- Linux-stable-mirror - lists.linaro.org

[Linux-stable-mirror] FAILED: patch "[PATCH] ext4: prevent data corruption with inline data + DAX" failed to apply to 4.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 559db4c6d784ceedc2a5418ced4d357cb843e221 Mon Sep 17 00:00:00 2001 From: Ross Zwisler <ross.zwisler(a)linux.intel.com> Date: Thu, 12 Oct 2017 11:52:34 -0400 Subject: [PATCH] ext4: prevent data corruption with inline data + DAX If an inode has inline data it is currently prevented from using DAX by a check in ext4_set_inode_flags(). When the inode grows inline data via ext4_create_inline_data() or removes its inline data via ext4_destroy_inline_data_nolock(), the value of S_DAX can change. Currently these changes are unsafe because we don't hold off page faults and I/O, write back dirty radix tree entries and invalidate all mappings. There are also issues with mm-level races when changing the value of S_DAX, as well as issues with the VM_MIXEDMAP flag: https://www.spinics.net/lists/linux-xfs/msg09859.html The unsafe transition of S_DAX can reliably cause data corruption, as shown by the following fstest: https://patchwork.kernel.org/patch/9948381/ Fix this issue by preventing the DAX mount option from being used on filesystems that were created to support inline data. Inline data is an option given to mkfs.ext4. Signed-off-by: Ross Zwisler <ross.zwisler(a)linux.intel.com> Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Reviewed-by: Jan Kara <jack(a)suse.cz> CC: stable(a)vger.kernel.org diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c index f0bbc8cb6555..1367553c43bb 100644 --- a/fs/ext4/inline.c +++ b/fs/ext4/inline.c @@ -303,11 +303,6 @@ static int ext4_create_inline_data(handle_t *handle, EXT4_I(inode)->i_inline_size = len + EXT4_MIN_INLINE_DATA_SIZE; ext4_clear_inode_flag(inode, EXT4_INODE_EXTENTS); ext4_set_inode_flag(inode, EXT4_INODE_INLINE_DATA); - /* - * Propagate changes to inode->i_flags as well - e.g. S_DAX may - * get cleared - */ - ext4_set_inode_flags(inode); get_bh(is.iloc.bh); error = ext4_mark_iloc_dirty(handle, inode, &is.iloc); @@ -452,11 +447,6 @@ static int ext4_destroy_inline_data_nolock(handle_t *handle, } } ext4_clear_inode_flag(inode, EXT4_INODE_INLINE_DATA); - /* - * Propagate changes to inode->i_flags as well - e.g. S_DAX may - * get set. - */ - ext4_set_inode_flags(inode); get_bh(is.iloc.bh); error = ext4_mark_iloc_dirty(handle, inode, &is.iloc); diff --git a/fs/ext4/super.c b/fs/ext4/super.c index b104096fce9e..986475c0d552 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -3708,6 +3708,11 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) } if (sbi->s_mount_opt & EXT4_MOUNT_DAX) { + if (ext4_has_feature_inline_data(sb)) { + ext4_msg(sb, KERN_ERR, "Cannot use DAX on a filesystem" + " that may contain inline data"); + goto failed_mount; + } err = bdev_dax_supported(sb, blocksize); if (err) goto failed_mount;

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] ext4: prevent data corruption with journaling + DAX" failed to apply to 4.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From e9072d859df3e0f2c3ba450f0d1739595c2d5d13 Mon Sep 17 00:00:00 2001 From: Ross Zwisler <ross.zwisler(a)linux.intel.com> Date: Thu, 12 Oct 2017 11:54:08 -0400 Subject: [PATCH] ext4: prevent data corruption with journaling + DAX The current code has the potential for data corruption when changing an inode's journaling mode, as that can result in a subsequent unsafe change in S_DAX. I've captured an instance of this data corruption in the following fstest: https://patchwork.kernel.org/patch/9948377/ Prevent this data corruption from happening by disallowing changes to the journaling mode if the '-o dax' mount option was used. This means that for a given filesystem we could have a mix of inodes using either DAX or data journaling, but whatever state the inodes are in will be held for the duration of the