Commits · 908ce21402426d8b844580363e6d6152a8849ee0 · linux / linux-davinci

14 Sep, 2009 6 commits

Credits where credits are due · 908ce214

Philipp Reisner authored Sep 14, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

908ce214

fix regression in 8.3.3rc, some typos: bio_flagged != bio_rw_flagged · c3361b7a

Lars Ellenberg authored Sep 14, 2009

This may have led to spurious "IO Errors" on a receiving node, when
barriers are not disabled, and not supported by the lower level device.

regression was introduced by:
commit a2ad6507a0b5a2dc1f5a6a823b9a28d9c4430200
Author: Lars Ellenberg <lars.ellenberg@linbit.com>
Date:   Thu Aug 27 20:40:39 2009 +0200

    fix potential endless retry loop on IO error
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

c3361b7a

give more detail about IO errors, if possible · 5910fc51

Lars Ellenberg authored Sep 14, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

5910fc51

Cosmetic changes · 240cbf21

Lars Ellenberg authored Sep 14, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

240cbf21

Add rate limit to error message, in case resync is aborted due to disk failure · a509c934
Lars Ellenberg authored Sep 14, 2009
```
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
```
a509c934

Preparing 8.3.3rc2 · 75841257

Philipp Reisner authored Sep 10, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

75841257

09 Sep, 2009 10 commits

Detect write errors of the sync_uuid during the next sync_handshake [Bugz 254 2/2] · 3b039b20
Philipp Reisner authored Sep 09, 2009
```
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
```
3b039b20

Abort waiting for sync_uuid if the disk goes away [Bugz 254 1/2] · 09b95ef0

Philipp Reisner authored Sep 09, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

09b95ef0

fix micro memleak in some error path · 9d374cd4

Lars Ellenberg authored Sep 08, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

9d374cd4

Fixing integer encoding in tag lists from kernel to userspace (reg in 8.3.2) [bugz 253] · 2a780782
Philipp Reisner authored Sep 08, 2009
```
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
```
2a780782

Send a ping if we have an empty (0 blocks) resync. [Bugz 251] · 4033e7f1

Philipp Reisner authored Sep 08, 2009

  That is necessary to detect the loss of connection before doing
  the after resync UUID modifications. The bug was usually triggered
  by failing before-resync-target handlers.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

4033e7f1

do not retry promotion after peer fencing failed once · d1bef1c4

Lars Ellenberg authored Sep 04, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

d1bef1c4

use GFP_NOIO when broadcasting events and state changes · ac2a2ea0

Lars Ellenberg authored Sep 01, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

ac2a2ea0

on disconnect, just give up all references on pages from net_ee · 5e8d125b

Lars Ellenberg authored Sep 01, 2009

no need to wait for the network stack to get its act together
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

5e8d125b

drbd_set_role needs to bail out if the state change failed · 5ae78ce7

Lars Ellenberg authored Aug 28, 2009

fix for bug#252
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

5ae78ce7

Preparing 8.3.3rc1 · 7cc08cdb

Philipp Reisner authored Aug 27, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

7cc08cdb

28 Aug, 2009 9 commits

micro spelling fixes · e6200c6b

Lars Ellenberg authored Aug 27, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

e6200c6b

wait for pending after-state-change work and similar before attaching · ae4c8d37

Lars Ellenberg authored Aug 27, 2009

Otherwise, if we do a very quick detach/attach, and the
after-state-change send_state() of the detach got delayed for some
reason, it may be processed after the new diskless->attaching state
change, probably confusing the peer.

Paranoia fix, not yet observed in real life.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

ae4c8d37

introduce drbd_flush_workqueue · e9df68b6

Lars Ellenberg authored Aug 27, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

e9df68b6

fix potential endless retry loop on IO error · 73e931cc

Lars Ellenberg authored Aug 27, 2009

On barrier request on receiving node.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

73e931cc

cleanup: Removed _drbd_clear_done_ee() · 2cbe330b

Philipp Reisner authored Aug 27, 2009

Because _drbd_clear_done_ee() was just an acient special case of
drbd_process_done_ee().
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

2cbe330b

Whitespace fixes · fcfbfec2

Philipp Reisner authored Aug 27, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

fcfbfec2

Do not consider peer dead when fence-peer handler can not reach it and we are < UpToDate [Bugz 198] · edf80e69

Philipp Reisner authored Aug 25, 2009

Unfortunately the use needs to give the dead-time of his cluster stack as
argument to the crm-fence-peer.sh script now.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

edf80e69

Removed drbd_io_error(), which predated the time of the worker based after_state_ch() [Bugz 224] · 8f5697e9

Philipp Reisner authored Aug 24, 2009

Besides removing some lines of code, this removes also a deadlock.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

