Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Sat, 31 Aug 2019 00:48:27 +0800
From:   Gao Xiang <gaoxiang25@huawei.com>
To:     Christoph Hellwig <hch@infradead.org>
CC:     Chao Yu <yuchao0@huawei.com>,
        Dan Carpenter <dan.carpenter@oracle.com>,
        Joe Perches <joe@perches.com>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        <devel@driverdev.osuosl.org>, LKML <linux-kernel@vger.kernel.org>,
        <linux-erofs@lists.ozlabs.org>, Chao Yu <chao@kernel.org>,
        Miao Xie <miaoxie@huawei.com>, <weidu.du@huawei.com>,
        "Fang Wei" <fangwei1@huawei.com>
Subject: Re: [PATCH v3 7/7] erofs: redundant assignment in
 __erofs_get_meta_page()
Message-ID: <20190830164827.GA107220@architecture4>
References: <20190830032006.GA20217@architecture4>
 <20190830033643.51019-1-gaoxiang25@huawei.com>
 <20190830033643.51019-7-gaoxiang25@huawei.com>
 <20190830162812.GA10694@infradead.org>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Disposition: inline
In-Reply-To: <20190830162812.GA10694@infradead.org>
User-Agent: Mutt/1.9.4 (2018-02-28)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi Christoph,

On Fri, Aug 30, 2019 at 09:28:12AM -0700, Christoph Hellwig wrote:
> > -		err = bio_add_page(bio, page, PAGE_SIZE, 0);
> > -		if (err != PAGE_SIZE) {
> > +		if (bio_add_page(bio, page, PAGE_SIZE, 0) != PAGE_SIZE) {
> >  			err = -EFAULT;
> >  			goto err_out;
> >  		}
> 
> This patch looks like an improvement.  But looking at that whole
> area just makes me cringe.

OK, I agree with you, I will improve it or just kill them all with
new iomap approach after it supports tail-end packing inline.

> 
> Why is there __erofs_get_meta_page with the two weird booleans instead
> of a single erofs_get_meta_page that gets and gfp_t for additional
> flags and an unsigned int for additional bio op flags.

I agree with you. Thanks for your suggestion.

> 
> Why do need ioprio support to start with?  Seeing that in a new
> fs look kinda odd.  Do you have benchmarks that show the difference?

I don't have some benchmark for all of these, can I just set
REQ_PRIO for all metadata? is that reasonable?
Could you kindly give some suggestion on this?

> 
> That function then calls erofs_grab_bio, which tries to handle a
> bio_alloc failure, except that the function will not actually fail
> due the mempool backing it.  It also seems like and awfully
> huge function to inline.

OK, I will simplify it. Thanks for your suggestion.

> 
> Why is there __submit_bio which really just obsfucates what is
> going on?  Also why is __submit_bio using bio_set_op_attrs instead
> of opencode it as the comment right next to it asks you to?

Originally, mainly due to backport consideration since some
of our smartphones use 3.x kernel as well...

> 
> Also I really don't understand why you can't just use read_cache_page
> or even read_cache_page_gfp instead of __erofs_get_meta_page.
> That function is a whole lot of duplication of functionality shared
> by a lot of other file systems.

OK, I have to admit, that code was originally just copied from f2fs
with some modification (maybe it's not a good example for us).

Thanks,
Gao Xiang