drbd: introduce stop-sector to online verify
We now can schedule only a specific range of sectors for online verify,
or interrupt a running verify without interrupting the connection.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h
index 9a6d3a4..3cce735 100644
--- a/drivers/block/drbd/drbd_int.h
+++ b/drivers/block/drbd/drbd_int.h
@@ -1052,6 +1052,7 @@
/* where does the admin want us to start? (sector) */
sector_t ov_start_sector;
+ sector_t ov_stop_sector;
/* where are we now? (sector) */
sector_t ov_position;
/* Start sector of out of sync range (to merge printk reporting). */
diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c
index dfa08b7..df9965d 100644
--- a/drivers/block/drbd/drbd_main.c
+++ b/drivers/block/drbd/drbd_main.c
@@ -1231,13 +1231,15 @@
wake_up(&mdev->misc_wait);
wake_up(&mdev->state_wait);
- /* aborted verify run. log the last position */
+ /* Aborted verify run, or we reached the stop sector.
+ * Log the last position, unless end-of-device. */
if ((os.conn == C_VERIFY_S || os.conn == C_VERIFY_T) &&
- ns.conn < C_CONNECTED) {
+ ns.conn <= C_CONNECTED) {
mdev->ov_start_sector =
BM_BIT_TO_SECT(drbd_bm_bits(mdev) - mdev->ov_left);
- dev_info(DEV, "Online Verify reached sector %llu\n",
- (unsigned long long)mdev->ov_start_sector);
+ if (mdev->ov_left)
+ dev_info(DEV, "Online Verify reached sector %llu\n",
+ (unsigned long long)mdev->ov_start_sector);
}
if ((os.conn == C_PAUSED_SYNC_T || os.conn == C_PAUSED_SYNC_S) &&
@@ -1703,6 +1705,13 @@
if (os.disk < D_UP_TO_DATE && os.conn >= C_SYNC_SOURCE && ns.conn == C_CONNECTED)
drbd_send_state(mdev, ns);
+ /* Verify finished, or reached stop sector. Peer did not know about
+ * the stop sector, and we may even have changed the stop sector during
+ * verify to interrupt/stop early. Send the new state. */
+ if (os.conn == C_VERIFY_S && ns.conn == C_CONNECTED
+ && mdev->agreed_pro_version >= 97)
+ drbd_send_state(mdev, ns);
+
/* Wake up role changes, that were delayed because of connection establishing */
if (os.conn == C_WF_REPORT_PARAMS && ns.conn != C_WF_REPORT_PARAMS) {
clear_bit(STATE_SENT, &mdev->flags);
diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c
index ab66055..e2d368f 100644
--- a/drivers/block/drbd/drbd_nl.c
+++ b/drivers/block/drbd/drbd_nl.c
@@ -2211,8 +2211,10 @@
struct drbd_nl_cfg_reply *reply)
{
/* default to resume from last known position, if possible */
- struct start_ov args =
- { .start_sector = mdev->ov_start_sector };
+ struct start_ov args = {
+ .start_sector = mdev->ov_start_sector,
+ .stop_sector = ULLONG_MAX,
+ };
if (!start_ov_from_tags(mdev, nlp->tag_list, &args)) {
reply->ret_code = ERR_MANDATORY_TAG;
@@ -2224,8 +2226,9 @@
drbd_suspend_io(mdev);
wait_event(mdev->misc_wait, !test_bit(BITMAP_IO, &mdev->flags));
- /* w_make_ov_request expects position to be aligned */
- mdev->ov_start_sector = args.start_sector & ~BM_SECT_PER_BIT;
+ /* w_make_ov_request expects start position to be aligned */
+ mdev->ov_start_sector = args.start_sector & ~(BM_SECT_PER_BIT-1);
+ mdev->ov_stop_sector = args.stop_sector;
reply->ret_code = drbd_request_state(mdev,NS(conn,C_VERIFY_S));
drbd_resume_io(mdev);
return 0;
diff --git a/drivers/block/drbd/drbd_proc.c b/drivers/block/drbd/drbd_proc.c
index 5496104..a5a453b 100644
--- a/drivers/block/drbd/drbd_proc.c
+++ b/drivers/block/drbd/drbd_proc.c
@@ -167,18 +167,24 @@
* we convert to sectors in the display below. */
unsigned long bm_bits = drbd_bm_bits(mdev);
unsigned long bit_pos;
+ unsigned long long stop_sector = 0;
if (mdev->state.conn == C_VERIFY_S ||
- mdev->state.conn == C_VERIFY_T)
+ mdev->state.conn == C_VERIFY_T) {
bit_pos = bm_bits - mdev->ov_left;
- else
+ if (mdev->agreed_pro_version >= 97)
+ stop_sector = mdev->ov_stop_sector;
+ } else
bit_pos = mdev->bm_resync_fo;
/* Total sectors may be slightly off for oddly
* sized devices. So what. */
seq_printf(seq,
- "\t%3d%% sector pos: %llu/%llu\n",
+ "\t%3d%% sector pos: %llu/%llu",
(int)(bit_pos / (bm_bits/100+1)),
(unsigned long long)bit_pos * BM_SECT_PER_BIT,
(unsigned long long)bm_bits * BM_SECT_PER_BIT);
+ if (stop_sector != 0 && stop_sector != ULLONG_MAX)
+ seq_printf(seq, " stop sector: %llu", stop_sector);
+ seq_printf(seq, "\n");
}
}
diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c
index 434adf7..280735d 100644
--- a/drivers/block/drbd/drbd_receiver.c
+++ b/drivers/block/drbd/drbd_receiver.c
@@ -3255,6 +3255,14 @@
}
}
+ /* explicit verify finished notification, stop sector reached. */
+ if (os.conn == C_VERIFY_T && os.disk == D_UP_TO_DATE &&
+ peer_state.conn == C_CONNECTED && real_peer_disk == D_UP_TO_DATE) {
+ ov_oos_print(mdev);
+ drbd_resync_finished(mdev);
+ return true;
+ }
+
/* peer says his disk is inconsistent, while we think it is uptodate,
* and this happens while the peer still thinks we have a sync going on,
* but we think we are already done with the sync.
