Revision df9576def004d2cd5beedc00cb6e8901427634b9 authored by Yang Shi on 03 August 2019, 04:48:37 UTC, committed by Linus Torvalds on 03 August 2019, 14:02:00 UTC
When running ltp's oom test with kmemleak enabled, the below warning was triggerred since kernel detects __GFP_NOFAIL & ~__GFP_DIRECT_RECLAIM is passed in: WARNING: CPU: 105 PID: 2138 at mm/page_alloc.c:4608 __alloc_pages_nodemask+0x1c31/0x1d50 Modules linked in: loop dax_pmem dax_pmem_core ip_tables x_tables xfs virtio_net net_failover virtio_blk failover ata_generic virtio_pci virtio_ring virtio libata CPU: 105 PID: 2138 Comm: oom01 Not tainted 5.2.0-next-20190710+ #7 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.10.2-0-g5f4c7b1-prebuilt.qemu-project.org 04/01/2014 RIP: 0010:__alloc_pages_nodemask+0x1c31/0x1d50 ... kmemleak_alloc+0x4e/0xb0 kmem_cache_alloc+0x2a7/0x3e0 mempool_alloc_slab+0x2d/0x40 mempool_alloc+0x118/0x2b0 bio_alloc_bioset+0x19d/0x350 get_swap_bio+0x80/0x230 __swap_writepage+0x5ff/0xb20 The mempool_alloc_slab() clears __GFP_DIRECT_RECLAIM, however kmemleak has __GFP_NOFAIL set all the time due to d9570ee3bd1d4f2 ("kmemleak: allow to coexist with fault injection"). But, it doesn't make any sense to have __GFP_NOFAIL and ~__GFP_DIRECT_RECLAIM specified at the same time. According to the discussion on the mailing list, the commit should be reverted for short term solution. Catalin Marinas would follow up with a better solution for longer term. The failure rate of kmemleak metadata allocation may increase in some circumstances, but this should be expected side effect. Link: http://lkml.kernel.org/r/1563299431-111710-1-git-send-email-yang.shi@linux.alibaba.com Fixes: d9570ee3bd1d4f2 ("kmemleak: allow to coexist with fault injection") Signed-off-by: Yang Shi <yang.shi@linux.alibaba.com> Suggested-by: Catalin Marinas <catalin.marinas@arm.com> Acked-by: Michal Hocko <mhocko@suse.com> Cc: Dmitry Vyukov <dvyukov@google.com> Cc: David Rientjes <rientjes@google.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Qian Cai <cai@lca.pw> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
1 parent 68d8681
blk-timeout.c
// SPDX-License-Identifier: GPL-2.0
/*
* Functions related to generic timeout handling of requests.
*/
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/blkdev.h>
#include <linux/fault-inject.h>
#include "blk.h"
#include "blk-mq.h"
#ifdef CONFIG_FAIL_IO_TIMEOUT
static DECLARE_FAULT_ATTR(fail_io_timeout);
static int __init setup_fail_io_timeout(char *str)
{
return setup_fault_attr(&fail_io_timeout, str);
}
__setup("fail_io_timeout=", setup_fail_io_timeout);
int blk_should_fake_timeout(struct request_queue *q)
{
if (!test_bit(QUEUE_FLAG_FAIL_IO, &q->queue_flags))
return 0;
return should_fail(&fail_io_timeout, 1);
}
static int __init fail_io_timeout_debugfs(void)
{
struct dentry *dir = fault_create_debugfs_attr("fail_io_timeout",
NULL, &fail_io_timeout);
return PTR_ERR_OR_ZERO(dir);
}
late_initcall(fail_io_timeout_debugfs);
ssize_t part_timeout_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
struct gendisk *disk = dev_to_disk(dev);
int set = test_bit(QUEUE_FLAG_FAIL_IO, &disk->queue->queue_flags);
return sprintf(buf, "%d\n", set != 0);
}
ssize_t part_timeout_store(struct device *dev, struct device_attribute *attr,
const char *buf, size_t count)
{
struct gendisk *disk = dev_to_disk(dev);
int val;
if (count) {
struct request_queue *q = disk->queue;
char *p = (char *) buf;
val = simple_strtoul(p, &p, 10);
if (val)
blk_queue_flag_set(QUEUE_FLAG_FAIL_IO, q);
else
blk_queue_flag_clear(QUEUE_FLAG_FAIL_IO, q);
}
return count;
}
#endif /* CONFIG_FAIL_IO_TIMEOUT */
/**
* blk_abort_request -- Request request recovery for the specified command
* @req: pointer to the request of interest
*
* This function requests that the block layer start recovery for the
* request by deleting the timer and calling the q's timeout function.
* LLDDs who implement their own error recovery MAY ignore the timeout
* event if they generated blk_abort_request.
*/
void blk_abort_request(struct request *req)
{
/*
* All we need to ensure is that timeout scan takes place
* immediately and that scan sees the new timeout value.
* No need for fancy synchronizations.
*/
WRITE_ONCE(req->deadline, jiffies);
kblockd_schedule_work(&req->q->timeout_work);
}
EXPORT_SYMBOL_GPL(blk_abort_request);
unsigned long blk_rq_timeout(unsigned long timeout)
{
unsigned long maxt;
maxt = round_jiffies_up(jiffies + BLK_MAX_TIMEOUT);
if (time_after(timeout, maxt))
timeout = maxt;
return timeout;
}
/**
* blk_add_timer - Start timeout timer for a single request
* @req: request that is about to start running.
*
* Notes:
* Each request has its own timer, and as it is added to the queue, we
* set up the timer. When the request completes, we cancel the timer.
*/
void blk_add_timer(struct request *req)
{
struct request_queue *q = req->q;
unsigned long expiry;
/*
* Some LLDs, like scsi, peek at the timeout to prevent a
* command from being retried forever.
*/
if (!req->timeout)
req->timeout = q->rq_timeout;
req->rq_flags &= ~RQF_TIMED_OUT;
expiry = jiffies + req->timeout;
WRITE_ONCE(req->deadline, expiry);
/*
* If the timer isn't already pending or this timeout is earlier
* than an existing one, modify the timer. Round up to next nearest
* second.
*/
expiry = blk_rq_timeout(round_jiffies_up(expiry));
if (!timer_pending(&q->timeout) ||
time_before(expiry, q->timeout.expires)) {
unsigned long diff = q->timeout.expires - expiry;
/*
* Due to added timer slack to group timers, the timer
* will often be a little in front of what we asked for.
* So apply some tolerance here too, otherwise we keep
* modifying the timer because expires for value X
* will be X + something.
*/
if (!timer_pending(&q->timeout) || (diff >= HZ / 2))
mod_timer(&q->timeout, expiry);
}
}
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...