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

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>

drivers/block/drbd/drbd_actlog.c

@@ -349,10 +349,8 @@ w_al_write_transaction(struct drbd_conf *mdev, struct drbd_work *w, int unused)
 	sector =  mdev->ldev->md.md_offset
 		+ mdev->ldev->md.al_offset + mdev->al_tr_pos;
 
-	if (!drbd_md_sync_page_io(mdev, mdev->ldev, sector, WRITE)) {
+	if (!drbd_md_sync_page_io(mdev, mdev->ldev, sector, WRITE))
 		drbd_chk_io_error(mdev, 1, TRUE);
-		drbd_io_error(mdev, TRUE);
-	}
 
 	if (++mdev->al_tr_pos >
 	    div_ceil(mdev->act_log->nr_elements, AL_EXTENTS_PT))
@@ -542,7 +540,6 @@ static void atodb_endio(struct bio *bio, int error)
 	if (!error && !uptodate)
 		error = -EIO;
 
-	/* corresponding drbd_io_error is in drbd_al_to_on_disk_bm */
 	drbd_chk_io_error(mdev, error, TRUE);
 	if (error && wc->error == 0)
 		wc->error = error;
@@ -715,8 +712,6 @@ void drbd_al_to_on_disk_bm(struct drbd_conf *mdev)
 
 	put_ldev(mdev);
 
-	if (wc.error)
-		drbd_io_error(mdev, TRUE);
 	kfree(bios);
 	return;
 

drivers/block/drbd/drbd_bitmap.c

@@ -838,7 +838,6 @@ static int bm_rw(struct drbd_conf *mdev, int rw) __must_hold(local)
 	if (test_bit(BM_MD_IO_ERROR, &b->bm_flags)) {
 		dev_alert(DEV, "we had at least one MD IO ERROR during bitmap IO\n");
 		drbd_chk_io_error(mdev, 1, TRUE);
-		drbd_io_error(mdev, TRUE);
 		err = -EIO;
 	}
 
@@ -911,7 +910,6 @@ int drbd_bm_write_sect(struct drbd_conf *mdev, unsigned long enr) __must_hold(lo
 		    "(meta-disk sector %llus)\n",
 		    enr, (unsigned long long)on_disk_sector);
 		drbd_chk_io_error(mdev, 1, TRUE);
-		drbd_io_error(mdev, TRUE);
 		for (i = 0; i < AL_EXT_PER_BM_SECT; i++)
 			drbd_bm_ALe_set_all(mdev, enr*AL_EXT_PER_BM_SECT+i);
 	}

drivers/block/drbd/drbd_int.h

@@ -1145,7 +1145,6 @@ extern int drbd_send_bitmap(struct drbd_conf *mdev);
 extern int _drbd_send_bitmap(struct drbd_conf *mdev);
 extern int drbd_send_sr_reply(struct drbd_conf *mdev, int retcode);
 extern void drbd_free_bc(struct drbd_backing_dev *ldev);
-extern int drbd_io_error(struct drbd_conf *mdev, int forcedetach);
 extern void drbd_mdev_cleanup(struct drbd_conf *mdev);
 
 /* drbd_meta-data.c (still in drbd_main.c) */
@@ -1653,7 +1652,7 @@ static inline void __drbd_chk_io_error(struct drbd_conf *mdev, int forcedetach)
  * @error:	 Error code passed to the IO completion callback
  * @forcedetach: Force detach. I.e. the error happened while accessing the meta data
  *
- * See also drbd_io_error().
+ * See also drbd_main.c:after_state_ch() if (os.disk > D_FAILED && ns.disk == D_FAILED)
  */
 static inline void drbd_chk_io_error(struct drbd_conf *mdev,
 	int error, int forcedetach)

drivers/block/drbd/drbd_main.c

@@ -416,61 +416,6 @@ void tl_clear(struct drbd_conf *mdev)
 	spin_unlock_irq(&mdev->req_lock);
 }
 
