Merge tag 'hwmon-for-v6.13-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git...

[linux.git] / fs / nfs / pagelist.c

// SPDX-License-Identifier: GPL-2.0-only
/*
 * linux/fs/nfs/pagelist.c
 *
 * A set of helper functions for managing NFS read and write requests.
 * The main purpose of these routines is to provide support for the
 * coalescing of several requests into a single RPC call.
 *
 * Copyright 2000, 2001 (c) Trond Myklebust <trond.myklebust@fys.uio.no>
 *
 */

#include <linux/slab.h>
#include <linux/file.h>
#include <linux/sched.h>
#include <linux/sunrpc/clnt.h>
#include <linux/nfs.h>
#include <linux/nfs3.h>
#include <linux/nfs4.h>
#include <linux/nfs_fs.h>
#include <linux/nfs_page.h>
#include <linux/nfs_mount.h>
#include <linux/export.h>
#include <linux/filelock.h>

#include "internal.h"
#include "pnfs.h"
#include "nfstrace.h"
#include "fscache.h"

#define NFSDBG_FACILITY         NFSDBG_PAGECACHE

static struct kmem_cache *nfs_page_cachep;
static const struct rpc_call_ops nfs_pgio_common_ops;

struct nfs_page_iter_page {
        const struct nfs_page *req;
        size_t count;
};

static void nfs_page_iter_page_init(struct nfs_page_iter_page *i,
                                    const struct nfs_page *req)
{
        i->req = req;
        i->count = 0;
}

static void nfs_page_iter_page_advance(struct nfs_page_iter_page *i, size_t sz)
{
        const struct nfs_page *req = i->req;
        size_t tmp = i->count + sz;

        i->count = (tmp < req->wb_bytes) ? tmp : req->wb_bytes;
}

static struct page *nfs_page_iter_page_get(struct nfs_page_iter_page *i)
{
        const struct nfs_page *req = i->req;
        struct page *page;

        if (i->count != req->wb_bytes) {
                size_t base = i->count + req->wb_pgbase;
                size_t len = PAGE_SIZE - offset_in_page(base);

                page = nfs_page_to_page(req, base);
                nfs_page_iter_page_advance(i, len);
                return page;
        }
        return NULL;
}

static struct nfs_pgio_mirror *
nfs_pgio_get_mirror(struct nfs_pageio_descriptor *desc, u32 idx)
{
        if (desc->pg_ops->pg_get_mirror)
                return desc->pg_ops->pg_get_mirror(desc, idx);
        return &desc->pg_mirrors[0];
}

struct nfs_pgio_mirror *
nfs_pgio_current_mirror(struct nfs_pageio_descriptor *desc)
{
        return nfs_pgio_get_mirror(desc, desc->pg_mirror_idx);
}
EXPORT_SYMBOL_GPL(nfs_pgio_current_mirror);

static u32
nfs_pgio_set_current_mirror(struct nfs_pageio_descriptor *desc, u32 idx)
{
        if (desc->pg_ops->pg_set_mirror)
                return desc->pg_ops->pg_set_mirror(desc, idx);
        return desc->pg_mirror_idx;
}

void nfs_pgheader_init(struct nfs_pageio_descriptor *desc,
                       struct nfs_pgio_header *hdr,
                       void (*release)(struct nfs_pgio_header *hdr))
{
        struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(desc);


        hdr->req = nfs_list_entry(mirror->pg_list.next);
        hdr->inode = desc->pg_inode;
        hdr->cred = nfs_req_openctx(hdr->req)->cred;
        hdr->io_start = req_offset(hdr->req);
        hdr->good_bytes = mirror->pg_count;
        hdr->io_completion = desc->pg_io_completion;
        hdr->dreq = desc->pg_dreq;
        nfs_netfs_set_pgio_header(hdr, desc);
        hdr->release = release;
        hdr->completion_ops = desc->pg_completion_ops;
        if (hdr->completion_ops->init_hdr)
                hdr->completion_ops->init_hdr(hdr);

        hdr->pgio_mirror_idx = desc->pg_mirror_idx;
}
EXPORT_SYMBOL_GPL(nfs_pgheader_init);

void nfs_set_pgio_error(struct nfs_pgio_header *hdr, int error, loff_t pos)
{
        unsigned int new = pos - hdr->io_start;

        trace_nfs_pgio_error(hdr, error, pos);
        if (hdr->good_bytes > new) {
                hdr->good_bytes = new;
                clear_bit(NFS_IOHDR_EOF, &hdr->flags);
                if (!test_and_set_bit(NFS_IOHDR_ERROR, &hdr->flags))
                        hdr->error = error;
        }
}

static inline struct nfs_page *nfs_page_alloc(void)
{
        struct nfs_page *p =
                kmem_cache_zalloc(nfs_page_cachep, nfs_io_gfp_mask());
        if (p)
                INIT_LIST_HEAD(&p->wb_list);
        return p;
}

static inline void
nfs_page_free(struct nfs_page *p)
{
        kmem_cache_free(nfs_page_cachep, p);
}

/**
 * nfs_iocounter_wait - wait for i/o to complete
 * @l_ctx: nfs_lock_context with io_counter to use
 *
 * returns -ERESTARTSYS if interrupted by a fatal signal.
 * Otherwise returns 0 once the io_count hits 0.
 */
int
nfs_iocounter_wait(struct nfs_lock_context *l_ctx)
{
        return wait_var_event_killable(&l_ctx->io_count,
                                       !atomic_read(&l_ctx->io_count));
}

/**
 * nfs_async_iocounter_wait - wait on a rpc_waitqueue for I/O
 * to complete
 * @task: the rpc_task that should wait
 * @l_ctx: nfs_lock_context with io_counter to check
 *
 * Returns true if there is outstanding I/O to wait on and the
 * task has been put to sleep.
 */
bool
nfs_async_iocounter_wait(struct rpc_task *task, struct nfs_lock_context *l_ctx)
{
        struct inode *inode = d_inode(l_ctx->open_context->dentry);
        bool ret = false;

        if (atomic_read(&l_ctx->io_count) > 0) {
                rpc_sleep_on(&NFS_SERVER(inode)->uoc_rpcwaitq, task, NULL);
                ret = true;
        }

        if (atomic_read(&l_ctx->io_count) == 0) {
                rpc_wake_up_queued_task(&NFS_SERVER(inode)->uoc_rpcwaitq, task);
                ret = false;
        }

        return ret;
}
EXPORT_SYMBOL_GPL(nfs_async_iocounter_wait);

