Introduction
Subsystems (4)
PATCH 1/1 Introduction
Following are 12 patches for nfsd in 2.6.11-rc2-mm2
- Fix minor problem with '..' access interacting with the "nohide" export option.
- reduce stack space usage in find_exported_dentry
- 6 patches from Bruce Fields to improve authentication handling for callbacks.
- Fix problem that stopped read access to append-only files.
- Fix bug where readdir replies are cache badly.
- Add lots of "static" declarations where appropriate
- Change nfscache to use list.h macros and smaller kmallocs
I believe they are all ready for ready for 2.6.11.
Number 10 (readdir fix) is probably the most critical (and most trivial). The append-only would be next in line.
The 6 from Bruce fix a real performance problem with locking.
The rest could be safely delayed if that seem appropriate, but they are either quite trivial or (for the last one) quite well tested.
NeilBrown
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 exportfs: reduce stack usage
find_exported_dentry() declares
char nbuf[NAME_MAX+1];
in 2 separate places, and gcc allocates space on the stack for both of them. Having just one of them will suffice, if we can put put with its scope.
Reduces function stack usage on x86-32 from 0x230 to 0x130.
### Diffstat output
./fs/exportfs/expfs.c | 4 +--- 1 files changed, 1 insertion(+), 3 deletions(-)
fs/exportfs/expfs.c renamed from fs/exportfs/expfs.c~current~ +1 -3
@@ -55,7 +55,7 @@ find_exported_dentry(struct super_blockstruct list_head *le, *head;struct dentry *toput =NULL ;int noprogress; - +char nbuf[NAME_MAX +1 ];/* * Attempt to find the inode.
@@ -176,7 +176,6 @@ find_exported_dentry(struct super_block*/ struct dentry *ppd;struct dentry *npd; -char nbuf[NAME_MAX +1 ];down (&pd->d_inode->i_sem); ppd =CALL (nops,get_parent)(pd);
@@ -241,7 +240,6 @@ find_exported_dentry(struct super_block/* if we weren't after a directory, have one more step to go */ if (result != target_dir) {struct dentry *nresult; -char nbuf[NAME_MAX +1 ]; err =CALL (nops,get_name)(target_dir, nbuf, result);if (!err) {down (&target_dir->d_inode->i_sem);
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 svcrpc: rename pg_authenticate
In reply to NeilBrownLater patches remove pg_authenticate and use the name for a different purpose; so rename it to pg_authenticate_obsolete for now.
### Diffstat output
./fs/nfs/callback.c | 2 +- ./include/linux/sunrpc/svc.h | 2 +- ./net/sunrpc/svc.c | 4 ++-- 3 files changed, 4 insertions(+), 4 deletions(-)
fs/nfs/callback.c renamed from fs/nfs/callback.c~current~ +1 -1
@@ -321,5 +321,5 @@ static struct svc_program nfs4_callback_ .pg_name = "NFSv4 callback", /* service name */ .pg_class = "nfs", /* authentication class */ .pg_stats = &nfs4_callback_stats, - .pg_authenticate = nfs_callback_auth, + .pg_authenticate_obsolete = nfs_callback_auth, };
include/linux/sunrpc/svc.h renamed from include/linux/sunrpc/svc.h~current~ +1 -1
@@ -274,7 +274,7 @@ struct svc_program {char * pg_class;/* class name: services sharing authentication */ struct svc_stat * pg_stats;/* rpc statistics */ /* Override authentication. NULL means use default */ -int (*pg_authenticate)(struct svc_rqst *,u32 *); +int (*pg_authenticate_obsolete)(struct svc_rqst *,u32 *); };/*
net/sunrpc/svc.c renamed from net/sunrpc/svc.c~current~ +2 -2
@@ -313,8 +313,8 @@ svc_process(struct svc_serv *serv, struc* We do this before anything else in order to get a decent * auth verifier. */ -if (progp->pg_authenticate !=NULL ) - auth_res = progp->pg_authenticate (rqstp, &auth_stat); +if (progp->pg_authenticate_obsolete !=NULL ) + auth_res = progp->pg_authenticate_obsolete (rqstp, &auth_stat);else auth_res =svc_authenticate (rqstp, &auth_stat);switch (auth_res) {
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 Allow read access over NFS to files with APPEND bit set.
In reply to NeilBrownWrite access cannot safely be allowed as NFS doesn't support append, but read access should be ok.
### Diffstat output
./fs/nfsd/vfs.c | 9 ++++++--- 1 files changed, 6 insertions(+), 3 deletions(-)
fs/nfsd/vfs.c renamed from fs/nfsd/vfs.c~current~ +6 -3
@@ -657,12 +657,15 @@ nfsd_open(struct svc_rqst *rqstp, struct dentry = fhp->fh_dentry; inode = dentry->d_inode; -/* Disallow access to files with the append-only bit set or -* with mandatory locking enabled +/* Disallow write access to files with the append-only bit set +* or any access when mandatory locking enabled */ err = nfserr_perm; -if (IS_APPEND (inode) ||IS_ISMNDLK (inode)) +if (IS_APPEND (inode) && (access &MAY_WRITE )) goto out; +if (IS_ISMNDLK (inode)) + goto out; +if (!inode->i_fop) goto out;
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 Subject: [PATCH 2.6] SGI 921857: find broken with nohide on NFSv3
In reply to NeilBrownThis patch makes "find" work when traversing nohide exports on NFSv3. The READDIRPLUS reply needs to not return a file handle for the ".." entry when the directory is a server side mountpoint, which would be the directory itself, otherwise the client remembers the wrong file handle and gets confused.
