suyu-github/horinux
0
0
Fork 0
mirror of https://github.com/suyu-emu/horinux.git synced 2026-09-21 14:15:21 -07:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'for-3.11/drivers' of git://git.kernel.dk/linux-block

Browse Source 
Pull block IO driver bits from Jens Axboe:
 "As I mentioned in the core block pull request, due to real life
  circumstances the driver pull request would be late.  Now it looks
  like -rc2 late...  On the plus side, apart form the rsxx update, these
  are all things that I could argue could go in later in the cycle as
  they are fixes and not features.  So even though things are late, it's
  not ALL bad.

  The pull request contains:

   - Updates to bcache, all bug fixes, from Kent.

   - A pile of drbd bug fixes (no big features this time!).

   - xen blk front/back fixes.

   - rsxx driver updates, some of them deferred form 3.10.  So should be
     well cooked by now"

* 'for-3.11/drivers' of git://git.kernel.dk/linux-block: (63 commits)
  bcache: Allocation kthread fixes
  bcache: Fix GC_SECTORS_USED() calculation
  bcache: Journal replay fix
  bcache: Shutdown fix
  bcache: Fix a sysfs splat on shutdown
  bcache: Advertise that flushes are supported
  bcache: check for allocation failures
  bcache: Fix a dumb race
  bcache: Use standard utility code
  bcache: Update email address
  bcache: Delete fuzz tester
  bcache: Document shrinker reserve better
  bcache: FUA fixes
  drbd: Allow online change of al-stripes and al-stripe-size
  drbd: Constants should be UPPERCASE
  drbd: Ignore the exit code of a fence-peer handler if it returns too late
  drbd: Fix rcu_read_lock balance on error path
  drbd: fix error return code in drbd_init()
  drbd: Do not sleep inside rcu
  bcache: Refresh usage docs
  ...
This commit is contained in:
Linus Torvalds
2013-07-22 19:02:52 -07:00
parent 3b2f64d00c 0878ae2db8
commit d4c90b1b9f
47 changed files with 3235 additions and 1602 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Documentation/ABI/testing/sysfs-driver-xen-blkback
+17
View File
@@ -0,0 +1,17 @@
What: /sys/module/xen_blkback/parameters/max_buffer_pages
Date: March 2013
KernelVersion: 3.11
Contact: Roger Pau Monné <roger.pau@citrix.com>
Description:
Maximum number of free pages to keep in each block
backend buffer.
What: /sys/module/xen_blkback/parameters/max_persistent_grants
Date: March 2013
KernelVersion: 3.11
Contact: Roger Pau Monné <roger.pau@citrix.com>
Description:
Maximum number of grants to map persistently in
blkback. If the frontend tries to use more than
max_persistent_grants, the LRU kicks in and starts
removing 5% of max_persistent_grants every 100ms.
Documentation/ABI/testing/sysfs-driver-xen-blkfront
+10
View File
@@ -0,0 +1,10 @@
What: /sys/module/xen_blkfront/parameters/max
Date: June 2013
KernelVersion: 3.11
Contact: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Description:
Maximum number of segments that the frontend will negotiate
with the backend for indirect descriptors. The default value
is 32 - higher value means more potential throughput but more
memory usage. The backend picks the minimum of the frontend
and its default backend value.
Documentation/bcache.txt
+24 -13
View File
@@ -46,29 +46,33 @@ you format your backing devices and cache device at the same time, you won't
have to manually attach:
make-bcache -B /dev/sda /dev/sdb -C /dev/sdc
To make bcache devices known to the kernel, echo them to /sys/fs/bcache/register:
bcache-tools now ships udev rules, and bcache devices are known to the kernel
immediately. Without udev, you can manually register devices like this:
echo /dev/sdb > /sys/fs/bcache/register
echo /dev/sdc > /sys/fs/bcache/register
To register your bcache devices automatically, you could add something like
this to an init script:
Registering the backing device makes the bcache device show up in /dev; you can
now format it and use it as normal. But the first time using a new bcache
device, it'll be running in passthrough mode until you attach it to a cache.
See the section on attaching.
echo /dev/sd* > /sys/fs/bcache/register_quiet
The devices show up as:
It'll look for bcache superblocks and ignore everything that doesn't have one.
/dev/bcache<N>
Registering the backing device makes the bcache show up in /dev; you can now
format it and use it as normal. But the first time using a new bcache device,
it'll be running in passthrough mode until you attach it to a cache. See the
section on attaching.
As well as (with udev):
The devices show up at /dev/bcacheN, and can be controlled via sysfs from
/sys/block/bcacheN/bcache:
/dev/bcache/by-uuid/<uuid>
/dev/bcache/by-label/<label>
To get started:
mkfs.ext4 /dev/bcache0
mount /dev/bcache0 /mnt
You can control bcache devices through sysfs at /sys/block/bcache<N>/bcache .
Cache devices are managed as sets; multiple caches per set isn't supported yet
but will allow for mirroring of metadata and dirty data in the future. Your new
cache set shows up as /sys/fs/bcache/<UUID>
@@ -80,11 +84,11 @@ must be attached to your cache set to enable caching. Attaching a backing
device to a cache set is done thusly, with the UUID of the cache set in
/sys/fs/bcache:
echo <UUID> > /sys/block/bcache0/bcache/attach
echo <CSET-UUID> > /sys/block/bcache0/bcache/attach
This only has to be done once. The next time you reboot, just reregister all
your bcache devices. If a backing device has data in a cache somewhere, the
/dev/bcache# device won't be created until the cache shows up - particularly
/dev/bcache<N> device won't be created until the cache shows up - particularly
important if you have writeback caching turned on.
If you're booting up and your cache device is gone and never coming back, you
@@ -191,6 +195,9 @@ want for getting the best possible numbers when benchmarking.
SYSFS - BACKING DEVICE:
Available at /sys/block/<bdev>/bcache, /sys/block/bcache*/bcache and
(if attached) /sys/fs/bcache/<cset-uuid>/bdev*
attach
Echo the UUID of a cache set to this file to enable caching.
@@ -300,6 +307,8 @@ cache_readaheads
SYSFS - CACHE SET:
Available at /sys/fs/bcache/<cset-uuid>
average_key_size
Average data per key in the btree.
@@ -390,6 +399,8 @@ trigger_gc
SYSFS - CACHE DEVICE:
Available at /sys/block/<cdev>/bcache
block_size
Minimum granularity of writes - should match hardware sector size.
MAINTAINERS
+2 -2
View File
@@ -1642,7 +1642,7 @@ S: Maintained
F: drivers/net/hamradio/baycom*
BCACHE (BLOCK LAYER CACHE)
M: Kent Overstreet <koverstreet@google.com>
M: Kent Overstreet <kmo@daterainc.com>
L: linux-bcache@vger.kernel.org
W: http://bcache.evilpiepirate.org
S: Maintained:
@@ -3346,7 +3346,7 @@ F: Documentation/firmware_class/
F: drivers/base/firmware*.c
F: include/linux/firmware.h
FLASHSYSTEM DRIVER (IBM FlashSystem 70/80 PCI SSD Flash Card)
FLASH ADAPTER DRIVER (IBM Flash Adapter 900GB Full Height PCI Flash Card)
M: Joshua Morris <josh.h.morris@us.ibm.com>
M: Philip Kelleher <pjk1939@linux.vnet.ibm.com>
S: Maintained
drivers/block/Kconfig
+2 -2
View File
@@ -532,11 +532,11 @@ config BLK_DEV_RBD
If unsure, say N.
config BLK_DEV_RSXX
tristate "IBM FlashSystem 70/80 PCIe SSD Device Driver"
tristate "IBM Flash Adapter 900GB Full Height PCIe Device Driver"
depends on PCI
help
Device driver for IBM's high speed PCIe SSD
storage devices: FlashSystem-70 and FlashSystem-80.
storage device: Flash Adapter 900GB Full Height.
To compile this driver as a module, choose M here: the
module will be called rsxx.
drivers/block/drbd/drbd_actlog.c
+21
View File
@@ -659,6 +659,27 @@ void drbd_al_shrink(struct drbd_conf *mdev)
wake_up(&mdev->al_wait);
}
int drbd_initialize_al(struct drbd_conf *mdev, void *buffer)
{
struct al_transaction_on_disk *al = buffer;
struct drbd_md *md = &mdev->ldev->md;
sector_t al_base = md->md_offset + md->al_offset;
int al_size_4k = md->al_stripes * md->al_stripe_size_4k;
int i;
memset(al, 0, 4096);
al->magic = cpu_to_be32(DRBD_AL_MAGIC);
al->transaction_type = cpu_to_be16(AL_TR_INITIALIZED);
al->crc32c = cpu_to_be32(crc32c(0, al, 4096));
for (i = 0; i < al_size_4k; i++) {
int err = drbd_md_sync_page_io(mdev, mdev->ldev, al_base + i * 8, WRITE);
if (err)
return err;
}
return 0;
}
static int w_update_odbm(struct drbd_work *w, int unused)
{
struct update_odbm_work *udw = container_of(w, struct update_odbm_work, w);
drivers/block/drbd/drbd_int.h
+13 -2
View File
@@ -832,6 +832,7 @@ struct drbd_tconn { /* is a resource from the config file */
unsigned susp_nod:1; /* IO suspended because no data */
unsigned susp_fen:1; /* IO suspended because fence peer handler runs */
struct mutex cstate_mutex; /* Protects graceful disconnects */
unsigned int connect_cnt; /* Inc each time a connection is established */
unsigned long flags;
struct net_conf *net_conf; /* content protected by rcu */
@@ -1132,6 +1133,7 @@ extern void drbd_mdev_cleanup(struct drbd_conf *mdev);
void drbd_print_uuids(struct drbd_conf *mdev, const char *text);
extern void conn_md_sync(struct drbd_tconn *tconn);
extern void drbd_md_write(struct drbd_conf *mdev, void *buffer);
extern void drbd_md_sync(struct drbd_conf *mdev);
extern int drbd_md_read(struct drbd_conf *mdev, struct drbd_backing_dev *bdev);
extern void drbd_uuid_set(struct drbd_conf *mdev, int idx, u64 val) __must_hold(local);
@@ -1466,8 +1468,16 @@ extern void drbd_suspend_io(struct drbd_conf *mdev);
extern void drbd_resume_io(struct drbd_conf *mdev);
extern char *ppsize(char *buf, unsigned long long size);
extern sector_t drbd_new_dev_size(struct drbd_conf *, struct drbd_backing_dev *, sector_t, int);
enum determine_dev_size { dev_size_error = -1, unchanged = 0, shrunk = 1, grew = 2 };
extern enum determine_dev_size drbd_determine_dev_size(struct drbd_conf *, enum dds_flags) __must_hold(local);
enum determine_dev_size {
DS_ERROR_SHRINK = -3,
DS_ERROR_SPACE_MD = -2,
DS_ERROR = -1,
DS_UNCHANGED = 0,
DS_SHRUNK = 1,
DS_GREW = 2
};
extern enum determine_dev_size
drbd_determine_dev_size(struct drbd_conf *, enum dds_flags, struct resize_parms *) __must_hold(local);
extern void resync_after_online_grow(struct drbd_conf *);
extern void drbd_reconsider_max_bio_size(struct drbd_conf *mdev);
extern enum drbd_state_rv drbd_set_role(struct drbd_conf *mdev,
@@ -1633,6 +1643,7 @@ extern int __drbd_set_out_of_sync(struct drbd_conf *mdev, sector_t sector,
#define drbd_set_out_of_sync(mdev, sector, size) \
__drbd_set_out_of_sync(mdev, sector, size, __FILE__, __LINE__)
extern void drbd_al_shrink(struct drbd_conf *mdev);
extern int drbd_initialize_al(struct drbd_conf *, void *);
/* drbd_nl.c */
/* state info broadcast */
drivers/block/drbd/drbd_main.c
+34 -27
View File
@@ -2762,8 +2762,6 @@ int __init drbd_init(void)
/*
* allocate all necessary structs
*/
err = -ENOMEM;
init_waitqueue_head(&drbd_pp_wait);
drbd_proc = NULL; /* play safe for drbd_cleanup */
@@ -2773,6 +2771,7 @@ int __init drbd_init(void)
if (err)
goto fail;
err = -ENOMEM;
drbd_proc = proc_create_data("drbd", S_IFREG | S_IRUGO , NULL, &drbd_proc_fops, NULL);
if (!drbd_proc) {
printk(KERN_ERR "drbd: unable to register proc file\n");
@@ -2803,7 +2802,6 @@ int __init drbd_init(void)
fail:
drbd_cleanup();
if (err == -ENOMEM)
/* currently always the case */
printk(KERN_ERR "drbd: ran out of memory\n");
else
printk(KERN_ERR "drbd: initialization failure\n");
@@ -2881,34 +2879,14 @@ struct meta_data_on_disk {
u8 reserved_u8[4096 - (7*8 + 10*4)];
} __packed;
/**
* drbd_md_sync() - Writes the meta data super block if the MD_DIRTY flag bit is set
* @mdev: DRBD device.