/*
 * nfs_page_set_headlock - set the request PG_HEADLOCK
 * @req: request that is to be locked
 *
 * this lock must be held when modifying req->wb_head
 *
 * return 0 on success, < 0 on error
 */
int
nfs_page_set_headlock(struct nfs_page *req)
{
        if (!test_and_set_bit(PG_HEADLOCK, &req->wb_flags))
                return 0;

        set_bit(PG_CONTENDED1, &req->wb_flags);
        smp_mb__after_atomic();
        return wait_on_bit_lock(&req->wb_flags, PG_HEADLOCK,
                                TASK_UNINTERRUPTIBLE);
}

/*
 * nfs_page_clear_headlock - clear the request PG_HEADLOCK
 * @req: request that is to be locked
 */
void
nfs_page_clear_headlock(struct nfs_page *req)
{
        clear_bit_unlock(PG_HEADLOCK, &req->wb_flags);
        smp_mb__after_atomic();
        if (!test_bit(PG_CONTENDED1, &req->wb_flags))
                return;
        wake_up_bit(&req->wb_flags, PG_HEADLOCK);
}

/*
 * nfs_page_group_lock - lock the head of the page group
 * @req: request in group that is to be locked
 *
 * this lock must be held when traversing or modifying the page
 * group list
 *
 * return 0 on success, < 0 on error
 */
int
nfs_page_group_lock(struct nfs_page *req)
{
        int ret;

        ret = nfs_page_set_headlock(req);
        if (ret || req->wb_head == req)
                return ret;
        return nfs_page_set_headlock(req->wb_head);
}

/*
 * nfs_page_group_unlock - unlock the head of the page group
 * @req: request in group that is to be unlocked
 */
void
nfs_page_group_unlock(struct nfs_page *req)
{
        if (req != req->wb_head)
                nfs_page_clear_headlock(req->wb_head);
        nfs_page_clear_headlock(req);
}

/*
 * nfs_page_group_sync_on_bit_locked
 *
 * must be called with page group lock held
 */
static bool
nfs_page_group_sync_on_bit_locked(struct nfs_page *req, unsigned int bit)
{
        struct nfs_page *head = req->wb_head;
        struct nfs_page *tmp;

        WARN_ON_ONCE(!test_bit(PG_HEADLOCK, &head->wb_flags));
        WARN_ON_ONCE(test_and_set_bit(bit, &req->wb_flags));

        tmp = req->wb_this_page;
        while (tmp != req) {
                if (!test_bit(bit, &tmp->wb_flags))
                        return false;
                tmp = tmp->wb_this_page;
        }

        /* true! reset all bits */
        tmp = req;
        do {
                clear_bit(bit, &tmp->wb_flags);
                tmp = tmp->wb_this_page;
        } while (tmp != req);

        return true;
}

/*
 * nfs_page_group_sync_on_bit - set bit on current request, but only
 *   return true if the bit is set for all requests in page group
 * @req - request in page group
 * @bit - PG_* bit that is used to sync page group
 */
bool nfs_page_group_sync_on_bit(struct nfs_page *req, unsigned int bit)
{
        bool ret;

        nfs_page_group_lock(req);
        ret = nfs_page_group_sync_on_bit_locked(req, bit);
        nfs_page_group_unlock(req);

        return ret;
}

/*
 * nfs_page_group_init - Initialize the page group linkage for @req
 * @req - a new nfs request
 * @prev - the previous request in page group, or NULL if @req is the first
 *         or only request in the group (the head).
 */
static inline void
nfs_page_group_init(struct nfs_page *req, struct nfs_page *prev)
{
        struct inode *inode;
        WARN_ON_ONCE(prev == req);

        if (!prev) {
                /* a head request */
                req->wb_head = req;
                req->wb_this_page = req;
        } else {
                /* a subrequest */
                WARN_ON_ONCE(prev->wb_this_page != prev->wb_head);
                WARN_ON_ONCE(!test_bit(PG_HEADLOCK, &prev->wb_head->wb_flags));
                req->wb_head = prev->wb_head;
                req->wb_this_page = prev->wb_this_page;
                prev->wb_this_page = req;

                /* All subrequests take a ref on the head request until
                 * nfs_page_group_destroy is called */
                kref_get(&req->wb_head->wb_kref);

                /* grab extra ref and bump the request count if head request
                 * has extra ref from the write/commit path to handle handoff
                 * between write and commit lists. */
                if (test_bit(PG_INODE_REF, &prev->wb_head->wb_flags)) {
                        inode = nfs_page_to_inode(req);
                        set_bit(PG_INODE_REF, &req->wb_flags);
                        kref_get(&req->wb_kref);
                        atomic_long_inc(&NFS_I(inode)->nrequests);
                }
        }
}

/*
 * nfs_page_group_destroy - sync the destruction of page groups
 * @req - request that no longer needs the page group
 *
 * releases the page group reference from each member once all
 * members have called this function.
 */
static void
nfs_page_group_destroy(struct kref *kref)
{
        struct nfs_page *req = container_of(kref, struct nfs_page, wb_kref);
        struct nfs_page *head = req->wb_head;
        struct nfs_page *tmp, *next;

        if (!nfs_page_group_sync_on_bit(req, PG_TEARDOWN))
                goto out;

        tmp = req;
        do {
                next = tmp->wb_this_page;
                /* unlink and free */
                tmp->wb_this_page = tmp;
                tmp->wb_head = tmp;
                nfs_free_request(tmp);
                tmp = next;
        } while (tmp != req);
out:
        /* subrequests must release the ref on the head request */
        if (head != req)
                nfs_release_request(head);
}

static struct nfs_page *nfs_page_create(struct nfs_lock_context *l_ctx,
                                        unsigned int pgbase, pgoff_t index,
                                        unsigned int offset, unsigned int count)
{
        struct nfs_page         *req;
        struct nfs_open_context *ctx = l_ctx->open_context;

        if (test_bit(NFS_CONTEXT_BAD, &ctx->flags))
                return ERR_PTR(-EBADF);
        /* try to allocate the request struct */
        req = nfs_page_alloc();
        if (req == NULL)
                return ERR_PTR(-ENOMEM);

        req->wb_lock_context = l_ctx;
        refcount_inc(&l_ctx->count);
        atomic_inc(&l_ctx->io_count);