8f5697e9

Fix spelling errors. · ff940fd4

Florian Haas authored Aug 19, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

ff940fd4

19 Aug, 2009 15 commits

linux-next: drbd tree build failure · 83f2029c

Stephen Rothwell authored Jul 29, 2009

Today's linux-next build (x86_64 allmodconfig) failed like this:

drivers/block/drbd/drbd_nl.c: In function 'drbd_setup_queue_param':
drivers/block/drbd/drbd_nl.c:707: error: implicit declaration of function 'blk_queue_stack_limits'

Caused by commit 6dc986e736ca1e76a45d025a920f3a66855fc2aa ("block:
Deprecate blk_queue_stack_limits") from the block tree.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>

83f2029c

cosmetic: add some static to match the respective declaration · b6a06601

Lars Ellenberg authored Aug 17, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

b6a06601

New protocol necessary, since drbd_uuid_compare() has new rules · 1e03325c

Philipp Reisner authored Aug 18, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

1e03325c

drbd_uuid_compare(): Handle loss of last P_WRITE_ACK packet of a resync right.... · bc9ef10b

Philipp Reisner authored Aug 18, 2009

drbd_uuid_compare(): Handle loss of last P_WRITE_ACK packet of a resync right. (Caused missing resyncs) Bugz 246

Connection drop while transmitting last ack:
SyncSource losses connection, SyncTarget sees the end of resync.

Aug 18 08:39:42 uml1 drbd0: Handshake successful: Agreed network protocol version 90
Aug 18 08:39:42 uml1 drbd0: conn( WFConnection -> WFReportParams )
Aug 18 08:39:42 uml1 drbd0: drbd_sync_handshake:
Aug 18 08:39:42 uml1 drbd0: self 81DAF2FF6134FC1E:16EF5753AD5FA994:95B9E9AD329C137B:A4B1B25AC5927436 bits:4255 flags:0
Aug 18 08:39:42 uml1 drbd0: peer 16EF5753AD5FA994:0000000000000000:95B9E9AD329C137A:A4B1B25AC5927436 bits:0 flags:0
Aug 18 08:39:42 uml1 drbd0: uuid_compare()=1 by rule 70
Aug 18 08:39:42 uml1 drbd0: peer( Unknown -> Secondary ) conn( WFReportParams -> WFBitMapS ) pdsk( Outdated -> UpToDate )
Aug 18 08:39:42 uml1 drbd0: conn( WFBitMapS -> SyncSource ) pdsk( UpToDate -> Inconsistent )
Aug 18 08:39:42 uml1 drbd0: Began resync as SyncSource (will sync 17020 KB [4255 bits set]).
Aug 18 08:39:43 uml1 drbd0: peer( Secondary -> Unknown ) conn( SyncSource -> Disconnecting )

Aug 18 08:39:42 uml2 drbd0: Handshake successful: Agreed network protocol version 90
Aug 18 08:39:42 uml2 drbd0: conn( WFConnection -> WFReportParams )
Aug 18 08:39:42 uml2 drbd0: drbd_sync_handshake:
Aug 18 08:39:42 uml2 drbd0: self 16EF5753AD5FA994:0000000000000000:95B9E9AD329C137A:A4B1B25AC5927436 bits:0 flags:0
Aug 18 08:39:42 uml2 drbd0: peer 81DAF2FF6134FC1E:16EF5753AD5FA994:95B9E9AD329C137B:A4B1B25AC5927436 bits:4255 flags:0
Aug 18 08:39:42 uml2 drbd0: uuid_compare()=-1 by rule 50
Aug 18 08:39:42 uml2 drbd0: peer( Unknown -> Secondary ) conn( WFReportParams -> WFBitMapT ) pdsk( DUnknown -> UpToDate )
Aug 18 08:39:42 uml2 drbd0: conn( WFBitMapT -> WFSyncUUID )
Aug 18 08:39:42 uml2 drbd0: conn( WFSyncUUID -> SyncTarget ) disk( UpToDate -> Inconsistent )
Aug 18 08:39:43 uml2 drbd0: conn( SyncTarget -> Connected ) disk( Inconsistent -> UpToDate )

Only uml2 recognised the end of resync.

Aug 18 09:49:51 uml1 drbd0: Handshake successful: Agreed network protocol version 90
Aug 18 09:49:51 uml1 drbd0: conn( WFConnection -> WFReportParams )
Aug 18 09:49:51 uml1 drbd0: drbd_sync_handshake:
Aug 18 09:49:51 uml1 drbd0: self 81DAF2FF6134FC1E:CB7A2BEB83B25C28:16EF5753AD5FA994:95B9E9AD329C137B bits:3 flags:0
Aug 18 09:49:51 uml1 drbd0: peer 81DAF2FF6134FC1E:0000000000000000:CB7A2BEB83B25C28:16EF5753AD5FA994 bits:0 flags:0
Aug 18 09:49:51 uml1 drbd0: uuid_compare()=0 by rule 40
Aug 18 09:49:51 uml1 drbd0: No resync, but 3 bits in bitmap!
Aug 18 09:49:51 uml1 drbd0: peer( Unknown -> Secondary ) conn( WFReportParams -> Connected ) pdsk( Inconsistent -> UpToDate )

Aug 18 09:49:51 uml2 drbd0: Handshake successful: Agreed network protocol version 90
Aug 18 09:49:51 uml2 drbd0: conn( WFConnection -> WFReportParams )
Aug 18 09:49:51 uml2 drbd0: drbd_sync_handshake:
Aug 18 09:49:51 uml2 drbd0: self 81DAF2FF6134FC1E:0000000000000000:CB7A2BEB83B25C28:16EF5753AD5FA994 bits:0 flags:0
Aug 18 09:49:51 uml2 drbd0: peer 81DAF2FF6134FC1E:CB7A2BEB83B25C28:16EF5753AD5FA994:95B9E9AD329C137B bits:3 flags:0
Aug 18 09:49:51 uml2 drbd0: uuid_compare()=0 by rule 40
Aug 18 09:49:51 uml2 drbd0: peer( Unknown -> Secondary ) conn( WFReportParams -> Connected ) pdsk( DUnknown -> UpToDate )

=> No resync, but 3 bits in bitmap! message on uml1.

rule 3.4:
  If Cs = Cp & Bs != 0 & Bp = 0 & Bs = H1p & H1s = H2p
 => I have not realized end of resync. I was SyncSource, target saw the end of resync.

    Correct my UUIDs: Bs = 0 (with rotate)

rule 3.5:
  If Cs = Cp & Bs = 0 & Bp != 0 & H1s = Bp & H2s = H1p
 => Peer has not realized end of resync. I was SyncTarget, resync is actually done.

    Correct peer's UUIDS: Bp = 0 (with rotate)
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

bc9ef10b

cosmetic: use container_of() instead of explicit type casts from (struct drbd_work*) · ff34dbb0
Lars Ellenberg authored Aug 17, 2009
```
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
```
ff34dbb0

fix ldev reference count imbalance on fault injection · eb9a85f8

Lars Ellenberg authored Aug 17, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

eb9a85f8

consolidate err_out path in receive_DataRequest · 08c8caf6

Lars Ellenberg authored Aug 17, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

08c8caf6

fix theoretical imbalance in ldev refcount · 8089af76

Lars Ellenberg authored Aug 17, 2009

never observed, unlikely to be real. Still a coding bug.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

8089af76

drbd_uuid_compare(): Also undo the changes of last unsuccessful start of... · 794112a0

Philipp Reisner authored Aug 17, 2009

drbd_uuid_compare(): Also undo the changes of last unsuccessful start of resync also on the peer's UUIDs
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

794112a0

drbd_uuid_compare(): Do not full sync in case a P_SYNC_UUID packet gets lost. · 5600d7d4

Philipp Reisner authored Aug 15, 2009

Aditional errata to
http://www.drbd.org/fileadmin/drbd/publications/drbd8.pdf

5 Data Generation UUIDs

Algorithm:

Rule 5a:
Cs = H1p & H1s = H2p ==>
Connection was lost before SyncUUID Packet came through. Become Sync target.

Rule 7a:
Cp = H1s & H1p = H2s ==>
Connection was lost before SyncUUID Packet came through. Correct my UUIDs:
B = H1
H1 = H2
H2 = 0
Become Sync source.

Here are the relevant log lines, showing the issue:

Aug 12 16:20:48 garcon1 kernel: [4941237.376998] drbd10: self 68CFBC7C5C0E6D4F:1C4BD1C2E7B77E75:F1812682E82B178A:3E381E47943D1E4B bits:85 flags:0
Aug 12 16:20:48 garcon1 kernel: [4941237.377003] drbd10: peer 1C4BD1C2E7B77E74:0000000000000000:F1812682E82B178A:3E381E47943D1E4B bits:0 flags:0
Aug 12 16:20:48 garcon1 kernel: [4941237.377006] drbd10: uuid_compare()=1 by rule 7
Aug 12 16:20:48 garcon1 kernel: [4941237.378131] drbd10: peer( Unknown -> Secondary ) conn( WFReportParams -> WFBitMapS ) pdsk( DUnknown -> UpToDate )
Aug 12 16:20:50 garcon1 kernel: [4941238.911877] drbd10: conn( WFBitMapS -> SyncSource ) pdsk( UpToDate -> Inconsistent )
Aug 12 16:20:50 garcon1 kernel: [4941238.911889] drbd10: Began resync as SyncSource (will sync 340 KB [85 bits set]).
Aug 12 16:21:10 garcon1 kernel: [4941259.500053] drbd10: peer( Secondary -> Unknown ) conn( SyncSource -> NetworkFailure )

