提交 6d22323b 编写于 作者: C Christoph Hellwig 提交者: Trond Myklebust

nfs: remove the objlayout driver

The objlayout code has been in the tree, but it's been unmaintained and
no server product for it actually ever shipped.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NTrond Myklebust <trond.myklebust@primarydata.com>
上级 260f32ad
......@@ -2419,12 +2419,6 @@
and gids from such clients. This is intended to ease
migration from NFSv2/v3.
objlayoutdriver.osd_login_prog=
[NFS] [OBJLAYOUT] sets the pathname to the program which
is used to automatically discover and login into new
osd-targets. Please see:
Documentation/filesystems/pnfs.txt for more explanations
nmi_debug= [KNL,AVR32,SH] Specify one or more actions to take
when a NMI is triggered.
Format: [state][,regs][,debounce][,die]
......
......@@ -64,46 +64,9 @@ table which are called by the nfs-client pnfs-core to implement the
different layout types.
Files-layout-driver code is in: fs/nfs/filelayout/.. directory
Objects-layout-driver code is in: fs/nfs/objlayout/.. directory
Blocks-layout-driver code is in: fs/nfs/blocklayout/.. directory
Flexfiles-layout-driver code is in: fs/nfs/flexfilelayout/.. directory
objects-layout setup
--------------------
As part of the full STD implementation the objlayoutdriver.ko needs, at times,
to automatically login to yet undiscovered iscsi/osd devices. For this the
driver makes up-calles to a user-mode script called *osd_login*
The path_name of the script to use is by default:
/sbin/osd_login.
This name can be overridden by the Kernel module parameter:
objlayoutdriver.osd_login_prog
If Kernel does not find the osd_login_prog path it will zero it out
and will not attempt farther logins. An admin can then write new value
to the objlayoutdriver.osd_login_prog Kernel parameter to re-enable it.
The /sbin/osd_login is part of the nfs-utils package, and should usually
be installed on distributions that support this Kernel version.
The API to the login script is as follows:
Usage: $0 -u <URI> -o <OSDNAME> -s <SYSTEMID>
Options:
-u target uri e.g. iscsi://<ip>:<port>
(always exists)
(More protocols can be defined in the future.
The client does not interpret this string it is
passed unchanged as received from the Server)
-o osdname of the requested target OSD
(Might be empty)
(A string which denotes the OSD name, there is a
limit of 64 chars on this string)
-s systemid of the requested target OSD
(Might be empty)
(This string, if not empty is always an hex
representation of the 20 bytes osd_system_id)
blocks-layout setup
-------------------
......
......@@ -123,11 +123,6 @@ config PNFS_BLOCK
depends on NFS_V4_1 && BLK_DEV_DM
default NFS_V4
config PNFS_OBJLAYOUT
tristate
depends on NFS_V4_1 && SCSI_OSD_ULD
default NFS_V4
config PNFS_FLEXFILE_LAYOUT
tristate
depends on NFS_V4_1 && NFS_V3
......
......@@ -31,6 +31,5 @@ nfsv4-$(CONFIG_NFS_V4_1) += pnfs.o pnfs_dev.o pnfs_nfs.o
nfsv4-$(CONFIG_NFS_V4_2) += nfs42proc.o
obj-$(CONFIG_PNFS_FILE_LAYOUT) += filelayout/
obj-$(CONFIG_PNFS_OBJLAYOUT) += objlayout/
obj-$(CONFIG_PNFS_BLOCK) += blocklayout/
obj-$(CONFIG_PNFS_FLEXFILE_LAYOUT) += flexfilelayout/
#
# Makefile for the pNFS Objects Layout Driver kernel module
#
objlayoutdriver-y := objio_osd.o pnfs_osd_xdr_cli.o objlayout.o
obj-$(CONFIG_PNFS_OBJLAYOUT) += objlayoutdriver.o
/*
* pNFS Objects layout implementation over open-osd initiator library
*
* Copyright (C) 2009 Panasas Inc. [year of first publication]
* All rights reserved.
*
* Benny Halevy <bhalevy@panasas.com>
* Boaz Harrosh <ooo@electrozaur.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2
* See the file COPYING included with this distribution for more details.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the Panasas company nor the names of its
* contributors may be used to endorse or promote products derived
* from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
* DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
* BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
* LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
* NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
* SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include <linux/module.h>
#include <scsi/osd_ore.h>
#include "objlayout.h"
#include "../internal.h"
#define NFSDBG_FACILITY NFSDBG_PNFS_LD
struct objio_dev_ent {
struct nfs4_deviceid_node id_node;
struct ore_dev od;
};
static void
objio_free_deviceid_node(struct nfs4_deviceid_node *d)
{
struct objio_dev_ent *de = container_of(d, struct objio_dev_ent, id_node);
dprintk("%s: free od=%p\n", __func__, de->od.od);
osduld_put_device(de->od.od);
kfree_rcu(d, rcu);
}
struct objio_segment {
struct pnfs_layout_segment lseg;
struct ore_layout layout;
struct ore_components oc;
};
static inline struct objio_segment *
OBJIO_LSEG(struct pnfs_layout_segment *lseg)
{
return container_of(lseg, struct objio_segment, lseg);
}
struct objio_state {
/* Generic layer */
struct objlayout_io_res oir;
bool sync;
/*FIXME: Support for extra_bytes at ore_get_rw_state() */
struct ore_io_state *ios;
};
/* Send and wait for a get_device_info of devices in the layout,
then look them up with the osd_initiator library */
struct nfs4_deviceid_node *
objio_alloc_deviceid_node(struct nfs_server *server, struct pnfs_device *pdev,
gfp_t gfp_flags)
{
struct pnfs_osd_deviceaddr *deviceaddr;
struct objio_dev_ent *ode = NULL;
struct osd_dev *od;
struct osd_dev_info odi;
bool retry_flag = true;
__be32 *p;
int err;
deviceaddr = kzalloc(sizeof(*deviceaddr), gfp_flags);
if (!deviceaddr)
return NULL;
p = page_address(pdev->pages[0]);
pnfs_osd_xdr_decode_deviceaddr(deviceaddr, p);
odi.systemid_len = deviceaddr->oda_systemid.len;
if (odi.systemid_len > sizeof(odi.systemid)) {
dprintk("%s: odi.systemid_len > sizeof(systemid=%zd)\n",
__func__, sizeof(odi.systemid));
err = -EINVAL;
goto out;
} else if (odi.systemid_len)
memcpy(odi.systemid, deviceaddr->oda_systemid.data,
odi.systemid_len);
odi.osdname_len = deviceaddr->oda_osdname.len;
odi.osdname = (u8 *)deviceaddr->oda_osdname.data;
if (!odi.osdname_len && !odi.systemid_len) {
dprintk("%s: !odi.osdname_len && !odi.systemid_len\n",
__func__);
err = -ENODEV;
goto out;
}
retry_lookup:
od = osduld_info_lookup(&odi);
if (IS_ERR(od)) {
err = PTR_ERR(od);
dprintk("%s: osduld_info_lookup => %d\n", __func__, err);
if (err == -ENODEV && retry_flag) {
err = objlayout_autologin(deviceaddr);
if (likely(!err)) {
retry_flag = false;
goto retry_lookup;
}
}
goto out;
}
dprintk("Adding new dev_id(%llx:%llx)\n",
_DEVID_LO(&pdev->dev_id), _DEVID_HI(&pdev->dev_id));
ode = kzalloc(sizeof(*ode), gfp_flags);
if (!ode) {
dprintk("%s: -ENOMEM od=%p\n", __func__, od);
goto out;
}
nfs4_init_deviceid_node(&ode->id_node, server, &pdev->dev_id);
kfree(deviceaddr);
ode->od.od = od;
return &ode->id_node;
out:
kfree(deviceaddr);
return NULL;
}
static void copy_single_comp(struct ore_components *oc, unsigned c,
struct pnfs_osd_object_cred *src_comp)
{
struct ore_comp *ocomp = &oc->comps[c];
WARN_ON(src_comp->oc_cap_key.cred_len > 0); /* libosd is NO_SEC only */
WARN_ON(src_comp->oc_cap.cred_len > sizeof(ocomp->cred));
ocomp->obj.partition = src_comp->oc_object_id.oid_partition_id;
ocomp->obj.id = src_comp->oc_object_id.oid_object_id;
memcpy(ocomp->cred, src_comp->oc_cap.cred, sizeof(ocomp->cred));
}
static int __alloc_objio_seg(unsigned numdevs, gfp_t gfp_flags,
struct objio_segment **pseg)
{
/* This is the in memory structure of the objio_segment
*
* struct __alloc_objio_segment {
* struct objio_segment olseg;
* struct ore_dev *ods[numdevs];
* struct ore_comp comps[numdevs];
* } *aolseg;
* NOTE: The code as above compiles and runs perfectly. It is elegant,
* type safe and compact. At some Past time Linus has decided he does not
* like variable length arrays, For the sake of this principal we uglify
* the code as below.