        /* Initialize the request struct. Initially, we assume a
         * long write-back delay. This will be adjusted in
         * update_nfs_request below if the region is not locked. */
        req->wb_pgbase = pgbase;
        req->wb_index = index;
        req->wb_offset = offset;
        req->wb_bytes = count;
        kref_init(&req->wb_kref);
        req->wb_nio = 0;
        return req;
}

static void nfs_page_assign_folio(struct nfs_page *req, struct folio *folio)
{
        if (folio != NULL) {
                req->wb_folio = folio;
                folio_get(folio);
                set_bit(PG_FOLIO, &req->wb_flags);
        }
}

static void nfs_page_assign_page(struct nfs_page *req, struct page *page)
{
        if (page != NULL) {
                req->wb_page = page;
                get_page(page);
        }
}

/**
 * nfs_page_create_from_page - Create an NFS read/write request.
 * @ctx: open context to use
 * @page: page to write
 * @pgbase: starting offset within the page for the write
 * @offset: file offset for the write
 * @count: number of bytes to read/write
 *
 * The page must be locked by the caller. This makes sure we never
 * create two different requests for the same page.
 * User should ensure it is safe to sleep in this function.
 */
struct nfs_page *nfs_page_create_from_page(struct nfs_open_context *ctx,
                                           struct page *page,
                                           unsigned int pgbase, loff_t offset,
                                           unsigned int count)
{
        struct nfs_lock_context *l_ctx = nfs_get_lock_context(ctx);
        struct nfs_page *ret;

        if (IS_ERR(l_ctx))
                return ERR_CAST(l_ctx);
        ret = nfs_page_create(l_ctx, pgbase, offset >> PAGE_SHIFT,
                              offset_in_page(offset), count);
        if (!IS_ERR(ret)) {
                nfs_page_assign_page(ret, page);
                nfs_page_group_init(ret, NULL);
        }
        nfs_put_lock_context(l_ctx);
        return ret;
}

/**
 * nfs_page_create_from_folio - Create an NFS read/write request.
 * @ctx: open context to use
 * @folio: folio to write
 * @offset: starting offset within the folio for the write
 * @count: number of bytes to read/write
 *
 * The page must be locked by the caller. This makes sure we never
 * create two different requests for the same page.
 * User should ensure it is safe to sleep in this function.
 */
struct nfs_page *nfs_page_create_from_folio(struct nfs_open_context *ctx,
                                            struct folio *folio,
                                            unsigned int offset,
                                            unsigned int count)
{
        struct nfs_lock_context *l_ctx = nfs_get_lock_context(ctx);
        struct nfs_page *ret;

        if (IS_ERR(l_ctx))
                return ERR_CAST(l_ctx);
        ret = nfs_page_create(l_ctx, offset, folio->index, offset, count);
        if (!IS_ERR(ret)) {
                nfs_page_assign_folio(ret, folio);
                nfs_page_group_init(ret, NULL);
        }
        nfs_put_lock_context(l_ctx);
        return ret;
}

static struct nfs_page *
nfs_create_subreq(struct nfs_page *req,
                  unsigned int pgbase,
                  unsigned int offset,
                  unsigned int count)
{
        struct nfs_page *last;
        struct nfs_page *ret;
        struct folio *folio = nfs_page_to_folio(req);
        struct page *page = nfs_page_to_page(req, pgbase);

        ret = nfs_page_create(req->wb_lock_context, pgbase, req->wb_index,
                              offset, count);
        if (!IS_ERR(ret)) {
                if (folio)
                        nfs_page_assign_folio(ret, folio);
                else
                        nfs_page_assign_page(ret, page);
                /* find the last request */
                for (last = req->wb_head;
                     last->wb_this_page != req->wb_head;
                     last = last->wb_this_page)
                        ;

                nfs_lock_request(ret);
                nfs_page_group_init(ret, last);
                ret->wb_nio = req->wb_nio;
        }
        return ret;
}

/**
 * nfs_unlock_request - Unlock request and wake up sleepers.
 * @req: pointer to request
 */
void nfs_unlock_request(struct nfs_page *req)
{
        clear_bit_unlock(PG_BUSY, &req->wb_flags);
        smp_mb__after_atomic();
        if (!test_bit(PG_CONTENDED2, &req->wb_flags))
                return;
        wake_up_bit(&req->wb_flags, PG_BUSY);
}

/**
 * nfs_unlock_and_release_request - Unlock request and release the nfs_page
 * @req: pointer to request
 */
void nfs_unlock_and_release_request(struct nfs_page *req)
{
        nfs_unlock_request(req);
        nfs_release_request(req);
}

/*
 * nfs_clear_request - Free up all resources allocated to the request
 * @req:
 *
 * Release page and open context resources associated with a read/write
 * request after it has completed.
 */
static void nfs_clear_request(struct nfs_page *req)
{
        struct folio *folio = nfs_page_to_folio(req);
        struct page *page = req->wb_page;
        struct nfs_lock_context *l_ctx = req->wb_lock_context;
        struct nfs_open_context *ctx;

        if (folio != NULL) {
                folio_put(folio);
                req->wb_folio = NULL;
                clear_bit(PG_FOLIO, &req->wb_flags);
        } else if (page != NULL) {
                put_page(page);
                req->wb_page = NULL;
        }
        if (l_ctx != NULL) {
                if (atomic_dec_and_test(&l_ctx->io_count)) {
                        wake_up_var(&l_ctx->io_count);
                        ctx = l_ctx->open_context;
                        if (test_bit(NFS_CONTEXT_UNLOCK, &ctx->flags))
                                rpc_wake_up(&NFS_SERVER(d_inode(ctx->dentry))->uoc_rpcwaitq);
                }
                nfs_put_lock_context(l_ctx);
                req->wb_lock_context = NULL;
        }
}

/**
 * nfs_free_request - Release the count on an NFS read/write request
 * @req: request to release
 *
 * Note: Should never be called with the spinlock held!
 */
void nfs_free_request(struct nfs_page *req)
{
        WARN_ON_ONCE(req->wb_this_page != req);

