Linux-BTRFS Archive on lore.kernel.org
 help / color / Atom feed
From: Qu Wenruo <quwenruo.btrfs@gmx.com>
To: dsterba@suse.cz, Qu Wenruo <wqu@suse.com>, linux-btrfs@vger.kernel.org
Subject: Re: [PATCH v4 04/12] btrfs: extent_io: Move the BUG_ON() in flush_write_bio() one level up
Date: Fri, 15 Feb 2019 17:58:26 +0800
Message-ID: <58b95afa-d627-6aa8-17a3-8e19d7eccafe@gmx.com> (raw)
In-Reply-To: <b61c2d29-888f-5fa3-bd64-59de8feabd10@gmx.com>

[-- Attachment #1.1: Type: text/plain, Size: 1884 bytes --]



On 2019/2/15 下午3:18, Qu Wenruo wrote:
> 
> 
> On 2019/1/31 下午10:36, David Sterba wrote:
>> On Thu, Jan 31, 2019 at 08:45:42AM +0800, Qu Wenruo wrote:
>>>
>>>
>>> On 2019/1/30 下午11:19, David Sterba wrote:
>>>> On Fri, Jan 25, 2019 at 01:09:17PM +0800, Qu Wenruo wrote:
>>>>> +static int __must_check flush_write_bio(struct extent_page_data *epd)
>>>>>  {
>>>>> -	if (epd->bio) {
>>>>> -		int ret;
>>>>> +	int ret = 0;
>>>>>  
>>>>> +	if (epd->bio) {
>>>>>  		ret = submit_one_bio(epd->bio, 0, 0);
>>>>> -		BUG_ON(ret < 0); /* -ENOMEM */
>>>>>  		epd->bio = NULL;
>>>>
>>>> I'm not sure if resetting epd->bio to NULL is all that needs to be done
>>>> here. With the BUG_ON the error case never happens so if all goes fine
>>>> it's also ok to set it to NULL and continue. But the callers might need
>>>> to send the flush again.
>>>
>>> If flush_write_bio() get called again on the failed one, it will just
>>> get skipped as  epd->bio is NULL, submit_one_bio() will not be triggered.
>>
>> Yes that's clear, but what is the state of the bio that went to
>> flush_write_bio and submit_one_bio failed?
>>
> 
> You got me.
> 
> In fact current simple "epd->bio = NULL" on failure will just leak epd->bio.
> 
> We need to at least call "bio_put(epd->bio);" before setting "epd->bio =
> NULL;"

Facepalm, I'm always too quick to draw a conclusion.

With proper test, the epd->bio is freed by its endio function.
So at flush_write_bio() call time, no matter whether submit_one_bio()
returns error or not, endio of epd->bio will be or has already been
triggered thus free the bio.

Either by successful bio submission, or by manual bio_endio(bio) call in
submit_bio_hook.

So in fact just setting "epd->bio = NULL" is completely fine in this
context.

Thanks,
Qu

> 
> Thanks for pointing this problem out,
> Qu
> 


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

  reply index

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-01-25  5:09 [PATCH v4 00/12] btrfs: Enhancement to tree block validation Qu Wenruo
2019-01-25  5:09 ` [PATCH v4 01/12] btrfs: Always output error message when key/level verification fails Qu Wenruo
2019-01-25  9:02   ` Johannes Thumshirn
2019-01-25  5:09 ` [PATCH v4 02/12] btrfs: extent_io: Kill the forward declaration of flush_write_bio() Qu Wenruo
2019-01-25  5:09 ` [PATCH v4 03/12] btrfs: disk-io: Show the timing of corrupted tree block explicitly Qu Wenruo
2019-01-25  9:03   ` Johannes Thumshirn
2019-01-30 14:57   ` David Sterba
2019-01-30 14:59     ` Nikolay Borisov
2019-01-31  0:03       ` Qu Wenruo
2019-01-31 14:20         ` David Sterba
2019-01-31 14:22           ` Nikolay Borisov
2019-02-07 17:27             ` David Sterba
2019-01-25  5:09 ` [PATCH v4 04/12] btrfs: extent_io: Move the BUG_ON() in flush_write_bio() one level up Qu Wenruo
2019-01-25  9:04   ` Johannes Thumshirn
2019-01-30 15:08   ` David Sterba
2019-01-30 15:19   ` David Sterba
2019-01-31  0:45     ` Qu Wenruo
2019-01-31 14:36       ` David Sterba
2019-02-15  7:18         ` Qu Wenruo
2019-02-15  9:58           ` Qu Wenruo [this message]
2019-01-25  5:09 ` [PATCH v4 05/12] btrfs: extent_io: Kill the BUG_ON() in extent_write_full_page() Qu Wenruo
2019-01-25  9:05   ` Johannes Thumshirn
2019-01-25  5:09 ` [PATCH v4 06/12] btrfs: extent_io: Kill the BUG_ON() in btree_write_cache_pages() Qu Wenruo
2019-01-25  9:11   ` Johannes Thumshirn
2019-01-25  5:09 ` [PATCH v4 07/12] btrfs: extent_io: Kill the dead branch in extent_write_cache_pages() Qu Wenruo
2019-01-25  9:11   ` Johannes Thumshirn
2019-01-25  5:09 ` [PATCH v4 08/12] btrfs: extent_io: Kill the BUG_ON() in extent_write_locked_range() Qu Wenruo
2019-01-25  5:09 ` [PATCH v4 09/12] btrfs: extent_io: Kill the BUG_ON() in lock_extent_buffer_for_io() Qu Wenruo
2019-01-25  5:09 ` [PATCH v4 10/12] btrfs: extent_io: Kill the BUG_ON() in extent_write_cache_pages() Qu Wenruo
2019-01-25  9:16   ` Johannes Thumshirn
2019-01-25  5:09 ` [PATCH v4 11/12] btrfs: extent_io: Kill the BUG_ON() in extent_writepages() Qu Wenruo
2019-01-25  9:16   ` Johannes Thumshirn
2019-01-25  5:09 ` [PATCH v4 12/12] btrfs: Do mandatory tree block check before submitting bio Qu Wenruo

Reply instructions:

You may reply publically to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=58b95afa-d627-6aa8-17a3-8e19d7eccafe@gmx.com \
    --to=quwenruo.btrfs@gmx.com \
    --cc=dsterba@suse.cz \
    --cc=linux-btrfs@vger.kernel.org \
    --cc=wqu@suse.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

Linux-BTRFS Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/linux-btrfs/0 linux-btrfs/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 linux-btrfs linux-btrfs/ https://lore.kernel.org/linux-btrfs \
		linux-btrfs@vger.kernel.org linux-btrfs@archiver.kernel.org
	public-inbox-index linux-btrfs


Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.linux-btrfs


AGPL code for this site: git clone https://public-inbox.org/ public-inbox