*/
struct objio_segment *lseg;
size_t lseg_size = sizeof(*lseg) +
numdevs * sizeof(lseg->oc.ods[0]) +
numdevs * sizeof(*lseg->oc.comps);
lseg = kzalloc(lseg_size, gfp_flags);
if (unlikely(!lseg)) {
dprintk("%s: Failed allocation numdevs=%d size=%zd\n", __func__,
numdevs, lseg_size);
return -ENOMEM;
}
lseg->oc.numdevs = numdevs;
lseg->oc.single_comp = EC_MULTPLE_COMPS;
lseg->oc.ods = (void *)(lseg + 1);
lseg->oc.comps = (void *)(lseg->oc.ods + numdevs);
*pseg = lseg;
return 0;
}
int objio_alloc_lseg(struct pnfs_layout_segment **outp,
struct pnfs_layout_hdr *pnfslay,
struct pnfs_layout_range *range,
struct xdr_stream *xdr,
gfp_t gfp_flags)
{
struct nfs_server *server = NFS_SERVER(pnfslay->plh_inode);
struct objio_segment *objio_seg;
struct pnfs_osd_xdr_decode_layout_iter iter;
struct pnfs_osd_layout layout;
struct pnfs_osd_object_cred src_comp;
unsigned cur_comp;
int err;
err = pnfs_osd_xdr_decode_layout_map(&layout, &iter, xdr);
if (unlikely(err))
return err;
err = __alloc_objio_seg(layout.olo_num_comps, gfp_flags, &objio_seg);
if (unlikely(err))
return err;
objio_seg->layout.stripe_unit = layout.olo_map.odm_stripe_unit;
objio_seg->layout.group_width = layout.olo_map.odm_group_width;
objio_seg->layout.group_depth = layout.olo_map.odm_group_depth;
objio_seg->layout.mirrors_p1 = layout.olo_map.odm_mirror_cnt + 1;
objio_seg->layout.raid_algorithm = layout.olo_map.odm_raid_algorithm;
err = ore_verify_layout(layout.olo_map.odm_num_comps,
&objio_seg->layout);
if (unlikely(err))
goto err;
objio_seg->oc.first_dev = layout.olo_comps_index;
cur_comp = 0;
while (pnfs_osd_xdr_decode_layout_comp(&src_comp, &iter, xdr, &err)) {
struct nfs4_deviceid_node *d;
struct objio_dev_ent *ode;
copy_single_comp(&objio_seg->oc, cur_comp, &src_comp);
d = nfs4_find_get_deviceid(server,
&src_comp.oc_object_id.oid_device_id,
pnfslay->plh_lc_cred, gfp_flags);
if (!d) {
err = -ENXIO;
goto err;
}
ode = container_of(d, struct objio_dev_ent, id_node);
objio_seg->oc.ods[cur_comp++] = &ode->od;
}
/* pnfs_osd_xdr_decode_layout_comp returns false on error */
if (unlikely(err))
goto err;
*outp = &objio_seg->lseg;
return 0;
err:
kfree(objio_seg);
dprintk("%s: Error: return %d\n", __func__, err);
*outp = NULL;
return err;
}
void objio_free_lseg(struct pnfs_layout_segment *lseg)
{
int i;
struct objio_segment *objio_seg = OBJIO_LSEG(lseg);
for (i = 0; i < objio_seg->oc.numdevs; i++) {
struct ore_dev *od = objio_seg->oc.ods[i];
struct objio_dev_ent *ode;
if (!od)
break;
ode = container_of(od, typeof(*ode), od);
nfs4_put_deviceid_node(&ode->id_node);
}
kfree(objio_seg);
}
static int
objio_alloc_io_state(struct pnfs_layout_hdr *pnfs_layout_type, bool is_reading,
struct pnfs_layout_segment *lseg, struct page **pages, unsigned pgbase,
loff_t offset, size_t count, void *rpcdata, gfp_t gfp_flags,
struct objio_state **outp)
{
struct objio_segment *objio_seg = OBJIO_LSEG(lseg);
struct ore_io_state *ios;
int ret;
struct __alloc_objio_state {
struct objio_state objios;
struct pnfs_osd_ioerr ioerrs[objio_seg->oc.numdevs];
} *aos;
aos = kzalloc(sizeof(*aos), gfp_flags);
if (unlikely(!aos))
return -ENOMEM;
objlayout_init_ioerrs(&aos->objios.oir, objio_seg->oc.numdevs,
aos->ioerrs, rpcdata, pnfs_layout_type);
ret = ore_get_rw_state(&objio_seg->layout, &objio_seg->oc, is_reading,
offset, count, &ios);
if (unlikely(ret)) {
kfree(aos);
return ret;
}
ios->pages = pages;
ios->pgbase = pgbase;
ios->private = aos;
BUG_ON(ios->nr_pages > (pgbase + count + PAGE_SIZE - 1) >> PAGE_SHIFT);
aos->objios.sync = 0;
aos->objios.ios = ios;
*outp = &aos->objios;
return 0;
}
void objio_free_result(struct objlayout_io_res *oir)
{
struct objio_state *objios = container_of(oir, struct objio_state, oir);
ore_put_io_state(objios->ios);
kfree(objios);
}
static enum pnfs_osd_errno osd_pri_2_pnfs_err(enum osd_err_priority oep)
{
switch (oep) {
case OSD_ERR_PRI_NO_ERROR:
return (enum pnfs_osd_errno)0;
case OSD_ERR_PRI_CLEAR_PAGES:
BUG_ON(1);
return 0;
case OSD_ERR_PRI_RESOURCE:
return PNFS_OSD_ERR_RESOURCE;
case OSD_ERR_PRI_BAD_CRED:
return PNFS_OSD_ERR_BAD_CRED;
case OSD_ERR_PRI_NO_ACCESS:
return PNFS_OSD_ERR_NO_ACCESS;
case OSD_ERR_PRI_UNREACHABLE:
return PNFS_OSD_ERR_UNREACHABLE;
case OSD_ERR_PRI_NOT_FOUND:
return PNFS_OSD_ERR_NOT_FOUND;
case OSD_ERR_PRI_NO_SPACE:
return PNFS_OSD_ERR_NO_SPACE;
default:
WARN_ON(1);
/* fallthrough */
case OSD_ERR_PRI_EIO:
return PNFS_OSD_ERR_EIO;
}
}
static void __on_dev_error(struct ore_io_state *ios,
struct ore_dev *od, unsigned dev_index, enum osd_err_priority oep,
u64 dev_offset, u64 dev_len)
{
struct objio_state *objios = ios->private;
struct pnfs_osd_objid pooid;
struct objio_dev_ent *ode = container_of(od, typeof(*ode), od);
/* FIXME: what to do with more-then-one-group layouts. We need to
* translate from ore_io_state index to oc->comps index
*/
unsigned comp = dev_index;
pooid.oid_device_id = ode->id_node.deviceid;
pooid.oid_partition_id = ios->oc->comps[comp].obj.partition;
pooid.oid_object_id = ios->oc->comps[comp].obj.id;
objlayout_io_set_result(&objios->oir, comp,
&pooid, osd_pri_2_pnfs_err(oep),
dev_offset, dev_len, !ios->reading);
}
/*
* read
*/
static void _read_done(struct ore_io_state *ios, void *private)
{
struct objio_state *objios = private;
ssize_t status;
int ret = ore_check_io(ios, &__on_dev_error);
/* FIXME: _io_free(ios) can we dealocate the libosd resources; */
if (likely(!ret))
status = ios->length;
else
status = ret;
objlayout_read_done(&objios->oir, status, objios->sync);
}
int objio_read_pagelist(struct nfs_pgio_header *hdr)
{
struct objio_state *objios;
int ret;
ret = objio_alloc_io_state(NFS_I(hdr->inode)->layout, true,
hdr->lseg, hdr->args.pages, hdr->args.pgbase,
hdr->args.offset, hdr->args.count, hdr,
GFP_KERNEL, &objios);
if (unlikely(ret))
return ret;
objios->ios->done = _read_done;
dprintk("%s: offset=0x%llx length=0x%x\n", __func__,
hdr->args.offset, hdr->args.count);
ret = ore_read(objios->ios);
if (unlikely(ret))
objio_free_result(&objios->oir);
return ret;
}
/*
* write
*/
static void _write_done(struct ore_io_state *ios, void *private)