        /* extra debug: make sure no sync bits are still set */
        WARN_ON_ONCE(test_bit(PG_TEARDOWN, &req->wb_flags));
        WARN_ON_ONCE(test_bit(PG_UNLOCKPAGE, &req->wb_flags));
        WARN_ON_ONCE(test_bit(PG_UPTODATE, &req->wb_flags));
        WARN_ON_ONCE(test_bit(PG_WB_END, &req->wb_flags));
        WARN_ON_ONCE(test_bit(PG_REMOVE, &req->wb_flags));

        /* Release struct file and open context */
        nfs_clear_request(req);
        nfs_page_free(req);
}

void nfs_release_request(struct nfs_page *req)
{
        kref_put(&req->wb_kref, nfs_page_group_destroy);
}
EXPORT_SYMBOL_GPL(nfs_release_request);

/*
 * nfs_generic_pg_test - determine if requests can be coalesced
 * @desc: pointer to descriptor
 * @prev: previous request in desc, or NULL
 * @req: this request
 *
 * Returns zero if @req cannot be coalesced into @desc, otherwise it returns
 * the size of the request.
 */
size_t nfs_generic_pg_test(struct nfs_pageio_descriptor *desc,
                           struct nfs_page *prev, struct nfs_page *req)
{
        struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(desc);


        if (mirror->pg_count > mirror->pg_bsize) {
                /* should never happen */
                WARN_ON_ONCE(1);
                return 0;
        }

        /*
         * Limit the request size so that we can still allocate a page array
         * for it without upsetting the slab allocator.
         */
        if (((mirror->pg_count + req->wb_bytes) >> PAGE_SHIFT) *
                        sizeof(struct page *) > PAGE_SIZE)
                return 0;

        return min(mirror->pg_bsize - mirror->pg_count, (size_t)req->wb_bytes);
}
EXPORT_SYMBOL_GPL(nfs_generic_pg_test);

struct nfs_pgio_header *nfs_pgio_header_alloc(const struct nfs_rw_ops *ops)
{
        struct nfs_pgio_header *hdr = ops->rw_alloc_header();

        if (hdr) {
                INIT_LIST_HEAD(&hdr->pages);
                hdr->rw_ops = ops;
        }
        return hdr;
}
EXPORT_SYMBOL_GPL(nfs_pgio_header_alloc);

/**
 * nfs_pgio_data_destroy - make @hdr suitable for reuse
 *
 * Frees memory and releases refs from nfs_generic_pgio, so that it may
 * be called again.
 *
 * @hdr: A header that has had nfs_generic_pgio called
 */
static void nfs_pgio_data_destroy(struct nfs_pgio_header *hdr)
{
        if (hdr->args.context)
                put_nfs_open_context(hdr->args.context);
        if (hdr->page_array.pagevec != hdr->page_array.page_array)
                kfree(hdr->page_array.pagevec);
}

/*
 * nfs_pgio_header_free - Free a read or write header
 * @hdr: The header to free
 */
void nfs_pgio_header_free(struct nfs_pgio_header *hdr)
{
        nfs_pgio_data_destroy(hdr);
        hdr->rw_ops->rw_free_header(hdr);
}
EXPORT_SYMBOL_GPL(nfs_pgio_header_free);

/**
 * nfs_pgio_rpcsetup - Set up arguments for a pageio call
 * @hdr: The pageio hdr
 * @pgbase: base
 * @count: Number of bytes to read
 * @how: How to commit data (writes only)
 * @cinfo: Commit information for the call (writes only)
 */
static void nfs_pgio_rpcsetup(struct nfs_pgio_header *hdr, unsigned int pgbase,
                              unsigned int count, int how,
                              struct nfs_commit_info *cinfo)
{
        struct nfs_page *req = hdr->req;

        /* Set up the RPC argument and reply structs
         * NB: take care not to mess about with hdr->commit et al. */

        hdr->args.fh     = NFS_FH(hdr->inode);
        hdr->args.offset = req_offset(req);
        /* pnfs_set_layoutcommit needs this */
        hdr->mds_offset = hdr->args.offset;
        hdr->args.pgbase = pgbase;
        hdr->args.pages  = hdr->page_array.pagevec;
        hdr->args.count  = count;
        hdr->args.context = get_nfs_open_context(nfs_req_openctx(req));
        hdr->args.lock_context = req->wb_lock_context;
        hdr->args.stable  = NFS_UNSTABLE;
        switch (how & (FLUSH_STABLE | FLUSH_COND_STABLE)) {
        case 0:
                break;
        case FLUSH_COND_STABLE:
                if (nfs_reqs_to_commit(cinfo))
                        break;
                fallthrough;
        default:
                hdr->args.stable = NFS_FILE_SYNC;
        }

        hdr->res.fattr   = &hdr->fattr;
        hdr->res.count   = 0;
        hdr->res.eof     = 0;
        hdr->res.verf    = &hdr->verf;
        nfs_fattr_init(&hdr->fattr);
}

/**
 * nfs_pgio_prepare - Prepare pageio hdr to go over the wire
 * @task: The current task
 * @calldata: pageio header to prepare
 */
static void nfs_pgio_prepare(struct rpc_task *task, void *calldata)
{
        struct nfs_pgio_header *hdr = calldata;
        int err;
        err = NFS_PROTO(hdr->inode)->pgio_rpc_prepare(task, hdr);
        if (err)
                rpc_exit(task, err);
}

int nfs_initiate_pgio(struct rpc_clnt *clnt, struct nfs_pgio_header *hdr,
                      const struct cred *cred, const struct nfs_rpc_ops *rpc_ops,
                      const struct rpc_call_ops *call_ops, int how, int flags,
                      struct nfsd_file *localio)
{
        struct rpc_task *task;
        struct rpc_message msg = {
                .rpc_argp = &hdr->args,
                .rpc_resp = &hdr->res,
                .rpc_cred = cred,
        };
        struct rpc_task_setup task_setup_data = {
                .rpc_client = clnt,
                .task = &hdr->task,
                .rpc_message = &msg,
                .callback_ops = call_ops,
                .callback_data = hdr,
                .workqueue = nfsiod_workqueue,
                .flags = RPC_TASK_ASYNC | flags,
        };

        if (nfs_server_capable(hdr->inode, NFS_CAP_MOVEABLE))
                task_setup_data.flags |= RPC_TASK_MOVEABLE;

        hdr->rw_ops->rw_initiate(hdr, &msg, rpc_ops, &task_setup_data, how);

        dprintk("NFS: initiated pgio call "
                "(req %s/%llu, %u bytes @ offset %llu)\n",
                hdr->inode->i_sb->s_id,
                (unsigned long long)NFS_FILEID(hdr->inode),
                hdr->args.count,
                (unsigned long long)hdr->args.offset);

        if (localio)
                return nfs_local_doio(NFS_SERVER(hdr->inode)->nfs_client,
                                      localio, hdr, call_ops);

        task = rpc_run_task(&task_setup_data);
        if (IS_ERR(task))
                return PTR_ERR(task);
        rpc_put_task(task);
        return 0;
}
EXPORT_SYMBOL_GPL(nfs_initiate_pgio);