mount. Signed-off-by: Ross Zwisler <ross.zwisler(a)linux.intel.com> Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Reviewed-by: Jan Kara <jack(a)suse.cz> Cc: stable(a)vger.kernel.org diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index edfe95f81274..350e0910ed32 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -6004,11 +6004,6 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) ext4_clear_inode_flag(inode, EXT4_INODE_JOURNAL_DATA); } ext4_set_aops(inode); - /* - * Update inode->i_flags after EXT4_INODE_JOURNAL_DATA was updated. - * E.g. S_DAX may get cleared / set. - */ - ext4_set_inode_flags(inode); jbd2_journal_unlock_updates(journal); percpu_up_write(&sbi->s_journal_flag_rwsem); diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index afb66d4ab5cf..b0b754b37c36 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -290,10 +290,20 @@ static int ext4_ioctl_setflags(struct inode *inode, if (err) goto flags_out; - if ((jflag ^ oldflags) & (EXT4_JOURNAL_DATA_FL)) + if ((jflag ^ oldflags) & (EXT4_JOURNAL_DATA_FL)) { + /* + * Changes to the journaling mode can cause unsafe changes to + * S_DAX if we are using the DAX mount option. + */ + if (test_opt(inode->i_sb, DAX)) { + err = -EBUSY; + goto flags_out; + } + err = ext4_change_inode_journal_flag(inode, jflag); - if (err) - goto flags_out; + if (err) + goto flags_out; + } if (migrate) { if (flags & EXT4_EXTENTS_FL) err = ext4_ext_migrate(inode);

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] ext4: prevent data corruption with journaling + DAX" failed to apply to 4.9-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.9-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From e9072d859df3e0f2c3ba450f0d1739595c2d5d13 Mon Sep 17 00:00:00 2001 From: Ross Zwisler <ross.zwisler(a)linux.intel.com> Date: Thu, 12 Oct 2017 11:54:08 -0400 Subject: [PATCH] ext4: prevent data corruption with journaling + DAX The current code has the potential for data corruption when changing an inode's journaling mode, as that can result in a subsequent unsafe change in S_DAX. I've captured an instance of this data corruption in the following fstest: https://patchwork.kernel.org/patch/9948377/ Prevent this data corruption from happening by disallowing changes to the journaling mode if the '-o dax' mount option was used. This means that for a given filesystem we could have a mix of inodes using either DAX or data journaling, but whatever state the inodes are in will be held for the duration of the mount. Signed-off-by: Ross Zwisler <ross.zwisler(a)linux.intel.com> Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Reviewed-by: Jan Kara <jack(a)suse.cz> Cc: stable(a)vger.kernel.org diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index edfe95f81274..350e0910ed32 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -6004,11 +6004,6 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) ext4_clear_inode_flag(inode, EXT4_INODE_JOURNAL_DATA); } ext4_set_aops(inode); - /* - * Update inode->i_flags after EXT4_INODE_JOURNAL_DATA was updated. - * E.g. S_DAX may get cleared / set. - */ - ext4_set_inode_flags(inode); jbd2_journal_unlock_updates(journal); percpu_up_write(&sbi->s_journal_flag_rwsem); diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index afb66d4ab5cf..b0b754b37c36 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -290,10 +290,20 @@ static int ext4_ioctl_setflags(struct inode *inode, if (err) goto flags_out; - if ((jflag ^ oldflags) & (EXT4_JOURNAL_DATA_FL)) + if ((jflag ^ oldflags) & (EXT4_JOURNAL_DATA_FL)) { + /* + * Changes to the journaling mode can cause unsafe changes to + * S_DAX if we are using the DAX mount option. + */ + if (test_opt(inode->i_sb, DAX)) { + err = -EBUSY; + goto flags_out; + } + err = ext4_change_inode_journal_flag(inode, jflag); - if (err) - goto flags_out; + if (err) + goto flags_out; + } if (migrate) { if (flags & EXT4_EXTENTS_FL) err = ext4_ext_migrate(inode);

