On 2011-04-20 22:53, Trond Myklebust wrote:
> On Wed, 2011-04-20 at 20:26 +0300, Benny Halevy wrote:
>> @@ -1424,9 +1424,10 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
>> */
>> void nfs4_evict_inode(struct inode *inode)
>> {
>> - pnfs_destroy_layout(NFS_I(inode));
>> + pnfs_return_layout(inode, NULL, true);
>
> Why does this want to come before the call to truncate_inode_pages()?
Actually, I don't see any good reason. :-/
> Is there any reason not to put pnfs_return_layout() and
> pnfs_destroy_layout into a single helper here?
>
Looks like an overkill to me for this one call site.
>> truncate_inode_pages(&inode->i_data, 0);
>> end_writeback(inode);
>> + pnfs_destroy_layout(NFS_I(inode));
>> /* If we are holding a delegation, return it! */
>> nfs_inode_return_delegation_noreclaim(inode);
>> /* First call standard NFS clear_inode() code */
>> diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
>> index 9bf41ea..b03defb 100644
>> --- a/fs/nfs/nfs4proc.c
>> +++ b/fs/nfs/nfs4proc.c
>> @@ -5662,6 +5662,103 @@ int nfs4_proc_layoutget(struct nfs4_layoutget *lgp)
>> return status;
>> }
>>
>> +static void
>> +nfs4_layoutreturn_prepare(struct rpc_task *task, void *calldata)
>> +{
>> + struct nfs4_layoutreturn *lrp = calldata;
>> +
>> + dprintk("--> %s\n", __func__);
>> + if (nfs41_setup_sequence(lrp->clp->cl_session, &lrp->args.seq_args,
>> + &lrp->res.seq_res, 0, task))
>> + return;
>> + rpc_call_start(task);
>> +}
>> +
>> +static void nfs4_layoutreturn_done(struct rpc_task *task, void *calldata)
>> +{
>> + struct nfs4_layoutreturn *lrp = calldata;
>> + struct nfs_server *server;
>> +
>> + dprintk("--> %s\n", __func__);
>> +
>> + if (!nfs4_sequence_done(task, &lrp->res.seq_res))
>> + return;
>> +
>> + if (lrp->args.return_type == RETURN_FILE)
>> + server = NFS_SERVER(lrp->args.inode);
>> + else
>> + server = NULL;
>> + if (nfs4_async_handle_error(task, server, NULL) == -EAGAIN) {
>> + nfs_restart_rpc(task, lrp->clp);
>> + return;
>> + }
>> + if ((task->tk_status == 0) && (lrp->args.return_type == RETURN_FILE)) {
>> + struct pnfs_layout_hdr *lo = NFS_I(lrp->args.inode)->layout;
>> +
>> + spin_lock(&lo->plh_inode->i_lock);
>> + if (lrp->res.lrs_present)
>> + pnfs_set_layout_stateid(lo, &lrp->res.stateid, true);
>> + else
>> + BUG_ON(!list_empty(&lo->plh_segs));
>> + spin_unlock(&lo->plh_inode->i_lock);
>> + }
>> + dprintk("<-- %s\n", __func__);
>> +}
>> +
>> +static void nfs4_layoutreturn_release(void *calldata)
>> +{
>> + struct nfs4_layoutreturn *lrp = calldata;
>> +
>> + dprintk("--> %s return_type %d\n", __func__, lrp->args.return_type);
>> + if (lrp->args.return_type == RETURN_FILE) {
>> + struct inode *ino = lrp->args.inode;
>> + struct pnfs_layout_hdr *lo = NFS_I(ino)->layout;
>> +
>> + put_layout_hdr(lo);
>> + }
>> + kfree(calldata);
>> + dprintk("<-- %s\n", __func__);
>> +}
>> +
>> +static const struct rpc_call_ops nfs4_layoutreturn_call_ops = {
>> + .rpc_call_prepare = nfs4_layoutreturn_prepare,
>> + .rpc_call_done = nfs4_layoutreturn_done,
>> + .rpc_release = nfs4_layoutreturn_release,
>> +};
>> +
>> +int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool issync)
>
> Why the 'issync' parameter?