{
struct objio_state *objios = private;
ssize_t status;
int ret = ore_check_io(ios, &__on_dev_error);
/* FIXME: _io_free(ios) can we dealocate the libosd resources; */
if (likely(!ret)) {
/* FIXME: should be based on the OSD's persistence model
* See OSD2r05 Section 4.13 Data persistence model */
objios->oir.committed = NFS_FILE_SYNC;
status = ios->length;
} else {
status = ret;
}
objlayout_write_done(&objios->oir, status, objios->sync);
}
static struct page *__r4w_get_page(void *priv, u64 offset, bool *uptodate)
{
struct objio_state *objios = priv;
struct nfs_pgio_header *hdr = objios->oir.rpcdata;
struct address_space *mapping = hdr->inode->i_mapping;
pgoff_t index = offset / PAGE_SIZE;
struct page *page;
loff_t i_size = i_size_read(hdr->inode);
if (offset >= i_size) {
*uptodate = true;
dprintk("%s: g_zero_page index=0x%lx\n", __func__, index);
return ZERO_PAGE(0);
}
page = find_get_page(mapping, index);
if (!page) {
page = find_or_create_page(mapping, index, GFP_NOFS);
if (unlikely(!page)) {
dprintk("%s: grab_cache_page Failed index=0x%lx\n",
__func__, index);
return NULL;
}
unlock_page(page);
}
*uptodate = PageUptodate(page);
dprintk("%s: index=0x%lx uptodate=%d\n", __func__, index, *uptodate);
return page;
}
static void __r4w_put_page(void *priv, struct page *page)
{
dprintk("%s: index=0x%lx\n", __func__,
(page == ZERO_PAGE(0)) ? -1UL : page->index);
if (ZERO_PAGE(0) != page)
put_page(page);
return;
}
static const struct _ore_r4w_op _r4w_op = {
.get_page = &__r4w_get_page,
.put_page = &__r4w_put_page,
};
int objio_write_pagelist(struct nfs_pgio_header *hdr, int how)
{
struct objio_state *objios;
int ret;
ret = objio_alloc_io_state(NFS_I(hdr->inode)->layout, false,
hdr->lseg, hdr->args.pages, hdr->args.pgbase,
hdr->args.offset, hdr->args.count, hdr, GFP_NOFS,
&objios);
if (unlikely(ret))
return ret;
objios->sync = 0 != (how & FLUSH_SYNC);
objios->ios->r4w = &_r4w_op;
if (!objios->sync)
objios->ios->done = _write_done;
dprintk("%s: offset=0x%llx length=0x%x\n", __func__,
hdr->args.offset, hdr->args.count);
ret = ore_write(objios->ios);
if (unlikely(ret)) {
objio_free_result(&objios->oir);
return ret;
}
if (objios->sync)
_write_done(objios->ios, objios);
return 0;
}
/*
* Return 0 if @req cannot be coalesced into @pgio, otherwise return the number
* of bytes (maximum @req->wb_bytes) that can be coalesced.
*/
static size_t objio_pg_test(struct nfs_pageio_descriptor *pgio,
struct nfs_page *prev, struct nfs_page *req)
{
struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(pgio);
unsigned int size;
size = pnfs_generic_pg_test(pgio, prev, req);
if (!size || mirror->pg_count + req->wb_bytes >
(unsigned long)pgio->pg_layout_private)
return 0;
return min(size, req->wb_bytes);
}
static void objio_init_read(struct nfs_pageio_descriptor *pgio, struct nfs_page *req)
{
pnfs_generic_pg_init_read(pgio, req);
if (unlikely(pgio->pg_lseg == NULL))
return; /* Not pNFS */
pgio->pg_layout_private = (void *)
OBJIO_LSEG(pgio->pg_lseg)->layout.max_io_length;
}
static bool aligned_on_raid_stripe(u64 offset, struct ore_layout *layout,
unsigned long *stripe_end)
{
u32 stripe_off;
unsigned stripe_size;
if (layout->raid_algorithm == PNFS_OSD_RAID_0)
return true;
stripe_size = layout->stripe_unit *
(layout->group_width - layout->parity);
div_u64_rem(offset, stripe_size, &stripe_off);
if (!stripe_off)
return true;
*stripe_end = stripe_size - stripe_off;
return false;
}
static void objio_init_write(struct nfs_pageio_descriptor *pgio, struct nfs_page *req)
{
unsigned long stripe_end = 0;
u64 wb_size;
if (pgio->pg_dreq == NULL)
wb_size = i_size_read(pgio->pg_inode) - req_offset(req);
else
wb_size = nfs_dreq_bytes_left(pgio->pg_dreq);
pnfs_generic_pg_init_write(pgio, req, wb_size);
if (unlikely(pgio->pg_lseg == NULL))
return; /* Not pNFS */
if (req->wb_offset ||
!aligned_on_raid_stripe(req->wb_index * PAGE_SIZE,
&OBJIO_LSEG(pgio->pg_lseg)->layout,
&stripe_end)) {
pgio->pg_layout_private = (void *)stripe_end;
} else {
pgio->pg_layout_private = (void *)
OBJIO_LSEG(pgio->pg_lseg)->layout.max_io_length;
}
}
static const struct nfs_pageio_ops objio_pg_read_ops = {
.pg_init = objio_init_read,
.pg_test = objio_pg_test,
.pg_doio = pnfs_generic_pg_readpages,
.pg_cleanup = pnfs_generic_pg_cleanup,
};
static const struct nfs_pageio_ops objio_pg_write_ops = {
.pg_init = objio_init_write,
.pg_test = objio_pg_test,
.pg_doio = pnfs_generic_pg_writepages,
.pg_cleanup = pnfs_generic_pg_cleanup,
};
static struct pnfs_layoutdriver_type objlayout_type = {
.id = LAYOUT_OSD2_OBJECTS,
.name = "LAYOUT_OSD2_OBJECTS",
.flags = PNFS_LAYOUTRET_ON_SETATTR |
PNFS_LAYOUTRET_ON_ERROR,
.max_deviceinfo_size = PAGE_SIZE,
.owner = THIS_MODULE,
.alloc_layout_hdr = objlayout_alloc_layout_hdr,
.free_layout_hdr = objlayout_free_layout_hdr,
.alloc_lseg = objlayout_alloc_lseg,
.free_lseg = objlayout_free_lseg,
.read_pagelist = objlayout_read_pagelist,
.write_pagelist = objlayout_write_pagelist,
.pg_read_ops = &objio_pg_read_ops,
.pg_write_ops = &objio_pg_write_ops,
.sync = pnfs_generic_sync,
.free_deviceid_node = objio_free_deviceid_node,
.encode_layoutcommit = objlayout_encode_layoutcommit,
.encode_layoutreturn = objlayout_encode_layoutreturn,
};
MODULE_DESCRIPTION("pNFS Layout Driver for OSD2 objects");
MODULE_AUTHOR("Benny Halevy <bhalevy@panasas.com>");
MODULE_LICENSE("GPL");
static int __init
objlayout_init(void)
{
int ret = pnfs_register_layoutdriver(&objlayout_type);
if (ret)
printk(KERN_INFO
"NFS: %s: Registering OSD pNFS Layout Driver failed: error=%d\n",
__func__, ret);
else
printk(KERN_INFO "NFS: %s: Registered OSD pNFS Layout Driver\n",
__func__);
return ret;
}
static void __exit
objlayout_exit(void)
{
pnfs_unregister_layoutdriver(&objlayout_type);
printk(KERN_INFO "NFS: %s: Unregistered OSD pNFS Layout Driver\n",
__func__);
}
MODULE_ALIAS("nfs-layouttype4-2");
module_init(objlayout_init);
module_exit(objlayout_exit);
/*
* pNFS Objects layout driver high level definitions
*
* Copyright (C) 2007 Panasas Inc. [year of first publication]
* All rights reserved.