*/
void drbd_md_sync(struct drbd_conf *mdev)
void drbd_md_write(struct drbd_conf *mdev, void *b)
{
struct meta_data_on_disk *buffer;
struct meta_data_on_disk *buffer = b;
sector_t sector;
int i;
/* Don't accidentally change the DRBD meta data layout. */
BUILD_BUG_ON(UI_SIZE != 4);
BUILD_BUG_ON(sizeof(struct meta_data_on_disk) != 4096);
del_timer(&mdev->md_sync_timer);
/* timer may be rearmed by drbd_md_mark_dirty() now. */
if (!test_and_clear_bit(MD_DIRTY, &mdev->flags))
return;
/* We use here D_FAILED and not D_ATTACHING because we try to write
* metadata even if we detach due to a disk failure! */
if (!get_ldev_if_state(mdev, D_FAILED))
return;
buffer = drbd_md_get_buffer(mdev);
if (!buffer)
goto out;
memset(buffer, 0, sizeof(*buffer));
buffer->la_size_sect = cpu_to_be64(drbd_get_capacity(mdev->this_bdev));
@@ -2937,6 +2915,35 @@ void drbd_md_sync(struct drbd_conf *mdev)
dev_err(DEV, "meta data update failed!\n");
drbd_chk_io_error(mdev, 1, DRBD_META_IO_ERROR);
}
}
/**
* drbd_md_sync() - Writes the meta data super block if the MD_DIRTY flag bit is set
* @mdev: DRBD device.
*/
void drbd_md_sync(struct drbd_conf *mdev)
{
struct meta_data_on_disk *buffer;
/* Don't accidentally change the DRBD meta data layout. */
BUILD_BUG_ON(UI_SIZE != 4);
BUILD_BUG_ON(sizeof(struct meta_data_on_disk) != 4096);
del_timer(&mdev->md_sync_timer);
/* timer may be rearmed by drbd_md_mark_dirty() now. */
if (!test_and_clear_bit(MD_DIRTY, &mdev->flags))
return;
/* We use here D_FAILED and not D_ATTACHING because we try to write
* metadata even if we detach due to a disk failure! */
if (!get_ldev_if_state(mdev, D_FAILED))
return;
buffer = drbd_md_get_buffer(mdev);
if (!buffer)
goto out;
drbd_md_write(mdev, buffer);
/* Update mdev->ldev->md.la_size_sect,
* since we updated it on metadata. */
drivers/block/drbd/drbd_nl.c
+143 -42
View File
@@ -417,6 +417,7 @@ static enum drbd_fencing_p highest_fencing_policy(struct drbd_tconn *tconn)
bool conn_try_outdate_peer(struct drbd_tconn *tconn)
{
unsigned int connect_cnt;
union drbd_state mask = { };
union drbd_state val = { };
enum drbd_fencing_p fp;
@@ -428,6 +429,10 @@ bool conn_try_outdate_peer(struct drbd_tconn *tconn)
return false;
}
spin_lock_irq(&tconn->req_lock);
connect_cnt = tconn->connect_cnt;
spin_unlock_irq(&tconn->req_lock);
fp = highest_fencing_policy(tconn);
switch (fp) {
case FP_NOT_AVAIL:
@@ -492,8 +497,14 @@ bool conn_try_outdate_peer(struct drbd_tconn *tconn)
here, because we might were able to re-establish the connection in the
meantime. */
spin_lock_irq(&tconn->req_lock);
if (tconn->cstate < C_WF_REPORT_PARAMS && !test_bit(STATE_SENT, &tconn->flags))
_conn_request_state(tconn, mask, val, CS_VERBOSE);
if (tconn->cstate < C_WF_REPORT_PARAMS && !test_bit(STATE_SENT, &tconn->flags)) {
if (tconn->connect_cnt != connect_cnt)
/* In case the connection was established and droped
while the fence-peer handler was running, ignore it */
conn_info(tconn, "Ignoring fence-peer exit code\n");
else
_conn_request_state(tconn, mask, val, CS_VERBOSE);
}
spin_unlock_irq(&tconn->req_lock);
return conn_highest_pdsk(tconn) <= D_OUTDATED;
@@ -816,15 +827,20 @@ void drbd_resume_io(struct drbd_conf *mdev)
* Returns 0 on success, negative return values indicate errors.
* You should call drbd_md_sync() after calling this function.
*/
enum determine_dev_size drbd_determine_dev_size(struct drbd_conf *mdev, enum dds_flags flags) __must_hold(local)
enum determine_dev_size
drbd_determine_dev_size(struct drbd_conf *mdev, enum dds_flags flags, struct resize_parms *rs) __must_hold(local)
{
sector_t prev_first_sect, prev_size; /* previous meta location */
sector_t la_size_sect, u_size;
struct drbd_md *md = &mdev->ldev->md;
u32 prev_al_stripe_size_4k;
u32 prev_al_stripes;
sector_t size;
char ppb[10];
void *buffer;
int md_moved, la_size_changed;
enum determine_dev_size rv = unchanged;
enum determine_dev_size rv = DS_UNCHANGED;
/* race:
* application request passes inc_ap_bio,
@@ -836,6 +852,11 @@ enum determine_dev_size drbd_determine_dev_size(struct drbd_conf *mdev, enum dds
* still lock the act_log to not trigger ASSERTs there.
