Hi
fs/nfsd/nfssvc.c: In function 'set_max_drc':
fs/nfsd/nfssvc.c:240: error: 'NFSD_DRC_SIZE_SHIFT' undeclared
CONFIG_NFSD_V4 is not set
On Apr. 06, 2009, 20:11 +0300, "J. Bruce Fields" <[email protected]> wrote:
> On Mon, Apr 06, 2009 at 11:53:28AM +0300, Benny Halevy wrote:
>> On Apr. 06, 2009, 10:27 +0300, Alexander Beregalov <[email protected]> wrote:
>>> Hi
>>>
>>> fs/nfsd/nfssvc.c: In function 'set_max_drc':
>>> fs/nfsd/nfssvc.c:240: error: 'NFSD_DRC_SIZE_SHIFT' undeclared
>>>
>>> CONFIG_NFSD_V4 is not set
>> Hi Alexander,
>>
>> Thanks for reporting this!
>>
>> Andy, Bruce: please see attached 2 patches fixing compile/link errors
>> with DRC under !defined(CONFIG_NFSD_V4):
>>
>> [PATCH 1/2] SQUASHEM: nfsd41: define NFSD_DRC_SIZE_SHIFT in set_max_drc
>> [PATCH 2/2] SQUASHME: nfsd41: define nfsd4_set_statp as noop for !CONFIG_NFSD_V4
>
> Thanks, applied.
>
> Committing on top (not squashing) since I'd rather not rewrite history
> on a branch I've already sent a pull request for.
Cool. Thanks!
I wasn't aware that you already sent a pull request already...
I got it from mainline now and I'm rebasing and testing
the rest of our stuff on top of it.
>
> (In fact, I'll try to stop rebasing those for-next branches at all this
> time around.)
I'm with you on that.
For 2.6.31, I'd like to send easy-to-swallow patch sets
that we can review and agree upon well ahead of the merge window
(i.e. start, e.g. with the backchannel stuff shortly after 2.6.30-rc1
is cut) and once we're in agreement on it we can put it in linux-next
to be visible to others and get some soak time. This will also
be a good time to freeze it, so it won't be rebased, unless something
important enough requires that. How does that sound?
I think that the same strategy should work for the client side too.
Trond what do you think?
Benny
>
> --b.
On Tue, Apr 7, 2009 at 2:53 PM, Benny Halevy <[email protected]> wrote:
> On Apr. 06, 2009, 20:11 +0300, "J. Bruce Fields" <[email protected]> wrote:
>> On Mon, Apr 06, 2009 at 11:53:28AM +0300, Benny Halevy wrote:
>>> On Apr. 06, 2009, 10:27 +0300, Alexander Beregalov <[email protected]> wrote:
>>>> Hi
>>>>
>>>> fs/nfsd/nfssvc.c: In function 'set_max_drc':
>>>> fs/nfsd/nfssvc.c:240: error: 'NFSD_DRC_SIZE_SHIFT' undeclared
>>>>
>>>> CONFIG_NFSD_V4 is not set
>>> Hi Alexander,
>>>
>>> Thanks for reporting this!
>>>
>>> Andy, Bruce: please see attached 2 patches fixing compile/link errors
>>> with DRC under !defined(CONFIG_NFSD_V4):
>>>
>>> [PATCH 1/2] SQUASHEM: nfsd41: define NFSD_DRC_SIZE_SHIFT in set_max_drc
>>> [PATCH 2/2] SQUASHME: nfsd41: define nfsd4_set_statp as noop for !CONFIG_NFSD_V4
>>
>> Thanks, applied.
>>
>> Committing on top (not squashing) since I'd rather not rewrite history
>> on a branch I've already sent a pull request for.
>
> Cool. Thanks!
> I wasn't aware that you already sent a pull request already...
> I got it from mainline now and I'm rebasing and testing
> the rest of our stuff on top of it.
>
>>
>> (In fact, I'll try to stop rebasing those for-next branches at all this
>> time around.)
>
> I'm with you on that.
>
> For 2.6.31, I'd like to send easy-to-swallow patch sets
> that we can review and agree upon well ahead of the merge window
> (i.e. start, e.g. with the backchannel stuff shortly after 2.6.30-rc1
> is cut) and once we're in agreement on it we can put it in linux-next
> to be visible to others and get some soak time.