### Diffstat output
./fs/nfsd/nfs3xdr.c | 5 +++++ 1 files changed, 5 insertions(+)
fs/nfsd/nfs3xdr.c renamed from fs/nfsd/nfs3xdr.c~current~ +5 -0
@@ -843,6 +843,11 @@ compose_entry_fh(struct nfsd3_readdirresif (isdotent (name, namlen)) {if (namlen ==2 ) { dchild =dget_parent (dparent); +if (dchild == dparent) { +/* filesystem root - cannot return filehandle for ".." */ +dput (dchild); +return 1 ; + } }else dchild =dget (dparent); }else
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 svcrpc: add a per-flavor set_client method
In reply to NeilBrownAdd a set_client method to the server rpc auth_ops struct, used to set the client (for the purposes of nfsd export authorization) using flavor-specific information.
### Diffstat output
./include/linux/sunrpc/svcauth.h | 2 ++ ./net/sunrpc/auth_gss/svcauth_gss.c | 14 ++++++++++++++ ./net/sunrpc/sunrpc_syms.c | 1 + ./net/sunrpc/svcauth.c | 5 +++++ ./net/sunrpc/svcauth_unix.c | 2 ++ 5 files changed, 24 insertions(+)
include/linux/sunrpc/svcauth.h renamed from include/linux/sunrpc/svcauth.h~current~ +2 -0
@@ -92,6 +92,7 @@ struct auth_ops {int (*accept)(struct svc_rqst *rq,u32 *authp);int (*release)(struct svc_rqst *rq);void (*domain_release)(struct auth_domain *); +int (*set_client)(struct svc_rqst *rq); };#define SVC_GARBAGE 1
@@ -107,6 +108,7 @@ struct auth_ops {extern int svc_authenticate (struct svc_rqst *rqstp,u32 *authp);extern int svc_authorise (struct svc_rqst *rqstp); +extern int svc_set_client (struct svc_rqst *rqstp);extern int svc_auth_register (rpc_authflavor_t flavor,struct auth_ops *aops);extern void svc_auth_unregister (rpc_authflavor_t flavor);
net/sunrpc/auth_gss/svcauth_gss.c renamed from net/sunrpc/auth_gss/svcauth_gss.c~current~ +14 -0
@@ -730,6 +730,19 @@ struct gss_svc_data {struct rsc *rsci; }; +static int +svcauth_gss_set_client (struct svc_rqst *rqstp) +{ +struct gss_svc_data *svcdata = rqstp->rq_auth_data; +struct rsc *rsci = svcdata->rsci; +struct rpc_gss_wire_cred *gc = &svcdata->clcred; + + rqstp->rq_client =find_gss_auth_domain (rsci->mechctx, gc->gc_svc); +if (rqstp->rq_client ==NULL ) +return SVC_DENIED ; +return SVC_OK ; +} +/* * Accept an rpcsec packet. * If context establishment, punt to user space
@@ -1052,6 +1065,7 @@ static struct auth_ops svcauthops_gss = .accept = svcauth_gss_accept, .release = svcauth_gss_release, .domain_release = svcauth_gss_domain_release, + .set_client = svcauth_gss_set_client, };int
net/sunrpc/sunrpc_syms.c renamed from net/sunrpc/sunrpc_syms.c~current~ +1 -0
@@ -91,6 +91,7 @@ EXPORT_SYMBOL(svc_reserve);EXPORT_SYMBOL (svc_auth_register);EXPORT_SYMBOL (auth_domain_lookup);EXPORT_SYMBOL (svc_authenticate); +EXPORT_SYMBOL (svc_set_client);/* RPC statistics */ #ifdef CONFIG_PROC_FS
net/sunrpc/svcauth.c renamed from net/sunrpc/svcauth.c~current~ +5 -0
@@ -59,6 +59,11 @@ svc_authenticate(struct svc_rqst *rqstp,return aops->accept (rqstp, authp); } +int svc_set_client (struct svc_rqst *rqstp) +{ +return rqstp->rq_authop->set_client (rqstp); +} +/* A request, which was authenticated, has now executed. * Time to finalise the the credentials and verifier * and release and resources
net/sunrpc/svcauth_unix.c renamed from net/sunrpc/svcauth_unix.c~current~ +2 -0
@@ -429,6 +429,7 @@ struct auth_ops svcauth_null = { .flavour = RPC_AUTH_NULL, .accept = svcauth_null_accept, .release = svcauth_null_release, + .set_client = svcauth_unix_set_client, };
@@ -510,5 +511,6 @@ struct auth_ops svcauth_unix = { .accept = svcauth_unix_accept, .release = svcauth_unix_release, .domain_release = svcauth_unix_domain_release, + .set_client = svcauth_unix_set_client, };
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 Don't try to cache reply to nfsv2 readdir.
In reply to NeilBrownAs readdir returns the reply in a separate page, the cache code cannot find the reply (and it would probably be too big anyway) so falg readdir or NOCACHE
### Diffstat output
./fs/nfsd/nfsproc.c | 2 +- 1 files changed, 1 insertion(+), 1 deletion(-)
fs/nfsd/nfsproc.c renamed from fs/nfsd/nfsproc.c~current~ +1 -1
@@ -540,7 +540,7 @@ static struct svc_procedure nfsd_procedPROC (symlink, symlinkargs, void, none,RC_REPLSTAT ,ST ),PROC (mkdir, createargs, diropres, fhandle,RC_REPLBUFF ,ST +FH +AT ),PROC (rmdir, diropargs, void, none,RC_REPLSTAT ,ST ), -PROC (readdir, readdirargs, readdirres, none,RC_REPLBUFF ,0 ), +PROC (readdir, readdirargs, readdirres, none,RC_NOCACHE ,0 ), PROC(statfs, fhandle, statfsres, none, RC_NOCACHE, ST+5), };
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 svcrpc: move export table checks to a per-program pg_add_client method
In reply to NeilBrownsvcauth_null_accept() and svcauth_unix_accept() are currently hard-wired to check the source ip address on an incoming request against the export table, which make sense for nfsd but not necessarily for other rpc-based services.
So instead we have the accept() method call a program-specific pg_authenticate() method. We also move the call to this method into svc_process instead of calling it from the flavor-specific accept() routines.