>
Hmm, you're right. Currently all call sites use it as sync.
>> +{
>> + struct rpc_task *task;
>> + struct rpc_message msg = {
>> + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LAYOUTRETURN],
>> + .rpc_argp = &lrp->args,
>> + .rpc_resp = &lrp->res,
>> + };
>> + struct rpc_task_setup task_setup_data = {
>> + .rpc_client = lrp->clp->cl_rpcclient,
>> + .rpc_message = &msg,
>> + .callback_ops = &nfs4_layoutreturn_call_ops,
>> + .callback_data = lrp,
>> + .flags = RPC_TASK_ASYNC,
>> + };
>> + int status = 0;
>> +
>> + dprintk("--> %s\n", __func__);
>> + task = rpc_run_task(&task_setup_data);
>> + if (IS_ERR(task))
>> + return PTR_ERR(task);
>> + if (!issync)
>> + goto out;
>> + status = nfs4_wait_for_completion_rpc_task(task);
>> + if (status != 0)
>> + goto out;
>> + status = task->tk_status;
>> +out:
>> + dprintk("<-- %s\n", __func__);
>> + rpc_put_task(task);
>> + return status;
>> +}
>> +
>> static int
>> _nfs4_proc_getdeviceinfo(struct nfs_server *server, struct pnfs_device *pdev)
>> {
>> diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
>> index dddfb57..53ea3e5 100644
>> --- a/fs/nfs/nfs4xdr.c
>> +++ b/fs/nfs/nfs4xdr.c
>> @@ -338,7 +338,12 @@ static int nfs4_stat_to_errno(int);
>> 1 /* layoutupdate4 layout type */ + \
>> 1 /* NULL filelayout layoutupdate4 payload */)
>> #define decode_layoutcommit_maxsz (op_decode_hdr_maxsz + 3)
>> -
>> +#define encode_layoutreturn_maxsz (8 + op_encode_hdr_maxsz + \
>> + encode_stateid_maxsz + \
>> + 1 /* FIXME: opaque lrf_body always empty at
>> + *the moment */)
>> +#define decode_layoutreturn_maxsz (op_decode_hdr_maxsz + \
>> + 1 + decode_stateid_maxsz)
>> #else /* CONFIG_NFS_V4_1 */
>> #define encode_sequence_maxsz 0
>> #define decode_sequence_maxsz 0
>> @@ -760,7 +765,14 @@ static int nfs4_stat_to_errno(int);
>> decode_putfh_maxsz + \
>> decode_layoutcommit_maxsz + \
>> decode_getattr_maxsz)
>> -
>> +#define NFS4_enc_layoutreturn_sz (compound_encode_hdr_maxsz + \
>> + encode_sequence_maxsz + \
>> + encode_putfh_maxsz + \
>> + encode_layoutreturn_maxsz)
>> +#define NFS4_dec_layoutreturn_sz (compound_decode_hdr_maxsz + \
>> + decode_sequence_maxsz + \
>> + decode_putfh_maxsz + \
>> + decode_layoutreturn_maxsz)
>>
>> const u32 nfs41_maxwrite_overhead = ((RPC_MAX_HEADER_WITH_AUTH +
>> compound_encode_hdr_maxsz +
>> @@ -1890,6 +1902,37 @@ encode_layoutcommit(struct xdr_stream *xdr,
>> hdr->replen += decode_layoutcommit_maxsz;
>> return 0;
>> }
>> +
>> +static void
