From: Jeff Layton <jlayton@primarydata.com>
To: bfields@fieldses.org
Cc: linux-nfs@vger.kernel.org, hch@infradead.org
Subject: [PATCH v2 4/4] nfsd: add appropriate __force directives to filehandle generation code
Date: Tue, 17 Jun 2014 07:44:14 -0400 [thread overview]
Message-ID: <1403005454-7319-5-git-send-email-jlayton@primarydata.com> (raw)
In-Reply-To: <1403005454-7319-1-git-send-email-jlayton@primarydata.com>
The filehandle structs all use host-endian values, but will sometimes
stuff big-endian values into those fields. This is OK since these
values are opaque to the client, but it confuses sparse. Add __force to
make it clear that we are doing this intentionally.
Signed-off-by: Jeff Layton <jlayton@primarydata.com>
---
fs/nfsd/nfsfh.c | 9 ++++++++-
fs/nfsd/nfsfh.h | 15 +++++++++++----
2 files changed, 19 insertions(+), 5 deletions(-)
diff --git a/fs/nfsd/nfsfh.c b/fs/nfsd/nfsfh.c
index ec8393418154..7e5b2d993372 100644
--- a/fs/nfsd/nfsfh.c
+++ b/fs/nfsd/nfsfh.c
@@ -162,7 +162,14 @@ static __be32 nfsd_set_fh_dentry(struct svc_rqst *rqstp, struct svc_fh *fhp)
/* deprecated, convert to type 3 */
len = key_len(FSID_ENCODE_DEV)/4;
fh->fh_fsid_type = FSID_ENCODE_DEV;
- fh->fh_fsid[0] = new_encode_dev(MKDEV(ntohl(fh->fh_fsid[0]), ntohl(fh->fh_fsid[1])));
+ /*
+ * struct knfsd_fh uses host-endian fields, which are
+ * sometimes used to hold net-endian values. This
+ * confuses sparse, so we must use __force here to
+ * keep it from complaining.
+ */
+ fh->fh_fsid[0] = new_encode_dev(MKDEV(ntohl((__force __be32)fh->fh_fsid[0]),
+ ntohl((__force __be32)fh->fh_fsid[1])));
fh->fh_fsid[1] = fh->fh_fsid[2];
}
data_left -= len;
diff --git a/fs/nfsd/nfsfh.h b/fs/nfsd/nfsfh.h
index 2e89e70ac15c..08236d70c667 100644
--- a/fs/nfsd/nfsfh.h
+++ b/fs/nfsd/nfsfh.h
@@ -73,8 +73,15 @@ enum fsid_source {
extern enum fsid_source fsid_source(struct svc_fh *fhp);
-/* This might look a little large to "inline" but in all calls except
+/*
+ * This might look a little large to "inline" but in all calls except
* one, 'vers' is constant so moste of the function disappears.
+ *
+ * In some cases the values are considered to be host endian and in
+ * others, net endian. fsidv is always considered to be u32 as the
+ * callers don't know which it will be. So we must use __force to keep
+ * sparse from complaining. Since these values are opaque to the
+ * client, that shouldn't be a problem.
*/
static inline void mk_fsid(int vers, u32 *fsidv, dev_t dev, ino_t ino,
u32 fsid, unsigned char *uuid)
@@ -82,7 +89,7 @@ static inline void mk_fsid(int vers, u32 *fsidv, dev_t dev, ino_t ino,
u32 *up;
switch(vers) {
case FSID_DEV:
- fsidv[0] = htonl((MAJOR(dev)<<16) |
+ fsidv[0] = (__force __u32)htonl((MAJOR(dev)<<16) |
MINOR(dev));
fsidv[1] = ino_t_to_u32(ino);
break;
@@ -90,8 +97,8 @@ static inline void mk_fsid(int vers, u32 *fsidv, dev_t dev, ino_t ino,
fsidv[0] = fsid;
break;
case FSID_MAJOR_MINOR:
- fsidv[0] = htonl(MAJOR(dev));
- fsidv[1] = htonl(MINOR(dev));
+ fsidv[0] = (__force __u32)htonl(MAJOR(dev));
+ fsidv[1] = (__force __u32)htonl(MINOR(dev));
fsidv[2] = ino_t_to_u32(ino);
break;
--
1.9.3
next prev parent reply other threads:[~2014-06-17 11:44 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-06-17 11:44 [PATCH v2 0/4] nfsd: clean up sparse endianness warnings Jeff Layton
2014-06-17 11:44 ` [PATCH v2 1/4] nfsd: add __force to opaque verifier field casts Jeff Layton
2014-06-17 13:52 ` Christoph Hellwig
2014-06-17 11:44 ` [PATCH v2 2/4] nfsd: clean up sparse endianness warnings in nfscache.c Jeff Layton
2014-06-17 13:52 ` Christoph Hellwig
2014-06-17 11:44 ` [PATCH v2 3/4] nfsd: nfsd_splice_read and nfsd_readv should return __be32 Jeff Layton
2014-06-17 13:53 ` Christoph Hellwig
2014-06-17 11:44 ` Jeff Layton [this message]
2014-06-17 13:57 ` [PATCH v2 4/4] nfsd: add appropriate __force directives to filehandle generation code Christoph Hellwig
2014-06-18 15:39 ` [PATCH v2 0/4] nfsd: clean up sparse endianness warnings J. Bruce Fields
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=1403005454-7319-5-git-send-email-jlayton@primarydata.com \
--to=jlayton@primarydata.com \
--cc=bfields@fieldses.org \
--cc=hch@infradead.org \
--cc=linux-nfs@vger.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