### Diffstat output
./fs/lockd/svc.c | 15 +++++++++++++++ ./fs/nfsd/nfssvc.c | 2 ++ ./include/linux/sunrpc/svc.h | 1 + ./net/sunrpc/auth_gss/svcauth_gss.c | 9 ++------- ./net/sunrpc/svc.c | 12 +++++++++++- ./net/sunrpc/svcauth_unix.c | 18 ++---------------- 6 files changed, 33 insertions(+), 24 deletions(-)
fs/lockd/svc.c renamed from fs/lockd/svc.c~current~ +15 -0
@@ -403,6 +403,20 @@ static int param_set_##name(const char *return 0 ; \ } +static int lockd_authenticate (struct svc_rqst *rqstp) +{ + rqstp->rq_client =NULL ; +switch (rqstp->rq_authop->flavour) { +case RPC_AUTH_NULL : +case RPC_AUTH_UNIX : +if (rqstp->rq_proc ==0 ) +return SVC_OK ; +return svc_set_client (rqstp); + } +return SVC_DENIED ; +} + + param_set_min_max(port, int, simple_strtol, 0, 65535) param_set_min_max(grace_period, unsigned long, simple_strtoul, nlm_grace_period_min, nlm_grace_period_max)
@@ -483,4 +497,5 @@ static struct svc_program nlmsvc_program .pg_name = "lockd", /* service name */ .pg_class = "nfsd", /* share authentication with nfsd */ .pg_stats = &nlmsvc_stats, /* stats table */ + .pg_authenticate = &lockd_authenticate /* export authentication */ };
fs/nfsd/nfssvc.c renamed from fs/nfsd/nfssvc.c~current~ +2 -0
@@ -402,4 +402,6 @@ struct svc_program nfsd_program = { .pg_name = "nfsd", /* program name */ .pg_class = "nfsd", /* authentication class */ .pg_stats = &nfsd_svcstats, /* version table */ + .pg_authenticate = &svc_set_client, /* export authentication */ + };
include/linux/sunrpc/svc.h renamed from include/linux/sunrpc/svc.h~current~ +1 -0
@@ -275,6 +275,7 @@ struct svc_program {struct svc_stat * pg_stats;/* rpc statistics */ /* Override authentication. NULL means use default */ int (*pg_authenticate_obsolete)(struct svc_rqst *,u32 *); +int (*pg_authenticate)(struct svc_rqst *); };/*
net/sunrpc/auth_gss/svcauth_gss.c renamed from net/sunrpc/auth_gss/svcauth_gss.c~current~ +2 -7
@@ -906,11 +906,6 @@ svcauth_gss_accept(struct svc_rqst *rqstsvc_putu32 (resv, rpc_success); goto complete;case RPC_GSS_PROC_DATA : - *authp = rpc_autherr_badcred; - rqstp->rq_client = -find_gss_auth_domain (rsci->mechctx, gc->gc_svc); -if (rqstp->rq_client ==NULL ) - goto auth_err; *authp = rpcsec_gsserr_ctxproblem;if (gss_write_verf (rqstp, rsci->mechctx, gc->gc_seq)) goto auth_err;
@@ -924,8 +919,6 @@ svcauth_gss_accept(struct svc_rqst *rqstif (unwrap_integ_data (&rqstp->rq_arg, gc->gc_seq, rsci->mechctx)) goto auth_err; - svcdata->rsci = rsci; -cache_get (&rsci->h);/* placeholders for length and seq. number: */ svcdata->body_start = resv->iov_base + resv->iov_len;svc_putu32 (resv,0 );
@@ -936,6 +929,8 @@ svcauth_gss_accept(struct svc_rqst *rqstdefault : goto auth_err; } + svcdata->rsci = rsci; +cache_get (&rsci->h); ret =SVC_OK ; goto out; }
net/sunrpc/svc.c renamed from net/sunrpc/svc.c~current~ +11 -1
@@ -266,6 +266,7 @@ svc_process(struct svc_serv *serv, strucu32 dir, prog, vers, proc, auth_stat, rpc_stat;int auth_res; +u32 *accept_statp; rpc_stat = rpc_success;
@@ -301,6 +302,9 @@ svc_process(struct svc_serv *serv, strucif (vers !=2 )/* RPC version number */ goto err_bad_rpc; +/* Save position in case we later decide to reject: */ + accept_statp = resv->iov_base + resv->iov_len; +svc_putu32 (resv, xdr_zero);/* ACCEPT */ rqstp->rq_prog = prog =ntohl (svc_getu32 (argv));/* program number */
@@ -317,6 +321,11 @@ svc_process(struct svc_serv *serv, struc auth_res = progp->pg_authenticate_obsolete (rqstp, &auth_stat); else auth_res =svc_authenticate (rqstp, &auth_stat); +/* Also give the program a chance to reject this call: */ +if (auth_res ==SVC_OK ) { + auth_stat = rpc_autherr_badcred; + auth_res = progp->pg_authenticate (rqstp); + }switch (auth_res) {case SVC_OK :break ;
@@ -442,7 +451,8 @@ err_bad_rpc: err_bad_auth:dprintk ("svc: authentication failed (%d)\n" ,ntohl (auth_stat)); serv->sv_stats->rpcbadauth++; - resv->iov_len -=4 ; +/* Restore write pointer to location of accept status: */ +xdr_ressize_check (rqstp, accept_statp);svc_putu32 (resv, xdr_one);/* REJECT */ svc_putu32 (resv, xdr_one);/* AUTH_ERROR */ svc_putu32 (resv, auth_stat);/* status */
net/sunrpc/svcauth_unix.c renamed from net/sunrpc/svcauth_unix.c~current~ +2 -16
@@ -368,7 +368,6 @@ svcauth_null_accept(struct svc_rqst *rqsstruct kvec *argv = &rqstp->rq_arg.head[0 ];struct kvec *resv = &rqstp->rq_res.head[0 ];struct svc_cred *cred = &rqstp->rq_cred; -int rv=0 ; cred->cr_group_info =NULL ; rqstp->rq_client =NULL ;
@@ -394,19 +393,11 @@ svcauth_null_accept(struct svc_rqst *rqsif (cred->cr_group_info ==NULL )return SVC_DROP ;/* kmalloc failure - client must retry */ - rv =svcauth_unix_set_client (rqstp); -if (rv ==SVC_DENIED ) - goto badcred; -/* Put NULL verifier */ svc_putu32 (resv,RPC_AUTH_NULL );svc_putu32 (resv,0 ); -return rv; - -badcred: - *authp = rpc_autherr_badcred; -return SVC_DENIED ; +return SVC_OK ; } static int
@@ -441,7 +432,6 @@ svcauth_unix_accept(struct svc_rqst *rqsstruct svc_cred *cred = &rqstp->rq_cred;u32 slen, i;int len = argv->iov_len; -int rv=0 ; cred->cr_group_info =NULL ; rqstp->rq_client =NULL ;
@@ -473,15 +463,11 @@ svcauth_unix_accept(struct svc_rqst *rqsreturn SVC_DENIED ; } - rv =svcauth_unix_set_client (rqstp); -if (rv ==SVC_DENIED ) - goto badcred; -/* Put NULL verifier */ svc_putu32 (resv,RPC_AUTH_NULL );svc_putu32 (resv,0 ); -return rv; +return SVC_OK ; badcred: *authp = rpc_autherr_badcred;
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 nfs4: use new pg_set_client method to simplify nfs4 callback authentication
In reply to NeilBrownUse new pg_authenticate method to simplify nfs4 callback authentication.