*
* Benny Halevy <bhalevy@panasas.com>
* Boaz Harrosh <ooo@electrozaur.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2
* See the file COPYING included with this distribution for more details.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the Panasas company nor the names of its
* contributors may be used to endorse or promote products derived
* from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
* DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
* BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
* LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
* NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
* SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include <linux/kmod.h>
#include <linux/moduleparam.h>
#include <linux/ratelimit.h>
#include <scsi/osd_initiator.h>
#include "objlayout.h"
#define NFSDBG_FACILITY NFSDBG_PNFS_LD
/*
* Create a objlayout layout structure for the given inode and return it.
*/
struct pnfs_layout_hdr *
objlayout_alloc_layout_hdr(struct inode *inode, gfp_t gfp_flags)
{
struct objlayout *objlay;
objlay = kzalloc(sizeof(struct objlayout), gfp_flags);
if (!objlay)
return NULL;
spin_lock_init(&objlay->lock);
INIT_LIST_HEAD(&objlay->err_list);
dprintk("%s: Return %p\n", __func__, objlay);
return &objlay->pnfs_layout;
}
/*
* Free an objlayout layout structure
*/
void
objlayout_free_layout_hdr(struct pnfs_layout_hdr *lo)
{
struct objlayout *objlay = OBJLAYOUT(lo);
dprintk("%s: objlay %p\n", __func__, objlay);
WARN_ON(!list_empty(&objlay->err_list));
kfree(objlay);
}
/*
* Unmarshall layout and store it in pnfslay.
*/
struct pnfs_layout_segment *
objlayout_alloc_lseg(struct pnfs_layout_hdr *pnfslay,
struct nfs4_layoutget_res *lgr,
gfp_t gfp_flags)
{
int status = -ENOMEM;
struct xdr_stream stream;
struct xdr_buf buf = {
.pages = lgr->layoutp->pages,
.page_len = lgr->layoutp->len,
.buflen = lgr->layoutp->len,
.len = lgr->layoutp->len,
};
struct page *scratch;
struct pnfs_layout_segment *lseg;
dprintk("%s: Begin pnfslay %p\n", __func__, pnfslay);
scratch = alloc_page(gfp_flags);
if (!scratch)
goto err_nofree;
xdr_init_decode(&stream, &buf, NULL);
xdr_set_scratch_buffer(&stream, page_address(scratch), PAGE_SIZE);
status = objio_alloc_lseg(&lseg, pnfslay, &lgr->range, &stream, gfp_flags);
if (unlikely(status)) {
dprintk("%s: objio_alloc_lseg Return err %d\n", __func__,
status);
goto err;
}
__free_page(scratch);
dprintk("%s: Return %p\n", __func__, lseg);
return lseg;
err:
__free_page(scratch);
err_nofree:
dprintk("%s: Err Return=>%d\n", __func__, status);
return ERR_PTR(status);
}
/*
* Free a layout segement
*/
void
objlayout_free_lseg(struct pnfs_layout_segment *lseg)
{
dprintk("%s: freeing layout segment %p\n", __func__, lseg);
if (unlikely(!lseg))
return;
objio_free_lseg(lseg);
}
/*
* I/O Operations
*/
static inline u64
end_offset(u64 start, u64 len)
{
u64 end;
end = start + len;
return end >= start ? end : NFS4_MAX_UINT64;
}
static void _fix_verify_io_params(struct pnfs_layout_segment *lseg,
struct page ***p_pages, unsigned *p_pgbase,
u64 offset, unsigned long count)
{
u64 lseg_end_offset;
BUG_ON(offset < lseg->pls_range.offset);
lseg_end_offset = end_offset(lseg->pls_range.offset,
lseg->pls_range.length);
BUG_ON(offset >= lseg_end_offset);
WARN_ON(offset + count > lseg_end_offset);
if (*p_pgbase > PAGE_SIZE) {
dprintk("%s: pgbase(0x%x) > PAGE_SIZE\n", __func__, *p_pgbase);
*p_pages += *p_pgbase >> PAGE_SHIFT;
*p_pgbase &= ~PAGE_MASK;
}
}
/*
* I/O done common code
*/
static void
objlayout_iodone(struct objlayout_io_res *oir)
{
if (likely(oir->status >= 0)) {
objio_free_result(oir);
} else {
struct objlayout *objlay = oir->objlay;
spin_lock(&objlay->lock);
objlay->delta_space_valid = OBJ_DSU_INVALID;
list_add(&objlay->err_list, &oir->err_list);
spin_unlock(&objlay->lock);
}
}
/*
* objlayout_io_set_result - Set an osd_error code on a specific osd comp.
*
* The @index component IO failed (error returned from target). Register
* the error for later reporting at layout-return.
*/
void
objlayout_io_set_result(struct objlayout_io_res *oir, unsigned index,
struct pnfs_osd_objid *pooid, int osd_error,
u64 offset, u64 length, bool is_write)
{
struct pnfs_osd_ioerr *ioerr = &oir->ioerrs[index];
BUG_ON(index >= oir->num_comps);
if (osd_error) {
ioerr->oer_component = *pooid;
ioerr->oer_comp_offset = offset;
ioerr->oer_comp_length = length;
ioerr->oer_iswrite = is_write;
ioerr->oer_errno = osd_error;
dprintk("%s: err[%d]: errno=%d is_write=%d dev(%llx:%llx) "
"par=0x%llx obj=0x%llx offset=0x%llx length=0x%llx\n",
__func__, index, ioerr->oer_errno,
ioerr->oer_iswrite,
_DEVID_LO(&ioerr->oer_component.oid_device_id),
_DEVID_HI(&ioerr->oer_component.oid_device_id),
ioerr->oer_component.oid_partition_id,
ioerr->oer_component.oid_object_id,
ioerr->oer_comp_offset,
ioerr->oer_comp_length);
} else {
/* User need not call if no error is reported */
ioerr->oer_errno = 0;
}
}
/* Function scheduled on rpc workqueue to call ->nfs_readlist_complete().
* This is because the osd completion is called with ints-off from
* the block layer
*/
static void _rpc_read_complete(struct work_struct *work)
{
struct rpc_task *task;
struct nfs_pgio_header *hdr;
dprintk("%s enter\n", __func__);
task = container_of(work, struct rpc_task, u.tk_work);
hdr = container_of(task, struct nfs_pgio_header, task);
pnfs_ld_read_done(hdr);
}
void
objlayout_read_done(struct objlayout_io_res *oir, ssize_t status, bool sync)
{
struct nfs_pgio_header *hdr = oir->rpcdata;
oir->status = hdr->task.tk_status = status;
if (status >= 0)
hdr->res.count = status;
else
hdr->pnfs_error = status;
objlayout_iodone(oir);
/* must not use oir after this point */
dprintk("%s: Return status=%zd eof=%d sync=%d\n", __func__,
status, hdr->res.eof, sync);
if (sync)
pnfs_ld_read_done(hdr);
else {
INIT_WORK(&hdr->task.u.tk_work, _rpc_read_complete);
schedule_work(&hdr->task.u.tk_work);
}
}
/*
* Perform sync or async reads.
