linux-kernel - Re: linux-next test error

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]

Message-ID: <CAFqt6za5OvHgONOgpmhxS+YsYZyiXUhzpmOgZYyHWPHEO34QwQ@mail.gmail.com>
Date:   Thu, 6 Sep 2018 00:54:50 +0530
From:   Souptick Joarder <jrdr.linux@...il.com>
To:     "Theodore Ts'o" <tytso@....edu>, Jan Kara <jack@...e.cz>,
        syzbot+87a05ae4accd500f5242@...kaller.appspotmail.com,
        ak@...ux.intel.com, Andrew Morton <akpm@...ux-foundation.org>,
        linux-kernel@...r.kernel.org, Linux-MM <linux-mm@...ck.org>,
        mawilcox@...rosoft.com, mgorman@...hsingularity.net,
        syzkaller-bugs@...glegroups.com, tim.c.chen@...ux.intel.com,
        zwisler@...nel.org, Matthew Wilcox <willy@...radead.org>
Subject: Re: linux-next test error

On Wed, Sep 5, 2018 at 7:05 PM Theodore Y. Ts'o <tytso@....edu> wrote:
>
> On Wed, Sep 05, 2018 at 03:20:16PM +0530, Souptick Joarder wrote:
> >
> > "fs: convert return type int to vm_fault_t" is still under
> > review/discusson and not yet merge
> > into linux-next. I am not seeing it into linux-next tree.Can you
> > please share the commit id ?
>
> It's at: 83c0adddcc6ed128168e7b87eaed0c21eac908e4 in the Linux Next
> branch.
>
> Dmitry, can you try reverting this commit and see if it makes the
> problem go away?
>
> Souptick, can we just NACK this patch and completely drop it from all
> trees?

Ok, I will correct it and post v3.

>
> I think we need to be a *lot* more careful about this vm_fault_t patch
> thing.  If you can't be bothered to run xfstests, we need to introduce
> a new function which replaces block_page_mkwrite() --- and then let
> each file system try to convert over to it at their own pace, after
> they've done regression testing.
>
>                                                 - Ted

Chris has his opinion,

block_page_mkwrite is only called by ext4 and nilfs2 anyway, so
converting both callers over should not be a problem, as long as
it actually is done properly.

Matthew's opinion in other mail thread -

> +vm_fault_t block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf,
> +                      get_block_t get_block, int *err)

I don't like returning both the errno and the vm_fault_t.  To me that's a
sign we need to rethink this interface.

I have two suggestions.  First, we could allocate a new VM_FAULT_NOSPC
bit.  Second, we could repurpose one of the existing bits, such as
VM_FAULT_RETRY for this purpose.

> -int ext4_page_mkwrite(struct vm_fault *vmf)
> +vm_fault_t ext4_page_mkwrite(struct vm_fault *vmf)

I also think perhaps we could start by _not_ converting block_page_mkwrite().
Just convert ext4_page_mkwrite(), and save converting block_page_mkwrite()
for later.

Which approach Shall I take ??