This sounds good to me. I'm continuing with Bruce's review comments WRT the DRC.
> This will also
> be a good time to freeze it, so it won't be rebased, unless something
> important enough requires that. How does that sound?
>
> I think that the same strategy should work for the client side too.
> Trond what do you think?
>
> Benny
>
>>
>> --b.
> _______________________________________________
> pNFS mailing list
> [email protected]
> http://linux-nfs.org/cgi-bin/mailman/listinfo/pnfs
>
On Tue, Apr 07, 2009 at 09:53:38PM +0300, Benny Halevy wrote:
> On Apr. 06, 2009, 20:11 +0300, "J. Bruce Fields" <[email protected]> wrote:
> > On Mon, Apr 06, 2009 at 11:53:28AM +0300, Benny Halevy wrote:
> >> On Apr. 06, 2009, 10:27 +0300, Alexander Beregalov <[email protected]> wrote:
> >>> Hi
> >>>
> >>> fs/nfsd/nfssvc.c: In function 'set_max_drc':
> >>> fs/nfsd/nfssvc.c:240: error: 'NFSD_DRC_SIZE_SHIFT' undeclared
> >>>
> >>> CONFIG_NFSD_V4 is not set
> >> Hi Alexander,
> >>
> >> Thanks for reporting this!
> >>
> >> Andy, Bruce: please see attached 2 patches fixing compile/link errors
> >> with DRC under !defined(CONFIG_NFSD_V4):
> >>
> >> [PATCH 1/2] SQUASHEM: nfsd41: define NFSD_DRC_SIZE_SHIFT in set_max_drc
> >> [PATCH 2/2] SQUASHME: nfsd41: define nfsd4_set_statp as noop for !CONFIG_NFSD_V4
> >
> > Thanks, applied.
> >
> > Committing on top (not squashing) since I'd rather not rewrite history
> > on a branch I've already sent a pull request for.
>
> Cool. Thanks!
> I wasn't aware that you already sent a pull request already...
> I got it from mainline now and I'm rebasing and testing
> the rest of our stuff on top of it.
>
> >
> > (In fact, I'll try to stop rebasing those for-next branches at all this
> > time around.)
>
> I'm with you on that.
>
> For 2.6.31, I'd like to send easy-to-swallow patch sets
> that we can review and agree upon well ahead of the merge window
> (i.e. start, e.g. with the backchannel stuff shortly after 2.6.30-rc1
> is cut) and once we're in agreement on it we can put it in linux-next
> to be visible to others and get some soak time. This will also
> be a good time to freeze it, so it won't be rebased,
Note that it's OK to rebase things in -next; so it doesn't have to be
frozen till it goes into one of our trees. (And I've tried once or
twice to maintain the discpline of not rebasing stuff in my tree and
have failed.... I'll make a better effort this time.)
> unless something
> important enough requires that. How does that sound?
Sounds fine. If you send in things as they're ready, I'll try to turn
them around a little faster (within a few days or a week anyway).
--b.
> I think that the same strategy should work for the client side too.
> Trond what do you think?
>
> Benny
>
> >
> > --b.
On Apr. 06, 2009, 10:27 +0300, Alexander Beregalov <[email protected]> wrote:
> Hi
>
> fs/nfsd/nfssvc.c: In function 'set_max_drc':
> fs/nfsd/nfssvc.c:240: error: 'NFSD_DRC_SIZE_SHIFT' undeclared
>
> CONFIG_NFSD_V4 is not set
Hi Alexander,
Thanks for reporting this!