*/
enum pnfs_try_status
objlayout_read_pagelist(struct nfs_pgio_header *hdr)
{
struct inode *inode = hdr->inode;
loff_t offset = hdr->args.offset;
size_t count = hdr->args.count;
int err;
loff_t eof;
eof = i_size_read(inode);
if (unlikely(offset + count > eof)) {
if (offset >= eof) {
err = 0;
hdr->res.count = 0;
hdr->res.eof = 1;
/*FIXME: do we need to call pnfs_ld_read_done() */
goto out;
}
count = eof - offset;
}
hdr->res.eof = (offset + count) >= eof;
_fix_verify_io_params(hdr->lseg, &hdr->args.pages,
&hdr->args.pgbase,
hdr->args.offset, hdr->args.count);
dprintk("%s: inode(%lx) offset 0x%llx count 0x%zx eof=%d\n",
__func__, inode->i_ino, offset, count, hdr->res.eof);
err = objio_read_pagelist(hdr);
out:
if (unlikely(err)) {
hdr->pnfs_error = err;
dprintk("%s: Returned Error %d\n", __func__, err);
return PNFS_NOT_ATTEMPTED;
}
return PNFS_ATTEMPTED;
}
/* Function scheduled on rpc workqueue to call ->nfs_writelist_complete().
* This is because the osd completion is called with ints-off from
* the block layer
*/
static void _rpc_write_complete(struct work_struct *work)
{
struct rpc_task *task;
struct nfs_pgio_header *hdr;
dprintk("%s enter\n", __func__);
task = container_of(work, struct rpc_task, u.tk_work);
hdr = container_of(task, struct nfs_pgio_header, task);
pnfs_ld_write_done(hdr);
}
void
objlayout_write_done(struct objlayout_io_res *oir, ssize_t status, bool sync)
{
struct nfs_pgio_header *hdr = oir->rpcdata;
oir->status = hdr->task.tk_status = status;
if (status >= 0) {
hdr->res.count = status;
hdr->verf.committed = oir->committed;
} else {
hdr->pnfs_error = status;
}
objlayout_iodone(oir);
/* must not use oir after this point */
dprintk("%s: Return status %zd committed %d sync=%d\n", __func__,
status, hdr->verf.committed, sync);
if (sync)
pnfs_ld_write_done(hdr);
else {
INIT_WORK(&hdr->task.u.tk_work, _rpc_write_complete);
schedule_work(&hdr->task.u.tk_work);
}
}
/*
* Perform sync or async writes.
*/
enum pnfs_try_status
objlayout_write_pagelist(struct nfs_pgio_header *hdr, int how)
{
int err;
_fix_verify_io_params(hdr->lseg, &hdr->args.pages,
&hdr->args.pgbase,
hdr->args.offset, hdr->args.count);
err = objio_write_pagelist(hdr, how);
if (unlikely(err)) {
hdr->pnfs_error = err;
dprintk("%s: Returned Error %d\n", __func__, err);
return PNFS_NOT_ATTEMPTED;
}
return PNFS_ATTEMPTED;
}
void
objlayout_encode_layoutcommit(struct pnfs_layout_hdr *pnfslay,
struct xdr_stream *xdr,
const struct nfs4_layoutcommit_args *args)
{
struct objlayout *objlay = OBJLAYOUT(pnfslay);
struct pnfs_osd_layoutupdate lou;
__be32 *start;
dprintk("%s: Begin\n", __func__);
spin_lock(&objlay->lock);
lou.dsu_valid = (objlay->delta_space_valid == OBJ_DSU_VALID);
lou.dsu_delta = objlay->delta_space_used;
objlay->delta_space_used = 0;
objlay->delta_space_valid = OBJ_DSU_INIT;
lou.olu_ioerr_flag = !list_empty(&objlay->err_list);
spin_unlock(&objlay->lock);
start = xdr_reserve_space(xdr, 4);
BUG_ON(pnfs_osd_xdr_encode_layoutupdate(xdr, &lou));
*start = cpu_to_be32((xdr->p - start - 1) * 4);
dprintk("%s: Return delta_space_used %lld err %d\n", __func__,
lou.dsu_delta, lou.olu_ioerr_flag);
}
static int
err_prio(u32 oer_errno)
{
switch (oer_errno) {
case 0:
return 0;
case PNFS_OSD_ERR_RESOURCE:
return OSD_ERR_PRI_RESOURCE;
case PNFS_OSD_ERR_BAD_CRED:
return OSD_ERR_PRI_BAD_CRED;
case PNFS_OSD_ERR_NO_ACCESS:
return OSD_ERR_PRI_NO_ACCESS;
case PNFS_OSD_ERR_UNREACHABLE:
return OSD_ERR_PRI_UNREACHABLE;
case PNFS_OSD_ERR_NOT_FOUND:
return OSD_ERR_PRI_NOT_FOUND;
case PNFS_OSD_ERR_NO_SPACE:
return OSD_ERR_PRI_NO_SPACE;
default:
WARN_ON(1);
/* fallthrough */
case PNFS_OSD_ERR_EIO:
return OSD_ERR_PRI_EIO;
}
}
static void
merge_ioerr(struct pnfs_osd_ioerr *dest_err,
const struct pnfs_osd_ioerr *src_err)
{
u64 dest_end, src_end;
if (!dest_err->oer_errno) {
*dest_err = *src_err;
/* accumulated device must be blank */
memset(&dest_err->oer_component.oid_device_id, 0,
sizeof(dest_err->oer_component.oid_device_id));
return;
}
if (dest_err->oer_component.oid_partition_id !=
src_err->oer_component.oid_partition_id)
dest_err->oer_component.oid_partition_id = 0;
if (dest_err->oer_component.oid_object_id !=
src_err->oer_component.oid_object_id)
dest_err->oer_component.oid_object_id = 0;
if (dest_err->oer_comp_offset > src_err->oer_comp_offset)
dest_err->oer_comp_offset = src_err->oer_comp_offset;
dest_end = end_offset(dest_err->oer_comp_offset,
dest_err->oer_comp_length);
src_end = end_offset(src_err->oer_comp_offset,
src_err->oer_comp_length);
if (dest_end < src_end)
dest_end = src_end;
dest_err->oer_comp_length = dest_end - dest_err->oer_comp_offset;
if ((src_err->oer_iswrite == dest_err->oer_iswrite) &&
(err_prio(src_err->oer_errno) > err_prio(dest_err->oer_errno))) {
dest_err->oer_errno = src_err->oer_errno;
} else if (src_err->oer_iswrite) {
dest_err->oer_iswrite = true;
dest_err->oer_errno = src_err->oer_errno;
}
}
static void
encode_accumulated_error(struct objlayout *objlay, __be32 *p)
{
struct objlayout_io_res *oir, *tmp;
struct pnfs_osd_ioerr accumulated_err = {.oer_errno = 0};
list_for_each_entry_safe(oir, tmp, &objlay->err_list, err_list) {
unsigned i;
for (i = 0; i < oir->num_comps; i++) {
struct pnfs_osd_ioerr *ioerr = &oir->ioerrs[i];
if (!ioerr->oer_errno)
continue;
printk(KERN_ERR "NFS: %s: err[%d]: errno=%d "
"is_write=%d dev(%llx:%llx) par=0x%llx "
"obj=0x%llx offset=0x%llx length=0x%llx\n",
__func__, i, ioerr->oer_errno,
ioerr->oer_iswrite,
_DEVID_LO(&ioerr->oer_component.oid_device_id),
_DEVID_HI(&ioerr->oer_component.oid_device_id),
ioerr->oer_component.oid_partition_id,
ioerr->oer_component.oid_object_id,
ioerr->oer_comp_offset,
ioerr->oer_comp_length);
merge_ioerr(&accumulated_err, ioerr);
}
list_del(&oir->err_list);
objio_free_result(oir);
}
pnfs_osd_xdr_encode_ioerr(p, &accumulated_err);
}
void
objlayout_encode_layoutreturn(struct xdr_stream *xdr,
const struct nfs4_layoutreturn_args *args)
{
struct pnfs_layout_hdr *pnfslay = args->layout;
struct objlayout *objlay = OBJLAYOUT(pnfslay);
struct objlayout_io_res *oir, *tmp;
__be32 *start;
dprintk("%s: Begin\n", __func__);
start = xdr_reserve_space(xdr, 4);
BUG_ON(!start);
spin_lock(&objlay->lock);
list_for_each_entry_safe(oir, tmp, &objlay->err_list, err_list) {
__be32 *last_xdr = NULL, *p;
unsigned i;
int res = 0;
for (i = 0; i < oir->num_comps; i++) {
struct pnfs_osd_ioerr *ioerr = &oir->ioerrs[i];
if (!ioerr->oer_errno)
continue;
dprintk("%s: err[%d]: errno=%d is_write=%d "
"dev(%llx:%llx) par=0x%llx obj=0x%llx "
"offset=0x%llx length=0x%llx\n",
__func__, i, ioerr->oer_errno,
ioerr->oer_iswrite,
_DEVID_LO(&ioerr->oer_component.oid_device_id),
_DEVID_HI(&ioerr->oer_component.oid_device_id),
ioerr->oer_component.oid_partition_id,
ioerr->oer_component.oid_object_id,
ioerr->oer_comp_offset,
ioerr->oer_comp_length);
p = pnfs_osd_xdr_ioerr_reserve_space(xdr);
if (unlikely(!p)) {
res = -E2BIG;
break; /* accumulated_error */
}
last_xdr = p;
pnfs_osd_xdr_encode_ioerr(p, &oir->ioerrs[i]);
}
/* TODO: use xdr_write_pages */
if (unlikely(res)) {
/* no space for even one error descriptor */
BUG_ON(!last_xdr);
/* we've encountered a situation with lots and lots of
* errors and no space to encode them all. Use the last
* available slot to report the union of all the
* remaining errors.