This also has the effect of changing the error return from rejectedcred to badcred. I believe the change is correct.
### Diffstat output
./fs/nfs/callback.c | 156 +++------------------------------------------------- 1 files changed, 9 insertions(+), 147 deletions(-)
fs/nfs/callback.c renamed from fs/nfs/callback.c~current~ +9 -147
@@ -139,133 +139,10 @@ out:return ret; } -/* -* AUTH_NULL authentication -*/ -static int nfs_callback_null_accept (struct svc_rqst *rqstp,u32 *authp) -{ -struct kvec *argv = &rqstp->rq_arg.head[0 ]; -struct kvec *resv = &rqstp->rq_res.head[0 ]; - -if (argv->iov_len <3 *4 ) -return SVC_GARBAGE ; - -if (svc_getu32 (argv) !=0 ) { -dprintk ("svc: bad null cred\n" ); - *authp = rpc_autherr_badcred; -return SVC_DENIED ; - } -if (svc_getu32 (argv) !=RPC_AUTH_NULL ||svc_getu32 (argv) !=0 ) { -dprintk ("svc: bad null verf\n" ); - *authp = rpc_autherr_badverf; -return SVC_DENIED ; - } - -/* Signal that mapping to nobody uid/gid is required */ - rqstp->rq_cred.cr_uid = (uid_t) -1 ; - rqstp->rq_cred.cr_gid = (gid_t) -1 ; - rqstp->rq_cred.cr_group_info =groups_alloc (0 ); -if (rqstp->rq_cred.cr_group_info ==NULL ) -return SVC_DROP ;/* kmalloc failure - client must retry */ - -/* Put NULL verifier */ -svc_putu32 (resv,RPC_AUTH_NULL ); -svc_putu32 (resv,0 ); -dprintk ("%s: success, returning %d!\n" , __FUNCTION__,SVC_OK ); -return SVC_OK ; -} - -static int nfs_callback_null_release (struct svc_rqst *rqstp) -{ -if (rqstp->rq_cred.cr_group_info) -put_group_info (rqstp->rq_cred.cr_group_info); - rqstp->rq_cred.cr_group_info =NULL ; -return 0 ;/* don't drop */ -} - -static struct auth_ops nfs_callback_auth_null = { - .name ="null" , - .flavour =RPC_AUTH_NULL , - .accept = nfs_callback_null_accept, - .release = nfs_callback_null_release, -}; - -/* -* AUTH_SYS authentication -*/ -static int nfs_callback_unix_accept (struct svc_rqst *rqstp,u32 *authp) -{ -struct kvec *argv = &rqstp->rq_arg.head[0 ]; -struct kvec *resv = &rqstp->rq_res.head[0 ]; -struct svc_cred *cred = &rqstp->rq_cred; -u32 slen, i; -int len = argv->iov_len; - -dprintk ("%s: start\n" , __FUNCTION__); - cred->cr_group_info =NULL ; - rqstp->rq_client =NULL ; -if ((len -=3 *4 ) <0 ) -return SVC_GARBAGE ; - -/* Get length, time stamp and machine name */ -svc_getu32 (argv); -svc_getu32 (argv); - slen =XDR_QUADLEN (ntohl (svc_getu32 (argv))); -if (slen >64 || (len -= (slen +3 )*4 ) <0 ) - goto badcred; - argv->iov_base = (void *)((u32 *)argv->iov_base + slen); - argv->iov_len -= slen*4 ; - - cred->cr_uid =ntohl (svc_getu32 (argv)); - cred->cr_gid =ntohl (svc_getu32 (argv)); - slen =ntohl (svc_getu32 (argv)); -if (slen >16 || (len -= (slen +2 )*4 ) <0 ) - goto badcred; - cred->cr_group_info =groups_alloc (slen); -if (cred->cr_group_info ==NULL ) -return SVC_DROP ; -for (i =0 ; i < slen; i++) -GROUP_AT (cred->cr_group_info, i) =ntohl (svc_getu32 (argv)); - -if (svc_getu32 (argv) !=RPC_AUTH_NULL ||svc_getu32 (argv) !=0 ) { - *authp = rpc_autherr_badverf; -return SVC_DENIED ; - } -/* Put NULL verifier */ -svc_putu32 (resv,RPC_AUTH_NULL ); -svc_putu32 (resv,0 ); -dprintk ("%s: success, returning %d!\n" , __FUNCTION__,SVC_OK ); -return SVC_OK ; -badcred: - *authp = rpc_autherr_badcred; -return SVC_DENIED ; -} - -static int nfs_callback_unix_release (struct svc_rqst *rqstp) -{ -if (rqstp->rq_cred.cr_group_info) -put_group_info (rqstp->rq_cred.cr_group_info); - rqstp->rq_cred.cr_group_info =NULL ; -return 0 ; -} - -static struct auth_ops nfs_callback_auth_unix = { - .name ="unix" , - .flavour =RPC_AUTH_UNIX , - .accept = nfs_callback_unix_accept, - .release = nfs_callback_unix_release, -}; - -/* -* Hook the authentication protocol -*/ -static int nfs_callback_auth (struct svc_rqst *rqstp,u32 *authp) +static int nfs_callback_authenticate (struct svc_rqst *rqstp) {struct in_addr *addr = &rqstp->rq_addr.sin_addr;struct nfs4_client *clp; -struct kvec *argv = &rqstp->rq_arg.head[0 ]; -int flavour; -int retval;/* Don't talk to strangers */ clp =nfs4_find_client (addr);