/**
 * nfs_pgio_error - Clean up from a pageio error
 * @hdr: pageio header
 */
static void nfs_pgio_error(struct nfs_pgio_header *hdr)
{
        set_bit(NFS_IOHDR_REDO, &hdr->flags);
        hdr->completion_ops->completion(hdr);
}

/**
 * nfs_pgio_release - Release pageio data
 * @calldata: The pageio header to release
 */
static void nfs_pgio_release(void *calldata)
{
        struct nfs_pgio_header *hdr = calldata;
        hdr->completion_ops->completion(hdr);
}

static void nfs_pageio_mirror_init(struct nfs_pgio_mirror *mirror,
                                   unsigned int bsize)
{
        INIT_LIST_HEAD(&mirror->pg_list);
        mirror->pg_bytes_written = 0;
        mirror->pg_count = 0;
        mirror->pg_bsize = bsize;
        mirror->pg_base = 0;
        mirror->pg_recoalesce = 0;
}

/**
 * nfs_pageio_init - initialise a page io descriptor
 * @desc: pointer to descriptor
 * @inode: pointer to inode
 * @pg_ops: pointer to pageio operations
 * @compl_ops: pointer to pageio completion operations
 * @rw_ops: pointer to nfs read/write operations
 * @bsize: io block size
 * @io_flags: extra parameters for the io function
 */
void nfs_pageio_init(struct nfs_pageio_descriptor *desc,
                     struct inode *inode,
                     const struct nfs_pageio_ops *pg_ops,
                     const struct nfs_pgio_completion_ops *compl_ops,
                     const struct nfs_rw_ops *rw_ops,
                     size_t bsize,
                     int io_flags)
{
        desc->pg_moreio = 0;
        desc->pg_inode = inode;
        desc->pg_ops = pg_ops;
        desc->pg_completion_ops = compl_ops;
        desc->pg_rw_ops = rw_ops;
        desc->pg_ioflags = io_flags;
        desc->pg_error = 0;
        desc->pg_lseg = NULL;
        desc->pg_io_completion = NULL;
        desc->pg_dreq = NULL;
        nfs_netfs_reset_pageio_descriptor(desc);
        desc->pg_bsize = bsize;

        desc->pg_mirror_count = 1;
        desc->pg_mirror_idx = 0;

        desc->pg_mirrors_dynamic = NULL;
        desc->pg_mirrors = desc->pg_mirrors_static;
        nfs_pageio_mirror_init(&desc->pg_mirrors[0], bsize);
        desc->pg_maxretrans = 0;
}

/**
 * nfs_pgio_result - Basic pageio error handling
 * @task: The task that ran
 * @calldata: Pageio header to check
 */
static void nfs_pgio_result(struct rpc_task *task, void *calldata)
{
        struct nfs_pgio_header *hdr = calldata;
        struct inode *inode = hdr->inode;

        if (hdr->rw_ops->rw_done(task, hdr, inode) != 0)
                return;
        if (task->tk_status < 0)
                nfs_set_pgio_error(hdr, task->tk_status, hdr->args.offset);
        else
                hdr->rw_ops->rw_result(task, hdr);
}

/*
 * Create an RPC task for the given read or write request and kick it.
 * The page must have been locked by the caller.
 *
 * It may happen that the page we're passed is not marked dirty.
 * This is the case if nfs_updatepage detects a conflicting request
 * that has been written but not committed.
 */
int nfs_generic_pgio(struct nfs_pageio_descriptor *desc,
                     struct nfs_pgio_header *hdr)
{
        struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(desc);

        struct nfs_page         *req;
        struct page             **pages,
                                *last_page;
        struct list_head *head = &mirror->pg_list;
        struct nfs_commit_info cinfo;
        struct nfs_page_array *pg_array = &hdr->page_array;
        unsigned int pagecount, pageused;
        unsigned int pg_base = offset_in_page(mirror->pg_base);
        gfp_t gfp_flags = nfs_io_gfp_mask();

        pagecount = nfs_page_array_len(pg_base, mirror->pg_count);
        pg_array->npages = pagecount;

        if (pagecount <= ARRAY_SIZE(pg_array->page_array))
                pg_array->pagevec = pg_array->page_array;
        else {
                pg_array->pagevec = kcalloc(pagecount, sizeof(struct page *), gfp_flags);
                if (!pg_array->pagevec) {
                        pg_array->npages = 0;
                        nfs_pgio_error(hdr);
                        desc->pg_error = -ENOMEM;
                        return desc->pg_error;
                }
        }

        nfs_init_cinfo(&cinfo, desc->pg_inode, desc->pg_dreq);
        pages = hdr->page_array.pagevec;
        last_page = NULL;
        pageused = 0;
        while (!list_empty(head)) {
                struct nfs_page_iter_page i;
                struct page *page;

                req = nfs_list_entry(head->next);
                nfs_list_move_request(req, &hdr->pages);

                if (req->wb_pgbase == 0)
                        last_page = NULL;

                nfs_page_iter_page_init(&i, req);
                while ((page = nfs_page_iter_page_get(&i)) != NULL) {
                        if (last_page != page) {
                                pageused++;
                                if (pageused > pagecount)
                                        goto full;
                                *pages++ = last_page = page;
                        }
                }
        }
full:
        if (WARN_ON_ONCE(pageused != pagecount)) {
                nfs_pgio_error(hdr);
                desc->pg_error = -EINVAL;
                return desc->pg_error;
        }

        if ((desc->pg_ioflags & FLUSH_COND_STABLE) &&
            (desc->pg_moreio || nfs_reqs_to_commit(&cinfo)))
                desc->pg_ioflags &= ~FLUSH_COND_STABLE;