>> +encode_layoutreturn(struct xdr_stream *xdr,
>> + const struct nfs4_layoutreturn_args *args,
>> + struct compound_hdr *hdr)
>> +{
>> + nfs4_stateid stateid;
>> + __be32 *p;
>> +
>> + p = reserve_space(xdr, 20);
>> + *p++ = cpu_to_be32(OP_LAYOUTRETURN);
>> + *p++ = cpu_to_be32(args->reclaim);
>> + *p++ = cpu_to_be32(args->layout_type);
>> + *p++ = cpu_to_be32(args->range.iomode);
>> + *p = cpu_to_be32(args->return_type);
>> + if (args->return_type == RETURN_FILE) {
>> + p = reserve_space(xdr, 16 + NFS4_STATEID_SIZE);
>> + p = xdr_encode_hyper(p, args->range.offset);
>> + p = xdr_encode_hyper(p, args->range.length);
>> + spin_lock(&args->inode->i_lock);
>> + memcpy(stateid.data, NFS_I(args->inode)->layout->plh_stateid.data,
>> + NFS4_STATEID_SIZE);
>> + spin_unlock(&args->inode->i_lock);
>> + p = xdr_encode_opaque_fixed(p, &stateid.data,
>> + NFS4_STATEID_SIZE);
>> + p = reserve_space(xdr, 4);
>> + *p = cpu_to_be32(0);
>> + }
>> + hdr->nops++;
>> + hdr->replen += decode_layoutreturn_maxsz;
>> +}
>> #endif /* CONFIG_NFS_V4_1 */
>>
>> /*
>> @@ -2707,9 +2750,9 @@ static void nfs4_xdr_enc_layoutget(struct rpc_rqst *req,
>> /*
>> * Encode LAYOUTCOMMIT request
>> */
>> -static int nfs4_xdr_enc_layoutcommit(struct rpc_rqst *req,
>> - struct xdr_stream *xdr,
>> - struct nfs4_layoutcommit_args *args)
>> +static void nfs4_xdr_enc_layoutcommit(struct rpc_rqst *req,
>> + struct xdr_stream *xdr,
>> + struct nfs4_layoutcommit_args *args)
>> {
>> struct compound_hdr hdr = {
>> .minorversion = nfs4_xdr_minorversion(&args->seq_args),
>> @@ -2721,7 +2764,24 @@ static int nfs4_xdr_enc_layoutcommit(struct rpc_rqst *req,
>> encode_layoutcommit(xdr, args, &hdr);
>> encode_getfattr(xdr, args->bitmask, &hdr);
>> encode_nops(&hdr);
>> - return 0;
>> +}
>> +
>> +/*
>> + * Encode LAYOUTRETURN request
>> + */
>> +static void nfs4_xdr_enc_layoutreturn(struct rpc_rqst *req,
>> + struct xdr_stream *xdr,
>> + struct nfs4_layoutreturn_args *args)
>> +{
>> + struct compound_hdr hdr = {
>> + .minorversion = nfs4_xdr_minorversion(&args->seq_args),
>> + };
>> +
>> + encode_compound_hdr(xdr, req, &hdr);
>> + encode_sequence(xdr, &args->seq_args, &hdr);
>> + encode_putfh(xdr, NFS_FH(args->inode), &hdr);
>> + encode_layoutreturn(xdr, args, &hdr);
>> + encode_nops(&hdr);
>> }
>> #endif /* CONFIG_NFS_V4_1 */
>>
>> @@ -5202,6 +5262,27 @@ out_overflow:
>> return -EIO;
>> }
>>
>> +static int decode_layoutreturn(struct xdr_stream *xdr,
>> + struct nfs4_layoutreturn_res *res)
>> +{