Andy, Bruce: please see attached 2 patches fixing compile/link errors
with DRC under !defined(CONFIG_NFSD_V4):
[PATCH 1/2] SQUASHEM: nfsd41: define NFSD_DRC_SIZE_SHIFT in set_max_drc
[PATCH 2/2] SQUASHME: nfsd41: define nfsd4_set_statp as noop for !CONFIG_NFSD_V4
Benny
Fixes the following compiler error:
fs/nfsd/nfssvc.c: In function 'set_max_drc':
fs/nfsd/nfssvc.c:240: error: 'NFSD_DRC_SIZE_SHIFT' undeclared
CONFIG_NFSD_V4 is not set
Reported-by: Alexander Beregalov <[email protected]>
Signed-off-by: Benny Halevy <[email protected]>
---
fs/nfsd/nfssvc.c | 2 ++
include/linux/nfsd/nfsd.h | 3 ---
2 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
index e9d5773..469c931 100644
--- a/fs/nfsd/nfssvc.c
+++ b/fs/nfsd/nfssvc.c
@@ -236,6 +236,8 @@ void nfsd_reset_versions(void)
*/
static void set_max_drc(void)
{
+ /* The percent of nr_free_buffer_pages used by the V4.1 server DRC */
+ #define NFSD_DRC_SIZE_SHIFT 7
nfsd_serv->sv_drc_max_pages = nr_free_buffer_pages()
>> NFSD_DRC_SIZE_SHIFT;
nfsd_serv->sv_drc_pages_used = 0;
diff --git a/include/linux/nfsd/nfsd.h b/include/linux/nfsd/nfsd.h
index 0ec4d14..2b49d67 100644
--- a/include/linux/nfsd/nfsd.h
+++ b/include/linux/nfsd/nfsd.h
@@ -333,9 +333,6 @@ extern struct timeval nfssvc_boot;
#define NFSD_LEASE_TIME (nfs4_lease_time())
#define NFSD_LAUNDROMAT_MINTIMEOUT 10 /* seconds */
-/* The percent of nr_free_buffer_pages used by the V4.1 server DRC */
-#define NFSD_DRC_SIZE_SHIFT 7
-
/*
* The following attributes are currently not supported by the NFSv4 server:
* ARCHIVE (deprecated anyway)
--
1.6.2.1
Fixes following modpost error:
ERROR: "nfsd4_set_statp" [fs/nfsd/nfsd.ko] undefined!
Signed-off-by: Benny Halevy <[email protected]>
---
include/linux/nfsd/cache.h | 7 +++++++
1 files changed, 7 insertions(+), 0 deletions(-)
diff --git a/include/linux/nfsd/cache.h b/include/linux/nfsd/cache.h
index a59a2df..5bccaab 100644
--- a/include/linux/nfsd/cache.h
+++ b/include/linux/nfsd/cache.h
@@ -75,6 +75,13 @@ int nfsd_reply_cache_init(void);
void nfsd_reply_cache_shutdown(void);
int nfsd_cache_lookup(struct svc_rqst *, int);
void nfsd_cache_update(struct svc_rqst *, int, __be32 *);
+
+#ifdef CONFIG_NFSD_V4
void nfsd4_set_statp(struct svc_rqst *rqstp, __be32 *statp);
+#else /* CONFIG_NFSD_V4 */
+static inline void nfsd4_set_statp(struct svc_rqst *rqstp, __be32 *statp)
+{
+}
+#endif /* CONFIG_NFSD_V4 */
#endif /* NFSCACHE_H */
--
1.6.2.1
On Mon, Apr 06, 2009 at 11:53:28AM +0300, Benny Halevy wrote:
> On Apr. 06, 2009, 10:27 +0300, Alexander Beregalov <[email protected]> wrote:
> > Hi
> >
> > fs/nfsd/nfssvc.c: In function 'set_max_drc':
> > fs/nfsd/nfssvc.c:240: error: 'NFSD_DRC_SIZE_SHIFT' undeclared
> >
> > CONFIG_NFSD_V4 is not set
>
> Hi Alexander,
>
> Thanks for reporting this!
>
> Andy, Bruce: please see attached 2 patches fixing compile/link errors
> with DRC under !defined(CONFIG_NFSD_V4):
>
> [PATCH 1/2] SQUASHEM: nfsd41: define NFSD_DRC_SIZE_SHIFT in set_max_drc
> [PATCH 2/2] SQUASHME: nfsd41: define nfsd4_set_statp as noop for !CONFIG_NFSD_V4
Thanks, applied.
Committing on top (not squashing) since I'd rather not rewrite history
on a branch I've already sent a pull request for.
(In fact, I'll try to stop rebasing those for-next branches at all this
time around.)
--b.