-/**
- * drbd_io_error() - Detach from the local disk of so configured with the on_io_error setting
- * @mdev:	DRBD device.
- * @force_detach: Detach no matter how on_io_error is set (meta data IO error)
- *
- * Should be called in the unlikely(!drbd_bio_uptodate(e->bio)) case from
- * kernel thread context.  See also drbd_chk_io_error().
- */
-int drbd_io_error(struct drbd_conf *mdev, int force_detach)
-{
-	enum drbd_io_error_p eh;
-	unsigned long flags;
-	int send;
-	int ok = 1;
-
-	eh = EP_PASS_ON;
-	if (get_ldev_if_state(mdev, D_FAILED)) {
-		eh = mdev->ldev->dc.on_io_error;
-		put_ldev(mdev);
-	}
-
-	if (!force_detach && eh == EP_PASS_ON)
-		return 1;
-
-	spin_lock_irqsave(&mdev->req_lock, flags);
-	send = (mdev->state.disk == D_FAILED);
-	if (send)
-		_drbd_set_state(_NS(mdev, disk, D_DISKLESS), CS_HARD, NULL);
-	spin_unlock_irqrestore(&mdev->req_lock, flags);
-
-	if (!send)
-		return ok;
-
-	if (mdev->state.conn >= C_CONNECTED) {
-		ok = drbd_send_state(mdev);
-		if (ok)
-			dev_warn(DEV, "Notified peer that my disk is broken.\n");
-		else
-			dev_err(DEV, "Sending state in drbd_io_error() failed\n");
-	}
-
-	/* Make sure we try to flush meta-data to disk - we come
-	 * in here because of a local disk error so it might fail
-	 * but we still need to try -- both because the error might
-	 * be in the data portion of the disk and because we need
-	 * to ensure the md-sync-timer is stopped if running. */
-	drbd_md_sync(mdev);
-
-	/* Releasing the backing device is done in after_state_ch() */
-
-	if (eh == EP_CALL_HELPER)
-		drbd_khelper(mdev, "local-io-error");
-
-	return ok;
-}
 
 /**
  * cl_wide_st_chg() - TRUE if the state change is a cluster wide one
@@ -1332,17 +1277,41 @@ static void after_state_ch(struct drbd_conf *mdev, union drbd_state os,
 	    os.disk > D_INCONSISTENT && ns.disk == D_INCONSISTENT)
 		drbd_queue_bitmap_io(mdev, &drbd_bmio_set_n_write, NULL, "set_n_write from invalidate");
 
-	if (os.disk > D_DISKLESS && ns.disk == D_DISKLESS) {
+	if (os.disk > D_FAILED && ns.disk == D_FAILED) {
+		enum drbd_io_error_p eh;
+
+		eh = EP_PASS_ON;
+		if (get_ldev_if_state(mdev, D_FAILED)) {
+			eh = mdev->ldev->dc.on_io_error;
+			put_ldev(mdev);
+		}
+
+		drbd_rs_cancel_all(mdev);
 		/* since get_ldev() only works as long as disk>=D_INCONSISTENT,
 		   and it is D_DISKLESS here, local_cnt can only go down, it can
 		   not increase... It will reach zero */
 		wait_event(mdev->misc_wait, !atomic_read(&mdev->local_cnt));
-
-		drbd_rs_cancel_all(mdev);
 		mdev->rs_total = 0;
 		mdev->rs_failed = 0;
 		atomic_set(&mdev->rs_pending_cnt, 0);
 
+		spin_lock_irq(&mdev->req_lock);
+		_drbd_set_state(_NS(mdev, disk, D_DISKLESS), CS_HARD, NULL);
+		spin_unlock_irq(&mdev->req_lock);
+
+		if (eh == EP_CALL_HELPER)
+			drbd_khelper(mdev, "local-io-error");
+	}
+
+	if (os.disk > D_DISKLESS && ns.disk == D_DISKLESS) {
+
+		if (os.disk == D_FAILED) /* && ns.disk == D_DISKLESS*/ {
+			if (drbd_send_state(mdev))
+				dev_warn(DEV, "Notified peer that my disk is broken.\n");
+			else
+				dev_err(DEV, "Sending state in drbd_io_error() failed\n");
+		}
+
 		lc_destroy(mdev->resync);
 		mdev->resync = NULL;
 		lc_destroy(mdev->act_log);
@@ -3305,7 +3274,6 @@ void drbd_md_sync(struct drbd_conf *mdev)
 		dev_err(DEV, "meta data update failed!\n");
 
 		drbd_chk_io_error(mdev, 1, TRUE);
-		drbd_io_error(mdev, TRUE);
 	}
 
 	/* Update mdev->ldev->md.la_size_sect,

drivers/block/drbd/drbd_receiver.c

@@ -1465,7 +1465,6 @@ static int e_end_resync_block(struct drbd_conf *mdev, struct drbd_work *w, int u
 		drbd_rs_failed_io(mdev, sector, e->size);
 
 		ok  = drbd_send_ack(mdev, P_NEG_ACK, e);
-		ok &= drbd_io_error(mdev, FALSE);
 	}
 	dec_unacked(mdev);
 