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] md: forbid a RAID5 from having both a bitmap and a journal." failed to apply to 4.14-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.14-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 230b55fa8d64007339319539f8f8e68114d08529 Mon Sep 17 00:00:00 2001 From: NeilBrown <neilb(a)suse.com> Date: Tue, 17 Oct 2017 14:24:09 +1100 Subject: [PATCH] md: forbid a RAID5 from having both a bitmap and a journal. Having both a bitmap and a journal is pointless. Attempting to do so can corrupt the bitmap if the journal replay happens before the bitmap is initialized. Rather than try to avoid this corruption, simply refuse to allow arrays with both a bitmap and a journal. So: - if raid5_run sees both are present, fail. - if adding a bitmap finds a journal is present, fail - if adding a journal finds a bitmap is present, fail. Cc: stable(a)vger.kernel.org (4.10+) Signed-off-by: NeilBrown <neilb(a)suse.com> Tested-by: Joshua Kinard <kumba(a)gentoo.org> Acked-by: Joshua Kinard <kumba(a)gentoo.org> Signed-off-by: Shaohua Li <shli(a)fb.com> diff --git a/drivers/md/md-bitmap.c b/drivers/md/md-bitmap.c index b843b53b0f65..d1b3b60669ea 100644 --- a/drivers/md/md-bitmap.c +++ b/drivers/md/md-bitmap.c @@ -1816,6 +1816,12 @@ struct bitmap *bitmap_create(struct mddev *mddev, int slot) BUG_ON(file && mddev->bitmap_info.offset); + if (test_bit(MD_HAS_JOURNAL, &mddev->flags)) { + pr_notice("md/raid:%s: array with journal cannot have bitmap\n", + mdname(mddev)); + return ERR_PTR(-EBUSY); + } + bitmap = kzalloc(sizeof(*bitmap), GFP_KERNEL); if (!bitmap) return ERR_PTR(-ENOMEM); diff --git a/drivers/md/md.c b/drivers/md/md.c index 97afb28c6f51..6f25e3f1a1cf 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -6362,7 +6362,7 @@ static int add_new_disk(struct mddev *mddev, mdu_disk_info_t *info) break; } } - if (has_journal) { + if (has_journal || mddev->bitmap) { export_rdev(rdev); return -EBUSY; } diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index a21dbd22a2fb..a8732955f130 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -7159,6 +7159,13 @@ static int raid5_run(struct mddev *mddev) min_offset_diff = diff; } + if ((test_bit(MD_HAS_JOURNAL, &mddev->flags) || journal_dev) && + (mddev->bitmap_info.offset || mddev->bitmap_info.file)) { + pr_notice("md/raid:%s: array cannot have both journal and bitmap\n", + mdname(mddev)); + return -EINVAL; + } + if (mddev->reshape_position != MaxSector) { /* Check that we can continue the reshape. * Difficulties arise if the stripe we would write to

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] NFS: revalidate "." etc correctly on "open"." failed to apply to 3.18-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 3.18-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From b688741cb06695312f18b730653d6611e1bad28d Mon Sep 17 00:00:00 2001 From: NeilBrown <neilb(a)suse.com> Date: Fri, 25 Aug 2017 17:34:41 +1000 Subject: [PATCH] NFS: revalidate "." etc correctly on "open". For correct close-to-open semantics, NFS must validate the change attribute of a directory (or file) on open. Since commit ecf3d1f1aa74 ("vfs: kill FS_REVAL_DOT by adding a d_weak_revalidate dentry op"), open() of "." or a path ending ".." is not revalidated reliably (except when that direct is a mount point). Prior to that commit, "." was revalidated using nfs_lookup_revalidate() which checks the LOOKUP_OPEN flag and forces revalidation if the flag is set. Since that commit, nfs_weak_revalidate() is used for NFSv3 (which ignores the flags) and nothing is used for NFSv4. This is fixed by using nfs_lookup_verify_inode() in nfs_weak_revalidate(). This does the revalidation exactly when needed. Also, add a definition of .d_weak_revalidate for NFSv4. The incorrect behavior is easily demonstrated by running "echo *" in some non-mountpoint NFS directory while watching network traffic. Without this