*/
drbd_suspend_io(mdev);
buffer = drbd_md_get_buffer(mdev); /* Lock meta-data IO */
if (!buffer) {
drbd_resume_io(mdev);
return DS_ERROR;
}
/* no wait necessary anymore, actually we could assert that */
wait_event(mdev->al_wait, lc_try_lock(mdev->act_log));
@@ -844,7 +865,17 @@ enum determine_dev_size drbd_determine_dev_size(struct drbd_conf *mdev, enum dds
prev_size = mdev->ldev->md.md_size_sect;
la_size_sect = mdev->ldev->md.la_size_sect;
/* TODO: should only be some assert here, not (re)init... */
if (rs) {
/* rs is non NULL if we should change the AL layout only */
prev_al_stripes = md->al_stripes;
prev_al_stripe_size_4k = md->al_stripe_size_4k;
md->al_stripes = rs->al_stripes;
md->al_stripe_size_4k = rs->al_stripe_size / 4;
md->al_size_4k = (u64)rs->al_stripes * rs->al_stripe_size / 4;
}
drbd_md_set_sector_offsets(mdev, mdev->ldev);
rcu_read_lock();
@@ -852,6 +883,21 @@ enum determine_dev_size drbd_determine_dev_size(struct drbd_conf *mdev, enum dds
rcu_read_unlock();
size = drbd_new_dev_size(mdev, mdev->ldev, u_size, flags & DDSF_FORCED);
if (size < la_size_sect) {
if (rs && u_size == 0) {
/* Remove "rs &&" later. This check should always be active, but
right now the receiver expects the permissive behavior */
dev_warn(DEV, "Implicit shrink not allowed. "
"Use --size=%llus for explicit shrink.\n",
(unsigned long long)size);
rv = DS_ERROR_SHRINK;
}
if (u_size > size)
rv = DS_ERROR_SPACE_MD;
if (rv != DS_UNCHANGED)
goto err_out;
}
if (drbd_get_capacity(mdev->this_bdev) != size ||
drbd_bm_capacity(mdev) != size) {
int err;
@@ -867,7 +913,7 @@ enum determine_dev_size drbd_determine_dev_size(struct drbd_conf *mdev, enum dds
"Leaving size unchanged at size = %lu KB\n",
(unsigned long)size);
}
rv = dev_size_error;
rv = DS_ERROR;
}
/* racy, see comments above. */
drbd_set_my_capacity(mdev, size);
@@ -875,38 +921,57 @@ enum determine_dev_size drbd_determine_dev_size(struct drbd_conf *mdev, enum dds
dev_info(DEV, "size = %s (%llu KB)\n", ppsize(ppb, size>>1),
(unsigned long long)size>>1);
}
if (rv == dev_size_error)
goto out;
if (rv <= DS_ERROR)
goto err_out;
la_size_changed = (la_size_sect != mdev->ldev->md.la_size_sect);
md_moved = prev_first_sect != drbd_md_first_sector(mdev->ldev)
|| prev_size != mdev->ldev->md.md_size_sect;
if (la_size_changed || md_moved) {
int err;
if (la_size_changed || md_moved || rs) {
u32 prev_flags;
drbd_al_shrink(mdev); /* All extents inactive. */
prev_flags = md->flags;
md->flags &= ~MDF_PRIMARY_IND;
drbd_md_write(mdev, buffer);
dev_info(DEV, "Writing the whole bitmap, %s\n",
la_size_changed && md_moved ? "size changed and md moved" :
la_size_changed ? "size changed" : "md moved");
/* next line implicitly does drbd_suspend_io()+drbd_resume_io() */
err = drbd_bitmap_io(mdev, md_moved ? &drbd_bm_write_all : &drbd_bm_write,
"size changed", BM_LOCKED_MASK);
if (err) {
rv = dev_size_error;
goto out;
}
drbd_md_mark_dirty(mdev);
drbd_bitmap_io(mdev, md_moved ? &drbd_bm_write_all : &drbd_bm_write,
"size changed", BM_LOCKED_MASK);
drbd_initialize_al(mdev, buffer);
md->flags = prev_flags;
drbd_md_write(mdev, buffer);
if (rs)
dev_info(DEV, "Changed AL layout to al-stripes = %d, al-stripe-size-kB = %d\n",
md->al_stripes, md->al_stripe_size_4k * 4);
}
if (size > la_size_sect)
rv = grew;
rv = DS_GREW;
if (size < la_size_sect)
rv = shrunk;
out:
rv = DS_SHRUNK;
if (0) {
err_out:
if (rs) {
md->al_stripes = prev_al_stripes;
md->al_stripe_size_4k = prev_al_stripe_size_4k;
md->al_size_4k = (u64)prev_al_stripes * prev_al_stripe_size_4k;
drbd_md_set_sector_offsets(mdev, mdev->ldev);
}
}
lc_unlock(mdev->act_log);
wake_up(&mdev->al_wait);
drbd_md_put_buffer(mdev);
drbd_resume_io(mdev);
return rv;
@@ -1607,11 +1672,11 @@ int drbd_adm_attach(struct sk_buff *skb, struct genl_info *info)
!drbd_md_test_flag(mdev->ldev, MDF_CONNECTED_IND))
set_bit(USE_DEGR_WFC_T, &mdev->flags);
dd = drbd_determine_dev_size(mdev, 0);
if (dd == dev_size_error) {
dd = drbd_determine_dev_size(mdev, 0, NULL);
if (dd <= DS_ERROR) {
retcode = ERR_NOMEM_BITMAP;
goto force_diskless_dec;
} else if (dd == grew)
} else if (dd == DS_GREW)
set_bit(RESYNC_AFTER_NEG, &mdev->flags);
if (drbd_md_test_flag(mdev->ldev, MDF_FULL_SYNC) ||
@@ -2305,6 +2370,7 @@ int drbd_adm_resize(struct sk_buff *skb, struct genl_info *info)
struct drbd_conf *mdev;
enum drbd_ret_code retcode;
enum determine_dev_size dd;
bool change_al_layout = false;
enum dds_flags ddsf;
sector_t u_size;
int err;
@@ -2315,31 +2381,33 @@ int drbd_adm_resize(struct sk_buff *skb, struct genl_info *info)
if (retcode != NO_ERROR)
goto fail;
mdev = adm_ctx.mdev;
if (!get_ldev(mdev)) {
retcode = ERR_NO_DISK;
goto fail;
}
memset(&rs, 0, sizeof(struct resize_parms));
rs.al_stripes = mdev->ldev->md.al_stripes;
rs.al_stripe_size = mdev->ldev->md.al_stripe_size_4k * 4;
if (info->attrs[DRBD_NLA_RESIZE_PARMS]) {
err = resize_parms_from_attrs(&rs, info);
if (err) {
retcode = ERR_MANDATORY_TAG;
drbd_msg_put_info(from_attrs_err_to_txt(err));
goto fail;
goto fail_ldev;
}
}
mdev = adm_ctx.mdev;
if (mdev->state.conn > C_CONNECTED) {
retcode = ERR_RESIZE_RESYNC;
goto fail;
goto fail_ldev;
}
if (mdev->state.role == R_SECONDARY &&
mdev->state.peer == R_SECONDARY) {
retcode = ERR_NO_PRIMARY;
goto fail;
}
if (!get_ldev(mdev)) {
retcode = ERR_NO_DISK;
goto fail;
goto fail_ldev;
}
if (rs.no_resync && mdev->tconn->agreed_pro_version < 93) {
@@ -2358,6 +2426,28 @@ int drbd_adm_resize(struct sk_buff *skb, struct genl_info *info)
}
}
if (mdev->ldev->md.al_stripes != rs.al_stripes ||
mdev->ldev->md.al_stripe_size_4k != rs.al_stripe_size / 4) {
u32 al_size_k = rs.al_stripes * rs.al_stripe_size;
if (al_size_k > (16 * 1024 * 1024)) {
retcode = ERR_MD_LAYOUT_TOO_BIG;
goto fail_ldev;
}
if (al_size_k < MD_32kB_SECT/2) {
retcode = ERR_MD_LAYOUT_TOO_SMALL;
goto fail_ldev;
}
if (mdev->state.conn != C_CONNECTED) {
retcode = ERR_MD_LAYOUT_CONNECTED;
goto fail_ldev;
}
change_al_layout = true;
}
if (mdev->ldev->known_size != drbd_get_capacity(mdev->ldev->backing_bdev))
mdev->ldev->known_size = drbd_get_capacity(mdev->ldev->backing_bdev);
@@ -2373,16 +2463,22 @@ int drbd_adm_resize(struct sk_buff *skb, struct genl_info *info)
}
ddsf = (rs.resize_force ? DDSF_FORCED : 0) | (rs.no_resync ? DDSF_NO_RESYNC : 0);
dd = drbd_determine_dev_size(mdev, ddsf);
dd = drbd_determine_dev_size(mdev, ddsf, change_al_layout ? &rs : NULL);
drbd_md_sync(mdev);
put_ldev(mdev);
if (dd == dev_size_error) {
if (dd == DS_ERROR) {
retcode = ERR_NOMEM_BITMAP;
goto fail;
} else if (dd == DS_ERROR_SPACE_MD) {
retcode = ERR_MD_LAYOUT_NO_FIT;
goto fail;
} else if (dd == DS_ERROR_SHRINK) {
retcode = ERR_IMPLICIT_SHRINK;
goto fail;
}
if (mdev->state.conn == C_CONNECTED) {
if (dd == grew)
if (dd == DS_GREW)
set_bit(RESIZE_PENDING, &mdev->flags);
drbd_send_uuids(mdev);
@@ -2658,7 +2754,6 @@ int nla_put_status_info(struct sk_buff *skb, struct drbd_conf *mdev,
const struct sib_info *sib)
{
struct state_info *si = NULL; /* for sizeof(si->member); */
struct net_conf *nc;
struct nlattr *nla;
int got_ldev;
int err = 0;
@@ -2688,13 +2783,19 @@ int nla_put_status_info(struct sk_buff *skb, struct drbd_conf *mdev,
goto nla_put_failure;
rcu_read_lock();
if (got_ldev)
if (disk_conf_to_skb(skb, rcu_dereference(mdev->ldev->disk_conf), exclude_sensitive))
goto nla_put_failure;
if (got_ldev) {
struct disk_conf *disk_conf;
nc = rcu_dereference(mdev->tconn->net_conf);
if (nc)
err = net_conf_to_skb(skb, nc, exclude_sensitive);
disk_conf = rcu_dereference(mdev->ldev->disk_conf);
err = disk_conf_to_skb(skb, disk_conf, exclude_sensitive);
}
if (!err) {
struct net_conf *nc;
nc = rcu_dereference(mdev->tconn->net_conf);
if (nc)
err = net_conf_to_skb(skb, nc, exclude_sensitive);
}
rcu_read_unlock();
if (err)
goto nla_put_failure;
drivers/block/drbd/drbd_receiver.c
+6 -6
View File
@@ -1039,6 +1039,8 @@ randomize:
rcu_read_lock();
idr_for_each_entry(&tconn->volumes, mdev, vnr) {
kref_get(&mdev->kref);
rcu_read_unlock();
/* Prevent a race between resync-handshake and
* being promoted to Primary.