*/
encode_accumulated_error(objlay, last_xdr);
goto loop_done;
}
list_del(&oir->err_list);
objio_free_result(oir);
}
loop_done:
spin_unlock(&objlay->lock);
*start = cpu_to_be32((xdr->p - start - 1) * 4);
dprintk("%s: Return\n", __func__);
}
enum {
OBJLAYOUT_MAX_URI_LEN = 256, OBJLAYOUT_MAX_OSDNAME_LEN = 64,
OBJLAYOUT_MAX_SYSID_HEX_LEN = OSD_SYSTEMID_LEN * 2 + 1,
OSD_LOGIN_UPCALL_PATHLEN = 256
};
static char osd_login_prog[OSD_LOGIN_UPCALL_PATHLEN] = "/sbin/osd_login";
module_param_string(osd_login_prog, osd_login_prog, sizeof(osd_login_prog),
0600);
MODULE_PARM_DESC(osd_login_prog, "Path to the osd_login upcall program");
struct __auto_login {
char uri[OBJLAYOUT_MAX_URI_LEN];
char osdname[OBJLAYOUT_MAX_OSDNAME_LEN];
char systemid_hex[OBJLAYOUT_MAX_SYSID_HEX_LEN];
};
static int __objlayout_upcall(struct __auto_login *login)
{
static char *envp[] = { "HOME=/",
"TERM=linux",
"PATH=/sbin:/usr/sbin:/bin:/usr/bin",
NULL
};
char *argv[8];
int ret;
if (unlikely(!osd_login_prog[0])) {
dprintk("%s: osd_login_prog is disabled\n", __func__);
return -EACCES;
}
dprintk("%s uri: %s\n", __func__, login->uri);
dprintk("%s osdname %s\n", __func__, login->osdname);
dprintk("%s systemid_hex %s\n", __func__, login->systemid_hex);
argv[0] = (char *)osd_login_prog;
argv[1] = "-u";
argv[2] = login->uri;
argv[3] = "-o";
argv[4] = login->osdname;
argv[5] = "-s";
argv[6] = login->systemid_hex;
argv[7] = NULL;
ret = call_usermodehelper(argv[0], argv, envp, UMH_WAIT_PROC);
/*
* Disable the upcall mechanism if we're getting an ENOENT or
* EACCES error. The admin can re-enable it on the fly by using
* sysfs to set the objlayoutdriver.osd_login_prog module parameter once
* the problem has been fixed.
*/
if (ret == -ENOENT || ret == -EACCES) {
printk(KERN_ERR "PNFS-OBJ: %s was not found please set "
"objlayoutdriver.osd_login_prog kernel parameter!\n",
osd_login_prog);
osd_login_prog[0] = '\0';
}
dprintk("%s %s return value: %d\n", __func__, osd_login_prog, ret);
return ret;
}
/* Assume dest is all zeros */
static void __copy_nfsS_and_zero_terminate(struct nfs4_string s,
char *dest, int max_len,
const char *var_name)
{
if (!s.len)
return;
if (s.len >= max_len) {
pr_warn_ratelimited(
"objlayout_autologin: %s: s.len(%d) >= max_len(%d)",
var_name, s.len, max_len);
s.len = max_len - 1; /* space for null terminator */
}
memcpy(dest, s.data, s.len);
}
/* Assume sysid is all zeros */
static void _sysid_2_hex(struct nfs4_string s,
char sysid[OBJLAYOUT_MAX_SYSID_HEX_LEN])
{
int i;
char *cur;
if (!s.len)
return;
if (s.len != OSD_SYSTEMID_LEN) {
pr_warn_ratelimited(
"objlayout_autologin: systemid_len(%d) != OSD_SYSTEMID_LEN",
s.len);
if (s.len > OSD_SYSTEMID_LEN)
s.len = OSD_SYSTEMID_LEN;
}
cur = sysid;
for (i = 0; i < s.len; i++)
cur = hex_byte_pack(cur, s.data[i]);
}
int objlayout_autologin(struct pnfs_osd_deviceaddr *deviceaddr)
{
int rc;
struct __auto_login login;
if (!deviceaddr->oda_targetaddr.ota_netaddr.r_addr.len)
return -ENODEV;
memset(&login, 0, sizeof(login));
__copy_nfsS_and_zero_terminate(
deviceaddr->oda_targetaddr.ota_netaddr.r_addr,
login.uri, sizeof(login.uri), "URI");
__copy_nfsS_and_zero_terminate(
deviceaddr->oda_osdname,
login.osdname, sizeof(login.osdname), "OSDNAME");
_sysid_2_hex(deviceaddr->oda_systemid, login.systemid_hex);
rc = __objlayout_upcall(&login);
if (rc > 0) /* script returns positive values */
rc = -ENODEV;
return rc;
}
/*
* Data types and function declerations for interfacing with the
* pNFS standard object layout driver.
*
* Copyright (C) 2007 Panasas Inc. [year of first publication]
* All rights reserved.
*
* Benny Halevy <bhalevy@panasas.com>
* Boaz Harrosh <ooo@electrozaur.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2
* See the file COPYING included with this distribution for more details.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the Panasas company nor the names of its
* contributors may be used to endorse or promote products derived
* from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
* DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
* BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
* LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
* NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
* SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#ifndef _OBJLAYOUT_H
#define _OBJLAYOUT_H
#include <linux/nfs_fs.h>
#include <linux/pnfs_osd_xdr.h>
#include "../pnfs.h"
/*
* per-inode layout
*/
struct objlayout {
struct pnfs_layout_hdr pnfs_layout;
/* for layout_commit */
enum osd_delta_space_valid_enum {
OBJ_DSU_INIT = 0,
OBJ_DSU_VALID,
OBJ_DSU_INVALID,
} delta_space_valid;
s64 delta_space_used; /* consumed by write ops */
/* for layout_return */
spinlock_t lock;
struct list_head err_list;
};
static inline struct objlayout *
OBJLAYOUT(struct pnfs_layout_hdr *lo)
{
return container_of(lo, struct objlayout, pnfs_layout);
}
/*
* per-I/O operation state
* embedded in objects provider io_state data structure
*/
struct objlayout_io_res {
struct objlayout *objlay;
void *rpcdata;
int status; /* res */
int committed; /* res */
/* Error reporting (layout_return) */
struct list_head err_list;
unsigned num_comps;
/* Pointer to array of error descriptors of size num_comps.
* It should contain as many entries as devices in the osd_layout
* that participate in the I/O. It is up to the io_engine to allocate
* needed space and set num_comps.