@@ -273,34 +150,19 @@ static int nfs_callback_auth(struct svc_return SVC_DROP ;dprintk ("%s: %u.%u.%u.%u NFSv4 callback!\n" , __FUNCTION__,NIPQUAD (addr));nfs4_put_client (clp); - flavour =ntohl (svc_getu32 (argv)); -switch (flavour) { +switch (rqstp->rq_authop->flavour) {case RPC_AUTH_NULL : -if (rqstp->rq_proc !=CB_NULL ) { - *authp = rpc_autherr_tooweak; - retval =SVC_DENIED ; -break ; - } - rqstp->rq_authop = &nfs_callback_auth_null; - retval =nfs_callback_null_accept (rqstp, authp); +if (rqstp->rq_proc !=CB_NULL ) +return SVC_DENIED ;break ;case RPC_AUTH_UNIX : -/* Eat the authentication flavour */ - rqstp->rq_authop = &nfs_callback_auth_unix; - retval =nfs_callback_unix_accept (rqstp, authp);break ; +case RPC_AUTH_GSS : +/* FIXME: RPCSEC_GSS handling? */ default : -/* FIXME: need to add RPCSEC_GSS upcalls */ -#if 0 -svc_ungetu32 (argv); - retval =svc_authenticate (rqstp, authp); -#else - *authp = rpc_autherr_rejectedcred; - retval =SVC_DENIED ; -#endif +return SVC_DENIED ; } -dprintk ("%s: flavour %d returning error %d\n" , __FUNCTION__, flavour, retval); -return retval; +return SVC_OK ; }/*
@@ -321,5 +183,5 @@ static struct svc_program nfs4_callback_ .pg_name = "NFSv4 callback", /* service name */ .pg_class = "nfs", /* authentication class */ .pg_stats = &nfs4_callback_stats, - .pg_authenticate_obsolete = nfs_callback_auth, + .pg_authenticate = nfs_callback_authenticate, };
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 lockd: don't try to match callback requests against export table
In reply to NeilBrownOn lockd callbacks, we're a client, and the source address is that of a server, so we shouldn't be trying to match the source address of the callback request against our export table.
### Diffstat output
./fs/lockd/svc.c | 18 ++++++++++++++++++ 1 files changed, 18 insertions(+)
fs/lockd/svc.c renamed from fs/lockd/svc.c~current~ +18 -0
@@ -403,6 +403,18 @@ static int param_set_##name(const char *return 0 ; \ } +static inline int is_callback (u32 proc) +{ +return proc ==NLMPROC_GRANTED + || proc ==NLMPROC_GRANTED_MSG + || proc ==NLMPROC_TEST_RES + || proc ==NLMPROC_LOCK_RES + || proc ==NLMPROC_CANCEL_RES + || proc ==NLMPROC_UNLOCK_RES + || proc ==NLMPROC_NSM_NOTIFY ; +} + +static int lockd_authenticate (struct svc_rqst *rqstp) { rqstp->rq_client =NULL ;
@@ -411,6 +423,12 @@ static int lockd_authenticate(struct svccase RPC_AUTH_UNIX :if (rqstp->rq_proc ==0 )return SVC_OK ; +if (is_callback (rqstp->rq_proc)) { +/* Leave it to individual procedures to +* call nlmsvc_lookup_host(rqstp) +*/ +return SVC_OK ; + }return svc_set_client (rqstp); }return SVC_DENIED ;
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 nfsd: remove pg_authenticate field
In reply to NeilBrownThe pg_authenticate (now pg_authenticate_obsolete) callback was only being used by the nfs4 client callback code to circumvent the svcauth_unix code's insistence on checking all requests against the export table. With that problem solved, we no longer need it.