@@ -1607,15 +1606,11 @@ static int e_end_block(struct drbd_conf *mdev, struct drbd_work *w, int unused)
 				drbd_set_in_sync(mdev, sector, e->size);
 		} else {
 			ok  = drbd_send_ack(mdev, P_NEG_ACK, e);
-			ok &= drbd_io_error(mdev, FALSE);
 			/* we expect it to be marked out of sync anyways...
 			 * maybe assert this?  */
 		}
 		dec_unacked(mdev);
-	} else if (unlikely(!drbd_bio_uptodate(e->private_bio))) {
-		ok = drbd_io_error(mdev, FALSE);
 	}
-
 	/* we delete from the conflict detection hash _after_ we sent out the
 	 * P_WRITE_ACK / P_NEG_ACK, to get the sequence number right.  */
 	if (mdev->net_conf->two_primaries) {

drivers/block/drbd/drbd_worker.c

@@ -257,7 +257,6 @@ void drbd_endio_pri(struct bio *bio, int error)
 int w_io_error(struct drbd_conf *mdev, struct drbd_work *w, int cancel)
 {
 	struct drbd_request *req = container_of(w, struct drbd_request, w);
-	int ok;
 
 	/* NOTE: mdev->ldev can be NULL by the time we get here! */
 	/* D_ASSERT(mdev->ldev->dc.on_io_error != EP_PASS_ON); */
@@ -266,10 +265,7 @@ int w_io_error(struct drbd_conf *mdev, struct drbd_work *w, int cancel)
 	 * when it is done and had a local write error, see comments there */
 	drbd_req_free(req);
 
-	ok = drbd_io_error(mdev, FALSE);
-	if (unlikely(!ok))
-		dev_err(DEV, "Sending in w_io_error() failed\n");
-	return ok;
+	return TRUE;
 }
 
 int w_read_retry_remote(struct drbd_conf *mdev, struct drbd_work *w, int cancel)
@@ -279,7 +275,6 @@ int w_read_retry_remote(struct drbd_conf *mdev, struct drbd_work *w, int cancel)
 	/* We should not detach for read io-error,
 	 * but try to WRITE the P_DATA_REPLY to the failed location,
 	 * to give the disk the chance to relocate that block */
-	drbd_io_error(mdev, FALSE); /* tries to schedule a detach and notifies peer */
 
 	spin_lock_irq(&mdev->req_lock);
 	if (cancel ||
@@ -354,10 +349,8 @@ static int w_e_send_csum(struct drbd_conf *mdev, struct drbd_work *w, int cancel
 			dev_err(DEV, "kmalloc() of digest failed.\n");
 			ok = 0;
 		}
-	} else {
-		drbd_io_error(mdev, FALSE);
+	} else
 		ok = 1;
-	}
 
 	drbd_free_ee(mdev, e);
 
@@ -849,8 +842,6 @@ int w_e_end_data_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel)
 			    (unsigned long long)e->sector);
 
 		ok = drbd_send_ack(mdev, P_NEG_DREPLY, e);
-
-		drbd_io_error(mdev, FALSE);
 	}
 
 	dec_unacked(mdev);
@@ -901,8 +892,6 @@ int w_e_end_rsdata_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel)
 
 		ok = drbd_send_ack(mdev, P_NEG_RS_DREPLY, e);
 
-		drbd_io_error(mdev, FALSE);
-
 		/* update resync data with failure */
 		drbd_rs_failed_io(mdev, e->sector, e->size);
 	}
@@ -962,7 +951,6 @@ int w_e_end_csum_rs_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel)
 		ok = drbd_send_ack(mdev, P_NEG_RS_DREPLY, e);
 		if (__ratelimit(&drbd_ratelimit_state))
 			dev_err(DEV, "Sending NegDReply. I guess it gets messy.\n");
-		drbd_io_error(mdev, FALSE);
 	}
 
 	dec_unacked(mdev);
@@ -1056,7 +1044,6 @@ int w_e_end_ov_reply(struct drbd_conf *mdev, struct drbd_work *w, int cancel)
 		ok = drbd_send_ack(mdev, P_NEG_RS_DREPLY, e);
 		if (__ratelimit(&drbd_ratelimit_state))
 			dev_err(DEV, "Sending NegDReply. I guess it gets messy.\n");
-		drbd_io_error(mdev, FALSE);
 	}
 
 	dec_unacked(mdev);