linux-kernel - Re: [LKP] [writeback] 8bc4ad9498: INFO: suspicious RCU usage. ]

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <4cace953-b5b9-4840-f3f2-b21bd4f30095@fb.com>
Date:   Thu, 8 Sep 2016 09:52:46 -0600
From:   Jens Axboe <axboe@...com>
To:     Ye Xiaolong <xiaolong.ye@...el.com>, Tejun Heo <tj@...nel.org>,
        "Fengguang Wu" <fengguang.wu@...el.com>
CC:     <lkp@...org>, Philip Li <philip.li@...el.com>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [LKP] [writeback] 8bc4ad9498: INFO: suspicious RCU usage. ]

On 09/07/2016 08:04 PM, Ye Xiaolong wrote:
> On 09/08, Fengguang Wu wrote:
>> Hi Jens and Tejun,
>>
>>> The command line was in the original email:
>>>
>>> $ qemu-system-x86_64 -enable-kvm -cpu Westmere -m 512M
>>>
>>> And agree, in general it'd be nice if there was a link to the image as
>>> well, so that folks can reproduce.
>>
>> Yes we have a reproduce script internally, however need to polish it
>> up and make it general usable by kernel developers.
>>
>> CC Philip. It'd be helpful to give more priority to the "lkp qemu"
>> reproduce command.
>>
>> For the warning here, it does not depend on any initrd, I just
>> confirmed this simple script can reproduce the problem.
>
> Yes, just confirmed that v4.8-rc5 doesn't have this warning, while kernel
> for the updated commit (fa0e2cb1af765691fabd329f03cad563a0eebf18) on branch
> wb-buf-throttle could reproduce it.

Anyone willing to try the below patch?

diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
index ef61bda76317..a7bb2e6bcaef 100644
--- a/block/cfq-iosched.c
+++ b/block/cfq-iosched.c
@@ -3764,9 +3764,11 @@ static void check_blkcg_changed(struct cfq_io_cq 
*cic, struct bio *bio)
  	struct cfq_data *cfqd = cic_to_cfqd(cic);
  	struct cfq_queue *cfqq;
  	uint64_t serial_nr;
+	bool nonroot_cg;

  	rcu_read_lock();
  	serial_nr = bio_blkcg(bio)->css.serial_nr;
+	nonroot_cg = bio_blkcg(bio) != &blkcg_root;
  	rcu_read_unlock();

  	/*
@@ -3781,7 +3783,7 @@ static void check_blkcg_changed(struct cfq_io_cq 
*cic, struct bio *bio)
  	 * do proper throttling of writes. Turn off wbt for that
  	 * case.
  	 */
-	if (bio_blkcg(bio) != &blkcg_root) {
+	if (nonroot_cg) {
  		struct request_queue *q = cfqd->queue;

  		if (q->rq_wb)

-- 
Jens Axboe