Received-SPF: pass (google.com: domain of linux-kernel+bounces-90607-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
From: John Garry <john.g.garry@oracle.com>
To: djwong@kernel.org, hch@lst.de, viro@zeniv.linux.org.uk, brauner@kernel.org,
        jack@suse.cz, chandan.babu@oracle.com, david@fromorbit.com,
        axboe@kernel.dk
Cc: martin.petersen@oracle.com, linux-kernel@vger.kernel.org,
        linux-xfs@vger.kernel.org, linux-fsdevel@vger.kernel.org,
        tytso@mit.edu, jbongio@google.com, ojaswin@linux.ibm.com,
        ritesh.list@gmail.com, linux-block@vger.kernel.org,
        John Garry <john.g.garry@oracle.com>
Subject: [PATCH v2 08/14] fs: xfs: iomap: Sub-extent zeroing
Date: Mon,  4 Mar 2024 13:04:22 +0000
Message-Id: <20240304130428.13026-9-john.g.garry@oracle.com>
In-Reply-To: <20240304130428.13026-1-john.g.garry@oracle.com>
References: <20240304130428.13026-1-john.g.garry@oracle.com>
Content-Transfer-Encoding: 8bit
Content-Type: text/plain
Precedence: bulk
MIME-Version: 1.0
X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1
X-MS-Exchange-AntiSpam-MessageData-0: 
	=?us-ascii?Q?OPGEKnwU5GVI87FnhPP9f6FnxVFNMKbjLuSNDcGcu6L09gCypcBNtcJM/ulT?=
 =?us-ascii?Q?dhmmd/oVircte1/C6tOQsYNKZTYfRhDiv+nZF2B62sF5wV5wfqlYSWjWA/lx?=
 =?us-ascii?Q?qXjYV3sGselAfNn2iLyrcmjEdWtgjfvYMfN4w89zWCIgojH7AqC4Yt0PxE9m?=
 =?us-ascii?Q?QGpOQvRguKFtLLdkt+8Jw4NbSIaH1bm5qw1gD1f6M9bh8Ota0CGW1AK9vsdS?=
 =?us-ascii?Q?50YpgeA2DUvboZ75Pe0RdQMcovEBXQD1TvU65rlD7rYaYO3dSCUEIH7EEZNh?=
 =?us-ascii?Q?aHTi6nANIalmpfVy+5+RlHKFseHr4sYu3d8jS0t6JdKg8iUctdp2EX3BNWeJ?=
 =?us-ascii?Q?NFYNg1oFgarSZ4uMnLdYb04vVTHYzL9m2KfSbf0FiKJ0AT+/VqbKnhPwJlou?=
 =?us-ascii?Q?EebO63HVIPevX0+XCOuE2Xvar7dT/k+rlbG+Dv3MHnUN2pYV+K1g7yhaL2NF?=
 =?us-ascii?Q?pRmpVgQ44Rsk1Q10f7vC3KxCqtv5SQuXwUoVGEhx1vFt7TN6mf3aXxbLzuId?=
 =?us-ascii?Q?R7dEgCBj8RecaDYPIYyNk+q5clRTcmuqEU7ss95G5S46uM1HJUPM8cMXM/Zd?=
 =?us-ascii?Q?W05/jrvf3rg4pg3OZvKx5GCwqw55XdAMo+gvAYVk36TnTD2s6nEdUWzkcPQ7?=
 =?us-ascii?Q?F5ZjjSQyl8b0tC78AsGI2QlbjwgJBSn+6nMkMaYo7MUm9skcsotEAAiCRQDc?=
 =?us-ascii?Q?vb7wIfREqOSt0A1PDiNd0wJB+fyssbZsFUN9rOCAizztCwbu2UjHahxPC00I?=
 =?us-ascii?Q?wp+S4Xh8zfu/nOq0hWXWO7XL2FGveTsALISJE0SagCX8oDtb0R95VhNzWFlF?=
 =?us-ascii?Q?YsNr9xW14R1nKwbs3Q/JNsROoINCAp/Qak3tvE8ci/dkTVCRfgw5ec/dxvlo?=
 =?us-ascii?Q?NrP28IpV09d89sws0z56coMJp+SyyuOYSBKJUx4XeZz0P6LW0IUMPUzTd884?=
 =?us-ascii?Q?oiPy8zAv76MWGsMvUqwQRLaK36UT0G7YL60RK4j5i0I7eSUNUf+keMHr+tat?=
 =?us-ascii?Q?HMmkR5889tu5jcSX0/unT6NfaFQ2rmd3sFtnOvN6gTFXXxWQR6YJKf9Gnz4v?=
 =?us-ascii?Q?Ieimp0YmzcSwsRA/Glpp+RfwN7aKXQ8OUyGL1D6quHcVxueVTtmUuw3Ou2NX?=