*/
struct pnfs_osd_ioerr *ioerrs;
};
static inline
void objlayout_init_ioerrs(struct objlayout_io_res *oir, unsigned num_comps,
struct pnfs_osd_ioerr *ioerrs, void *rpcdata,
struct pnfs_layout_hdr *pnfs_layout_type)
{
oir->objlay = OBJLAYOUT(pnfs_layout_type);
oir->rpcdata = rpcdata;
INIT_LIST_HEAD(&oir->err_list);
oir->num_comps = num_comps;
oir->ioerrs = ioerrs;
}
/*
* Raid engine I/O API
*/
extern int objio_alloc_lseg(struct pnfs_layout_segment **outp,
struct pnfs_layout_hdr *pnfslay,
struct pnfs_layout_range *range,
struct xdr_stream *xdr,
gfp_t gfp_flags);
extern void objio_free_lseg(struct pnfs_layout_segment *lseg);
/* objio_free_result will free these @oir structs received from
* objlayout_{read,write}_done
*/
extern void objio_free_result(struct objlayout_io_res *oir);
extern int objio_read_pagelist(struct nfs_pgio_header *rdata);
extern int objio_write_pagelist(struct nfs_pgio_header *wdata, int how);
/*
* callback API
*/
extern void objlayout_io_set_result(struct objlayout_io_res *oir,
unsigned index, struct pnfs_osd_objid *pooid,
int osd_error, u64 offset, u64 length, bool is_write);
static inline void
objlayout_add_delta_space_used(struct objlayout *objlay, s64 space_used)
{
/* If one of the I/Os errored out and the delta_space_used was
* invalid we render the complete report as invalid. Protocol mandate
* the DSU be accurate or not reported.
*/
spin_lock(&objlay->lock);
if (objlay->delta_space_valid != OBJ_DSU_INVALID) {
objlay->delta_space_valid = OBJ_DSU_VALID;
objlay->delta_space_used += space_used;
}
spin_unlock(&objlay->lock);
}
extern void objlayout_read_done(struct objlayout_io_res *oir,
ssize_t status, bool sync);
extern void objlayout_write_done(struct objlayout_io_res *oir,
ssize_t status, bool sync);
/*
* exported generic objects function vectors
*/
extern struct pnfs_layout_hdr *objlayout_alloc_layout_hdr(struct inode *, gfp_t gfp_flags);
extern void objlayout_free_layout_hdr(struct pnfs_layout_hdr *);
extern struct pnfs_layout_segment *objlayout_alloc_lseg(
struct pnfs_layout_hdr *,
struct nfs4_layoutget_res *,
gfp_t gfp_flags);
extern void objlayout_free_lseg(struct pnfs_layout_segment *);
extern enum pnfs_try_status objlayout_read_pagelist(
struct nfs_pgio_header *);
extern enum pnfs_try_status objlayout_write_pagelist(
struct nfs_pgio_header *,
int how);
extern void objlayout_encode_layoutcommit(
struct pnfs_layout_hdr *,
struct xdr_stream *,
const struct nfs4_layoutcommit_args *);
extern void objlayout_encode_layoutreturn(
struct xdr_stream *,
const struct nfs4_layoutreturn_args *);
extern int objlayout_autologin(struct pnfs_osd_deviceaddr *deviceaddr);
#endif /* _OBJLAYOUT_H */
/*
* Object-Based pNFS Layout XDR layer
*
* Copyright (C) 2007 Panasas Inc. [year of first publication]
* All rights reserved.
*
* Benny Halevy <bhalevy@panasas.com>
* Boaz Harrosh <ooo@electrozaur.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2
* See the file COPYING included with this distribution for more details.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the Panasas company nor the names of its
* contributors may be used to endorse or promote products derived
* from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
* DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
* BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
* LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
* NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
* SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include <linux/pnfs_osd_xdr.h>
#define NFSDBG_FACILITY NFSDBG_PNFS_LD
/*
* The following implementation is based on RFC5664
*/
/*
* struct pnfs_osd_objid {
* struct nfs4_deviceid oid_device_id;
* u64 oid_partition_id;
* u64 oid_object_id;
* }; // xdr size 32 bytes
*/
static __be32 *
_osd_xdr_decode_objid(__be32 *p, struct pnfs_osd_objid *objid)
{
p = xdr_decode_opaque_fixed(p, objid->oid_device_id.data,
sizeof(objid->oid_device_id.data));
p = xdr_decode_hyper(p, &objid->oid_partition_id);
p = xdr_decode_hyper(p, &objid->oid_object_id);
return p;
}
/*
* struct pnfs_osd_opaque_cred {
* u32 cred_len;
* void *cred;
* }; // xdr size [variable]
* The return pointers are from the xdr buffer
*/
static int
_osd_xdr_decode_opaque_cred(struct pnfs_osd_opaque_cred *opaque_cred,
struct xdr_stream *xdr)
{
__be32 *p = xdr_inline_decode(xdr, 1);
if (!p)
return -EINVAL;
opaque_cred->cred_len = be32_to_cpu(*p++);
p = xdr_inline_decode(xdr, opaque_cred->cred_len);
if (!p)
return -EINVAL;
opaque_cred->cred = p;
return 0;
}
/*
* struct pnfs_osd_object_cred {
* struct pnfs_osd_objid oc_object_id;
* u32 oc_osd_version;
* u32 oc_cap_key_sec;
* struct pnfs_osd_opaque_cred oc_cap_key
* struct pnfs_osd_opaque_cred oc_cap;
* }; // xdr size 32 + 4 + 4 + [variable] + [variable]
*/
static int
_osd_xdr_decode_object_cred(struct pnfs_osd_object_cred *comp,
struct xdr_stream *xdr)
{
__be32 *p = xdr_inline_decode(xdr, 32 + 4 + 4);
int ret;
if (!p)
return -EIO;
p = _osd_xdr_decode_objid(p, &comp->oc_object_id);
comp->oc_osd_version = be32_to_cpup(p++);
comp->oc_cap_key_sec = be32_to_cpup(p);
ret = _osd_xdr_decode_opaque_cred(&comp->oc_cap_key, xdr);
if (unlikely(ret))
return ret;
ret = _osd_xdr_decode_opaque_cred(&comp->oc_cap, xdr);
return ret;
}
/*
* struct pnfs_osd_data_map {
* u32 odm_num_comps;
* u64 odm_stripe_unit;
* u32 odm_group_width;
* u32 odm_group_depth;
* u32 odm_mirror_cnt;
* u32 odm_raid_algorithm;
* }; // xdr size 4 + 8 + 4 + 4 + 4 + 4
*/
static inline int
_osd_data_map_xdr_sz(void)
{
return 4 + 8 + 4 + 4 + 4 + 4;
}
static __be32 *
_osd_xdr_decode_data_map(__be32 *p, struct pnfs_osd_data_map *data_map)
{
data_map->odm_num_comps = be32_to_cpup(p++);
p = xdr_decode_hyper(p, &data_map->odm_stripe_unit);
data_map->odm_group_width = be32_to_cpup(p++);
data_map->odm_group_depth = be32_to_cpup(p++);
data_map->odm_mirror_cnt = be32_to_cpup(p++);
data_map->odm_raid_algorithm = be32_to_cpup(p++);
dprintk("%s: odm_num_comps=%u odm_stripe_unit=%llu odm_group_width=%u "
"odm_group_depth=%u odm_mirror_cnt=%u odm_raid_algorithm=%u\n",
__func__,
data_map->odm_num_comps,
(unsigned long long)data_map->odm_stripe_unit,
data_map->odm_group_width,
data_map->odm_group_depth,
data_map->odm_mirror_cnt,
data_map->odm_raid_algorithm);
return p;
}
int pnfs_osd_xdr_decode_layout_map(struct pnfs_osd_layout *layout,
struct pnfs_osd_xdr_decode_layout_iter *iter, struct xdr_stream *xdr)
{
__be32 *p;
memset(iter, 0, sizeof(*iter));
p = xdr_inline_decode(xdr, _osd_data_map_xdr_sz() + 4 + 4);
if (unlikely(!p))
return -EINVAL;
p = _osd_xdr_decode_data_map(p, &layout->olo_map);
layout->olo_comps_index = be32_to_cpup(p++);
layout->olo_num_comps = be32_to_cpup(p++);
dprintk("%s: olo_comps_index=%d olo_num_comps=%d\n", __func__,
layout->olo_comps_index, layout->olo_num_comps);
iter->total_comps = layout->olo_num_comps;
return 0;
}
bool pnfs_osd_xdr_decode_layout_comp(struct pnfs_osd_object_cred *comp,
struct pnfs_osd_xdr_decode_layout_iter *iter, struct xdr_stream *xdr,
int *err)
{
BUG_ON(iter->decoded_comps > iter->total_comps);
if (iter->decoded_comps == iter->total_comps)
return false;
*err = _osd_xdr_decode_object_cred(comp, xdr);
if (unlikely(*err)) {
dprintk("%s: _osd_xdr_decode_object_cred=>%d decoded_comps=%d "
"total_comps=%d\n", __func__, *err,
iter->decoded_comps, iter->total_comps);
return false; /* stop the loop */
}
dprintk("%s: dev(%llx:%llx) par=0x%llx obj=0x%llx "
"key_len=%u cap_len=%u\n",
__func__,
_DEVID_LO(&comp->oc_object_id.oid_device_id),
_DEVID_HI(&comp->oc_object_id.oid_device_id),
comp->oc_object_id.oid_partition_id,
comp->oc_object_id.oid_object_id,
comp->oc_cap_key.cred_len, comp->oc_cap.cred_len);
iter->decoded_comps++;
return true;
}
/*
* Get Device Information Decoding
*
* Note: since Device Information is currently done synchronously, all
* variable strings fields are left inside the rpc buffer and are only
* pointed to by the pnfs_osd_deviceaddr members. So the read buffer
* should not be freed while the returned information is in use.
