Merge tag 'perf-core-for-mingo-5.4-20190814' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core

Pull perf/core improvements and fixes from Arnaldo:

Intel PT:

  Adrian Hunter:

  - Add PEBS via Intel PT support, the kernel bits went via PeterZ.

perf record:

  Alexander Shishkin:

  - Add an option to take an AUX snapshot on exit.

  Tan Xiaojun:

  - Support aarch64 random socket_id assignment, just like was fixed for S/390.

tools:

  Andy Shevchenko:

  - Keep list of tools in alphabetical order on 'make -C tools help'.

perf session:

  Arnaldo Carvalho de Melo:

  - Avoid infinite loop when seeing invalid header.size, reported by
    Vince Weaver using a perf.data fuzzer.

Documentation:

  Vince Weaver:

  - Clarify HEADER_SAMPLE_TOPOLOGY format in the perf.data spec.

perf config:

  Arnaldo Carvalho de Melo:

  - Honour $PERF_CONFIG env var to specify alternate .perfconfig.

perf test:

  Arnaldo Carvalho de Melo:

  - Disable ~/.perfconfig to get default output in 'perf trace' tests.

perf top:

  Arnaldo Carvalho de Melo:

  - Set display thread COMM to help with debugging.

  - Collapse and resort evsels in a group, so that we have output
    similar to 'perf report' when using event groups, i.e.

      perf top -e '{cycles,instructions}'

    Will have two columns, and the instructions one will work.

core:

  Igor Lubashev:

  - Detect if libcap development files are available so that we
    can use capabilities to match the checks made by the kernel instead
    of using plain (geteuid() == 0).

Intel:

  Haiyan Song:

  - Add Icelake V1.00 event file.

perf trace:

  Leo Yan:

  - Fix segmentation fault when access syscall info on arm64.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

1 files changed, 73 insertions, 40 deletions

diff --git a/fs/block_dev.c b/fs/block_dev.c
index c2a85b587922..eb657ab94060 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -349,7 +349,7 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages)
 	loff_t pos = iocb->ki_pos;
 	blk_qc_t qc = BLK_QC_T_NONE;
 	gfp_t gfp;
-	ssize_t ret;
+	int ret;
 
 	if ((pos | iov_iter_alignment(iter)) &
 	    (bdev_logical_block_size(bdev) - 1))
@@ -386,8 +386,6 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages)
 
 	ret = 0;
 	for (;;) {
-		int err;
-
 		bio_set_dev(bio, bdev);
 		bio->bi_iter.bi_sector = pos >> 9;
 		bio->bi_write_hint = iocb->ki_hint;
@@ -395,10 +393,8 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages)
 		bio->bi_end_io = blkdev_bio_end_io;
 		bio->bi_ioprio = iocb->ki_ioprio;
 
-		err = bio_iov_iter_get_pages(bio, iter);
-		if (unlikely(err)) {
-			if (!ret)
-				ret = err;
+		ret = bio_iov_iter_get_pages(bio, iter);
+		if (unlikely(ret)) {
 			bio->bi_status = BLK_STS_IOERR;
 			bio_endio(bio);
 			break;
@@ -421,7 +417,6 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages)
 		if (nowait)
 			bio->bi_opf |= (REQ_NOWAIT | REQ_NOWAIT_INLINE);
 
-		dio->size += bio->bi_iter.bi_size;
 		pos += bio->bi_iter.bi_size;
 
 		nr_pages = iov_iter_npages(iter, BIO_MAX_PAGES);
@@ -433,10 +428,11 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages)
 				polled = true;
 			}
 
+			dio->size += bio->bi_iter.bi_size;
 			qc = submit_bio(bio);
 			if (qc == BLK_QC_T_EAGAIN) {
-				if (!ret)
-					ret = -EAGAIN;
+				dio->size -= bio->bi_iter.bi_size;
+				ret = -EAGAIN;
 				goto error;
 			}
 
@@ -459,18 +455,17 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages)
 			atomic_inc(&dio->ref);
 		}
 