 =?us-ascii?Q?iBWUFbhvXqje33+6/CcG00J/H2qfCpx9s2xOK+opfzkxttQOTxCcyvrOp3Aj?=
 =?us-ascii?Q?zob24pdK6gH442hb14MczJM4i/FKiAHZvR5bZ+STqGwHJ0jOsENbi9yiZXq7?=
 =?us-ascii?Q?rL0XA2BhlvC1viv8C3TLJekeolRuKKgJ8z5kvdWE8iJ/mm+FJkQIasvAGRDx?=
 =?us-ascii?Q?q1Nrq2sjmECT2QUXpj+i7bi+dChSUxlq5jHBALaWHxEQCyxNam86/jc4UoSe?=
 =?us-ascii?Q?NwA5pGdDiV8w+LR1xNU7N4dE/n4KKikYrIs5Ng6BgMtMFzu3NEAIVprNBXCo?=
 =?us-ascii?Q?8Juh7voO+3FqLRDNXTzA05beR/9Rq4c6cYghXOkQDG4bygDPMfydejmWRkIo?=
 =?us-ascii?Q?S9rdE2sHs/fEXkLsf2DZLWj7RUX4UqWwn2qmbHp+WF2Qy9lhcx4l2kWi8a/C?=
 =?us-ascii?Q?CgDkirNgsQf7djl3/iR7n3qR/2eUL5MxtN/rA/kwWVRgoseTSIQ39Cn5NCJK?=
 =?us-ascii?Q?aw=3D=3D?=
X-MS-Exchange-AntiSpam-ExternalHop-MessageData-ChunkCount: 1
X-MS-Exchange-AntiSpam-ExternalHop-MessageData-0: 
	7iVSIzggI0FKDcDoDbj8+ddm3s3bbY/GGiFvHGkHK6z7Yj8c+fN/YWbePKqEkRSUV3dGKGqCYIF8beRB8YnHoi/Qx+ha72jj7tFqJajXJGDKeHpzx40vFPR1tvvjamPYViePvk+p/UkqRT+UwydO1CyQAQ/ztp7Y7eIvkF1ONo07gC7nfcooLdGsZg1a3Ik7lM5lhDijcYd8PF2R8Ck1GRDigKdjBVyKw8veyZIYt/2Qji9ZCotpHjKbiNUWTjKoa6aAdUB33HEjgdEG02TPmAw3XH/cr68p4iV/CippIFIs/HFLX6sFD7yoP7LXU9uR2v8YtNS/iHvr3hhM07jJPWtuQomtqV5D6T5f6dggbWl3DLsTm/701JUZDx0A9nOnbM6L/9+DFErSO+jKvpv+aBhiGdHHO2pztcy9Ye8NFbZ0RNE49mzwEW61X2U3o/AU3jG7VVTUCFSgDGr6SuXouCOqjjbnE1nfgswXmy+UjkQeCQj5ED4ozYQRmBPOnThEypXed+zAxsAT51sF0FV2ruekoyEnKm25rEs9iSNixfdMQ+Jk1LF7f4GtvvWhvb6CT82iu2InuEtYWM6bIxg/JjNSK/QFZXpIHeLtNeidlhc=
X-OriginatorOrg: oracle.com
X-MS-Exchange-CrossTenant-Network-Message-Id: 84bf7950-e820-4e23-84dc-08dc3c4bbb9d
X-MS-Exchange-CrossTenant-AuthSource: DM6PR10MB4313.namprd10.prod.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 04 Mar 2024 13:05:14.8995
 (UTC)
X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted
X-MS-Exchange-CrossTenant-Id: 4e2c6054-71cb-48f1-bd6c-3a9705aca71b
X-MS-Exchange-CrossTenant-MailboxType: HOSTED
X-MS-Exchange-CrossTenant-UserPrincipalName: KqNGMyOf/Ah9kXQ3rXeKgi8SGnW0lTztFlFSAgyc/6Q7apEn1iOJWcWkPqbNOW/vgjUJyuhTYfjjOTGcZxmnDw==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: CH3PR10MB7805
X-Proofpoint-Virus-Version: vendor=baseguard
 engine=ICAP:2.0.272,Aquarius:18.0.1011,Hydra:6.0.619,FMLib:17.11.176.26
 definitions=2024-03-04_09,2024-03-04_01,2023-05-22_02
X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 phishscore=0 suspectscore=0
 adultscore=0 bulkscore=0 spamscore=0 mlxlogscore=999 malwarescore=0
 mlxscore=0 classifier=spam adjust=0 reason=mlx scancount=1
 engine=8.12.0-2311290000 definitions=main-2403040098
X-Proofpoint-ORIG-GUID: e8lg22jpYYc021IiyK1Mvmy4VjYVbQzX
X-Proofpoint-GUID: e8lg22jpYYc021IiyK1Mvmy4VjYVbQzX