        /* Set up the argument struct */
        nfs_pgio_rpcsetup(hdr, pg_base, mirror->pg_count, desc->pg_ioflags,
                          &cinfo);
        desc->pg_rpc_callops = &nfs_pgio_common_ops;
        return 0;
}
EXPORT_SYMBOL_GPL(nfs_generic_pgio);

static int nfs_generic_pg_pgios(struct nfs_pageio_descriptor *desc)
{
        struct nfs_pgio_header *hdr;
        int ret;
        unsigned short task_flags = 0;

        hdr = nfs_pgio_header_alloc(desc->pg_rw_ops);
        if (!hdr) {
                desc->pg_error = -ENOMEM;
                return desc->pg_error;
        }
        nfs_pgheader_init(desc, hdr, nfs_pgio_header_free);
        ret = nfs_generic_pgio(desc, hdr);
        if (ret == 0) {
                struct nfs_client *clp = NFS_SERVER(hdr->inode)->nfs_client;

                struct nfsd_file *localio =
                        nfs_local_open_fh(clp, hdr->cred,
                                          hdr->args.fh, hdr->args.context->mode);

                if (NFS_SERVER(hdr->inode)->nfs_client->cl_minorversion)
                        task_flags = RPC_TASK_MOVEABLE;
                ret = nfs_initiate_pgio(NFS_CLIENT(hdr->inode),
                                        hdr,
                                        hdr->cred,
                                        NFS_PROTO(hdr->inode),
                                        desc->pg_rpc_callops,
                                        desc->pg_ioflags,
                                        RPC_TASK_CRED_NOREF | task_flags,
                                        localio);
        }
        return ret;
}

static struct nfs_pgio_mirror *
nfs_pageio_alloc_mirrors(struct nfs_pageio_descriptor *desc,
                unsigned int mirror_count)
{
        struct nfs_pgio_mirror *ret;
        unsigned int i;

        kfree(desc->pg_mirrors_dynamic);
        desc->pg_mirrors_dynamic = NULL;
        if (mirror_count == 1)
                return desc->pg_mirrors_static;
        ret = kmalloc_array(mirror_count, sizeof(*ret), nfs_io_gfp_mask());
        if (ret != NULL) {
                for (i = 0; i < mirror_count; i++)
                        nfs_pageio_mirror_init(&ret[i], desc->pg_bsize);
                desc->pg_mirrors_dynamic = ret;
        }
        return ret;
}

/*
 * nfs_pageio_setup_mirroring - determine if mirroring is to be used
 *                              by calling the pg_get_mirror_count op
 */
static void nfs_pageio_setup_mirroring(struct nfs_pageio_descriptor *pgio,
                                       struct nfs_page *req)
{
        unsigned int mirror_count = 1;

        if (pgio->pg_ops->pg_get_mirror_count)
                mirror_count = pgio->pg_ops->pg_get_mirror_count(pgio, req);
        if (mirror_count == pgio->pg_mirror_count || pgio->pg_error < 0)
                return;

        if (!mirror_count || mirror_count > NFS_PAGEIO_DESCRIPTOR_MIRROR_MAX) {
                pgio->pg_error = -EINVAL;
                return;
        }

        pgio->pg_mirrors = nfs_pageio_alloc_mirrors(pgio, mirror_count);
        if (pgio->pg_mirrors == NULL) {
                pgio->pg_error = -ENOMEM;
                pgio->pg_mirrors = pgio->pg_mirrors_static;
                mirror_count = 1;
        }
        pgio->pg_mirror_count = mirror_count;
}

static void nfs_pageio_cleanup_mirroring(struct nfs_pageio_descriptor *pgio)
{
        pgio->pg_mirror_count = 1;
        pgio->pg_mirror_idx = 0;
        pgio->pg_mirrors = pgio->pg_mirrors_static;
        kfree(pgio->pg_mirrors_dynamic);
        pgio->pg_mirrors_dynamic = NULL;
}

static bool nfs_match_lock_context(const struct nfs_lock_context *l1,
                const struct nfs_lock_context *l2)
{
        return l1->lockowner == l2->lockowner;
}

static bool nfs_page_is_contiguous(const struct nfs_page *prev,
                                   const struct nfs_page *req)
{
        size_t prev_end = prev->wb_pgbase + prev->wb_bytes;

        if (req_offset(req) != req_offset(prev) + prev->wb_bytes)
                return false;
        if (req->wb_pgbase == 0)
                return prev_end == nfs_page_max_length(prev);
        if (req->wb_pgbase == prev_end) {
                struct folio *folio = nfs_page_to_folio(req);
                if (folio)
                        return folio == nfs_page_to_folio(prev);
                return req->wb_page == prev->wb_page;
        }
        return false;
}

/**
 * nfs_coalesce_size - test two requests for compatibility
 * @prev: pointer to nfs_page
 * @req: pointer to nfs_page
 * @pgio: pointer to nfs_pagio_descriptor
 *
 * The nfs_page structures 'prev' and 'req' are compared to ensure that the
 * page data area they describe is contiguous, and that their RPC
 * credentials, NFSv4 open state, and lockowners are the same.
 *
 * Returns size of the request that can be coalesced
 */
static unsigned int nfs_coalesce_size(struct nfs_page *prev,
                                      struct nfs_page *req,
                                      struct nfs_pageio_descriptor *pgio)
{
        struct file_lock_context *flctx;

        if (prev) {
                if (!nfs_match_open_context(nfs_req_openctx(req), nfs_req_openctx(prev)))
                        return 0;
                flctx = locks_inode_context(d_inode(nfs_req_openctx(req)->dentry));
                if (flctx != NULL &&
                    !(list_empty_careful(&flctx->flc_posix) &&
                      list_empty_careful(&flctx->flc_flock)) &&
                    !nfs_match_lock_context(req->wb_lock_context,
                                            prev->wb_lock_context))
                        return 0;
                if (!nfs_page_is_contiguous(prev, req))
                        return 0;
        }
        return pgio->pg_ops->pg_test(pgio, prev, req);
}

