linux-kernel - Re: [RFC PATCH 1/3] filemap: convert page_endio to folio

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: <04047489-e528-4451-4af2-c19bd3635e7e@samsung.com>
Date:   Fri, 17 Mar 2023 17:14:28 +0100
From:   Pankaj Raghav <p.raghav@...sung.com>
To:     Matthew Wilcox <willy@...radead.org>
CC:     Christoph Hellwig <hch@...radead.org>, <hubcap@...ibond.com>,
        <senozhatsky@...omium.org>, <martin@...ibond.com>,
        <minchan@...nel.org>, <viro@...iv.linux.org.uk>,
        <brauner@...nel.org>, <axboe@...nel.dk>,
        <akpm@...ux-foundation.org>, <linux-kernel@...r.kernel.org>,
        <linux-block@...r.kernel.org>, <linux-fsdevel@...r.kernel.org>,
        <linux-mm@...ck.org>, <gost.dev@...sung.com>, <mcgrof@...nel.org>,
        <devel@...ts.orangefs.org>
Subject: Re: [RFC PATCH 1/3] filemap: convert page_endio to folio_endio

On 2023-03-17 16:31, Matthew Wilcox wrote:
>> +
>> +       while ((folio = readahead_folio(rac))) {
>> +               folio_mark_uptodate(folio);
>> +               folio_unlock(folio);
>>         }
> 
> readahead_folio() is a bit too heavy-weight for that, IMO.  I'd do this
> as;
> 
> 	while ((folio = readahead_folio(rac))) {
> 		if (!ret)
> 			folio_mark_uptodate(folio);
> 		folio_unlock(folio);
> 	}
> 

This looks good.

> (there's no need to call folio_set_error(), nor folio_clear_uptodate())

I am trying to understand why these calls are not needed for the error case.
I see similar pattern, for e.g. in iomap_finish_folio_read() where we call these
functions for the error case.

If we don't need to call these anymore, can the mpage code also be shortened like this:

-static void mpage_end_io(struct bio *bio)
+static void mpage_read_end_io(struct bio *bio)
 {
-       struct bio_vec *bv;
-       struct bvec_iter_all iter_all;
+       struct folio_iter fi;
+       int err = blk_status_to_errno(bio->bi_status);

-       bio_for_each_segment_all(bv, bio, iter_all) {
-               struct page *page = bv->bv_page;
-               page_endio(page, bio_op(bio),
-                          blk_status_to_errno(bio->bi_status));
+       bio_for_each_folio_all(fi, bio) {
+               struct folio *folio = fi.folio;
+
+               if (!err)
+                       folio_mark_uptodate(folio);
+               folio_unlock(folio);
+       }
+
+       bio_put(bio);
+}
+
+static void mpage_write_end_io(struct bio *bio)
+{
....
+