*
@@ -1049,8 +1051,6 @@ randomize:
mutex_lock(mdev->state_mutex);
mutex_unlock(mdev->state_mutex);
rcu_read_unlock();
if (discard_my_data)
set_bit(DISCARD_MY_DATA, &mdev->flags);
else
@@ -3545,7 +3545,7 @@ static int receive_sizes(struct drbd_tconn *tconn, struct packet_info *pi)
{
struct drbd_conf *mdev;
struct p_sizes *p = pi->data;
enum determine_dev_size dd = unchanged;
enum determine_dev_size dd = DS_UNCHANGED;
sector_t p_size, p_usize, my_usize;
int ldsc = 0; /* local disk size changed */
enum dds_flags ddsf;
@@ -3617,9 +3617,9 @@ static int receive_sizes(struct drbd_tconn *tconn, struct packet_info *pi)
ddsf = be16_to_cpu(p->dds_flags);
if (get_ldev(mdev)) {
dd = drbd_determine_dev_size(mdev, ddsf);
dd = drbd_determine_dev_size(mdev, ddsf, NULL);
put_ldev(mdev);
if (dd == dev_size_error)
if (dd == DS_ERROR)
return -EIO;
drbd_md_sync(mdev);
} else {
@@ -3647,7 +3647,7 @@ static int receive_sizes(struct drbd_tconn *tconn, struct packet_info *pi)
drbd_send_sizes(mdev, 0, ddsf);
}
if (test_and_clear_bit(RESIZE_PENDING, &mdev->flags) ||
(dd == grew && mdev->state.conn == C_CONNECTED)) {
(dd == DS_GREW && mdev->state.conn == C_CONNECTED)) {
if (mdev->state.pdsk >= D_INCONSISTENT &&
mdev->state.disk >= D_INCONSISTENT) {
if (ddsf & DDSF_NO_RESYNC)
drivers/block/drbd/drbd_state.c
+3 -1
View File
@@ -1115,8 +1115,10 @@ __drbd_set_state(struct drbd_conf *mdev, union drbd_state ns,
drbd_thread_restart_nowait(&mdev->tconn->receiver);
/* Resume AL writing if we get a connection */
if (os.conn < C_CONNECTED && ns.conn >= C_CONNECTED)
if (os.conn < C_CONNECTED && ns.conn >= C_CONNECTED) {
drbd_resume_al(mdev);
mdev->tconn->connect_cnt++;
}
/* remember last attach time so request_timer_fn() won't
* kill newly established sessions while we are still trying to thaw
drivers/block/rsxx/core.c
+348 -11
View File
@@ -31,6 +31,8 @@
#include <linux/slab.h>
#include <linux/bitops.h>
#include <linux/delay.h>
#include <linux/debugfs.h>
#include <linux/seq_file.h>
#include <linux/genhd.h>
#include <linux/idr.h>
@@ -39,8 +41,9 @@
#include "rsxx_cfg.h"
#define NO_LEGACY 0
#define SYNC_START_TIMEOUT (10 * 60) /* 10 minutes */
MODULE_DESCRIPTION("IBM FlashSystem 70/80 PCIe SSD Device Driver");
MODULE_DESCRIPTION("IBM Flash Adapter 900GB Full Height Device Driver");
MODULE_AUTHOR("Joshua Morris/Philip Kelleher, IBM");
MODULE_LICENSE("GPL");
MODULE_VERSION(DRIVER_VERSION);
@@ -49,9 +52,282 @@ static unsigned int force_legacy = NO_LEGACY;
module_param(force_legacy, uint, 0444);
MODULE_PARM_DESC(force_legacy, "Force the use of legacy type PCI interrupts");
static unsigned int sync_start = 1;
module_param(sync_start, uint, 0444);
MODULE_PARM_DESC(sync_start, "On by Default: Driver load will not complete "
"until the card startup has completed.");
static DEFINE_IDA(rsxx_disk_ida);
static DEFINE_SPINLOCK(rsxx_ida_lock);
/* --------------------Debugfs Setup ------------------- */
struct rsxx_cram {
u32 f_pos;
u32 offset;
void *i_private;
};
static int rsxx_attr_pci_regs_show(struct seq_file *m, void *p)
{
struct rsxx_cardinfo *card = m->private;
seq_printf(m, "HWID 0x%08x\n",
ioread32(card->regmap + HWID));
seq_printf(m, "SCRATCH 0x%08x\n",
ioread32(card->regmap + SCRATCH));
seq_printf(m, "IER 0x%08x\n",
ioread32(card->regmap + IER));
seq_printf(m, "IPR 0x%08x\n",
ioread32(card->regmap + IPR));
seq_printf(m, "CREG_CMD 0x%08x\n",
ioread32(card->regmap + CREG_CMD));
seq_printf(m, "CREG_ADD 0x%08x\n",
ioread32(card->regmap + CREG_ADD));
seq_printf(m, "CREG_CNT 0x%08x\n",
ioread32(card->regmap + CREG_CNT));
seq_printf(m, "CREG_STAT 0x%08x\n",
ioread32(card->regmap + CREG_STAT));
seq_printf(m, "CREG_DATA0 0x%08x\n",
ioread32(card->regmap + CREG_DATA0));
seq_printf(m, "CREG_DATA1 0x%08x\n",
ioread32(card->regmap + CREG_DATA1));
seq_printf(m, "CREG_DATA2 0x%08x\n",
ioread32(card->regmap + CREG_DATA2));
seq_printf(m, "CREG_DATA3 0x%08x\n",
ioread32(card->regmap + CREG_DATA3));
seq_printf(m, "CREG_DATA4 0x%08x\n",
ioread32(card->regmap + CREG_DATA4));
seq_printf(m, "CREG_DATA5 0x%08x\n",
ioread32(card->regmap + CREG_DATA5));
seq_printf(m, "CREG_DATA6 0x%08x\n",
ioread32(card->regmap + CREG_DATA6));
seq_printf(m, "CREG_DATA7 0x%08x\n",
ioread32(card->regmap + CREG_DATA7));
seq_printf(m, "INTR_COAL 0x%08x\n",
ioread32(card->regmap + INTR_COAL));
seq_printf(m, "HW_ERROR 0x%08x\n",
ioread32(card->regmap + HW_ERROR));
seq_printf(m, "DEBUG0 0x%08x\n",
ioread32(card->regmap + PCI_DEBUG0));
seq_printf(m, "DEBUG1 0x%08x\n",
ioread32(card->regmap + PCI_DEBUG1));
seq_printf(m, "DEBUG2 0x%08x\n",
ioread32(card->regmap + PCI_DEBUG2));
seq_printf(m, "DEBUG3 0x%08x\n",
ioread32(card->regmap + PCI_DEBUG3));
seq_printf(m, "DEBUG4 0x%08x\n",
ioread32(card->regmap + PCI_DEBUG4));
seq_printf(m, "DEBUG5 0x%08x\n",
ioread32(card->regmap + PCI_DEBUG5));
seq_printf(m, "DEBUG6 0x%08x\n",
ioread32(card->regmap + PCI_DEBUG6));
seq_printf(m, "DEBUG7 0x%08x\n",
ioread32(card->regmap + PCI_DEBUG7));
seq_printf(m, "RECONFIG 0x%08x\n",
ioread32(card->regmap + PCI_RECONFIG));
return 0;
}
static int rsxx_attr_stats_show(struct seq_file *m, void *p)
{
struct rsxx_cardinfo *card = m->private;
int i;
for (i = 0; i < card->n_targets; i++) {
seq_printf(m, "Ctrl %d CRC Errors = %d\n",
i, card->ctrl[i].stats.crc_errors);
seq_printf(m, "Ctrl %d Hard Errors = %d\n",
i, card->ctrl[i].stats.hard_errors);
seq_printf(m, "Ctrl %d Soft Errors = %d\n",
i, card->ctrl[i].stats.soft_errors);
seq_printf(m, "Ctrl %d Writes Issued = %d\n",
i, card->ctrl[i].stats.writes_issued);
seq_printf(m, "Ctrl %d Writes Failed = %d\n",
i, card->ctrl[i].stats.writes_failed);
seq_printf(m, "Ctrl %d Reads Issued = %d\n",
i, card->ctrl[i].stats.reads_issued);
seq_printf(m, "Ctrl %d Reads Failed = %d\n",
i, card->ctrl[i].stats.reads_failed);
seq_printf(m, "Ctrl %d Reads Retried = %d\n",
i, card->ctrl[i].stats.reads_retried);
seq_printf(m, "Ctrl %d Discards Issued = %d\n",
i, card->ctrl[i].stats.discards_issued);
seq_printf(m, "Ctrl %d Discards Failed = %d\n",
i, card->ctrl[i].stats.discards_failed);
seq_printf(m, "Ctrl %d DMA SW Errors = %d\n",
i, card->ctrl[i].stats.dma_sw_err);
seq_printf(m, "Ctrl %d DMA HW Faults = %d\n",
i, card->ctrl[i].stats.dma_hw_fault);
seq_printf(m, "Ctrl %d DMAs Cancelled = %d\n",
i, card->ctrl[i].stats.dma_cancelled);
seq_printf(m, "Ctrl %d SW Queue Depth = %d\n",
i, card->ctrl[i].stats.sw_q_depth);
seq_printf(m, "Ctrl %d HW Queue Depth = %d\n",
i, atomic_read(&card->ctrl[i].stats.hw_q_depth));
}
return 0;
}
static int rsxx_attr_stats_open(struct inode *inode, struct file *file)
{
return single_open(file, rsxx_attr_stats_show, inode->i_private);
}
static int rsxx_attr_pci_regs_open(struct inode *inode, struct file *file)
{
return single_open(file, rsxx_attr_pci_regs_show, inode->i_private);
}
static ssize_t rsxx_cram_read(struct file *fp, char __user *ubuf,
size_t cnt, loff_t *ppos)
{
struct rsxx_cram *info = fp->private_data;
struct rsxx_cardinfo *card = info->i_private;
char *buf;
int st;
buf = kzalloc(sizeof(*buf) * cnt, GFP_KERNEL);
if (!buf)
return -ENOMEM;
info->f_pos = (u32)*ppos + info->offset;
st = rsxx_creg_read(card, CREG_ADD_CRAM + info->f_pos, cnt, buf, 1);
if (st)
return st;
st = copy_to_user(ubuf, buf, cnt);
if (st)
return st;
info->offset += cnt;
kfree(buf);
return cnt;
}
static ssize_t rsxx_cram_write(struct file *fp, const char __user *ubuf,
size_t cnt, loff_t *ppos)
{
struct rsxx_cram *info = fp->private_data;
struct rsxx_cardinfo *card = info->i_private;
char *buf;
int st;
buf = kzalloc(sizeof(*buf) * cnt, GFP_KERNEL);
if (!buf)
return -ENOMEM;
st = copy_from_user(buf, ubuf, cnt);
if (st)
return st;
info->f_pos = (u32)*ppos + info->offset;
st = rsxx_creg_write(card, CREG_ADD_CRAM + info->f_pos, cnt, buf, 1);
if (st)
return st;
info->offset += cnt;
kfree(buf);
return cnt;
}
static int rsxx_cram_open(struct inode *inode, struct file *file)
{
struct rsxx_cram *info = kzalloc(sizeof(*info), GFP_KERNEL);
if (!info)
return -ENOMEM;
info->i_private = inode->i_private;
info->f_pos = file->f_pos;
file->private_data = info;
return 0;
}
static int rsxx_cram_release(struct inode *inode, struct file *file)
{
struct rsxx_cram *info = file->private_data;
if (!info)
return 0;
kfree(info);
file->private_data = NULL;
return 0;
}
static const struct file_operations debugfs_cram_fops = {
.owner = THIS_MODULE,
.open = rsxx_cram_open,
.read = rsxx_cram_read,
.write = rsxx_cram_write,
.release = rsxx_cram_release,
};
static const struct file_operations debugfs_stats_fops = {
.owner = THIS_MODULE,
.open = rsxx_attr_stats_open,
.read = seq_read,
.llseek = seq_lseek,
.release = single_release,
};
static const struct file_operations debugfs_pci_regs_fops = {
.owner = THIS_MODULE,
.open = rsxx_attr_pci_regs_open,
.read = seq_read,
.llseek = seq_lseek,
.release = single_release,
};
static void rsxx_debugfs_dev_new(struct rsxx_cardinfo *card)
{
struct dentry *debugfs_stats;
struct dentry *debugfs_pci_regs;
struct dentry *debugfs_cram;
card->debugfs_dir = debugfs_create_dir(card->gendisk->disk_name, NULL);
if (IS_ERR_OR_NULL(card->debugfs_dir))
goto failed_debugfs_dir;
debugfs_stats = debugfs_create_file("stats", S_IRUGO,
card->debugfs_dir, card,