patch, "echo *" sometimes doesn't produce any traffic. With the patch it always does. Fixes: ecf3d1f1aa74 ("vfs: kill FS_REVAL_DOT by adding a d_weak_revalidate dentry op") cc: stable(a)vger.kernel.org (3.9+) Signed-off-by: NeilBrown <neilb(a)suse.com> Signed-off-by: Anna Schumaker <Anna.Schumaker(a)Netapp.com> diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index db482d4c15d5..c583093a066b 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1241,8 +1241,7 @@ static int nfs_weak_revalidate(struct dentry *dentry, unsigned int flags) return 0; } - if (nfs_mapping_need_revalidate_inode(inode)) - error = __nfs_revalidate_inode(NFS_SERVER(inode), inode); + error = nfs_lookup_verify_inode(inode, flags); dfprintk(LOOKUPCACHE, "NFS: %s: inode %lu is %s\n", __func__, inode->i_ino, error ? "invalid" : "valid"); return !error; @@ -1393,6 +1392,7 @@ static int nfs4_lookup_revalidate(struct dentry *, unsigned int); const struct dentry_operations nfs4_dentry_operations = { .d_revalidate = nfs4_lookup_revalidate, + .d_weak_revalidate = nfs_weak_revalidate, .d_delete = nfs_dentry_delete, .d_iput = nfs_dentry_iput, .d_automount = nfs_d_automount,

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] NFS: revalidate "." etc correctly on "open"." failed to apply to 4.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From b688741cb06695312f18b730653d6611e1bad28d Mon Sep 17 00:00:00 2001 From: NeilBrown <neilb(a)suse.com> Date: Fri, 25 Aug 2017 17:34:41 +1000 Subject: [PATCH] NFS: revalidate "." etc correctly on "open". For correct close-to-open semantics, NFS must validate the change attribute of a directory (or file) on open. Since commit ecf3d1f1aa74 ("vfs: kill FS_REVAL_DOT by adding a d_weak_revalidate dentry op"), open() of "." or a path ending ".." is not revalidated reliably (except when that direct is a mount point). Prior to that commit, "." was revalidated using nfs_lookup_revalidate() which checks the LOOKUP_OPEN flag and forces revalidation if the flag is set. Since that commit, nfs_weak_revalidate() is used for NFSv3 (which ignores the flags) and nothing is used for NFSv4. This is fixed by using nfs_lookup_verify_inode() in nfs_weak_revalidate(). This does the revalidation exactly when needed. Also, add a definition of .d_weak_revalidate for NFSv4. The incorrect behavior is easily demonstrated by running "echo *" in some non-mountpoint NFS directory while watching network traffic. Without this patch, "echo *" sometimes doesn't produce any traffic. With the patch it always does. Fixes: ecf3d1f1aa74 ("vfs: kill FS_REVAL_DOT by adding a d_weak_revalidate dentry op") cc: stable(a)vger.kernel.org (3.9+) Signed-off-by: NeilBrown <neilb(a)suse.com> Signed-off-by: Anna Schumaker <Anna.Schumaker(a)Netapp.com> diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index db482d4c15d5..c583093a066b 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1241,8 +1241,7 @@ static int nfs_weak_revalidate(struct dentry *dentry, unsigned int flags) return 0; } - if (nfs_mapping_need_revalidate_inode(inode)) - error = __nfs_revalidate_inode(NFS_SERVER(inode), inode); + error = nfs_lookup_verify_inode(inode, flags); dfprintk(LOOKUPCACHE, "NFS: %s: inode %lu is %s\n", __func__, inode->i_ino, error ? "invalid" : "valid"); return !error; @@ -1393,6 +1392,7 @@ static int nfs4_lookup_revalidate(struct dentry *, unsigned int); const struct dentry_operations nfs4_dentry_operations = { .d_revalidate = nfs4_lookup_revalidate, + .d_weak_revalidate = nfs_weak_revalidate, .d_delete = nfs_dentry_delete, .d_iput = nfs_dentry_iput, .d_automount = nfs_d_automount,