>> + __be32 *p;
>> + int status;
>> +
>> + status = decode_op_hdr(xdr, OP_LAYOUTRETURN);
>> + if (status)
>> + return status;
>> + p = xdr_inline_decode(xdr, 4);
>> + if (unlikely(!p))
>> + goto out_overflow;
>> + res->lrs_present = be32_to_cpup(p);
>> + if (res->lrs_present)
>> + status = decode_stateid(xdr, &res->stateid);
>> + return status;
>> +out_overflow:
>> + print_overflow_msg(__func__, xdr);
>> + return -EIO;
>> +}
>> +
>> static int decode_layoutcommit(struct xdr_stream *xdr,
>> struct rpc_rqst *req,
>> struct nfs4_layoutcommit_res *res)
>> @@ -6319,6 +6400,30 @@ out:
>> }
>>
>> /*
>> + * Decode LAYOUTRETURN response
>> + */
>> +static int nfs4_xdr_dec_layoutreturn(struct rpc_rqst *rqstp,
>> + struct xdr_stream *xdr,
>> + struct nfs4_layoutreturn_res *res)
>> +{
>> + struct compound_hdr hdr;
>> + int status;
>> +
>> + status = decode_compound_hdr(xdr, &hdr);
>> + if (status)
>> + goto out;
>> + status = decode_sequence(xdr, &res->seq_res, rqstp);
>> + if (status)
>> + goto out;
>> + status = decode_putfh(xdr);
>> + if (status)
>> + goto out;
>> + status = decode_layoutreturn(xdr, res);
>> +out:
>> + return status;
>> +}
>> +
>> +/*
>> * Decode LAYOUTCOMMIT response
>> */
>> static int nfs4_xdr_dec_layoutcommit(struct rpc_rqst *rqstp,
>> @@ -6544,6 +6649,7 @@ struct rpc_procinfo nfs4_procedures[] = {
>> PROC(GETDEVICEINFO, enc_getdeviceinfo, dec_getdeviceinfo),
>> PROC(LAYOUTGET, enc_layoutget, dec_layoutget),
>> PROC(LAYOUTCOMMIT, enc_layoutcommit, dec_layoutcommit),
>> + PROC(LAYOUTRETURN, enc_layoutreturn, dec_layoutreturn),
>> #endif /* CONFIG_NFS_V4_1 */
>> };
>>
>> diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
>> index d9ab972..89e7725 100644
>> --- a/fs/nfs/pnfs.c
>> +++ b/fs/nfs/pnfs.c
>> @@ -321,6 +321,36 @@ mark_matching_lsegs_invalid(struct pnfs_layout_hdr *lo,
>> return invalid - removed;
>> }
>>
>> +/* Returns false if there was nothing to do, true otherwise */
>> +static bool
>> +pnfs_clear_lseg_list(struct pnfs_layout_hdr *lo, struct list_head *tmp_list,
>> + struct pnfs_layout_range *range)
>> +{
>> + struct pnfs_layout_segment *lseg, *next;
>> + bool rv = false;
>> +
>> + dprintk("%s:Begin lo %p offset %llu length %llu iomode %d\n",
>> + __func__, lo, range->offset, range->length, range->iomode);
>> + assert_spin_locked(&lo->plh_inode->i_lock);
>
> Not needed in the case of nfs4_evict_inode.
>
True. I can remove the assert
but the lock better be taken if called while the inode is alive.