&debugfs_stats_fops);
if (IS_ERR_OR_NULL(debugfs_stats))
goto failed_debugfs_stats;
debugfs_pci_regs = debugfs_create_file("pci_regs", S_IRUGO,
card->debugfs_dir, card,
&debugfs_pci_regs_fops);
if (IS_ERR_OR_NULL(debugfs_pci_regs))
goto failed_debugfs_pci_regs;
debugfs_cram = debugfs_create_file("cram", S_IRUGO | S_IWUSR,
card->debugfs_dir, card,
&debugfs_cram_fops);
if (IS_ERR_OR_NULL(debugfs_cram))
goto failed_debugfs_cram;
return;
failed_debugfs_cram:
debugfs_remove(debugfs_pci_regs);
failed_debugfs_pci_regs:
debugfs_remove(debugfs_stats);
failed_debugfs_stats:
debugfs_remove(card->debugfs_dir);
failed_debugfs_dir:
card->debugfs_dir = NULL;
}
/*----------------- Interrupt Control & Handling -------------------*/
static void rsxx_mask_interrupts(struct rsxx_cardinfo *card)
@@ -163,12 +439,13 @@ static irqreturn_t rsxx_isr(int irq, void *pdata)
}
if (isr & CR_INTR_CREG) {
schedule_work(&card->creg_ctrl.done_work);
queue_work(card->creg_ctrl.creg_wq,
&card->creg_ctrl.done_work);
handled++;
}
if (isr & CR_INTR_EVENT) {
schedule_work(&card->event_work);
queue_work(card->event_wq, &card->event_work);
rsxx_disable_ier_and_isr(card, CR_INTR_EVENT);
handled++;
}
@@ -329,7 +606,7 @@ static int rsxx_eeh_frozen(struct pci_dev *dev)
int i;
int st;
dev_warn(&dev->dev, "IBM FlashSystem PCI: preparing for slot reset.\n");
dev_warn(&dev->dev, "IBM Flash Adapter PCI: preparing for slot reset.\n");
card->eeh_state = 1;
rsxx_mask_interrupts(card);
@@ -367,15 +644,26 @@ static void rsxx_eeh_failure(struct pci_dev *dev)
{
struct rsxx_cardinfo *card = pci_get_drvdata(dev);
int i;
int cnt = 0;
dev_err(&dev->dev, "IBM FlashSystem PCI: disabling failed card.\n");
dev_err(&dev->dev, "IBM Flash Adapter PCI: disabling failed card.\n");
card->eeh_state = 1;
card->halt = 1;
for (i = 0; i < card->n_targets; i++)
del_timer_sync(&card->ctrl[i].activity_timer);
for (i = 0; i < card->n_targets; i++) {
spin_lock_bh(&card->ctrl[i].queue_lock);
cnt = rsxx_cleanup_dma_queue(&card->ctrl[i],
&card->ctrl[i].queue);
spin_unlock_bh(&card->ctrl[i].queue_lock);
rsxx_eeh_cancel_dmas(card);
cnt += rsxx_dma_cancel(&card->ctrl[i]);
if (cnt)
dev_info(CARD_TO_DEV(card),
"Freed %d queued DMAs on channel %d\n",
cnt, card->ctrl[i].id);
}
}
static int rsxx_eeh_fifo_flush_poll(struct rsxx_cardinfo *card)
@@ -432,7 +720,7 @@ static pci_ers_result_t rsxx_slot_reset(struct pci_dev *dev)
int st;
dev_warn(&dev->dev,
"IBM FlashSystem PCI: recovering from slot reset.\n");
"IBM Flash Adapter PCI: recovering from slot reset.\n");
st = pci_enable_device(dev);
if (st)
@@ -485,7 +773,7 @@ static pci_ers_result_t rsxx_slot_reset(struct pci_dev *dev)
&card->ctrl[i].issue_dma_work);
}
dev_info(&dev->dev, "IBM FlashSystem PCI: recovery complete.\n");
dev_info(&dev->dev, "IBM Flash Adapter PCI: recovery complete.\n");
return PCI_ERS_RESULT_RECOVERED;
@@ -528,6 +816,7 @@ static int rsxx_pci_probe(struct pci_dev *dev,
{
struct rsxx_cardinfo *card;
int st;
unsigned int sync_timeout;
dev_info(&dev->dev, "PCI-Flash SSD discovered\n");
@@ -610,7 +899,11 @@ static int rsxx_pci_probe(struct pci_dev *dev,
}
/************* Setup Processor Command Interface *************/
rsxx_creg_setup(card);
st = rsxx_creg_setup(card);
if (st) {
dev_err(CARD_TO_DEV(card), "Failed to setup creg interface.\n");
goto failed_creg_setup;
}
spin_lock_irq(&card->irq_lock);
rsxx_enable_ier_and_isr(card, CR_INTR_CREG);
@@ -650,6 +943,12 @@ static int rsxx_pci_probe(struct pci_dev *dev,
}
/************* Setup Card Event Handler *************/
card->event_wq = create_singlethread_workqueue(DRIVER_NAME"_event");
if (!card->event_wq) {
dev_err(CARD_TO_DEV(card), "Failed card event setup.\n");
goto failed_event_handler;
}
INIT_WORK(&card->event_work, card_event_handler);
st = rsxx_setup_dev(card);
@@ -676,6 +975,33 @@ static int rsxx_pci_probe(struct pci_dev *dev,
if (st)
dev_crit(CARD_TO_DEV(card),
"Failed issuing card startup\n");
if (sync_start) {
sync_timeout = SYNC_START_TIMEOUT;
dev_info(CARD_TO_DEV(card),
"Waiting for card to startup\n");
do {
ssleep(1);
sync_timeout--;
rsxx_get_card_state(card, &card->state);
} while (sync_timeout &&
(card->state == CARD_STATE_STARTING));
if (card->state == CARD_STATE_STARTING) {
dev_warn(CARD_TO_DEV(card),
"Card startup timed out\n");
card->size8 = 0;
} else {
dev_info(CARD_TO_DEV(card),
"card state: %s\n",
rsxx_card_state_to_str(card->state));
st = rsxx_get_card_size8(card, &card->size8);
if (st)
card->size8 = 0;
}
}
} else if (card->state == CARD_STATE_GOOD ||
card->state == CARD_STATE_RD_ONLY_FAULT) {
st = rsxx_get_card_size8(card, &card->size8);
@@ -685,12 +1011,21 @@ static int rsxx_pci_probe(struct pci_dev *dev,
rsxx_attach_dev(card);
/************* Setup Debugfs *************/
rsxx_debugfs_dev_new(card);
return 0;
failed_create_dev:
destroy_workqueue(card->event_wq);
card->event_wq = NULL;
failed_event_handler:
rsxx_dma_destroy(card);
failed_dma_setup:
failed_compatiblity_check:
destroy_workqueue(card->creg_ctrl.creg_wq);
card->creg_ctrl.creg_wq = NULL;
failed_creg_setup:
spin_lock_irq(&card->irq_lock);
rsxx_disable_ier_and_isr(card, CR_INTR_ALL);
spin_unlock_irq(&card->irq_lock);
@@ -756,6 +1091,8 @@ static void rsxx_pci_remove(struct pci_dev *dev)
/* Prevent work_structs from re-queuing themselves. */
card->halt = 1;
debugfs_remove_recursive(card->debugfs_dir);
free_irq(dev->irq, card);
if (!force_legacy)
drivers/block/rsxx/cregs.c
+14
View File
@@ -431,6 +431,15 @@ static int __issue_creg_rw(struct rsxx_cardinfo *card,
*hw_stat = completion.creg_status;
if (completion.st) {
/*
* This read is needed to verify that there has not been any
* extreme errors that might have occurred, i.e. EEH. The
* function iowrite32 will not detect EEH errors, so it is
* necessary that we recover if such an error is the reason
* for the timeout. This is a dummy read.
*/
ioread32(card->regmap + SCRATCH);
dev_warn(CARD_TO_DEV(card),
"creg command failed(%d x%08x)\n",
completion.st, addr);
@@ -727,6 +736,11 @@ int rsxx_creg_setup(struct rsxx_cardinfo *card)
{
card->creg_ctrl.active_cmd = NULL;
card->creg_ctrl.creg_wq =
create_singlethread_workqueue(DRIVER_NAME"_creg");
if (!card->creg_ctrl.creg_wq)
return -ENOMEM;
INIT_WORK(&card->creg_ctrl.done_work, creg_cmd_done);
mutex_init(&card->creg_ctrl.reset_lock);
INIT_LIST_HEAD(&card->creg_ctrl.queue);
drivers/block/rsxx/dev.c
+12 -21
View File
@@ -155,7 +155,8 @@ static void bio_dma_done_cb(struct rsxx_cardinfo *card,
atomic_set(&meta->error, 1);
if (atomic_dec_and_test(&meta->pending_dmas)) {
disk_stats_complete(card, meta->bio, meta->start_time);
if (!card->eeh_state && card->gendisk)
disk_stats_complete(card, meta->bio, meta->start_time);
bio_endio(meta->bio, atomic_read(&meta->error) ? -EIO : 0);
kmem_cache_free(bio_meta_pool, meta);
@@ -170,6 +171,12 @@ static void rsxx_make_request(struct request_queue *q, struct bio *bio)
might_sleep();
if (!card)
goto req_err;
if (bio->bi_sector + (bio->bi_size >> 9) > get_capacity(card->gendisk))
goto req_err;
if (unlikely(card->halt)) {
st = -EFAULT;
goto req_err;
@@ -196,7 +203,8 @@ static void rsxx_make_request(struct request_queue *q, struct bio *bio)
atomic_set(&bio_meta->pending_dmas, 0);
bio_meta->start_time = jiffies;
disk_stats_start(card, bio);
if (!unlikely(card->halt))
disk_stats_start(card, bio);
dev_dbg(CARD_TO_DEV(card), "BIO[%c]: meta: %p addr8: x%llx size: %d\n",
bio_data_dir(bio) ? 'W' : 'R', bio_meta,
@@ -225,24 +233,6 @@ static bool rsxx_discard_supported(struct rsxx_cardinfo *card)
return (pci_rev >= RSXX_DISCARD_SUPPORT);
}
static unsigned short rsxx_get_logical_block_size(
struct rsxx_cardinfo *card)
{
u32 capabilities = 0;
int st;
st = rsxx_get_card_capabilities(card, &capabilities);
if (st)
dev_warn(CARD_TO_DEV(card),
"Failed reading card capabilities register\n");
/* Earlier firmware did not have support for 512 byte accesses */
if (capabilities & CARD_CAP_SUBPAGE_WRITES)
return 512;
else
return RSXX_HW_BLK_SIZE;
}
int rsxx_attach_dev(struct rsxx_cardinfo *card)
{
mutex_lock(&card->dev_lock);
@@ -305,7 +295,7 @@ int rsxx_setup_dev(struct rsxx_cardinfo *card)
return -ENOMEM;
}
blk_size = rsxx_get_logical_block_size(card);
blk_size = card->config.data.block_size;
blk_queue_make_request(card->queue, rsxx_make_request);
blk_queue_bounce_limit(card->queue, BLK_BOUNCE_ANY);
@@ -347,6 +337,7 @@ void rsxx_destroy_dev(struct rsxx_cardinfo *card)
card->gendisk = NULL;
blk_cleanup_queue(card->queue);
card->queue->queuedata = NULL;
unregister_blkdev(card->major, DRIVER_NAME);
}
drivers/block/rsxx/dma.c
+96 -89
View File
@@ -245,6 +245,22 @@ static void rsxx_complete_dma(struct rsxx_dma_ctrl *ctrl,
kmem_cache_free(rsxx_dma_pool, dma);
}
int rsxx_cleanup_dma_queue(struct rsxx_dma_ctrl *ctrl,
struct list_head *q)
{
struct rsxx_dma *dma;
struct rsxx_dma *tmp;
int cnt = 0;
list_for_each_entry_safe(dma, tmp, q, list) {
list_del(&dma->list);
rsxx_complete_dma(ctrl, dma, DMA_CANCELLED);
cnt++;
}
return cnt;
}
static void rsxx_requeue_dma(struct rsxx_dma_ctrl *ctrl,
struct rsxx_dma *dma)
{
@@ -252,9 +268,10 @@ static void rsxx_requeue_dma(struct rsxx_dma_ctrl *ctrl,
* Requeued DMAs go to the front of the queue so they are issued
* first.
