle+0x13/0x20
default_idle_call+0x29/0xf0
do_idle+0x1cd/0x230
cpu_startup_entry+0x29/0x30
start_secondary+0x114/0x140
common_startup_64+0x13e/0x148
value changed: 0x1474 -> 0x1476
Reported by Kernel Concurrency Sanitizer on:
CPU: 2 UID: 0 PID: 0 Comm: swapper/2 Not tainted 6.15.4 #14 PREEMPT(voluntary)
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.3-32-g9029a010ec41 04/01/2014
BUG: KCSAN: data-race in vring_interrupt+0xdc/0x160
race at unknown origin, with read to 0xffff888111a03242 of 2 bytes by interrupt on cpu 4:
vring_interrupt+0xdc/0x160
__handle_irq_event_percpu+0x7e/0x240
handle_irq_event+0x81/0x100
handle_edge_irq+0x136/0x430
__common_interrupt+0x3e/0xa0
common_interrupt+0x80/0xa0
asm_common_interrupt+0x26/0x40
pv_native_safe_halt+0xf/0x20
default_idle+0x13/0x20
default_idle_call+0x29/0xf0
do_idle+0x1cd/0x230
cpu_startup_entry+0x29/0x30
start_secondary+0x114/0x140
common_startup_64+0x13e/0x148
value changed: 0xd4e5 -> 0xd4e6
Reported by Kernel Concurrency Sanitizer on:
CPU: 4 UID: 0 PID: 0 Comm: swapper/4 Not tainted 6.15.6 #15 PREEMPT(voluntary)
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.3-32-g9029a010ec41 04/01/2014
BUG: KCSAN: data-race in _find_next_bit+0x37/0xb0
race at unknown origin, with read to 0xffffffff83242880 of 8 bytes by interrupt on cpu 0:
_find_next_bit+0x37/0xb0
_nohz_idle_balance.isra.0+0x113/0x380
handle_softirqs+0xe3/0x2a0
irq_exit_rcu+0x9a/0xc0
sysvec_call_function_single+0x71/0x90
asm_sysvec_call_function_single+0x1a/0x20
pv_native_safe_halt+0xf/0x20
default_idle+0x13/0x20
default_idle_call+0x29/0xf0
do_idle+0x1cd/0x230
cpu_startup_entry+0x29/0x30
rest_init+0x106/0x110
start_kernel+0x936/0x940
x86_64_start_reservations+0x24/0x30
x86_64_start_kernel+0x8b/0x90
common_startup_64+0x13e/0x148
value changed: 0x00000000000000f7 -> 0x00000000000000ff
Reported by Kernel Concurrency Sanitizer on:
CPU: 0 UID: 0 PID: 0 Comm: swapper/0 Not tainted 6.15.4 #14 PREEMPT(voluntary)
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.3-32-g9029a010ec41 04/01/2014
本站所有文章转发 CSDN 将按侵权追究法律责任，其它情况随意。