>> + if (list_empty(&lo->plh_segs)) {
>> + if (!test_and_set_bit(NFS_LAYOUT_DESTROYED, &lo->plh_flags))
>> + put_layout_hdr_locked(lo);
>> + return 0;
>> + }
>> + list_for_each_entry_safe(lseg, next, &lo->plh_segs, pls_list)
>> + if (should_free_lseg(lseg->pls_range.iomode, range->iomode)) {
>> + dprintk("%s: freeing lseg %p iomode %d "
>> + "offset %llu length %llu\n", __func__,
>> + lseg, lseg->pls_range.iomode,
>> + lseg->pls_range.offset,
>> + lseg->pls_range.length);
>> + mark_lseg_invalid(lseg, tmp_list);
>> + rv = true;
>> + }
>> + dprintk("%s:Return %d\n", __func__, rv);
>> + return rv;
>> +}
>> +
>> /* note free_me must contain lsegs from a single layout_hdr */
>> void
>> pnfs_free_lseg_list(struct list_head *free_me)
>> @@ -539,6 +569,72 @@ out_err_free:
>> return NULL;
>> }
>>
>> +static int
>> +return_layout(struct inode *ino, struct pnfs_layout_range *range, bool wait)
>> +{
>> + struct nfs4_layoutreturn *lrp;
>> + struct nfs_server *server = NFS_SERVER(ino);
>> + int status = -ENOMEM;
>> +
>> + dprintk("--> %s\n", __func__);
>> +
>> + lrp = kzalloc(sizeof(*lrp), GFP_KERNEL);
>> + if (lrp == NULL) {
>> + put_layout_hdr(NFS_I(ino)->layout);
>> + goto out;
>> + }
>> + lrp->args.reclaim = 0;
>> + lrp->args.layout_type = server->pnfs_curr_ld->id;
>> + lrp->args.return_type = RETURN_FILE;
>> + lrp->args.range = *range;
>> + lrp->args.inode = ino;
>> + lrp->clp = server->nfs_client;
>> +
>> + status = nfs4_proc_layoutreturn(lrp, wait);
>> +out:
>> + dprintk("<-- %s status: %d\n", __func__, status);
>> + return status;
>> +}
>> +
>> +/* Initiates a LAYOUTRETURN(FILE) */
>> +int
>> +_pnfs_return_layout(struct inode *ino, struct pnfs_layout_range *range, bool wait)
>
> What are the 'range' and 'wait' parameters for? We don't have any users
> other than nfs4_evict_inode.
>
OK, I'll introduce these with their usage.
>> +{
>> + struct pnfs_layout_hdr *lo = NULL;
>> + struct nfs_inode *nfsi = NFS_I(ino);
>> + struct pnfs_layout_range arg;
>> + LIST_HEAD(tmp_list);
>> + int status = 0;
>> +
>> + dprintk("--> %s\n", __func__);
>> +
>> + arg.iomode = range ? range->iomode : IOMODE_ANY;
>> + arg.offset = 0;
>> + arg.length = NFS4_MAX_UINT64;
>> +
>> + spin_lock(&ino->i_lock);
>> + lo = nfsi->layout;
>> + if (!lo || !pnfs_clear_lseg_list(lo, &tmp_list, &arg)) {
>> + spin_unlock(&ino->i_lock);
>> + dprintk("%s: no layout segments to return\n", __func__);
>> + goto out;
>> + }
>> + /* Reference matched in nfs4_layoutreturn_release */
>> + get_layout_hdr(lo);
>> + spin_unlock(&ino->i_lock);
>> + pnfs_free_lseg_list(&tmp_list);
>> +
>> + /* Return layout even if layoutcommit fails */
>> + status = pnfs_layoutcommit_inode(ino, wait);
>
> Why is this needed? Again, by the time we get to nfs4_evict_inode, the
> inode is guaranteed to be clean.
>
OK, let me just add a WARN_ON if we get here and layout commit is
required. Does that work for you?