### Diffstat output
./include/linux/sunrpc/svc.h | 2 -- ./net/sunrpc/svc.c | 5 +---- 2 files changed, 1 insertion(+), 6 deletions(-)
include/linux/sunrpc/svc.h renamed from include/linux/sunrpc/svc.h~current~ +0 -2
@@ -273,8 +273,6 @@ struct svc_program {char * pg_name;/* service name */ char * pg_class;/* class name: services sharing authentication */ struct svc_stat * pg_stats;/* rpc statistics */ -/* Override authentication. NULL means use default */ -int (*pg_authenticate_obsolete)(struct svc_rqst *,u32 *);int (*pg_authenticate)(struct svc_rqst *); };
net/sunrpc/svc.c renamed from net/sunrpc/svc.c~current~ +1 -4
@@ -317,10 +317,7 @@ svc_process(struct svc_serv *serv, struc* We do this before anything else in order to get a decent * auth verifier. */ -if (progp->pg_authenticate_obsolete !=NULL ) - auth_res = progp->pg_authenticate_obsolete (rqstp, &auth_stat); -else - auth_res =svc_authenticate (rqstp, &auth_stat); + auth_res =svc_authenticate (rqstp, &auth_stat);/* Also give the program a chance to reject this call: */ if (auth_res ==SVC_OK ) { auth_stat = rpc_autherr_badcred;
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 global/static cleanups for nfsd
In reply to NeilBrownThe patch below contains the following cleanups:
- make some needlessly global code static
### Diffstat output
./fs/nfsd/export.c | 22 +++++++++++----------- ./fs/nfsd/lockd.c | 2 +- ./fs/nfsd/nfsfh.c | 11 ++++++----- ./fs/nfsd/nfssvc.c | 2 +- ./fs/nfsd/vfs.c | 8 ++++---- 5 files changed, 23 insertions(+), 22 deletions(-)
fs/nfsd/export.c renamed from fs/nfsd/export.c~current~ +11 -11
@@ -79,9 +79,9 @@ void expkey_put(struct cache_head *item, } } -void expkey_request (struct cache_detail *cd, -struct cache_head *h, -char **bpp,int *blen) +static void expkey_request (struct cache_detail *cd, +struct cache_head *h, +char **bpp,int *blen) {/* client fsidtype \xfsid */ struct svc_expkey *ek = container_of(h, struct svc_expkey, h);
@@ -95,7 +95,7 @@ void expkey_request(struct cache_detail }static struct svc_expkey *svc_expkey_lookup (struct svc_expkey *,int ); -int expkey_parse (struct cache_detail *cd,char *mesg,int mlen) +static int expkey_parse (struct cache_detail *cd,char *mesg,int mlen) {/* client fsidtype fsid [path] */ char *buf;
@@ -284,9 +284,9 @@ void svc_export_put(struct cache_head *i } } -void svc_export_request (struct cache_detail *cd, -struct cache_head *h, -char **bpp,int *blen) +static void svc_export_request (struct cache_detail *cd, +struct cache_head *h, +char **bpp,int *blen) {/* client path */ struct svc_export *exp = container_of(h, struct svc_export, h);
@@ -340,7 +340,7 @@ static int check_export(struct inode *in } -int svc_export_parse (struct cache_detail *cd,char *mesg,int mlen) +static int svc_export_parse (struct cache_detail *cd,char *mesg,int mlen) {/* client path expiry [flags anonuid anongid fsid] */ char *buf;
@@ -510,8 +510,8 @@ exp_find_key(svc_client *clp, int fsid_treturn ek; } -int exp_set_key (svc_client *clp,int fsid_type,u32 *fsidv, -struct svc_export *exp) +static int exp_set_key (svc_client *clp,int fsid_type,u32 *fsidv, +struct svc_export *exp) {struct svc_expkey key, *ek;
@@ -999,7 +999,7 @@ static void e_stop(struct seq_file *m, vexp_readunlock (); } -struct flags { +static struct flags {int flag;char *name[2 ]; } expflags[] = {
fs/nfsd/lockd.c renamed from fs/nfsd/lockd.c~current~ +1 -1
@@ -60,7 +60,7 @@ nlm_fclose(struct file *filp)fput (filp); } -struct nlmsvc_binding nfsd_nlm_ops = { +static struct nlmsvc_binding nfsd_nlm_ops = { .fopen = nlm_fopen,/* open file for locking */ .fclose = nlm_fclose,/* close file */ };
fs/nfsd/nfsfh.c renamed from fs/nfsd/nfsfh.c~current~ +6 -5
@@ -41,7 +41,7 @@ extern struct export_operations export_o* if not, require that we can walk up to exp->ex_dentry * doing some checks on the 'x' bits */ -int nfsd_acceptable (void *expv,struct dentry *dentry) +static int nfsd_acceptable (void *expv,struct dentry *dentry) {struct svc_export *exp = expv;int rv;
@@ -280,8 +280,8 @@ out:* an inode. In this case a call to fh_update should be made * before the fh goes out on the wire ... */ -inline int _fh_update (struct dentry *dentry,struct svc_export *exp, -__u32 *datap,int *maxsize) +static inline int _fh_update (struct dentry *dentry,struct svc_export *exp, +__u32 *datap,int *maxsize) {struct export_operations *nop = exp->ex_mnt->mnt_sb->s_export_op;
@@ -297,8 +297,9 @@ inline int _fh_update(struct dentry *den/* * for composing old style file handles */ -inline void _fh_update_old (struct dentry *dentry,struct svc_export *exp, -struct knfsd_fh *fh) +static inline void _fh_update_old (struct dentry *dentry, +struct svc_export *exp, +struct knfsd_fh *fh) { fh->ofh_ino =ino_t_to_u32 (dentry->d_inode->i_ino); fh->ofh_generation = dentry->d_inode->i_generation;
fs/nfsd/nfssvc.c renamed from fs/nfsd/nfssvc.c~current~ +1 -1
@@ -63,7 +63,7 @@ struct nfsd_list {struct list_head list;struct task_struct *task; }; -struct list_head nfsd_list =LIST_HEAD_INIT (nfsd_list); +static struct list_head nfsd_list =LIST_HEAD_INIT (nfsd_list);/* * Maximum number of nfsd processes
fs/nfsd/vfs.c renamed from fs/nfsd/vfs.c~current~ +4 -4
@@ -708,8 +708,8 @@ nfsd_close(struct file *filp)* As this calls fsync (not fdatasync) there is no need for a write_inode * after it. */ -inline void nfsd_dosync (struct file *filp,struct dentry *dp, -struct file_operations *fop) +static inline void nfsd_dosync (struct file *filp,struct dentry *dp, +struct file_operations *fop) {struct inode *inode = dp->d_inode;int (*fsync) (struct file *,struct dentry *,int );
@@ -721,7 +721,7 @@ inline void nfsd_dosync(struct file *fil } -void +static void nfsd_sync (struct file *filp) {struct inode *inode = filp->f_dentry->d_inode;
@@ -731,7 +731,7 @@ nfsd_sync(struct file *filp)up (&inode->i_sem); } -void +static void nfsd_sync_dir (struct dentry *dp) {nfsd_dosync (NULL , dp, dp->d_inode->i_fop);
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
PATCH 1/1 change nfsd reply cache to use list.h lists
In reply to NeilBrownalso kmalloc the cache one entry at a time, instead of in one big slab.