Set iomap->extent_shift when sub-extent zeroing is required.

We treat a sub-extent write same as an unaligned write, so we can leverage
the existing sub-FSblock unaligned write support, i.e. try a shared lock
with IOMAP_DIO_OVERWRITE_ONLY flag, if this fails then try the exclusive
lock.

In xfs_iomap_write_unwritten(), FSB calcs are now based on the extsize.

Signed-off-by: John Garry <john.g.garry@oracle.com>
---
 fs/xfs/xfs_file.c  | 28 +++++++++++++++-------------
 fs/xfs/xfs_iomap.c | 15 +++++++++++++--
 2 files changed, 28 insertions(+), 15 deletions(-)

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index e33e5e13b95f..d0bd9d5f596c 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -617,18 +617,19 @@ xfs_file_dio_write_aligned(
  * Handle block unaligned direct I/O writes
  *
  * In most cases direct I/O writes will be done holding IOLOCK_SHARED, allowing
- * them to be done in parallel with reads and other direct I/O writes.  However,
- * if the I/O is not aligned to filesystem blocks, the direct I/O layer may need
- * to do sub-block zeroing and that requires serialisation against other direct
- * I/O to the same block.  In this case we need to serialise the submission of
- * the unaligned I/O so that we don't get racing block zeroing in the dio layer.
- * In the case where sub-block zeroing is not required, we can do concurrent
- * sub-block dios to the same block successfully.
+ * them to be done in parallel with reads and other direct I/O writes.
+ * However if the I/O is not aligned to filesystem blocks/extent, the direct
+ * I/O layer may need to do sub-block/extent zeroing and that requires
+ * serialisation against other direct I/O to the same block/extent.  In this
+ * case we need to serialise the submission of the unaligned I/O so that we
+ * don't get racing block/extent zeroing in the dio layer.
+ * In the case where sub-block/extent zeroing is not required, we can do
+ * concurrent sub-block/extent dios to the same block/extent successfully.
  *
  * Optimistically submit the I/O using the shared lock first, but use the
  * IOMAP_DIO_OVERWRITE_ONLY flag to tell the lower layers to return -EAGAIN
- * if block allocation or partial block zeroing would be required.  In that case
- * we try again with the exclusive lock.
+ * if block/extent allocation or partial block/extent zeroing would be
+ * required.  In that case we try again with the exclusive lock.
  */
 static noinline ssize_t
 xfs_file_dio_write_unaligned(
@@ -643,9 +644,9 @@ xfs_file_dio_write_unaligned(
 	ssize_t			ret;
 
