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: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:   Tue, 24 Jul 2018 19:42:53 +0800
From:   Yunlong Song <yunlong.song@...wei.com>
To:     Chao Yu <yuchao0@...wei.com>, <jaegeuk@...nel.org>,
        <chao@...nel.org>, <yunlong.song@...oud.com>
CC:     <miaoxie@...wei.com>, <bintian.wang@...wei.com>,
        <shengyong1@...wei.com>, <heyunlei@...wei.com>,
        <linux-f2fs-devel@...ts.sourceforge.net>,
        <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2] f2fs: clear victim_secmap when section has full valid
 blocks

Just in case, or maybe this patch should be put after [patch 2/5],which 
let BG_GC avoids
skipping BG_GC victim, then SSR can also select the BG_GC victim to 
allocate data blocks,
which can make bggc selected section get back to full state.

On 2018/7/24 17:36, Chao Yu wrote:
> On 2018/7/24 17:27, Yunlong Song wrote:
>> Without this patch, f2fs only clears victim_secmap when it finds out
>> that the section has no valid blocks at all, but forgets to clear the
>> victim_secmap when the whole section has full valid blocks.
> Look this patch again, I have a question, why bggc selected section can
> get back to full state?
>
> Thanks,
>
>> Signed-off-by: Yunlong Song <yunlong.song@...wei.com>
>> ---
>>   fs/f2fs/segment.c | 3 ++-
>>   1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
>> index cfff7cf..0a79554 100644
>> --- a/fs/f2fs/segment.c
>> +++ b/fs/f2fs/segment.c
>> @@ -776,7 +776,8 @@ static void __remove_dirty_segment(struct f2fs_sb_info *sbi, unsigned int segno,
>>   		if (test_and_clear_bit(segno, dirty_i->dirty_segmap[t]))
>>   			dirty_i->nr_dirty[t]--;
>>   
>> -		if (get_valid_blocks(sbi, segno, true) == 0)
>> +		if (get_valid_blocks(sbi, segno, true) == 0 ||
>> +			get_valid_blocks(sbi, segno, true) == BLKS_PER_SEC(sbi))
>>   			clear_bit(GET_SEC_FROM_SEG(sbi, segno),
>>   						dirty_i->victim_secmap);
>>   	}
>>
>
> .
>

-- 
Thanks,
Yunlong Song


Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