### Diffstat output
./fs/nfsd/nfscache.c | 131 ++++++++++++++----------------------------- ./include/linux/nfsd/cache.h | 7 -- 2 files changed, 46 insertions(+), 92 deletions(-)
fs/nfsd/nfscache.c renamed from fs/nfsd/nfscache.c~current~ +43 -88
@@ -15,6 +15,7 @@#include <linux/slab.h> #include <linux/string.h> #include <linux/spinlock.h> +#include <linux/list.h> #include <linux/sunrpc/svc.h> #include <linux/nfsd/nfsd.h>
@@ -30,15 +31,8 @@#define HASHSIZE 64#define REQHASH (xid) ((((xid) >> 24) ^ (xid)) & (HASHSIZE-1)) -struct nfscache_head { -struct svc_cacherep * next; -struct svc_cacherep * prev; -}; - -static struct nfscache_head * hash_list; -static struct svc_cacherep * lru_head; -static struct svc_cacherep * lru_tail; -static struct svc_cacherep * nfscache; +static struct hlist_head * hash_list; +static struct list_head lru_head;static int cache_disabled =1 ;static int nfsd_cache_append (struct svc_rqst *rqstp,struct kvec *vec);
@@ -54,46 +48,32 @@ voidnfsd_cache_init (void) {struct svc_cacherep *rp; -struct nfscache_head *rh; -size_t i; -unsigned long order; +int i; - - i =CACHESIZE *sizeof (struct svc_cacherep ); -for (order =0 ; (PAGE_SIZE << order) < i; order++) - ; - nfscache = (struct svc_cacherep *) -__get_free_pages (GFP_KERNEL , order); -if (!nfscache) { -printk (KERN_ERR "nfsd: cannot allocate %Zd bytes for reply cache\n" , i); -return ; +INIT_LIST_HEAD (&lru_head); + i =CACHESIZE ; +while (i) { + rp =kmalloc (sizeof (*rp),GFP_KERNEL ); +if (!rp)break ; +list_add (&rp->c_lru, &lru_head); + rp->c_state =RC_UNUSED ; + rp->c_type =RC_NOCACHE ; +INIT_HLIST_NODE (&rp->c_hash); + i--; } -memset (nfscache,0 , i); - i =HASHSIZE *sizeof (struct nfscache_head ); - hash_list =kmalloc (i,GFP_KERNEL ); +if (i) +printk (KERN_ERR "nfsd: cannot allocate all %d cache entries, only got %d\n" , +CACHESIZE ,CACHESIZE -i); + + hash_list =kmalloc (HASHSIZE *sizeof (struct hlist_head ),GFP_KERNEL );if (!hash_list) { -free_pages ((unsigned long )nfscache, order); - nfscache =NULL ; -printk (KERN_ERR "nfsd: cannot allocate %Zd bytes for hash list\n" , i); +nfsd_cache_shutdown (); +printk (KERN_ERR "nfsd: cannot allocate %Zd bytes for hash list\n" , +HASHSIZE *sizeof (struct hlist_head ));return ; } - -for (i =0 , rh = hash_list; i <HASHSIZE ; i++, rh++) - rh->next = rh->prev = (struct svc_cacherep *) rh; - -for (i =0 , rp = nfscache; i <CACHESIZE ; i++, rp++) { - rp->c_state =RC_UNUSED ; - rp->c_type =RC_NOCACHE ; - rp->c_hash_next = - rp->c_hash_prev = rp; - rp->c_lru_next = rp +1 ; - rp->c_lru_prev = rp -1 ; - } - lru_head = nfscache; - lru_tail = nfscache +CACHESIZE -1 ; - lru_head->c_lru_prev =NULL ; - lru_tail->c_lru_next =NULL ; +memset (hash_list,0 ,HASHSIZE *sizeof (struct hlist_head )); cache_disabled =0 ; }
@@ -102,48 +82,30 @@ voidnfsd_cache_shutdown (void) {struct svc_cacherep *rp; -size_t i; -unsigned long order; -for (rp = lru_head; rp; rp = rp->c_lru_next) { +while (!list_empty (&lru_head)) { + rp = list_entry(lru_head.next, struct svc_cacherep, c_lru);if (rp->c_state ==RC_DONE && rp->c_type ==RC_REPLBUFF )kfree (rp->c_replvec.iov_base); +list_del (&rp->c_lru); +kfree (rp); } cache_disabled =1 ; - i =CACHESIZE *sizeof (struct svc_cacherep ); -for (order =0 ; (PAGE_SIZE << order) < i; order++) - ; -free_pages ((unsigned long )nfscache, order); - nfscache =NULL ; -kfree (hash_list); +if (hash_list) +kfree (hash_list); hash_list =NULL ; }/* -* Move cache entry to front of LRU list +* Move cache entry to end of LRU list */ static void -lru_put_front (struct svc_cacherep *rp) +lru_put_end (struct svc_cacherep *rp) { -struct svc_cacherep *prev = rp->c_lru_prev, - *next = rp->c_lru_next; - -if (prev) - prev->c_lru_next = next; -else - lru_head = next; -if (next) - next->c_lru_prev = prev; -else - lru_tail = prev; - - rp->c_lru_next = lru_head; - rp->c_lru_prev =NULL ; -if (lru_head) - lru_head->c_lru_prev = rp; - lru_head = rp; +list_del (&rp->c_lru); +list_add_tail (&rp->c_lru, &lru_head); }/*
@@ -152,17 +114,8 @@ lru_put_front(struct svc_cacherep *rp)static void hash_refile (struct svc_cacherep *rp) { -struct svc_cacherep *prev = rp->c_hash_prev, - *next = rp->c_hash_next; -struct nfscache_head *head = hash_list +REQHASH (rp->c_xid); - - prev->c_hash_next = next; - next->c_hash_prev = prev; - - rp->c_hash_next = head->next; - rp->c_hash_prev = (struct svc_cacherep *) head; - head->next->c_hash_prev = rp; - head->next = rp; +hlist_del_init (&rp->c_hash); +hlist_add_head (&rp->c_hash, hash_list +REQHASH (rp->c_xid)); }/*
@@ -173,7 +126,9 @@ hash_refile(struct svc_cacherep *rp)int nfsd_cache_lookup (struct svc_rqst *rqstp,int type) { -struct svc_cacherep *rh, *rp; +struct hlist_node *hn; +struct hlist_head *rh; +struct svc_cacherep *rp;u32 xid = rqstp->rq_xid, proto = rqstp->rq_prot, vers = rqstp->rq_vers,