/**
 * nfs_pageio_do_add_request - Attempt to coalesce a request into a page list.
 * @desc: destination io descriptor
 * @req: request
 *
 * If the request 'req' was successfully coalesced into the existing list
 * of pages 'desc', it returns the size of req.
 */
static unsigned int
nfs_pageio_do_add_request(struct nfs_pageio_descriptor *desc,
                struct nfs_page *req)
{
        struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(desc);
        struct nfs_page *prev = NULL;
        unsigned int size;

        if (list_empty(&mirror->pg_list)) {
                if (desc->pg_ops->pg_init)
                        desc->pg_ops->pg_init(desc, req);
                if (desc->pg_error < 0)
                        return 0;
                mirror->pg_base = req->wb_pgbase;
                mirror->pg_count = 0;
                mirror->pg_recoalesce = 0;
        } else
                prev = nfs_list_entry(mirror->pg_list.prev);

        if (desc->pg_maxretrans && req->wb_nio > desc->pg_maxretrans) {
                if (NFS_SERVER(desc->pg_inode)->flags & NFS_MOUNT_SOFTERR)
                        desc->pg_error = -ETIMEDOUT;
                else
                        desc->pg_error = -EIO;
                return 0;
        }

        size = nfs_coalesce_size(prev, req, desc);
        if (size < req->wb_bytes)
                return size;
        nfs_list_move_request(req, &mirror->pg_list);
        mirror->pg_count += req->wb_bytes;
        return req->wb_bytes;
}

/*
 * Helper for nfs_pageio_add_request and nfs_pageio_complete
 */
static void nfs_pageio_doio(struct nfs_pageio_descriptor *desc)
{
        struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(desc);

        if (!list_empty(&mirror->pg_list)) {
                int error = desc->pg_ops->pg_doio(desc);
                if (error < 0)
                        desc->pg_error = error;
                if (list_empty(&mirror->pg_list))
                        mirror->pg_bytes_written += mirror->pg_count;
        }
}

static void
nfs_pageio_cleanup_request(struct nfs_pageio_descriptor *desc,
                struct nfs_page *req)
{
        LIST_HEAD(head);

        nfs_list_move_request(req, &head);
        desc->pg_completion_ops->error_cleanup(&head, desc->pg_error);
}

/**
 * __nfs_pageio_add_request - Attempt to coalesce a request into a page list.
 * @desc: destination io descriptor
 * @req: request
 *
 * This may split a request into subrequests which are all part of the
 * same page group. If so, it will submit @req as the last one, to ensure
 * the pointer to @req is still valid in case of failure.
 *
 * Returns true if the request 'req' was successfully coalesced into the
 * existing list of pages 'desc'.
 */
static int __nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
                           struct nfs_page *req)
{
        struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(desc);
        struct nfs_page *subreq;
        unsigned int size, subreq_size;

        nfs_page_group_lock(req);

        subreq = req;
        subreq_size = subreq->wb_bytes;
        for(;;) {
                size = nfs_pageio_do_add_request(desc, subreq);
                if (size == subreq_size) {
                        /* We successfully submitted a request */
                        if (subreq == req)
                                break;
                        req->wb_pgbase += size;
                        req->wb_bytes -= size;
                        req->wb_offset += size;
                        subreq_size = req->wb_bytes;
                        subreq = req;
                        continue;
                }
                if (WARN_ON_ONCE(subreq != req)) {
                        nfs_page_group_unlock(req);
                        nfs_pageio_cleanup_request(desc, subreq);
                        subreq = req;
                        subreq_size = req->wb_bytes;
                        nfs_page_group_lock(req);
                }
                if (!size) {
                        /* Can't coalesce any more, so do I/O */
                        nfs_page_group_unlock(req);
                        desc->pg_moreio = 1;
                        nfs_pageio_doio(desc);
                        if (desc->pg_error < 0 || mirror->pg_recoalesce)
                                return 0;
                        /* retry add_request for this subreq */
                        nfs_page_group_lock(req);
                        continue;
                }
                subreq = nfs_create_subreq(req, req->wb_pgbase,
                                req->wb_offset, size);
                if (IS_ERR(subreq))
                        goto err_ptr;
                subreq_size = size;
        }

        nfs_page_group_unlock(req);
        return 1;
err_ptr:
        desc->pg_error = PTR_ERR(subreq);
        nfs_page_group_unlock(req);
        return 0;
}

static int nfs_do_recoalesce(struct nfs_pageio_descriptor *desc)
{
        struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(desc);
        LIST_HEAD(head);

        do {
                list_splice_init(&mirror->pg_list, &head);
                mirror->pg_recoalesce = 0;

                while (!list_empty(&head)) {
                        struct nfs_page *req;

                        req = list_first_entry(&head, struct nfs_page, wb_list);
                        if (__nfs_pageio_add_request(desc, req))
                                continue;
                        if (desc->pg_error < 0) {
                                list_splice_tail(&head, &mirror->pg_list);
                                mirror->pg_recoalesce = 1;
                                return 0;
                        }
                        break;
                }
        } while (mirror->pg_recoalesce);
        return 1;
}

static int nfs_pageio_add_request_mirror(struct nfs_pageio_descriptor *desc,
                struct nfs_page *req)
{
        int ret;

        do {
                ret = __nfs_pageio_add_request(desc, req);
                if (ret)
                        break;
                if (desc->pg_error < 0)
                        break;
                ret = nfs_do_recoalesce(desc);
        } while (ret);

        return ret;
}

static void nfs_pageio_error_cleanup(struct nfs_pageio_descriptor *desc)
{
        u32 midx;
        struct nfs_pgio_mirror *mirror;

        if (!desc->pg_error)
                return;

        for (midx = 0; midx < desc->pg_mirror_count; midx++) {
                mirror = nfs_pgio_get_mirror(desc, midx);
                desc->pg_completion_ops->error_cleanup(&mirror->pg_list,
                                desc->pg_error);
        }
}

int nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
                           struct nfs_page *req)
{
        u32 midx;
        unsigned int pgbase, offset, bytes;
        struct nfs_page *dupreq;

        pgbase = req->wb_pgbase;
        offset = req->wb_offset;
        bytes = req->wb_bytes;

        nfs_pageio_setup_mirroring(desc, req);
        if (desc->pg_error < 0)
                goto out_failed;