*/
/*
*struct nfs4_string {
* unsigned int len;
* char *data;
*}; // size [variable]
* NOTE: Returned string points to inside the XDR buffer
*/
static __be32 *
__read_u8_opaque(__be32 *p, struct nfs4_string *str)
{
str->len = be32_to_cpup(p++);
str->data = (char *)p;
p += XDR_QUADLEN(str->len);
return p;
}
/*
* struct pnfs_osd_targetid {
* u32 oti_type;
* struct nfs4_string oti_scsi_device_id;
* };// size 4 + [variable]
*/
static __be32 *
__read_targetid(__be32 *p, struct pnfs_osd_targetid* targetid)
{
u32 oti_type;
oti_type = be32_to_cpup(p++);
targetid->oti_type = oti_type;
switch (oti_type) {
case OBJ_TARGET_SCSI_NAME:
case OBJ_TARGET_SCSI_DEVICE_ID:
p = __read_u8_opaque(p, &targetid->oti_scsi_device_id);
}
return p;
}
/*
* struct pnfs_osd_net_addr {
* struct nfs4_string r_netid;
* struct nfs4_string r_addr;
* };
*/
static __be32 *
__read_net_addr(__be32 *p, struct pnfs_osd_net_addr* netaddr)
{
p = __read_u8_opaque(p, &netaddr->r_netid);
p = __read_u8_opaque(p, &netaddr->r_addr);
return p;
}
/*
* struct pnfs_osd_targetaddr {
* u32 ota_available;
* struct pnfs_osd_net_addr ota_netaddr;
* };
*/
static __be32 *
__read_targetaddr(__be32 *p, struct pnfs_osd_targetaddr *targetaddr)
{
u32 ota_available;
ota_available = be32_to_cpup(p++);
targetaddr->ota_available = ota_available;
if (ota_available)
p = __read_net_addr(p, &targetaddr->ota_netaddr);
return p;
}
/*
* struct pnfs_osd_deviceaddr {
* struct pnfs_osd_targetid oda_targetid;
* struct pnfs_osd_targetaddr oda_targetaddr;
* u8 oda_lun[8];
* struct nfs4_string oda_systemid;
* struct pnfs_osd_object_cred oda_root_obj_cred;
* struct nfs4_string oda_osdname;
* };
*/
/* We need this version for the pnfs_osd_xdr_decode_deviceaddr which does
* not have an xdr_stream
*/
static __be32 *
__read_opaque_cred(__be32 *p,
struct pnfs_osd_opaque_cred *opaque_cred)
{
opaque_cred->cred_len = be32_to_cpu(*p++);
opaque_cred->cred = p;
return p + XDR_QUADLEN(opaque_cred->cred_len);
}
static __be32 *
__read_object_cred(__be32 *p, struct pnfs_osd_object_cred *comp)
{
p = _osd_xdr_decode_objid(p, &comp->oc_object_id);
comp->oc_osd_version = be32_to_cpup(p++);
comp->oc_cap_key_sec = be32_to_cpup(p++);
p = __read_opaque_cred(p, &comp->oc_cap_key);
p = __read_opaque_cred(p, &comp->oc_cap);
return p;
}
void pnfs_osd_xdr_decode_deviceaddr(
struct pnfs_osd_deviceaddr *deviceaddr, __be32 *p)
{
p = __read_targetid(p, &deviceaddr->oda_targetid);
p = __read_targetaddr(p, &deviceaddr->oda_targetaddr);
p = xdr_decode_opaque_fixed(p, deviceaddr->oda_lun,
sizeof(deviceaddr->oda_lun));
p = __read_u8_opaque(p, &deviceaddr->oda_systemid);
p = __read_object_cred(p, &deviceaddr->oda_root_obj_cred);
p = __read_u8_opaque(p, &deviceaddr->oda_osdname);
/* libosd likes this terminated in dbg. It's last, so no problems */
deviceaddr->oda_osdname.data[deviceaddr->oda_osdname.len] = 0;
}
/*
* struct pnfs_osd_layoutupdate {
* u32 dsu_valid;
* s64 dsu_delta;
* u32 olu_ioerr_flag;
* }; xdr size 4 + 8 + 4
*/
int
pnfs_osd_xdr_encode_layoutupdate(struct xdr_stream *xdr,
struct pnfs_osd_layoutupdate *lou)
{
__be32 *p = xdr_reserve_space(xdr, 4 + 8 + 4);
if (!p)
return -E2BIG;
*p++ = cpu_to_be32(lou->dsu_valid);
if (lou->dsu_valid)
p = xdr_encode_hyper(p, lou->dsu_delta);
*p++ = cpu_to_be32(lou->olu_ioerr_flag);
return 0;
}
/*
* struct pnfs_osd_objid {
* struct nfs4_deviceid oid_device_id;
* u64 oid_partition_id;
* u64 oid_object_id;
* }; // xdr size 32 bytes
*/
static inline __be32 *
pnfs_osd_xdr_encode_objid(__be32 *p, struct pnfs_osd_objid *object_id)
{
p = xdr_encode_opaque_fixed(p, &object_id->oid_device_id.data,
sizeof(object_id->oid_device_id.data));
p = xdr_encode_hyper(p, object_id->oid_partition_id);
p = xdr_encode_hyper(p, object_id->oid_object_id);
return p;
}
/*
* struct pnfs_osd_ioerr {
* struct pnfs_osd_objid oer_component;
* u64 oer_comp_offset;
* u64 oer_comp_length;
* u32 oer_iswrite;
* u32 oer_errno;
* }; // xdr size 32 + 24 bytes
*/
void pnfs_osd_xdr_encode_ioerr(__be32 *p, struct pnfs_osd_ioerr *ioerr)
{
p = pnfs_osd_xdr_encode_objid(p, &ioerr->oer_component);
p = xdr_encode_hyper(p, ioerr->oer_comp_offset);
p = xdr_encode_hyper(p, ioerr->oer_comp_length);
*p++ = cpu_to_be32(ioerr->oer_iswrite);
*p = cpu_to_be32(ioerr->oer_errno);
}
__be32 *pnfs_osd_xdr_ioerr_reserve_space(struct xdr_stream *xdr)
{
__be32 *p;
p = xdr_reserve_space(xdr, 32 + 24);
if (unlikely(!p))
dprintk("%s: out of xdr space\n", __func__);
return p;
}
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册