*/
spin_lock(&ctrl->queue_lock);
spin_lock_bh(&ctrl->queue_lock);
ctrl->stats.sw_q_depth++;
list_add(&dma->list, &ctrl->queue);
spin_unlock(&ctrl->queue_lock);
spin_unlock_bh(&ctrl->queue_lock);
}
static void rsxx_handle_dma_error(struct rsxx_dma_ctrl *ctrl,
@@ -329,6 +346,7 @@ static void rsxx_handle_dma_error(struct rsxx_dma_ctrl *ctrl,
static void dma_engine_stalled(unsigned long data)
{
struct rsxx_dma_ctrl *ctrl = (struct rsxx_dma_ctrl *)data;
int cnt;
if (atomic_read(&ctrl->stats.hw_q_depth) == 0 ||
unlikely(ctrl->card->eeh_state))
@@ -349,18 +367,28 @@ static void dma_engine_stalled(unsigned long data)
"DMA channel %d has stalled, faulting interface.\n",
ctrl->id);
ctrl->card->dma_fault = 1;
/* Clean up the DMA queue */
spin_lock(&ctrl->queue_lock);
cnt = rsxx_cleanup_dma_queue(ctrl, &ctrl->queue);
spin_unlock(&ctrl->queue_lock);
cnt += rsxx_dma_cancel(ctrl);
if (cnt)
dev_info(CARD_TO_DEV(ctrl->card),
"Freed %d queued DMAs on channel %d\n",
cnt, ctrl->id);
}
}
static void rsxx_issue_dmas(struct work_struct *work)
static void rsxx_issue_dmas(struct rsxx_dma_ctrl *ctrl)
{
struct rsxx_dma_ctrl *ctrl;
struct rsxx_dma *dma;
int tag;
int cmds_pending = 0;
struct hw_cmd *hw_cmd_buf;
ctrl = container_of(work, struct rsxx_dma_ctrl, issue_dma_work);
hw_cmd_buf = ctrl->cmd.buf;
if (unlikely(ctrl->card->halt) ||
@@ -368,22 +396,22 @@ static void rsxx_issue_dmas(struct work_struct *work)
return;
while (1) {
spin_lock(&ctrl->queue_lock);
spin_lock_bh(&ctrl->queue_lock);
if (list_empty(&ctrl->queue)) {
spin_unlock(&ctrl->queue_lock);
spin_unlock_bh(&ctrl->queue_lock);
break;
}
spin_unlock(&ctrl->queue_lock);
spin_unlock_bh(&ctrl->queue_lock);
tag = pop_tracker(ctrl->trackers);
if (tag == -1)
break;
spin_lock(&ctrl->queue_lock);
spin_lock_bh(&ctrl->queue_lock);
dma = list_entry(ctrl->queue.next, struct rsxx_dma, list);
list_del(&dma->list);
ctrl->stats.sw_q_depth--;
spin_unlock(&ctrl->queue_lock);
spin_unlock_bh(&ctrl->queue_lock);
/*
* This will catch any DMAs that slipped in right before the
@@ -440,9 +468,8 @@ static void rsxx_issue_dmas(struct work_struct *work)
}
}
static void rsxx_dma_done(struct work_struct *work)
static void rsxx_dma_done(struct rsxx_dma_ctrl *ctrl)
{
struct rsxx_dma_ctrl *ctrl;
struct rsxx_dma *dma;
unsigned long flags;
u16 count;
@@ -450,7 +477,6 @@ static void rsxx_dma_done(struct work_struct *work)
u8 tag;
struct hw_status *hw_st_buf;
ctrl = container_of(work, struct rsxx_dma_ctrl, dma_done_work);
hw_st_buf = ctrl->status.buf;
if (unlikely(ctrl->card->halt) ||
@@ -520,33 +546,32 @@ static void rsxx_dma_done(struct work_struct *work)
rsxx_enable_ier(ctrl->card, CR_INTR_DMA(ctrl->id));
spin_unlock_irqrestore(&ctrl->card->irq_lock, flags);
spin_lock(&ctrl->queue_lock);
spin_lock_bh(&ctrl->queue_lock);
if (ctrl->stats.sw_q_depth)
queue_work(ctrl->issue_wq, &ctrl->issue_dma_work);
spin_unlock(&ctrl->queue_lock);
spin_unlock_bh(&ctrl->queue_lock);
}
static int rsxx_cleanup_dma_queue(struct rsxx_cardinfo *card,
struct list_head *q)
static void rsxx_schedule_issue(struct work_struct *work)
{
struct rsxx_dma *dma;
struct rsxx_dma *tmp;
int cnt = 0;
struct rsxx_dma_ctrl *ctrl;
list_for_each_entry_safe(dma, tmp, q, list) {
list_del(&dma->list);
ctrl = container_of(work, struct rsxx_dma_ctrl, issue_dma_work);
if (dma->dma_addr)
pci_unmap_page(card->dev, dma->dma_addr,
get_dma_size(dma),
(dma->cmd == HW_CMD_BLK_WRITE) ?
PCI_DMA_TODEVICE :
PCI_DMA_FROMDEVICE);
kmem_cache_free(rsxx_dma_pool, dma);
cnt++;
}
mutex_lock(&ctrl->work_lock);
rsxx_issue_dmas(ctrl);
mutex_unlock(&ctrl->work_lock);
}
return cnt;
static void rsxx_schedule_done(struct work_struct *work)
{
struct rsxx_dma_ctrl *ctrl;
ctrl = container_of(work, struct rsxx_dma_ctrl, dma_done_work);
mutex_lock(&ctrl->work_lock);
rsxx_dma_done(ctrl);
mutex_unlock(&ctrl->work_lock);
}
static int rsxx_queue_discard(struct rsxx_cardinfo *card,
@@ -698,10 +723,10 @@ int rsxx_dma_queue_bio(struct rsxx_cardinfo *card,
for (i = 0; i < card->n_targets; i++) {
if (!list_empty(&dma_list[i])) {
spin_lock(&card->ctrl[i].queue_lock);
spin_lock_bh(&card->ctrl[i].queue_lock);
card->ctrl[i].stats.sw_q_depth += dma_cnt[i];
list_splice_tail(&dma_list[i], &card->ctrl[i].queue);
spin_unlock(&card->ctrl[i].queue_lock);
spin_unlock_bh(&card->ctrl[i].queue_lock);
queue_work(card->ctrl[i].issue_wq,
&card->ctrl[i].issue_dma_work);
@@ -711,8 +736,11 @@ int rsxx_dma_queue_bio(struct rsxx_cardinfo *card,
return 0;
bvec_err:
for (i = 0; i < card->n_targets; i++)
rsxx_cleanup_dma_queue(card, &dma_list[i]);
for (i = 0; i < card->n_targets; i++) {
spin_lock_bh(&card->ctrl[i].queue_lock);
rsxx_cleanup_dma_queue(&card->ctrl[i], &dma_list[i]);
spin_unlock_bh(&card->ctrl[i].queue_lock);
}
return st;
}
@@ -780,6 +808,7 @@ static int rsxx_dma_ctrl_init(struct pci_dev *dev,
spin_lock_init(&ctrl->trackers->lock);
spin_lock_init(&ctrl->queue_lock);
mutex_init(&ctrl->work_lock);
INIT_LIST_HEAD(&ctrl->queue);
setup_timer(&ctrl->activity_timer, dma_engine_stalled,
@@ -793,8 +822,8 @@ static int rsxx_dma_ctrl_init(struct pci_dev *dev,
if (!ctrl->done_wq)
return -ENOMEM;
INIT_WORK(&ctrl->issue_dma_work, rsxx_issue_dmas);
INIT_WORK(&ctrl->dma_done_work, rsxx_dma_done);
INIT_WORK(&ctrl->issue_dma_work, rsxx_schedule_issue);
INIT_WORK(&ctrl->dma_done_work, rsxx_schedule_done);
st = rsxx_hw_buffers_init(dev, ctrl);
if (st)
@@ -918,13 +947,30 @@ failed_dma_setup:
return st;
}
int rsxx_dma_cancel(struct rsxx_dma_ctrl *ctrl)
{
struct rsxx_dma *dma;
int i;
int cnt = 0;
/* Clean up issued DMAs */
for (i = 0; i < RSXX_MAX_OUTSTANDING_CMDS; i++) {
dma = get_tracker_dma(ctrl->trackers, i);
if (dma) {
atomic_dec(&ctrl->stats.hw_q_depth);
rsxx_complete_dma(ctrl, dma, DMA_CANCELLED);
push_tracker(ctrl->trackers, i);
cnt++;
}
}
return cnt;
}
void rsxx_dma_destroy(struct rsxx_cardinfo *card)
{
struct rsxx_dma_ctrl *ctrl;
struct rsxx_dma *dma;
int i, j;
int cnt = 0;
int i;
for (i = 0; i < card->n_targets; i++) {
ctrl = &card->ctrl[i];
@@ -943,33 +989,11 @@ void rsxx_dma_destroy(struct rsxx_cardinfo *card)
del_timer_sync(&ctrl->activity_timer);
/* Clean up the DMA queue */
spin_lock(&ctrl->queue_lock);
cnt = rsxx_cleanup_dma_queue(card, &ctrl->queue);
spin_unlock(&ctrl->queue_lock);
spin_lock_bh(&ctrl->queue_lock);
rsxx_cleanup_dma_queue(ctrl, &ctrl->queue);
spin_unlock_bh(&ctrl->queue_lock);
if (cnt)
dev_info(CARD_TO_DEV(card),
"Freed %d queued DMAs on channel %d\n",
cnt, i);
/* Clean up issued DMAs */
for (j = 0; j < RSXX_MAX_OUTSTANDING_CMDS; j++) {
dma = get_tracker_dma(ctrl->trackers, j);
if (dma) {
pci_unmap_page(card->dev, dma->dma_addr,
get_dma_size(dma),
(dma->cmd == HW_CMD_BLK_WRITE) ?