@@ -190,8 +145,8 @@ nfsd_cache_lookup(struct svc_rqst *rqstpspin_lock (&cache_lock); rtn =RC_DOIT ; - rp = rh = (struct svc_cacherep *) &hash_list[REQHASH (xid)]; -while ((rp = rp->c_hash_next) != rh) { + rh = &hash_list[REQHASH (xid)]; +hlist_for_each_entry (rp, hn, rh, c_hash) { if (rp->c_state != RC_UNUSED && xid == rp->c_xid && proc == rp->c_proc && proto == rp->c_prot && vers == rp->c_vers &&
@@ -206,7 +161,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp/* This loop shouldn't take more than a few iterations normally */ {int safe =0 ; -for (rp = lru_tail; rp; rp = rp->c_lru_prev) { +list_for_each_entry (rp, &lru_head, c_lru) {if (rp->c_state !=RC_INPROG )break ;if (safe++ >CACHESIZE ) {
@@ -254,7 +209,7 @@ found_entry:/* We found a matching entry which is either in progress or done. */ age = jiffies - rp->c_timestamp; rp->c_timestamp = jiffies; -lru_put_front (rp); +lru_put_end (rp); rtn =RC_DROPIT ;/* Request being processed or excessive rexmits */
@@ -343,7 +298,7 @@ nfsd_cache_update(struct svc_rqst *rqstpbreak ; }spin_lock (&cache_lock); -lru_put_front (rp); +lru_put_end (rp); rp->c_secure = rqstp->rq_secure; rp->c_type = cachetype; rp->c_state =RC_DONE ;
include/linux/nfsd/cache.h renamed from include/linux/nfsd/cache.h~current~ +3 -4
@@ -19,10 +19,9 @@* be hash_next and hash_prev. */ struct svc_cacherep { -struct svc_cacherep * c_hash_next; -struct svc_cacherep * c_hash_prev; -struct svc_cacherep * c_lru_next; -struct svc_cacherep * c_lru_prev; +struct hlist_node c_hash; +struct list_head c_lru; +unsigned char c_state,/* unused, inprog, done */ c_type,/* status, buffer */ c_secure :1 ;/* req came from port < 1024 */
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
Quoted 12 lines · NeilBrown
Following are 12 patches for nfsd in 2.6.11-rc2-mm2
- Fix minor problem with '..' access interacting with the "nohide" export option.
- reduce stack space usage infind_exported_dentry
- 6 patches from Bruce Fields to improve authentication handling for callbacks.
- Fix problem that stopped read access to append-only files.
- Fix bug where readdir replies are cache badly.
- Add lots of "static" declarations where appropriate
- Change nfscache to uselist.hmacros and smaller kmallocs
Quoted 1 line
I believe they are all ready for ready for 2.6.11.
Yeah, I do worry about making these changes at this time.
Quoted 2 lines
Number 10 (readdir fix) is probably the most critical (and most trivial). The append-only would be next in line.
OK.
Quoted 1 line
The 6 from Bruce fix a real performance problem with locking.
That one depends on a bunch of the preceding ones, yes?
Quoted 2 lines
The rest could be safely delayed if that seem appropriate, but they are either quite trivial or (for the last one) quite well tested.
Maybe just the readdir and append patches for 2.6.11??
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
Quoted 4 lines · On Monday February 7, [email protected] wrote:
Quoted 2 lines
The 6 from Bruce fix a real performance problem with locking.
That one depends on a bunch of the preceding ones, yes?
The set of six (numbered 3 to 8) should be independent of the rest. I believe Bruce has done substantial testing on them. I'll leave him to make a case or not.
Quoted 5 lines
Quoted 3 lines
The rest could be safely delayed if that seem appropriate, but they are either quite trivial or (for the last one) quite well tested.
Maybe just the readdir and append patches for 2.6.11??
I'm OK with that if Bruce is.
Thanks,
NeilBrown
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
Quoted 9 lines · Neil Brown
Quoted 4 lines · On Monday February 7, [email protected] wrote:
Quoted 2 lines
The 6 from Bruce fix a real performance problem with locking.
That one depends on a bunch of the preceding ones, yes?
The set of six (numbered 3 to 8) should be independent of the rest. I believe Bruce has done substantial testing on them. I'll leave him to make a case or not.
Well, they've been waiting around longer than I'd like, but they're also not completely trivial, so I hate to drop them in right before 2.6.11.
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
The rest could be safely delayed if that seem appropriate, but they are either quite trivial or (for the last one) quite well tested.
Maybe just the readdir and append patches for 2.6.11??
I'm OK with that if Bruce is.
OK. If someone else cares passionately about getting the lockd callback problem fixed as soon as possible, they can test and speak up. They seem fine to me, but I haven't gotten much feedback on them.
--b.
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs