linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Dan Schatzberg <dschatzberg@fb.com>
To: Miklos Szeredi <miklos@szeredi.hu>
Cc: Tejun Heo <tj@kernel.org>, Kernel Team <Kernel-team@fb.com>,
	"linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] fuse: enable caching of symlinks
Date: Tue, 2 Oct 2018 21:33:41 +0000	[thread overview]
Message-ID: <FEA49C2E-639E-4B1E-A4A1-0E0879F1214A@fb.com> (raw)
In-Reply-To: <CAJfpegsRNv3Uby2w-B0+d4fE8ecSRkT0ef2AfQ_QnnEVT-nkmA@mail.gmail.com>

On 9/27/18, 9:59 AM, "Miklos Szeredi" <miklos@szeredi.hu> wrote:

>  On Thu, Sep 27, 2018 at 3:52 PM, Dan Schatzberg <dschatzberg@fb.com> wrote:
>> FUSE file reads are cached in the page cache, but symlink reads are
>> not. This patch enables FUSE READLINK operations to be cached which
>> can improve performance of some FUSE workloads.
>>
>> In particular, I'm working on a FUSE filesystem for access to source
>> code and discovered that about a 10% improvement to build times is
>> achieved with this patch (there are a lot of symlinks in the source
>> tree).
>>
>> Please provide feedback, I'm looking to flesh this out more.
> 
> Need to think about how/when to invalidate the cached symlink
> contents.  I think treating it like metadata (i.e. look at
> attr_timeout for validity) would be the simplest.
> 
> Thanks,
> Miklos
> 

Any further thoughts on this? Just so I can confirm my understanding,
the attached patch will unconditionally cache READLINK responses
However, based on the entry timeout provided by the LOOKUP response,
Userspace can decide to invalidate this by providing a different inode on
a subsequent LOOKUP (e.g. during fuse_dentry_revalidate())

Would you like there to be a way to invalidate the cached symlink without
 changing the inode via LOOKUP?
>>
>> Signed-off-by: Dan Schatzberg <dschatzberg@fb.com>
>> ---
>>  fs/fuse/dir.c | 70 +++++++++++++++++++++++++++++++--------------------
>>  1 file changed, 43 insertions(+), 27 deletions(-)
>>
>> diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
>> index 0979609d6eba..3c0a81ef5bca 100644
>> --- a/fs/fuse/dir.c
>> +++ b/fs/fuse/dir.c
>> @@ -1398,38 +1398,46 @@ static int fuse_readdir(struct file *file, struct dir_context *ctx)
>>         return err;
>>  }
>>
>> -static const char *fuse_get_link(struct dentry *dentry,
>> -                                struct inode *inode,
>> -                                struct delayed_call *done)
>> +static int fuse_symlink_readpage(struct file *file, struct page *page)
>>  {
>> +       struct inode *inode = page->mapping->host;
>>         struct fuse_conn *fc = get_fuse_conn(inode);
>> -       FUSE_ARGS(args);
>> -       char *link;
>> -       ssize_t ret;
>> -
>> -       if (!dentry)
>> -               return ERR_PTR(-ECHILD);
>> +       struct fuse_req *req;
>> +       int err;
>> +       size_t num_read;
>>
>> -       link = kmalloc(PAGE_SIZE, GFP_KERNEL);
>> -       if (!link)
>> -               return ERR_PTR(-ENOMEM);
>> +       err = -EIO;
>> +       if (is_bad_inode(inode))
>> +               goto out;
>>
>> -       args.in.h.opcode = FUSE_READLINK;
>> -       args.in.h.nodeid = get_node_id(inode);
>> -       args.out.argvar = 1;
>> -       args.out.numargs = 1;
>> -       args.out.args[0].size = PAGE_SIZE - 1;
>> -       args.out.args[0].value = link;
>> -       ret = fuse_simple_request(fc, &args);
>> -       if (ret < 0) {
>> -               kfree(link);
>> -               link = ERR_PTR(ret);
>> -       } else {
>> -               link[ret] = '\0';
>> -               set_delayed_call(done, kfree_link, link);
>> +       req = fuse_get_req(fc, 1);
>> +       if (IS_ERR(req)) {
>> +               err = PTR_ERR(req);
>> +               goto out;
>>         }
>> +
>> +       req->out.page_zeroing = 1;
>> +       req->out.argpages = 1;
>> +       req->num_pages = 1;
>> +       req->pages[0] = page;
>> +       req->page_descs[0].length = PAGE_SIZE - 1;
>> +       req->in.h.opcode = FUSE_READLINK;
>> +       req->in.h.nodeid = get_node_id(inode);
>> +       req->out.argvar = 1;
>> +       req->out.numargs = 1;
>> +       req->out.args[0].size = PAGE_SIZE - 1;
>> +       fuse_request_send(fc, req);
>> +       num_read = req->out.args[0].size;
>> +       err = req->out.h.error;
>> +
>> +       if (!err)
>> +               SetPageUptodate(page);
>> +
>> +       fuse_put_request(fc, req);
>>         fuse_invalidate_atime(inode);
>> -       return link;
>> +out:
>> +       unlock_page(page);
>> +       return err;
>>  }
>>
>>  static int fuse_dir_open(struct inode *inode, struct file *file)
>> @@ -1855,7 +1863,7 @@ static const struct inode_operations fuse_common_inode_operations = {
>>
>>  static const struct inode_operations fuse_symlink_inode_operations = {
>>         .setattr        = fuse_setattr,
>> -       .get_link       = fuse_get_link,
>> +       .get_link       = page_get_link,
>>         .getattr        = fuse_getattr,
>>         .listxattr      = fuse_listxattr,
>>  };
>> @@ -1871,7 +1879,15 @@ void fuse_init_dir(struct inode *inode)
>>         inode->i_fop = &fuse_dir_operations;
>>  }
>>
>> +static const struct address_space_operations fuse_symlink_aops = {
>> +       .readpage       = fuse_symlink_readpage,
>> +};
>> +
>>  void fuse_init_symlink(struct inode *inode)
>>  {
>>         inode->i_op = &fuse_symlink_inode_operations;
>> +       inode->i_data.a_ops = &fuse_symlink_aops;
>> +       mapping_set_gfp_mask(inode->i_mapping,
>> +                       mapping_gfp_constraint(inode->i_mapping,
>> +                                       ~(__GFP_FS | __GFP_HIGHMEM)));
>>  }
>> --
>> 2.17.1


  reply	other threads:[~2018-10-03  4:19 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-09-27 13:52 [PATCH] fuse: enable caching of symlinks Dan Schatzberg
2018-09-27 13:58 ` Miklos Szeredi
2018-10-02 21:33   ` Dan Schatzberg [this message]
2018-10-10  8:29     ` Miklos Szeredi
2018-10-09 14:29   ` Dan Schatzberg

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=FEA49C2E-639E-4B1E-A4A1-0E0879F1214A@fb.com \
    --to=dschatzberg@fb.com \
    --cc=Kernel-team@fb.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=miklos@szeredi.hu \
    --cc=tj@kernel.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).