From: Jens Axboe <axboe@suse.de>
To: Andrew Morton <akpm@osdl.org>
Cc: linux-kernel@vger.kernel.org, torvalds@osdl.org
Subject: Re: [PATCH] splice support #2
Date: Thu, 30 Mar 2006 14:30:52 +0200 [thread overview]
Message-ID: <20060330123051.GC13476@suse.de> (raw)
In-Reply-To: <20060330123011.GB13476@suse.de>
On Thu, Mar 30 2006, Jens Axboe wrote:
> On Thu, Mar 30 2006, Jens Axboe wrote:
> > On Thu, Mar 30 2006, Andrew Morton wrote:
> > > Jens Axboe <axboe@suse.de> wrote:
> > > >
> > > > > > +static void *page_cache_pipe_buf_map(struct file *file,
> > > > > > + struct pipe_inode_info *info,
> > > > > > + struct pipe_buffer *buf)
> > > > > > +{
> > > > > > + struct page *page = buf->page;
> > > > > > +
> > > > > > + lock_page(page);
> > > > > > +
> > > > > > + if (!PageUptodate(page)) {
> > > > >
> > > > > || page->mapping == NULL
> > > >
> > > > Fixed
> > >
> > > Actually that wasn't right. If the page was truncated then we shouldn't
> > > return -EIO to userspace. We should return zero, as this is the first
> > > page.
> >
> > _if_ this is the first page, then agree.
> >
> > > Which means either returning an ERR_PTR here for -EIO or re-checking i_size
> > > in the caller. Maybe the latter?
> >
> > The latter sounds better, I'll add that.
>
> Actually that's a little hard, since you don't know what state buf->page
> is in - it might be page cache, it might not be. So I opted for the
> ERR_PTR approach, just returning -ENODATA on !page->mapping.
Ala
diff --git a/fs/splice.c b/fs/splice.c
index 7927dff..68b98bb 100644
--- a/fs/splice.c
+++ b/fs/splice.c
@@ -47,9 +47,14 @@ static void *page_cache_pipe_buf_map(str
lock_page(page);
- if (!PageUptodate(page) || !page->mapping) {
+ if (!PageUptodate(page)) {
+ unlock_page(page);
+ return ERR_PTR(-EIO);
+ }
+
+ if (!page->mapping) {
unlock_page(page);
- return NULL;
+ return ERR_PTR(-ENODATA);
}
return kmap(buf->page);
@@ -279,6 +293,7 @@ static int pipe_to_sendpage(struct pipe_
loff_t pos = sd->pos;
unsigned int offset;
ssize_t ret;
+ void *ptr;
/*
* sub-optimal, but we are limited by the pipe ->map. we don't
@@ -286,8 +301,9 @@ static int pipe_to_sendpage(struct pipe_
* have the page pinned if the pipe page originates from the
* page cache
*/
- if (!buf->ops->map(file, info, buf))
- return -EIO;
+ ptr = buf->ops->map(file, info, buf);
+ if (IS_ERR(ptr))
+ return PTR_ERR(ptr);
offset = pos & ~PAGE_CACHE_MASK;
@@ -334,8 +350,8 @@ static int pipe_to_file(struct pipe_inod
* after this, page will be locked and unmapped
*/
src = buf->ops->map(file, info, buf);
- if (!src)
- return -EIO;
+ if (IS_ERR(src))
+ return PTR_ERR(src);
index = sd->pos >> PAGE_CACHE_SHIFT;
offset = sd->pos & ~PAGE_CACHE_MASK;
@@ -436,7 +452,7 @@ static ssize_t move_from_pipe(struct ino
err = actor(info, buf, &sd);
if (err) {
- if (!ret)
+ if (!ret && err != -ENODATA)
ret = err;
break;
--
Jens Axboe
next prev parent reply other threads:[~2006-03-30 12:30 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-03-30 10:06 [PATCH] splice support #2 Jens Axboe
2006-03-30 10:16 ` Andrew Morton
2006-03-30 10:24 ` Jens Axboe
2006-03-30 11:16 ` Andrew Morton
2006-03-30 11:55 ` Jens Axboe
2006-03-30 12:30 ` Jens Axboe
2006-03-30 12:30 ` Jens Axboe [this message]
2006-03-30 19:19 ` Andrew Morton
2006-03-30 12:00 ` Ingo Molnar
2006-03-30 12:05 ` Jens Axboe
2006-03-30 12:10 ` Ingo Molnar
2006-03-30 12:16 ` Jens Axboe
2006-03-30 12:38 ` Ingo Molnar
2006-03-30 12:42 ` Jens Axboe
2006-03-30 12:42 ` Ingo Molnar
2006-03-30 13:02 ` Jens Axboe
2006-03-30 14:20 ` Christoph Hellwig
2006-03-30 17:02 ` Linus Torvalds
2006-03-30 17:17 ` Linus Torvalds
2006-03-31 20:38 ` Hua Zhong
2006-03-31 20:49 ` Linus Torvalds
2006-03-30 20:48 ` Jeff Garzik
2006-03-30 21:16 ` Linus Torvalds
2006-03-31 0:59 ` Nick Piggin
2006-03-31 2:43 ` Andrew Morton
2006-03-31 2:51 ` Andrew Morton
2006-03-31 3:20 ` Nick Piggin
2006-03-31 6:35 ` Christoph Hellwig
2006-03-31 7:09 ` Ingo Molnar
2006-04-02 22:33 ` Pavel Machek
2006-03-31 12:46 ` Bernd Petrovitsch
2006-03-31 9:56 ` Jens Axboe
2006-03-31 12:18 ` Ingo Molnar
2006-03-31 12:23 ` Jens Axboe
2006-03-31 12:26 ` Jens Axboe
2006-03-31 12:47 ` Ingo Molnar
2006-03-31 18:18 ` Jens Axboe
2006-03-31 12:27 ` Ingo Molnar
-- strict thread matches above, loose matches on Subject: below --
2006-03-31 0:03 linux
2006-03-31 6:06 tridge
2006-03-31 6:59 ` Antonio Vargas
2006-03-31 7:37 ` tridge
2006-03-31 9:57 ` Jens Axboe
2006-03-31 19:11 ` Linus Torvalds
2006-03-31 19:40 ` Jens Axboe
2006-04-04 17:16 ` Andy Lutomirski
2006-04-04 17:34 ` Jens Axboe
[not found] <5W2gv-Tp-19@gated-at.bofh.it>
[not found] ` <5W48C-3KW-17@gated-at.bofh.it>
[not found] ` <5W48D-3KW-21@gated-at.bofh.it>
[not found] ` <5W8OT-2ms-17@gated-at.bofh.it>
[not found] ` <5WcfS-7x9-23@gated-at.bofh.it>
[not found] ` <5WcIT-8nr-13@gated-at.bofh.it>
[not found] ` <5Wm5I-53z-7@gated-at.bofh.it>
[not found] ` <5XjoS-8t9-11@gated-at.bofh.it>
2006-04-03 12:39 ` Bodo Eggert
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20060330123051.GC13476@suse.de \
--to=axboe@suse.de \
--cc=akpm@osdl.org \
--cc=linux-kernel@vger.kernel.org \
--cc=torvalds@osdl.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).