linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Al Viro <viro@ZenIV.linux.org.uk>
To: Tomeu Vizoso <tomeu@tomeuvizoso.net>
Cc: "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Neil Brown <neilb@suse.com>,
	linux-fsdevel@vger.kernel.org
Subject: Re: [PATCH v2 06/11] don't put symlink bodies in pagecache into highmem
Date: Thu, 14 Jan 2016 21:02:23 +0000	[thread overview]
Message-ID: <20160114210223.GA17997@ZenIV.linux.org.uk> (raw)
In-Reply-To: <CAAObsKCz_KubDyfVYfogCenC4XCziKJcaQYEgj4E_Bir+v517A@mail.gmail.com>

On Thu, Jan 14, 2016 at 08:15:57PM +0100, Tomeu Vizoso wrote:
> On 14 January 2016 at 18:13, Al Viro <viro@zeniv.linux.org.uk> wrote:
> > On Thu, Jan 14, 2016 at 05:57:42PM +0100, Tomeu Vizoso wrote:
> >> Here it is:
> >>
> >> [  170.715356] inode: ec8c30b0, pages: 1
> >> [  170.719014] page_address:   (null)
> >>
> >> https://lava.collabora.co.uk/scheduler/job/127698/log_file
> >
> >         Lovely...  And that looks like the first time that inode hits
> > nfs_get_link().  Ho-hum...
> >
> >         Could you add WARN_ON(inode->i_mapping.nrpages) in inode_nohighmem()
> > and see if that triggers?  It really shouldn't (we hit it after iget5_locked()
> 
> Indeed :(
> 
> https://lava.collabora.co.uk/scheduler/job/127782/log_file

OK...  Unless I'm misreading that, we have
	* inode->i_data.flags set to GFP_USER, with no pages present in
there.
	* at some later point nfs_get_link() is called on that inode
(for the first time) and sees a page with logical offset 0 already present
in there, that page being a highmem one.

That would certainly suffice for the things to blow up...

Let's try this: in the beginning of __add_to_page_cache_locked() add
	VM_BUG_ON_PAGE(PageHighMem(page) & !(mapping->flags & __GFP_HIGHMEM),
			page);
and see if that triggers.

<pokes around>

Arrrgh.  Try this:

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index ce5a218..8a05309 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1894,15 +1894,14 @@ int nfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname)
 	attr.ia_mode = S_IFLNK | S_IRWXUGO;
 	attr.ia_valid = ATTR_MODE;
 
-	page = alloc_page(GFP_HIGHUSER);
+	page = alloc_page(GFP_USER);
 	if (!page)
 		return -ENOMEM;
 
-	kaddr = kmap_atomic(page);
+	kaddr = page_address(page);
 	memcpy(kaddr, symname, pathlen);
 	if (pathlen < PAGE_SIZE)
 		memset(kaddr + pathlen, 0, PAGE_SIZE - pathlen);
-	kunmap_atomic(kaddr);
 
 	trace_nfs_symlink_enter(dir, dentry);
 	error = NFS_PROTO(dir)->symlink(dir, dentry, page, pathlen, &attr);

  reply	other threads:[~2016-01-14 21:02 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-11-17 22:57 [PATCHSET] ->follow_link() without dropping from RCU mode Al Viro
2015-11-17 23:00 ` [PATCH 01/10] switch befs long symlinks to page_symlink_operations Al Viro
2015-11-17 23:00 ` [PATCH 02/10] logfs: don't duplicate page_symlink_inode_operations Al Viro
2015-11-17 23:00 ` [PATCH 03/10] udf: " Al Viro
2015-11-17 23:00 ` [PATCH 04/10] ufs: get rid of ->setattr() for symlinks Al Viro
2015-11-17 23:00 ` [PATCH 05/10] namei: page_getlink() and page_follow_link_light() are the same thing Al Viro
2015-11-17 23:00 ` [PATCH 06/10] [vfs] don't put symlink bodies in pagecache into highmem Al Viro
2015-11-19 23:02   ` Dave Chinner
2015-11-17 23:00 ` [PATCH 07/10] [vfs] replace ->follow_link() with new method that could stay in RCU mode Al Viro
2015-11-17 23:00 ` [PATCH 08/10] teach page_get_link() to work " Al Viro
2015-11-17 23:00 ` [PATCH 09/10] teach shmem_get_link() " Al Viro
2015-11-17 23:00 ` [PATCH 10/10] teach proc_self_get_link()/proc_thread_self_get_link() " Al Viro
2015-12-09  5:32 ` [PATCHSET v2] ->follow_link() without dropping from " Al Viro
2015-12-09  5:34   ` [PATCH v2 01/11] switch befs long symlinks to page_symlink_operations Al Viro
2015-12-09  5:34   ` [PATCH v2 02/11] logfs: don't duplicate page_symlink_inode_operations Al Viro
2015-12-09  5:34   ` [PATCH v2 03/11] udf: " Al Viro
2015-12-09  5:34   ` [PATCH v2 04/11] ufs: get rid of ->setattr() for symlinks Al Viro
2015-12-09  5:34   ` [PATCH v2 05/11] namei: page_getlink() and page_follow_link_light() are the same thing Al Viro
2015-12-09  5:34   ` [PATCH v2 06/11] don't put symlink bodies in pagecache into highmem Al Viro
2016-01-14 13:22     ` Tomeu Vizoso
2016-01-14 15:25       ` Al Viro
2016-01-14 15:58         ` Tomeu Vizoso
2016-01-14 16:23           ` Al Viro
2016-01-14 16:57             ` Tomeu Vizoso
2016-01-14 17:13               ` Al Viro
2016-01-14 19:15                 ` Tomeu Vizoso
2016-01-14 21:02                   ` Al Viro [this message]
2016-01-14 21:40                     ` Linus Torvalds
2016-01-14 22:25                       ` Al Viro
2016-01-14 23:33                         ` Al Viro
2016-01-14 23:58                         ` Linus Torvalds
2016-01-15  0:05                           ` Al Viro
2015-12-09  5:34   ` [PATCH v2 07/11] replace ->follow_link() with new method that could stay in RCU mode Al Viro
2015-12-09  5:34   ` [PATCH v2 08/11] teach page_get_link() to work " Al Viro
2015-12-09  5:34   ` [PATCH v2 09/11] teach shmem_get_link() " Al Viro
2015-12-09  5:34   ` [PATCH v2 10/11] teach proc_self_get_link()/proc_thread_self_get_link() " Al Viro
2015-12-09  5:34   ` [PATCH v2 11/11] teach nfs_get_link() " Al Viro
2015-12-09 17:24   ` [PATCHSET v2] ->follow_link() without dropping from " Linus Torvalds
2015-12-09 18:23     ` Al Viro
2015-12-10  0:10       ` Al Viro
2015-12-10  2:40         ` Al Viro
2015-12-11  1:54           ` Al Viro
2015-12-11  7:49             ` Rasmus Villemoes
2015-12-11 23:16               ` Al Viro
2015-12-12  2:00                 ` Al Viro
2015-12-13 18:43                   ` Rasmus Villemoes
2015-12-13  3:47             ` Al Viro
2015-12-09 21:57   ` NeilBrown

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=20160114210223.GA17997@ZenIV.linux.org.uk \
    --to=viro@zeniv.linux.org.uk \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=neilb@suse.com \
    --cc=tomeu@tomeuvizoso.net \
    --cc=torvalds@linux-foundation.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).