!76 fix RAID0 becoming unusable after setting fault of member disks

From: @wguanghao 
Reviewed-by: @swf504 
Signed-off-by: @swf504

0011-Manage-Block-unsafe-member-failing.patch

@@ -0,0 +1,91 @@
+From fc6fd4063769f4194c3fb8f77b32b2819e140fb9 Mon Sep 17 00:00:00 2001
+From: Mateusz Kusiak <mateusz.kusiak@intel.com>
+Date: Thu, 18 Aug 2022 11:47:21 +0200
+Subject: [PATCH] Manage: Block unsafe member failing
+
+Kernel may or may not block mdadm from removing member device if it
+will cause arrays failed state. It depends on raid personality
+implementation in kernel.
+Add verification on requested removal path (#mdadm --set-faulty
+command).
+
+Signed-off-by: Mateusz Kusiak <mateusz.kusiak@intel.com>
+Signed-off-by: Jes Sorensen <jsorensen@fb.com>
+---
+ Manage.c | 53 ++++++++++++++++++++++++++++++++++++++++++++++++++++-
+ 1 file changed, 52 insertions(+), 1 deletion(-)
+
+diff --git a/Manage.c b/Manage.c
+index a142f8bd..b1d0e630 100644
+--- a/Manage.c
++++ b/Manage.c
+@@ -1285,6 +1285,50 @@ int Manage_with(struct supertype *tst, int fd, struct mddev_dev *dv,
+ 	return -1;
+ }
+ 
++/**
++ * is_remove_safe() - Check if remove is safe.
++ * @array: Array info.
++ * @fd: Array file descriptor.
++ * @devname: Name of device to remove.
++ * @verbose: Verbose.
++ *
++ * The function determines if array will be operational
++ * after removing &devname.
++ *
++ * Return: True if array will be operational, false otherwise.
++ */
++bool is_remove_safe(mdu_array_info_t *array, const int fd, char *devname, const int verbose)
++{
++	dev_t devid = devnm2devid(devname + 5);
++	struct mdinfo *mdi = sysfs_read(fd, NULL, GET_DEVS | GET_DISKS | GET_STATE);
++
++	if (!mdi) {
++		if (verbose)
++			pr_err("Failed to read sysfs attributes for %s\n", devname);
++		return false;
++	}
++
++	char *avail = xcalloc(array->raid_disks, sizeof(char));
++
++	for (mdi = mdi->devs; mdi; mdi = mdi->next) {
++		if (mdi->disk.raid_disk < 0)
++			continue;
++		if (!(mdi->disk.state & (1 << MD_DISK_SYNC)))
++			continue;
++		if (makedev(mdi->disk.major, mdi->disk.minor) == devid)
++			continue;
++		avail[mdi->disk.raid_disk] = 1;
++	}
++	sysfs_free(mdi);
++
++	bool is_enough = enough(array->level, array->raid_disks,
++				array->layout, (array->state & 1),
++				avail);
++
++	free(avail);
++	return is_enough;
++}
++
+ int Manage_subdevs(char *devname, int fd,
+ 		   struct mddev_dev *devlist, int verbose, int test,
+ 		   char *update, int force)
+@@ -1598,7 +1642,14 @@ int Manage_subdevs(char *devname, int fd,
+ 			break;
+ 
+ 		case 'f': /* set faulty */
+-			/* FIXME check current member */
++			if (!is_remove_safe(&array, fd, dv->devname, verbose)) {
++				pr_err("Cannot remove %s from %s, array will be failed.\n",
++				       dv->devname, devname);
++				if (sysfd >= 0)
++					close(sysfd);
++				goto abort;
++			}
++
+ 			if ((sysfd >= 0 && write(sysfd, "faulty", 6) != 6) ||
+ 			    (sysfd < 0 && ioctl(fd, SET_DISK_FAULTY,
+ 						rdev))) {
+-- 
+2.33.0
+

0012-Manage-do-not-check-array-state-when-drive-is-remove.patch

@@ -0,0 +1,33 @@
+From b3e7b7eb1dfedd7cbd9a3800e884941f67d94c96 Mon Sep 17 00:00:00 2001
+From: Kinga Tanska <kinga.tanska@intel.com>
+Date: Tue, 27 Dec 2022 06:50:42 +0100
+Subject: [PATCH] Manage: do not check array state when drive is removed
+
+Array state doesn't need to be checked when drive is
+removed, but until now clean state was required. Result
+of the is_remove_safe() function will be independent
+from array state.
+
+Signed-off-by: Kinga Tanska <kinga.tanska@intel.com>
+Signed-off-by: Jes Sorensen <jes@trained-monkey.org>
+---
+ Manage.c | 3 +--
+ 1 file changed, 1 insertion(+), 2 deletions(-)
+
+diff --git a/Manage.c b/Manage.c
+index 594e3d2c..4d6e54b1 100644
+--- a/Manage.c
++++ b/Manage.c
+@@ -1321,8 +1321,7 @@ bool is_remove_safe(mdu_array_info_t *array, const int fd, char *devname, const
+ 	sysfs_free(mdi);
+ 
+ 	bool is_enough = enough(array->level, array->raid_disks,
+-				array->layout, (array->state & 1),
+-				avail);
++				array->layout, 1, avail);
+ 
+ 	free(avail);
+ 	return is_enough;
+-- 
+2.33.0
+

0013-incremental-manage-do-not-verify-if-remove-is-safe.patch

@@ -0,0 +1,59 @@
+From 461fae7e7809670d286cc19aac5bfa861c29f93a Mon Sep 17 00:00:00 2001
+From: Kinga Tanska <kinga.tanska@intel.com>
+Date: Tue, 27 Dec 2022 06:50:43 +0100
+Subject: [PATCH] incremental, manage: do not verify if remove is safe
+
+Function is_remove_safe() was introduced to verify if removing
+member device won't cause failed state of the array. This
+verification should be used only with set-faulty command. Add
+special mode indicating that Incremental removal was executed.
+If this mode is used do not execute is_remove_safe() routine.
+
+Signed-off-by: Kinga Tanska <kinga.tanska@intel.com>
+Signed-off-by: Jes Sorensen <jes@trained-monkey.org>
+---
+ Incremental.c | 2 +-
+ Manage.c      | 7 ++++---
+ 2 files changed, 5 insertions(+), 4 deletions(-)
+
+diff --git a/Incremental.c b/Incremental.c
+index ff3548c0..09b94b9f 100644
+--- a/Incremental.c
++++ b/Incremental.c
+@@ -1744,7 +1744,7 @@ int IncrementalRemove(char *devname, char *id_path, int verbose)
+ 
+ 	memset(&devlist, 0, sizeof(devlist));
+ 	devlist.devname = devname;
+-	devlist.disposition = 'f';
++	devlist.disposition = 'I';
+ 	/* for a container, we must fail each member array */
+ 	if (ent->metadata_version &&
+ 	    strncmp(ent->metadata_version, "external:", 9) == 0) {
+diff --git a/Manage.c b/Manage.c
+index 4d6e54b1..6184d3f7 100644
+--- a/Manage.c
++++ b/Manage.c
+@@ -1494,8 +1494,9 @@ int Manage_subdevs(char *devname, int fd,
+ 			/* Assume this is a kernel-internal name like 'sda1' */
+ 			int found = 0;
+ 			char dname[55];
+-			if (dv->disposition != 'r' && dv->disposition != 'f') {
+-				pr_err("%s only meaningful with -r or -f, not -%c\n",
++			if (dv->disposition != 'r' && dv->disposition != 'f' &&
++			    dv->disposition != 'I') {
++				pr_err("%s only meaningful with -r, -f or -I, not -%c\n",
+ 					dv->devname, dv->disposition);
+ 				goto abort;
+ 			}
+@@ -1647,7 +1648,7 @@ int Manage_subdevs(char *devname, int fd,
+ 					close(sysfd);
+ 				goto abort;
+ 			}
+-
++		case 'I': /* incremental fail */
+ 			if ((sysfd >= 0 && write(sysfd, "faulty", 6) != 6) ||
+ 			    (sysfd < 0 && ioctl(fd, SET_DISK_FAULTY,
+ 						rdev))) {
+-- 
+2.33.0
+

mdadm.spec

@@ -1,6 +1,6 @@
 Name:     mdadm
 Version:  4.2
-Release:  9
+Release:  10
 Summary:  The software RAID arrays user manage tools
 License:  GPLv2+
 URL:      http://www.kernel.org/pub/linux/utils/raid/mdadm/
@@ -20,6 +20,9 @@ Patch7:  0007-DDF-Fix-NULL-pointer-dereference-in-validate_geometr.patch
 Patch8:  0008-fix-NULL-dereference-in-super_by_fd.patch
 Patch9:  0009-fix-mdmonitor-oneshot.service-start-error.patch
 Patch10: 0010-Fix-null-pointer-for-incremental-in-mdadm.patch
+Patch11: 0011-Manage-Block-unsafe-member-failing.patch
+Patch12: 0012-Manage-do-not-check-array-state-when-drive-is-remove.patch
+Patch13: 0013-incremental-manage-do-not-verify-if-remove-is-safe.patch
 
 BuildRequires:    systemd gcc binutils libudev-devel
 Requires(post):   systemd coreutils
@@ -85,6 +88,9 @@ install -d -m 710 %{buildroot}/var/run/mdadm/
 %{_mandir}/man*/*
 
 %changelog
+* Tue Nov 7 2023 wuguanghao <wuguanghao3@huawei.com> - 4.2-10
+* fix RAID0 becoming unusable after setting fault of member disks
+
 * Wed Sep 13 2023 miaoguanqin <miaoguanqin@huawei.com> - 4.2-9
 * Fix null pointer for incremental in mdadm