>> + if (status)
>> + dprintk("%s: layoutcommit failed, status=%d. Returning layout anyway\n",
>> + __func__, status);
>> + status = return_layout(ino, &arg, wait);
>> +out:
>> + dprintk("<-- %s status: %d\n", __func__, status);
>> + return status;
>> +}
>> +
>> bool pnfs_roc(struct inode *ino)
>> {
>> struct pnfs_layout_hdr *lo;
>> diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
>> index 4cb0a0d..a308f3c 100644
>> --- a/fs/nfs/pnfs.h
>> +++ b/fs/nfs/pnfs.h
>> @@ -123,6 +123,7 @@ extern void pnfs_unregister_layoutdriver(struct pnfs_layoutdriver_type *);
>> extern int nfs4_proc_getdeviceinfo(struct nfs_server *server,
>> struct pnfs_device *dev);
>> extern int nfs4_proc_layoutget(struct nfs4_layoutget *lgp);
>> +extern int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool wait);
>>
>> /* pnfs.c */
>> void get_layout_hdr(struct pnfs_layout_hdr *lo);
>> @@ -158,6 +159,7 @@ void pnfs_roc_set_barrier(struct inode *ino, u32 barrier);
>> bool pnfs_roc_drain(struct inode *ino, u32 *barrier);
>> void pnfs_set_layoutcommit(struct nfs_write_data *wdata);
>> int pnfs_layoutcommit_inode(struct inode *inode, bool sync);
>> +int _pnfs_return_layout(struct inode *, struct pnfs_layout_range *, bool wait);
>>
>> static inline int lo_fail_bit(u32 iomode)
>> {
>> @@ -226,6 +228,19 @@ static inline void pnfs_clear_request_commit(struct nfs_page *req)
>> put_lseg(req->wb_commit_lseg);
>> }
>>
>> +static inline int pnfs_return_layout(struct inode *ino,
>> + struct pnfs_layout_range *range,
>> + bool wait)
>> +{
>> + struct nfs_inode *nfsi = NFS_I(ino);
>> + struct nfs_server *nfss = NFS_SERVER(ino);
>> +
>> + if (pnfs_enabled_sb(nfss) && nfsi->layout)
>> + return _pnfs_return_layout(ino, range, wait);
>> +
>> + return 0;
>> +}
>> +
>> #else /* CONFIG_NFS_V4_1 */
>>
>> static inline void pnfs_destroy_all_layouts(struct nfs_client *clp)
>> @@ -267,6 +282,13 @@ pnfs_try_to_write_data(struct nfs_write_data *data,
>> return PNFS_NOT_ATTEMPTED;
>> }
>>
>> +static inline int pnfs_return_layout(struct inode *ino,
>> + struct pnfs_layout_range *range,
>> + bool wait)
>> +{
>> + return 0;
>> +}
>> +
>> static inline bool
>> pnfs_roc(struct inode *ino)
>> {
>> diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h
>> index 178fafe..9376eaf 100644
>> --- a/include/linux/nfs4.h
>> +++ b/include/linux/nfs4.h
>> @@ -562,6 +562,7 @@ enum {
>> NFSPROC4_CLNT_LAYOUTGET,
>> NFSPROC4_CLNT_GETDEVICEINFO,
>> NFSPROC4_CLNT_LAYOUTCOMMIT,
>> + NFSPROC4_CLNT_LAYOUTRETURN,
>> };
>>
>> /* nfs41 types */
>> diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
>> index 78b101e..455ddfb 100644
>> --- a/include/linux/nfs_xdr.h
>> +++ b/include/linux/nfs_xdr.h
>> @@ -266,6 +266,29 @@ struct nfs4_layoutcommit_data {
>> struct nfs4_layoutcommit_res res;
>> };
>>
>> +struct nfs4_layoutreturn_args {
>> + __u32 reclaim;
>> + __u32 layout_type;
>> + __u32 return_type;
>
> Why do we need a 'return_type'? As far as I can see, this will always be
> RETURN_FILE.
>
Yeah, we can remove this for now.
Benny
>> + struct pnfs_layout_range range;
>> + struct inode *inode;
>> + struct nfs4_sequence_args seq_args;
>> +};
>> +
>> +struct nfs4_layoutreturn_res {
>> + struct nfs4_sequence_res seq_res;
>> + u32 lrs_present;
>> + nfs4_stateid stateid;
>> +};
>> +
>> +struct nfs4_layoutreturn {
>> + struct nfs4_layoutreturn_args args;
>> + struct nfs4_layoutreturn_res res;
>> + struct rpc_cred *cred;
>> + struct nfs_client *clp;
>> + int rpc_status;
>> +};
>> +
>> /*
>> * Arguments to the open call.
>> */
>
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