PCI_DMA_TODEVICE :
PCI_DMA_FROMDEVICE);
kmem_cache_free(rsxx_dma_pool, dma);
cnt++;
}
}
if (cnt)
dev_info(CARD_TO_DEV(card),
"Freed %d pending DMAs on channel %d\n",
cnt, i);
rsxx_dma_cancel(ctrl);
vfree(ctrl->trackers);
@@ -1013,7 +1037,7 @@ int rsxx_eeh_save_issued_dmas(struct rsxx_cardinfo *card)
cnt++;
}
spin_lock(&card->ctrl[i].queue_lock);
spin_lock_bh(&card->ctrl[i].queue_lock);
list_splice(&issued_dmas[i], &card->ctrl[i].queue);
atomic_sub(cnt, &card->ctrl[i].stats.hw_q_depth);
@@ -1028,7 +1052,7 @@ int rsxx_eeh_save_issued_dmas(struct rsxx_cardinfo *card)
PCI_DMA_TODEVICE :
PCI_DMA_FROMDEVICE);
}
spin_unlock(&card->ctrl[i].queue_lock);
spin_unlock_bh(&card->ctrl[i].queue_lock);
}
kfree(issued_dmas);
@@ -1036,30 +1060,13 @@ int rsxx_eeh_save_issued_dmas(struct rsxx_cardinfo *card)
return 0;
}
void rsxx_eeh_cancel_dmas(struct rsxx_cardinfo *card)
{
struct rsxx_dma *dma;
struct rsxx_dma *tmp;
int i;
for (i = 0; i < card->n_targets; i++) {
spin_lock(&card->ctrl[i].queue_lock);
list_for_each_entry_safe(dma, tmp, &card->ctrl[i].queue, list) {
list_del(&dma->list);
rsxx_complete_dma(&card->ctrl[i], dma, DMA_CANCELLED);
}
spin_unlock(&card->ctrl[i].queue_lock);
}
}
int rsxx_eeh_remap_dmas(struct rsxx_cardinfo *card)
{
struct rsxx_dma *dma;
int i;
for (i = 0; i < card->n_targets; i++) {
spin_lock(&card->ctrl[i].queue_lock);
spin_lock_bh(&card->ctrl[i].queue_lock);
list_for_each_entry(dma, &card->ctrl[i].queue, list) {
dma->dma_addr = pci_map_page(card->dev, dma->page,
dma->pg_off, get_dma_size(dma),
@@ -1067,12 +1074,12 @@ int rsxx_eeh_remap_dmas(struct rsxx_cardinfo *card)
PCI_DMA_TODEVICE :
PCI_DMA_FROMDEVICE);
if (!dma->dma_addr) {
spin_unlock(&card->ctrl[i].queue_lock);
spin_unlock_bh(&card->ctrl[i].queue_lock);
kmem_cache_free(rsxx_dma_pool, dma);
return -ENOMEM;
}
}
spin_unlock(&card->ctrl[i].queue_lock);
spin_unlock_bh(&card->ctrl[i].queue_lock);
}
return 0;
drivers/block/rsxx/rsxx_priv.h
+9 -1
View File
@@ -39,6 +39,7 @@
#include <linux/vmalloc.h>
#include <linux/timer.h>
#include <linux/ioctl.h>
#include <linux/delay.h>
#include "rsxx.h"
#include "rsxx_cfg.h"
@@ -114,6 +115,7 @@ struct rsxx_dma_ctrl {
struct timer_list activity_timer;
struct dma_tracker_list *trackers;
struct rsxx_dma_stats stats;
struct mutex work_lock;
};
struct rsxx_cardinfo {
@@ -134,6 +136,7 @@ struct rsxx_cardinfo {
spinlock_t lock;
bool active;
struct creg_cmd *active_cmd;
struct workqueue_struct *creg_wq;
struct work_struct done_work;
struct list_head queue;
unsigned int q_depth;
@@ -154,6 +157,7 @@ struct rsxx_cardinfo {
int buf_len;
} log;
struct workqueue_struct *event_wq;
struct work_struct event_work;
unsigned int state;
u64 size8;
@@ -181,6 +185,8 @@ struct rsxx_cardinfo {
int n_targets;
struct rsxx_dma_ctrl *ctrl;
struct dentry *debugfs_dir;
};
enum rsxx_pci_regmap {
@@ -283,6 +289,7 @@ enum rsxx_creg_addr {
CREG_ADD_CAPABILITIES = 0x80001050,
CREG_ADD_LOG = 0x80002000,
CREG_ADD_NUM_TARGETS = 0x80003000,
CREG_ADD_CRAM = 0xA0000000,
CREG_ADD_CONFIG = 0xB0000000,
};
@@ -372,6 +379,8 @@ typedef void (*rsxx_dma_cb)(struct rsxx_cardinfo *card,
int rsxx_dma_setup(struct rsxx_cardinfo *card);
void rsxx_dma_destroy(struct rsxx_cardinfo *card);
int rsxx_dma_init(void);
int rsxx_cleanup_dma_queue(struct rsxx_dma_ctrl *ctrl, struct list_head *q);
int rsxx_dma_cancel(struct rsxx_dma_ctrl *ctrl);
void rsxx_dma_cleanup(void);
void rsxx_dma_queue_reset(struct rsxx_cardinfo *card);
int rsxx_dma_configure(struct rsxx_cardinfo *card);
@@ -382,7 +391,6 @@ int rsxx_dma_queue_bio(struct rsxx_cardinfo *card,
void *cb_data);
int rsxx_hw_buffers_init(struct pci_dev *dev, struct rsxx_dma_ctrl *ctrl);
int rsxx_eeh_save_issued_dmas(struct rsxx_cardinfo *card);
void rsxx_eeh_cancel_dmas(struct rsxx_cardinfo *card);
int rsxx_eeh_remap_dmas(struct rsxx_cardinfo *card);
/***** cregs.c *****/
drivers/block/xen-blkback/blkback.c
+557 -325
View File
File diff suppressed because it is too large Load Diff
drivers/block/xen-blkback/common.h
+145 -2
View File
@@ -50,6 +50,19 @@
__func__, __LINE__, ##args)
/*
* This is the maximum number of segments that would be allowed in indirect
* requests. This value will also be passed to the frontend.
*/
#define MAX_INDIRECT_SEGMENTS 256
#define SEGS_PER_INDIRECT_FRAME \
(PAGE_SIZE/sizeof(struct blkif_request_segment_aligned))
#define MAX_INDIRECT_PAGES \
((MAX_INDIRECT_SEGMENTS + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)
#define INDIRECT_PAGES(_segs) \
((_segs + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)
/* Not a real protocol. Used to generate ring structs which contain
* the elements common to all protocols only. This way we get a
* compiler-checkable way to use common struct elements, so we can
@@ -83,12 +96,31 @@ struct blkif_x86_32_request_other {
uint64_t id; /* private guest value, echoed in resp */
} __attribute__((__packed__));
struct blkif_x86_32_request_indirect {
uint8_t indirect_op;
uint16_t nr_segments;
uint64_t id;
blkif_sector_t sector_number;
blkif_vdev_t handle;
uint16_t _pad1;
grant_ref_t indirect_grefs[BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST];
/*
* The maximum number of indirect segments (and pages) that will
* be used is determined by MAX_INDIRECT_SEGMENTS, this value
* is also exported to the guest (via xenstore
* feature-max-indirect-segments entry), so the frontend knows how
* many indirect segments the backend supports.
*/
uint64_t _pad2; /* make it 64 byte aligned */
} __attribute__((__packed__));
struct blkif_x86_32_request {
uint8_t operation; /* BLKIF_OP_??? */
union {
struct blkif_x86_32_request_rw rw;
struct blkif_x86_32_request_discard discard;
struct blkif_x86_32_request_other other;
struct blkif_x86_32_request_indirect indirect;
} u;
} __attribute__((__packed__));
@@ -127,12 +159,32 @@ struct blkif_x86_64_request_other {
uint64_t id; /* private guest value, echoed in resp */
} __attribute__((__packed__));
struct blkif_x86_64_request_indirect {
uint8_t indirect_op;
uint16_t nr_segments;
uint32_t _pad1; /* offsetof(blkif_..,u.indirect.id)==8 */
uint64_t id;
blkif_sector_t sector_number;
blkif_vdev_t handle;
uint16_t _pad2;
grant_ref_t indirect_grefs[BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST];
/*
* The maximum number of indirect segments (and pages) that will
* be used is determined by MAX_INDIRECT_SEGMENTS, this value
* is also exported to the guest (via xenstore
* feature-max-indirect-segments entry), so the frontend knows how
* many indirect segments the backend supports.
*/
uint32_t _pad3; /* make it 64 byte aligned */
} __attribute__((__packed__));
struct blkif_x86_64_request {
uint8_t operation; /* BLKIF_OP_??? */
union {
struct blkif_x86_64_request_rw rw;
struct blkif_x86_64_request_discard discard;
struct blkif_x86_64_request_other other;
struct blkif_x86_64_request_indirect indirect;
} u;
} __attribute__((__packed__));
@@ -182,12 +234,26 @@ struct xen_vbd {
struct backend_info;
/* Number of available flags */
#define PERSISTENT_GNT_FLAGS_SIZE 2
/* This persistent grant is currently in use */
#define PERSISTENT_GNT_ACTIVE 0
/*
* This persistent grant has been used, this flag is set when we remove the
* PERSISTENT_GNT_ACTIVE, to know that this grant has been used recently.