diff --git a/drivers/block/drbd/drbd_worker.c b/drivers/block/drbd/drbd_worker.c
index 6bce2cc..1352455 100644
--- a/drivers/block/drbd/drbd_worker.c
+++ b/drivers/block/drbd/drbd_worker.c
@@ -691,6 +691,7 @@
int number, i, size;
sector_t sector;
const sector_t capacity = drbd_get_capacity(mdev->this_bdev);
+ bool stop_sector_reached = false;
if (unlikely(cancel))
return 1;
@@ -699,9 +700,17 @@
sector = mdev->ov_position;
for (i = 0; i < number; i++) {
- if (sector >= capacity) {
+ if (sector >= capacity)
return 1;
- }
+
+ /* We check for "finished" only in the reply path:
+ * w_e_end_ov_reply().
+ * We need to send at least one request out. */
+ stop_sector_reached = i > 0
+ && mdev->agreed_pro_version >= 97
+ && sector >= mdev->ov_stop_sector;
+ if (stop_sector_reached)
+ break;
size = BM_BLOCK_SIZE;
@@ -725,7 +734,8 @@
requeue:
mdev->rs_in_flight += (i << (BM_BLOCK_SHIFT - 9));
- mod_timer(&mdev->resync_timer, jiffies + SLEEP_TIME);
+ if (i == 0 || !stop_sector_reached)
+ mod_timer(&mdev->resync_timer, jiffies + SLEEP_TIME);
return 1;
}
@@ -808,7 +818,12 @@
dt = (jiffies - mdev->rs_start - mdev->rs_paused) / HZ;
if (dt <= 0)
dt = 1;
+
db = mdev->rs_total;
+ /* adjust for verify start and stop sectors, respective reached position */
+ if (mdev->state.conn == C_VERIFY_S || mdev->state.conn == C_VERIFY_T)
+ db -= mdev->ov_left;
+
dbdt = Bit2KB(db/dt);
mdev->rs_paused /= HZ;
@@ -831,7 +846,7 @@
ns.conn = C_CONNECTED;
dev_info(DEV, "%s done (total %lu sec; paused %lu sec; %lu K/sec)\n",
- verify_done ? "Online verify " : "Resync",
+ verify_done ? "Online verify" : "Resync",
dt + mdev->rs_paused, mdev->rs_paused, dbdt);
n_oos = drbd_bm_total_weight(mdev);
@@ -912,7 +927,9 @@
mdev->rs_total = 0;
mdev->rs_failed = 0;
mdev->rs_paused = 0;
- if (verify_done)
+
+ /* reset start sector, if we reached end of device */
+ if (verify_done && mdev->ov_left == 0)
mdev->ov_start_sector = 0;
drbd_md_sync(mdev);
@@ -1158,6 +1175,7 @@
unsigned int size = e->size;
int digest_size;
int ok, eq = 0;
+ bool stop_sector_reached = false;
if (unlikely(cancel)) {
drbd_free_ee(mdev, e);
@@ -1208,7 +1226,10 @@
if ((mdev->ov_left & 0x200) == 0x200)
drbd_advance_rs_marks(mdev, mdev->ov_left);
- if (mdev->ov_left == 0) {
+ stop_sector_reached = mdev->agreed_pro_version >= 97 &&
+ (sector + (size>>9)) >= mdev->ov_stop_sector;
+
+ if (mdev->ov_left == 0 || stop_sector_reached) {
ov_oos_print(mdev);
drbd_resync_finished(mdev);
}
diff --git a/include/linux/drbd.h b/include/linux/drbd.h
index 47e3d48..4a7eccb 100644
--- a/include/linux/drbd.h
+++ b/include/linux/drbd.h
@@ -56,7 +56,7 @@
#define REL_VERSION "8.3.13"
#define API_VERSION 88
#define PRO_VERSION_MIN 86
-#define PRO_VERSION_MAX 96
+#define PRO_VERSION_MAX 97
enum drbd_io_error_p {
diff --git a/include/linux/drbd_nl.h b/include/linux/drbd_nl.h
index a8706f0..f6a576d 100644
--- a/include/linux/drbd_nl.h
+++ b/include/linux/drbd_nl.h
@@ -145,6 +145,7 @@
NL_PACKET(start_ov, 25,
NL_INT64( 66, T_MAY_IGNORE, start_sector)
+ NL_INT64( 90, T_MANDATORY, stop_sector)
)
NL_PACKET(new_c_uuid, 26,