        /* Create the mirror instances first, and fire them off */
        for (midx = 1; midx < desc->pg_mirror_count; midx++) {
                nfs_page_group_lock(req);

                dupreq = nfs_create_subreq(req,
                                pgbase, offset, bytes);

                nfs_page_group_unlock(req);
                if (IS_ERR(dupreq)) {
                        desc->pg_error = PTR_ERR(dupreq);
                        goto out_failed;
                }

                nfs_pgio_set_current_mirror(desc, midx);
                if (!nfs_pageio_add_request_mirror(desc, dupreq))
                        goto out_cleanup_subreq;
        }

        nfs_pgio_set_current_mirror(desc, 0);
        if (!nfs_pageio_add_request_mirror(desc, req))
                goto out_failed;

        return 1;

out_cleanup_subreq:
        nfs_pageio_cleanup_request(desc, dupreq);
out_failed:
        nfs_pageio_error_cleanup(desc);
        return 0;
}

/*
 * nfs_pageio_complete_mirror - Complete I/O on the current mirror of an
 *                              nfs_pageio_descriptor
 * @desc: pointer to io descriptor
 * @mirror_idx: pointer to mirror index
 */
static void nfs_pageio_complete_mirror(struct nfs_pageio_descriptor *desc,
                                       u32 mirror_idx)
{
        struct nfs_pgio_mirror *mirror;
        u32 restore_idx;

        restore_idx = nfs_pgio_set_current_mirror(desc, mirror_idx);
        mirror = nfs_pgio_current_mirror(desc);

        for (;;) {
                nfs_pageio_doio(desc);
                if (desc->pg_error < 0 || !mirror->pg_recoalesce)
                        break;
                if (!nfs_do_recoalesce(desc))
                        break;
        }
        nfs_pgio_set_current_mirror(desc, restore_idx);
}

/*
 * nfs_pageio_resend - Transfer requests to new descriptor and resend
 * @hdr - the pgio header to move request from
 * @desc - the pageio descriptor to add requests to
 *
 * Try to move each request (nfs_page) from @hdr to @desc then attempt
 * to send them.
 *
 * Returns 0 on success and < 0 on error.
 */
int nfs_pageio_resend(struct nfs_pageio_descriptor *desc,
                      struct nfs_pgio_header *hdr)
{
        LIST_HEAD(pages);

        desc->pg_io_completion = hdr->io_completion;
        desc->pg_dreq = hdr->dreq;
        nfs_netfs_set_pageio_descriptor(desc, hdr);
        list_splice_init(&hdr->pages, &pages);
        while (!list_empty(&pages)) {
                struct nfs_page *req = nfs_list_entry(pages.next);

                if (!nfs_pageio_add_request(desc, req))
                        break;
        }
        nfs_pageio_complete(desc);
        if (!list_empty(&pages)) {
                int err = desc->pg_error < 0 ? desc->pg_error : -EIO;
                hdr->completion_ops->error_cleanup(&pages, err);
                nfs_set_pgio_error(hdr, err, hdr->io_start);
                return err;
        }
        return 0;
}
EXPORT_SYMBOL_GPL(nfs_pageio_resend);

/**
 * nfs_pageio_complete - Complete I/O then cleanup an nfs_pageio_descriptor
 * @desc: pointer to io descriptor
 */
void nfs_pageio_complete(struct nfs_pageio_descriptor *desc)
{
        u32 midx;

        for (midx = 0; midx < desc->pg_mirror_count; midx++)
                nfs_pageio_complete_mirror(desc, midx);

        if (desc->pg_error < 0)
                nfs_pageio_error_cleanup(desc);
        if (desc->pg_ops->pg_cleanup)
                desc->pg_ops->pg_cleanup(desc);
        nfs_pageio_cleanup_mirroring(desc);
}

/**
 * nfs_pageio_cond_complete - Conditional I/O completion
 * @desc: pointer to io descriptor
 * @index: page index
 *
 * It is important to ensure that processes don't try to take locks
 * on non-contiguous ranges of pages as that might deadlock. This
 * function should be called before attempting to wait on a locked
 * nfs_page. It will complete the I/O if the page index 'index'
 * is not contiguous with the existing list of pages in 'desc'.
 */
void nfs_pageio_cond_complete(struct nfs_pageio_descriptor *desc, pgoff_t index)
{
        struct nfs_pgio_mirror *mirror;
        struct nfs_page *prev;
        struct folio *folio;
        u32 midx;

        for (midx = 0; midx < desc->pg_mirror_count; midx++) {
                mirror = nfs_pgio_get_mirror(desc, midx);
                if (!list_empty(&mirror->pg_list)) {
                        prev = nfs_list_entry(mirror->pg_list.prev);
                        folio = nfs_page_to_folio(prev);
                        if (folio) {
                                if (index == folio_next_index(folio))
                                        continue;
                        } else if (index == prev->wb_index + 1)
                                continue;
                        /*
                         * We will submit more requests after these. Indicate
                         * this to the underlying layers.
                         */
                        desc->pg_moreio = 1;
                        nfs_pageio_complete(desc);
                        break;
                }
        }
}

/*
 * nfs_pageio_stop_mirroring - stop using mirroring (set mirror count to 1)
 */
void nfs_pageio_stop_mirroring(struct nfs_pageio_descriptor *pgio)
{
        nfs_pageio_complete(pgio);
}

int __init nfs_init_nfspagecache(void)
{
        nfs_page_cachep = kmem_cache_create("nfs_page",
                                            sizeof(struct nfs_page),
                                            0, SLAB_HWCACHE_ALIGN,
                                            NULL);
        if (nfs_page_cachep == NULL)
                return -ENOMEM;

        return 0;
}

void nfs_destroy_nfspagecache(void)
{
        kmem_cache_destroy(nfs_page_cachep);
}

static const struct rpc_call_ops nfs_pgio_common_ops = {
        .rpc_call_prepare = nfs_pgio_prepare,
        .rpc_call_done = nfs_pgio_result,
        .rpc_release = nfs_pgio_release,
};

const struct nfs_pageio_ops nfs_pgio_rw_ops = {
        .pg_test = nfs_generic_pg_test,
        .pg_doio = nfs_generic_pg_pgios,
};