*/
#define PERSISTENT_GNT_WAS_ACTIVE 1
/* Number of requests that we can fit in a ring */
#define XEN_BLKIF_REQS 32
struct persistent_gnt {
struct page *page;
grant_ref_t gnt;
grant_handle_t handle;
DECLARE_BITMAP(flags, PERSISTENT_GNT_FLAGS_SIZE);
struct rb_node node;
struct list_head remove_node;
};
struct xen_blkif {
@@ -219,6 +285,23 @@ struct xen_blkif {
/* tree to store persistent grants */
struct rb_root persistent_gnts;
unsigned int persistent_gnt_c;
atomic_t persistent_gnt_in_use;
unsigned long next_lru;
/* used by the kworker that offload work from the persistent purge */
struct list_head persistent_purge_list;
struct work_struct persistent_purge_work;
/* buffer of free pages to map grant refs */
spinlock_t free_pages_lock;
int free_pages_num;
struct list_head free_pages;
/* List of all 'pending_req' available */
struct list_head pending_free;
/* And its spinlock. */
spinlock_t pending_free_lock;
wait_queue_head_t pending_free_wq;
/* statistics */
unsigned long st_print;
@@ -231,6 +314,41 @@ struct xen_blkif {
unsigned long long st_wr_sect;
wait_queue_head_t waiting_to_free;
/* Thread shutdown wait queue. */
wait_queue_head_t shutdown_wq;
};
struct seg_buf {
unsigned long offset;
unsigned int nsec;
};
struct grant_page {
struct page *page;
struct persistent_gnt *persistent_gnt;
grant_handle_t handle;
grant_ref_t gref;
};
/*
* Each outstanding request that we've passed to the lower device layers has a
* 'pending_req' allocated to it. Each buffer_head that completes decrements
* the pendcnt towards zero. When it hits zero, the specified domain has a
* response queued for it, with the saved 'id' passed back.
*/
struct pending_req {
struct xen_blkif *blkif;
u64 id;
int nr_pages;
atomic_t pendcnt;
unsigned short operation;
int status;
struct list_head free_list;
struct grant_page *segments[MAX_INDIRECT_SEGMENTS];
/* Indirect descriptors */
struct grant_page *indirect_pages[MAX_INDIRECT_PAGES];
struct seg_buf seg[MAX_INDIRECT_SEGMENTS];
struct bio *biolist[MAX_INDIRECT_SEGMENTS];
};
@@ -257,6 +375,7 @@ int xen_blkif_xenbus_init(void);
irqreturn_t xen_blkif_be_int(int irq, void *dev_id);
int xen_blkif_schedule(void *arg);
int xen_blkif_purge_persistent(void *arg);
int xen_blkbk_flush_diskcache(struct xenbus_transaction xbt,
struct backend_info *be, int state);
@@ -268,7 +387,7 @@ struct xenbus_device *xen_blkbk_xenbus(struct backend_info *be);
static inline void blkif_get_x86_32_req(struct blkif_request *dst,
struct blkif_x86_32_request *src)
{
int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST;
int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j;
dst->operation = src->operation;
switch (src->operation) {
case BLKIF_OP_READ:
@@ -291,6 +410,18 @@ static inline void blkif_get_x86_32_req(struct blkif_request *dst,
dst->u.discard.sector_number = src->u.discard.sector_number;
dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
break;
case BLKIF_OP_INDIRECT:
dst->u.indirect.indirect_op = src->u.indirect.indirect_op;
dst->u.indirect.nr_segments = src->u.indirect.nr_segments;
dst->u.indirect.handle = src->u.indirect.handle;
dst->u.indirect.id = src->u.indirect.id;
dst->u.indirect.sector_number = src->u.indirect.sector_number;
barrier();
j = min(MAX_INDIRECT_PAGES, INDIRECT_PAGES(dst->u.indirect.nr_segments));
for (i = 0; i < j; i++)
dst->u.indirect.indirect_grefs[i] =
src->u.indirect.indirect_grefs[i];
break;
default:
/*
* Don't know how to translate this op. Only get the
@@ -304,7 +435,7 @@ static inline void blkif_get_x86_32_req(struct blkif_request *dst,
static inline void blkif_get_x86_64_req(struct blkif_request *dst,
struct blkif_x86_64_request *src)
{
int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST;
int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j;
dst->operation = src->operation;
switch (src->operation) {
case BLKIF_OP_READ:
@@ -327,6 +458,18 @@ static inline void blkif_get_x86_64_req(struct blkif_request *dst,
dst->u.discard.sector_number = src->u.discard.sector_number;
dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
break;
case BLKIF_OP_INDIRECT:
dst->u.indirect.indirect_op = src->u.indirect.indirect_op;
dst->u.indirect.nr_segments = src->u.indirect.nr_segments;
dst->u.indirect.handle = src->u.indirect.handle;
dst->u.indirect.id = src->u.indirect.id;
dst->u.indirect.sector_number = src->u.indirect.sector_number;
barrier();
j = min(MAX_INDIRECT_PAGES, INDIRECT_PAGES(dst->u.indirect.nr_segments));
for (i = 0; i < j; i++)
dst->u.indirect.indirect_grefs[i] =
src->u.indirect.indirect_grefs[i];
break;
default:
/*
* Don't know how to translate this op. Only get the
drivers/block/xen-blkback/xenbus.c
+85
View File
@@ -98,12 +98,17 @@ static void xen_update_blkif_status(struct xen_blkif *blkif)
err = PTR_ERR(blkif->xenblkd);
blkif->xenblkd = NULL;
xenbus_dev_error(blkif->be->dev, err, "start xenblkd");
return;
}
}
static struct xen_blkif *xen_blkif_alloc(domid_t domid)
{
struct xen_blkif *blkif;
struct pending_req *req, *n;
int i, j;
BUILD_BUG_ON(MAX_INDIRECT_PAGES > BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST);
blkif = kmem_cache_zalloc(xen_blkif_cachep, GFP_KERNEL);
if (!blkif)
@@ -118,8 +123,57 @@ static struct xen_blkif *xen_blkif_alloc(domid_t domid)
blkif->st_print = jiffies;
init_waitqueue_head(&blkif->waiting_to_free);
blkif->persistent_gnts.rb_node = NULL;
spin_lock_init(&blkif->free_pages_lock);
INIT_LIST_HEAD(&blkif->free_pages);
blkif->free_pages_num = 0;
atomic_set(&blkif->persistent_gnt_in_use, 0);
INIT_LIST_HEAD(&blkif->pending_free);
for (i = 0; i < XEN_BLKIF_REQS; i++) {
req = kzalloc(sizeof(*req), GFP_KERNEL);
if (!req)
goto fail;
list_add_tail(&req->free_list,
&blkif->pending_free);
for (j = 0; j < MAX_INDIRECT_SEGMENTS; j++) {
req->segments[j] = kzalloc(sizeof(*req->segments[0]),
GFP_KERNEL);
if (!req->segments[j])
goto fail;
}
for (j = 0; j < MAX_INDIRECT_PAGES; j++) {
req->indirect_pages[j] = kzalloc(sizeof(*req->indirect_pages[0]),
GFP_KERNEL);
if (!req->indirect_pages[j])
goto fail;
}
}
spin_lock_init(&blkif->pending_free_lock);
init_waitqueue_head(&blkif->pending_free_wq);
init_waitqueue_head(&blkif->shutdown_wq);
return blkif;
fail:
list_for_each_entry_safe(req, n, &blkif->pending_free, free_list) {
list_del(&req->free_list);
for (j = 0; j < MAX_INDIRECT_SEGMENTS; j++) {
if (!req->segments[j])
break;
kfree(req->segments[j]);
}
for (j = 0; j < MAX_INDIRECT_PAGES; j++) {
if (!req->indirect_pages[j])
break;
kfree(req->indirect_pages[j]);
}
kfree(req);
}
kmem_cache_free(xen_blkif_cachep, blkif);
return ERR_PTR(-ENOMEM);
}
static int xen_blkif_map(struct xen_blkif *blkif, unsigned long shared_page,
@@ -178,6 +232,7 @@ static void xen_blkif_disconnect(struct xen_blkif *blkif)
{
if (blkif->xenblkd) {
kthread_stop(blkif->xenblkd);
wake_up(&blkif->shutdown_wq);
blkif->xenblkd = NULL;
}
@@ -198,8 +253,28 @@ static void xen_blkif_disconnect(struct xen_blkif *blkif)
static void xen_blkif_free(struct xen_blkif *blkif)
{
struct pending_req *req, *n;
int i = 0, j;
if (!atomic_dec_and_test(&blkif->refcnt))
BUG();
/* Check that there is no request in use */
list_for_each_entry_safe(req, n, &blkif->pending_free, free_list) {
list_del(&req->free_list);
for (j = 0; j < MAX_INDIRECT_SEGMENTS; j++)
kfree(req->segments[j]);
for (j = 0; j < MAX_INDIRECT_PAGES; j++)
kfree(req->indirect_pages[j]);
kfree(req);
i++;
}
WARN_ON(i != XEN_BLKIF_REQS);
kmem_cache_free(xen_blkif_cachep, blkif);
}
@@ -678,6 +753,11 @@ again:
dev->nodename);
goto abort;
}
err = xenbus_printf(xbt, dev->nodename, "feature-max-indirect-segments", "%u",
MAX_INDIRECT_SEGMENTS);
if (err)
dev_warn(&dev->dev, "writing %s/feature-max-indirect-segments (%d)",
dev->nodename, err);
err = xenbus_printf(xbt, dev->nodename, "sectors", "%llu",
(unsigned long long)vbd_sz(&be->blkif->vbd));
@@ -704,6 +784,11 @@ again:
dev->nodename);
goto abort;
}
err = xenbus_printf(xbt, dev->nodename, "physical-sector-size", "%u",
bdev_physical_block_size(be->blkif->vbd.bdev));
if (err)
xenbus_dev_error(dev, err, "writing %s/physical-sector-size",
dev->nodename);
err = xenbus_transaction_end(xbt, 0);
if (err == -EAGAIN)
drivers/block/xen-blkfront.c
+432 -100
View File
File diff suppressed because it is too large Load Diff

Some files were not shown because too many files have changed in this diff Show More