Aug 12 07:20:48 portman kernel: [11341861.890379] drbd10: self 1C4BD1C2E7B77E74:0000000000000000:F1812682E82B178A:3E381E47943D1E4B bits:0 flags:0
Aug 12 07:20:48 portman kernel: [11341861.890379] drbd10: peer 68CFBC7C5C0E6D4F:1C4BD1C2E7B77E75:F1812682E82B178A:3E381E47943D1E4B bits:85 flags:0
Aug 12 07:20:48 portman kernel: [11341861.890379] drbd10: uuid_compare()=-1 by rule 5
Aug 12 07:20:48 portman kernel: [11341861.890379] drbd10: peer( Unknown -> Primary ) conn( WFReportParams -> WFBitMapT ) pdsk( DUnknown -> UpToDate )
Aug 12 07:20:48 portman kernel: [11341862.146623] drbd10: conn( WFBitMapT -> WFSyncUUID )
Aug 12 07:21:10 portman kernel: [11341884.923517] drbd10: peer( Primary -> Unknown ) conn( WFSyncUUID -> NetworkFailure ) pdsk( UpToDate -> DUnknown )

Aug 12 16:21:34 garcon1 kernel: [4941283.720095] drbd10: self 68CFBC7C5C0E6D4F:6AA2B905AD177D20:1C4BD1C2E7B77E75:F1812682E82B178A bits:91 flags:0
Aug 12 16:21:34 garcon1 kernel: [4941283.720099] drbd10: peer 1C4BD1C2E7B77E74:0000000000000000:F1812682E82B178A:3E381E47943D1E4B bits:85 flags:0
Aug 12 16:21:34 garcon1 kernel: [4941283.720102] drbd10: uuid_compare()=2 by rule 8
Aug 12 16:21:34 garcon1 kernel: [4941283.720105] drbd10: Writing the whole bitmap, full sync required after drbd_sync_handshake.

Aug 12 07:22:08 portman kernel: [11341943.908341] drbd10: self 1C4BD1C2E7B77E74:0000000000000000:F1812682E82B178A:3E381E47943D1E4B bits:85 flags:0
Aug 12 07:22:08 portman kernel: [11341943.908390] drbd10: peer 68CFBC7C5C0E6D4F:6AA2B905AD177D20:1C4BD1C2E7B77E75:F1812682E82B178A bits:7864320 flags:0
Aug 12 07:22:08 portman kernel: [11341943.908437] drbd10: uuid_compare()=-2 by rule 6
Aug 12 07:22:08 portman kernel: [11341943.908461] drbd10: Writing the whole bitmap, full sync required after drbd_sync_handshake.
Aug 12 07:22:08 portman kernel: [11341943.925199] drbd10: 30 GB (7864320 bits) marked out-of-sync by on disk bit-map.
Aug 12 07:22:08 portman kernel: [11341943.925199] drbd10: peer( Unknown -> Primary ) conn( WFReportParams -> WFBitMapT ) pdsk( DUnknown -> UpToDate )
Aug 12 07:22:09 portman kernel: [11341944.896914] drbd10: conn( WFBitMapT -> WFSyncUUID )
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

5600d7d4

update comment: you must not hold the req_lock in drbd_free_ee · 07e7f673

Lars Ellenberg authored Aug 14, 2009

You do not NEED to hold the req_lock since
107a7ca67f9fffd0c0c94018f5a1f61a0afe7bf8 (2005-03-15).
Since the drbd_pp_lock is not spin_lock_irqsave anymore, you MUST NOT
hold the req_lock when trying to aquire the drbd_pp_lock.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

07e7f673

move drbd_free_ee outside of req_lock · ebfcb47b

Lars Ellenberg authored Aug 14, 2009

fixes recently introduced potential spinlock deadlock.
reduces the size of various critical sections.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

ebfcb47b

don't hardcode the timeout for the handshake packet · 1cec97ee

Lars Ellenberg authored Aug 13, 2009

2 seconds may be too small for high latency long distance medium packet loss links
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

1cec97ee

fix potential socket leak, addendum · bc870666

Lars Ellenberg authored Aug 12, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

bc870666

out-of-spinlock completion of master bio · 22a15d49

Lars Ellenberg authored Aug 12, 2009

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

22a15d49