Re: Bug?: unlink cause btrfs error but other fs don't

[Hongzhi, Song]

Hi Nikolay,

>

There were multiple fixes from Josef recently improving btrfs enospc
handling with tiny filesystems (which is generally not the targeted use
case of btrfs). The code lives in
https://github.com/kdave/btrfs-devel/commits/misc-next  should you want
to test it. Otherwise re-test after next merge windows when those
patches are supposed to be merged for 5.4

>


Thank for your reply, I will keep eyes on the branch.

ps: this email is my simply testcase from ltp


--Hongzhi


On 9/4/19 4:02 PM, Hongzhi, Song wrote:
> Hi ,
>
>
> *Kernel:*
>
> ÂÂÂ After v5.2-rc1, qemux86-64
>
> ÂÂÂ make -j40 ARCH=x86_64 CROSS_COMPILE=x86-64-gcc
> ÂÂÂ use qemu to bootup kernel
>
>
> *Reproduce:*
>
> ÂÂÂ There is a test case failed on btrfs but success on other 
> fs(ext4,ext3), see attachment.
>
>
> ÂÂÂ Download attachments:
>
> ÂÂÂ ÂÂÂ gcc test.c -o myout -Wall -lpthread
>
> ÂÂÂ ÂÂÂ copy myout and run.sh to your qemu same directory.
>
> ÂÂÂ ÂÂÂ on qemu:
>
> ÂÂÂ ÂÂÂ ÂÂÂ ./run.sh
>
>
> ÂÂÂ I found the block device size with btrfs set 512M will cause the 
> error.
> ÂÂÂ 256M and 1G all success.
>
>
> *Error info:*
>
> ÂÂÂ "BTRFS warning (device loop0): could not allocate space for a 
> delete; will truncate on mount"
>
>
> *Related patch:*
>
> ÂÂÂ I use git bisect to find the following patch introduces the issue.
>
> ÂÂÂ commit c8eaeac7b734347c3afba7008b7af62f37b9c140
> ÂÂÂ Author: Josef Bacik <josef@xxxxxxxxxxxxxx>
> ÂÂÂ Date:ÂÂ Wed Apr 10 15:56:10 2019 -0400
>
> ÂÂÂÂÂÂÂ btrfs: reserve delalloc metadata differently
> ÂÂÂÂÂÂÂ ...
>
>
> Anyone's reply will be appreciated.
>
> --Hongzhi