linux-ext4 - RE: Re: [PATCH] ext4: change sequential discard handling on commit complete phase into parallel manner

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:   Wed, 31 May 2017 00:49:26 +0000
From:   Daeho Jeong <daeho.jeong@...sung.com>
To:     Jan Kara <jack@...e.cz>, Daeho Jeong <daeho.jeong@...sung.com>
CC:     "jack@...e.com" <jack@...e.com>,
        "hch@...radead.org" <hch@...radead.org>,
        "tytso@....edu" <tytso@....edu>,
        "linux-ext4@...r.kernel.org" <linux-ext4@...r.kernel.org>
Subject: RE: Re: [PATCH] ext4: change sequential discard handling on commit
 complete phase into parallel manner

Hi Jan,

> Hum, these games with several callbacks, lists, etc. look awkward and
> unnecessary. It think they mostly come from the fact that we call separate
> freeing callback for each extent to free which doesn't fit the needs of
> async discard well.
 
> So instead of adding post_cb_list and several callback functions, it would
> seem easier to have just one callback structure instead of one for every
> extent. Then the structure would contain a list of extents that need to be
> freed freed. So something like:
 
> struct ext4_free_data {
>         struct ext4_journal_cb_entry efd_jce;
>         struct list_head efd_extents;
> }
 
> struct ext4_freed_extent {
>         struct list_head efe_list;
>         struct rb_node efe_node;
>         ext4_group_t efe_group;
>         ext4_grpblk_t efe_start_cluster;
>         ext4_grpblk_t efe_count;
>         tid_t efe_tid;
> }
 
> When commit happens, we can just walk the efd_extents list while efe_tid is
> equal tid of the transaction for which the callback was called and submit all
> discard requests. You can use bio chaining implemented in
> __blkdev_issue_discard() which XFS already uses and so the result of all
> the discards you submit will be just one bio. Then you walk the list of
> extents again and free them in the buddy bitmaps. And finally, you wait for
> the bio to complete. All will be then happening in one function and it will
> be much easier to understand.

It's right. the patch didn't look neat because of a few callbacks and the
post callback list. I will modify the patch as your suggestion. It will
look better.

Thank you very much. :-)