+		dio->size += bio->bi_iter.bi_size;
 		qc = submit_bio(bio);
 		if (qc == BLK_QC_T_EAGAIN) {
-			if (!ret)
-				ret = -EAGAIN;
+			dio->size -= bio->bi_iter.bi_size;
+			ret = -EAGAIN;
 			goto error;
 		}
-		ret += bio->bi_iter.bi_size;
 
 		bio = bio_alloc(gfp, nr_pages);
 		if (!bio) {
-			if (!ret)
-				ret = -EAGAIN;
+			ret = -EAGAIN;
 			goto error;
 		}
 	}
@@ -495,6 +490,8 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages)
 out:
 	if (!ret)
 		ret = blk_status_to_errno(dio->bio.bi_status);
+	if (likely(!ret))
+		ret = dio->size;
 
 	bio_put(&dio->bio);
 	return ret;
@@ -1181,8 +1178,7 @@ static struct gendisk *bdev_get_gendisk(struct block_device *bdev, int *partno)
  * Pointer to the block device containing @bdev on success, ERR_PTR()
  * value on failure.
  */
-static struct block_device *bd_start_claiming(struct block_device *bdev,
-					      void *holder)
+struct block_device *bd_start_claiming(struct block_device *bdev, void *holder)
 {
 	struct gendisk *disk;
 	struct block_device *whole;
@@ -1229,6 +1225,62 @@ static struct block_device *bd_start_claiming(struct block_device *bdev,
 		return ERR_PTR(err);
 	}
 }
+EXPORT_SYMBOL(bd_start_claiming);
+
+static void bd_clear_claiming(struct block_device *whole, void *holder)
+{
+	lockdep_assert_held(&bdev_lock);
+	/* tell others that we're done */
+	BUG_ON(whole->bd_claiming != holder);
+	whole->bd_claiming = NULL;
+	wake_up_bit(&whole->bd_claiming, 0);
+}
+
+/**
+ * bd_finish_claiming - finish claiming of a block device
+ * @bdev: block device of interest
+ * @whole: whole block device (returned from bd_start_claiming())
+ * @holder: holder that has claimed @bdev
+ *
+ * Finish exclusive open of a block device. Mark the device as exlusively
+ * open by the holder and wake up all waiters for exclusive open to finish.
+ */
+void bd_finish_claiming(struct block_device *bdev, struct block_device *whole,
+			void *holder)
+{
+	spin_lock(&bdev_lock);
+	BUG_ON(!bd_may_claim(bdev, whole, holder));
+	/*
+	 * Note that for a whole device bd_holders will be incremented twice,
+	 * and bd_holder will be set to bd_may_claim before being set to holder
+	 */
+	whole->bd_holders++;
+	whole->bd_holder = bd_may_claim;
+	bdev->bd_holders++;
+	bdev->bd_holder = holder;
+	bd_clear_claiming(whole, holder);
+	spin_unlock(&bdev_lock);
+}
+EXPORT_SYMBOL(bd_finish_claiming);
+
+/**
+ * bd_abort_claiming - abort claiming of a block device
+ * @bdev: block device of interest
+ * @whole: whole block device (returned from bd_start_claiming())
+ * @holder: holder that has claimed @bdev
+ *
+ * Abort claiming of a block device when the exclusive open failed. This can be
+ * also used when exclusive open is not actually desired and we just needed
+ * to block other exclusive openers for a while.
+ */
+void bd_abort_claiming(struct block_device *bdev, struct block_device *whole,
+		       void *holder)
+{
+	spin_lock(&bdev_lock);
+	bd_clear_claiming(whole, holder);
+	spin_unlock(&bdev_lock);
+}
+EXPORT_SYMBOL(bd_abort_claiming);
 
 #ifdef CONFIG_SYSFS
 struct bd_holder_disk {
@@ -1698,29 +1750,10 @@ int blkdev_get(struct block_device *bdev, fmode_t mode, void *holder)
 
 		/* finish claiming */
 		mutex_lock(&bdev->bd_mutex);
-		spin_lock(&bdev_lock);
-
-		if (!res) {
-			BUG_ON(!bd_may_claim(bdev, whole, holder));
-			/*
-			 * Note that for a whole device bd_holders
-			 * will be incremented twice, and bd_holder
-			 * will be set to bd_may_claim before being
-			 * set to holder
-			 */
-			whole->bd_holders++;
-			whole->bd_holder = bd_may_claim;
-			bdev->bd_holders++;
-			bdev->bd_holder = holder;
-		}
-
-		/* tell others that we're done */
-		BUG_ON(whole->bd_claiming != holder);
-		whole->bd_claiming = NULL;
-		wake_up_bit(&whole->bd_claiming, 0);
-
-		spin_unlock(&bdev_lock);
-
+		if (!res)
+			bd_finish_claiming(bdev, whole, holder);
+		else
+			bd_abort_claiming(bdev, whole, holder);
 		/*
 		 * Block event polling for write claims if requested.  Any
 		 * write holder makes the write_holder state stick until