 	/*
-	 * Extending writes need exclusivity because of the sub-block zeroing
-	 * that the DIO code always does for partial tail blocks beyond EOF, so
-	 * don't even bother trying the fast path in this case.
+	 * Extending writes need exclusivity because of the sub-block/extent
+	 * zeroing that the DIO code always does for partial tail blocks
+	 * beyond EOF, so don't even bother trying the fast path in this case.
 	 */
 	if (iocb->ki_pos > isize || iocb->ki_pos + count >= isize) {
 		if (iocb->ki_flags & IOCB_NOWAIT)
@@ -709,13 +710,14 @@ xfs_file_dio_write(
 	struct iov_iter		*from)
 {
 	struct xfs_inode	*ip = XFS_I(file_inode(iocb->ki_filp));
+	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_buftarg      *target = xfs_inode_buftarg(ip);
 	size_t			count = iov_iter_count(from);
 
 	/* direct I/O must be aligned to device logical sector size */
 	if ((iocb->ki_pos | count) & target->bt_logical_sectormask)
 		return -EINVAL;
-	if ((iocb->ki_pos | count) & ip->i_mount->m_blockmask)
+	if ((iocb->ki_pos | count) & (XFS_FSB_TO_B(mp, xfs_get_extsz(ip)) - 1))
 		return xfs_file_dio_write_unaligned(ip, iocb, from);
 	return xfs_file_dio_write_aligned(ip, iocb, from);
 }
diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c
index 70fe873951f3..88cc20bb19c9 100644
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -98,6 +98,7 @@ xfs_bmbt_to_iomap(
 {
 	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_buftarg	*target = xfs_inode_buftarg(ip);
+	xfs_extlen_t		extsz = xfs_get_extsz(ip);
 
 	if (unlikely(!xfs_valid_startblock(ip, imap->br_startblock)))
 		return xfs_alert_fsblock_zero(ip, imap);
@@ -134,6 +135,8 @@ xfs_bmbt_to_iomap(
 
 	iomap->validity_cookie = sequence_cookie;
 	iomap->folio_ops = &xfs_iomap_folio_ops;
+	if (extsz > 1)
+		iomap->extent_shift = ffs(extsz) - 1;
 	return 0;
 }
 
@@ -563,11 +566,19 @@ xfs_iomap_write_unwritten(
 	xfs_fsize_t	i_size;
 	uint		resblks;
 	int		error;
+	xfs_extlen_t	extsz = xfs_get_extsz(ip);
 
 	trace_xfs_unwritten_convert(ip, offset, count);
 
-	offset_fsb = XFS_B_TO_FSBT(mp, offset);
-	count_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)offset + count);
+	if (extsz > 1) {
+		xfs_extlen_t extsize_bytes = XFS_FSB_TO_B(mp, extsz);
+
+		offset_fsb = XFS_B_TO_FSBT(mp, round_down(offset, extsize_bytes));
+		count_fsb = XFS_B_TO_FSB(mp, round_up(offset + count, extsize_bytes));
+	} else {
+		offset_fsb = XFS_B_TO_FSBT(mp, offset);
+		count_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)offset + count);
+	}
 	count_fsb = (xfs_filblks_t)(count_fsb - offset_fsb);
 
 	/*
-- 
2.31.1