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] NFS: revalidate "." etc correctly on "open"." failed to apply to 4.9-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.9-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From b688741cb06695312f18b730653d6611e1bad28d Mon Sep 17 00:00:00 2001 From: NeilBrown <neilb(a)suse.com> Date: Fri, 25 Aug 2017 17:34:41 +1000 Subject: [PATCH] NFS: revalidate "." etc correctly on "open". For correct close-to-open semantics, NFS must validate the change attribute of a directory (or file) on open. Since commit ecf3d1f1aa74 ("vfs: kill FS_REVAL_DOT by adding a d_weak_revalidate dentry op"), open() of "." or a path ending ".." is not revalidated reliably (except when that direct is a mount point). Prior to that commit, "." was revalidated using nfs_lookup_revalidate() which checks the LOOKUP_OPEN flag and forces revalidation if the flag is set. Since that commit, nfs_weak_revalidate() is used for NFSv3 (which ignores the flags) and nothing is used for NFSv4. This is fixed by using nfs_lookup_verify_inode() in nfs_weak_revalidate(). This does the revalidation exactly when needed. Also, add a definition of .d_weak_revalidate for NFSv4. The incorrect behavior is easily demonstrated by running "echo *" in some non-mountpoint NFS directory while watching network traffic. Without this patch, "echo *" sometimes doesn't produce any traffic. With the patch it always does. Fixes: ecf3d1f1aa74 ("vfs: kill FS_REVAL_DOT by adding a d_weak_revalidate dentry op") cc: stable(a)vger.kernel.org (3.9+) Signed-off-by: NeilBrown <neilb(a)suse.com> Signed-off-by: Anna Schumaker <Anna.Schumaker(a)Netapp.com> diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index db482d4c15d5..c583093a066b 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1241,8 +1241,7 @@ static int nfs_weak_revalidate(struct dentry *dentry, unsigned int flags) return 0; } - if (nfs_mapping_need_revalidate_inode(inode)) - error = __nfs_revalidate_inode(NFS_SERVER(inode), inode); + error = nfs_lookup_verify_inode(inode, flags); dfprintk(LOOKUPCACHE, "NFS: %s: inode %lu is %s\n", __func__, inode->i_ino, error ? "invalid" : "valid"); return !error; @@ -1393,6 +1392,7 @@ static int nfs4_lookup_revalidate(struct dentry *, unsigned int); const struct dentry_operations nfs4_dentry_operations = { .d_revalidate = nfs4_lookup_revalidate, + .d_weak_revalidate = nfs_weak_revalidate, .d_delete = nfs_dentry_delete, .d_iput = nfs_dentry_iput, .d_automount = nfs_d_automount,

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] NFS: Avoid RCU usage in tracepoints" failed to apply to 3.18-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 3.18-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 3944369db701f075092357b511fd9f5755771585 Mon Sep 17 00:00:00 2001 From: Anna Schumaker <Anna.Schumaker(a)Netapp.com> Date: Wed, 1 Nov 2017 15:48:43 -0400 Subject: [PATCH] NFS: Avoid RCU usage in tracepoints There isn't an obvious way to acquire and release the RCU lock during a tracepoint, so we can't use the rpc_peeraddr2str() function here. Instead, rely on the client's cl_hostname, which should have similar enough information without needing an rcu_dereference(). Reported-by: Dave Jones <davej(a)codemonkey.org.uk> Cc: stable(a)vger.kernel.org # v3.12 Signed-off-by: Anna Schumaker <Anna.Schumaker(a)Netapp.com> diff --git a/fs/nfs/nfs4trace.h b/fs/nfs/nfs4trace.h index be1da19c65d6..7a5588c25f70 100644 --- a/fs/nfs/nfs4trace.h +++ b/fs/nfs/nfs4trace.h @@ -201,17 +201,13 @@ DECLARE_EVENT_CLASS(nfs4_clientid_event, TP_ARGS(clp, error), TP_STRUCT__entry( - __string(dstaddr, - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR)) + __string(dstaddr, clp->cl_hostname) __field(int, error) ), TP_fast_assign( __entry->error = error; - __assign_str(dstaddr, - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR)); + __assign_str(dstaddr, clp->cl_hostname); ), TP_printk( @@ -1132,9 +1128,7 @@ DECLARE_EVENT_CLASS(nfs4_inode_callback_event, __field(dev_t, dev) __field(u32, fhandle) __field(u64, fileid) - __string(dstaddr, clp ? - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR) : "unknown") + __string(dstaddr, clp ? clp->cl_hostname : "unknown") ), TP_fast_assign( @@ -1147,9 +1141,7 @@ DECLARE_EVENT_CLASS(nfs4_inode_callback_event, __entry->fileid = 0; __entry->dev = 0; } - __assign_str(dstaddr, clp ? - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR) : "unknown") + __assign_str(dstaddr, clp ? clp->cl_hostname : "unknown") ), TP_printk( @@ -1191,9 +1183,7 @@ DECLARE_EVENT_CLASS(nfs4_inode_stateid_callback_event, __field(dev_t, dev) __field(u32, fhandle) __field(u64, fileid) - __string(dstaddr, clp ? - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR) : "unknown") + __string(dstaddr, clp ? clp->cl_hostname : "unknown") __field(int, stateid_seq) __field(u32, stateid_hash) ), @@ -1208,9 +1198,7 @@ DECLARE_EVENT_CLASS(nfs4_inode_stateid_callback_event, __entry->fileid = 0; __entry->dev = 0; } - __assign_str(dstaddr, clp ? - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR) : "unknown") + __assign_str(dstaddr, clp ? clp->cl_hostname : "unknown") __entry->stateid_seq = be32_to_cpu(stateid->seqid); __entry->stateid_hash =

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] NFS: Avoid RCU usage in tracepoints" failed to apply to 4.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 3944369db701f075092357b511fd9f5755771585 Mon Sep 17 00:00:00 2001 From: Anna Schumaker <Anna.Schumaker(a)Netapp.com> Date: Wed, 1 Nov 2017 15:48:43 -0400 Subject: [PATCH] NFS: Avoid RCU usage in tracepoints There isn't an obvious way to acquire and release the RCU lock during a tracepoint, so we can't use the rpc_peeraddr2str() function here. Instead, rely on the client's cl_hostname, which should have similar enough information without needing an rcu_dereference(). Reported-by: Dave Jones <davej(a)codemonkey.org.uk> Cc: stable(a)vger.kernel.org # v3.12 Signed-off-by: Anna Schumaker <Anna.Schumaker(a)Netapp.com> diff --git a/fs/nfs/nfs4trace.h b/fs/nfs/nfs4trace.h index be1da19c65d6..7a5588c25f70 100644 --- a/fs/nfs/nfs4trace.h +++ b/fs/nfs/nfs4trace.h @@ -201,17 +201,13 @@ DECLARE_EVENT_CLASS(nfs4_clientid_event, TP_ARGS(clp, error), TP_STRUCT__entry( - __string(dstaddr, - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR)) + __string(dstaddr, clp->cl_hostname) __field(int, error) ), TP_fast_assign( __entry->error = error; - __assign_str(dstaddr, - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR)); + __assign_str(dstaddr, clp->cl_hostname); ), TP_printk( @@ -1132,9 +1128,7 @@ DECLARE_EVENT_CLASS(nfs4_inode_callback_event, __field(dev_t, dev) __field(u32, fhandle) __field(u64, fileid) - __string(dstaddr, clp ? - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR) : "unknown") + __string(dstaddr, clp ? clp->cl_hostname : "unknown") ), TP_fast_assign( @@ -1147,9 +1141,7 @@ DECLARE_EVENT_CLASS(nfs4_inode_callback_event, __entry->fileid = 0; __entry->dev = 0; } - __assign_str(dstaddr, clp ? - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR) : "unknown") + __assign_str(dstaddr, clp ? clp->cl_hostname : "unknown") ), TP_printk( @@ -1191,9 +1183,7 @@ DECLARE_EVENT_CLASS(nfs4_inode_stateid_callback_event, __field(dev_t, dev) __field(u32, fhandle) __field(u64, fileid) - __string(dstaddr, clp ? - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR) : "unknown") + __string(dstaddr, clp ? clp->cl_hostname : "unknown") __field(int, stateid_seq) __field(u32, stateid_hash) ), @@ -1208,9 +1198,7 @@ DECLARE_EVENT_CLASS(nfs4_inode_stateid_callback_event, __entry->fileid = 0; __entry->dev = 0; } - __assign_str(dstaddr, clp ? - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR) : "unknown") + __assign_str(dstaddr, clp ? clp->cl_hostname : "unknown") __entry->stateid_seq = be32_to_cpu(stateid->seqid); __entry->stateid_hash =

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] btrfs: change how we decide to commit transactions during" failed to apply to 4.9-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.9-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 996478ca9c460886ac147eb0d00e99841b71d31b Mon Sep 17 00:00:00 2001 From: Josef Bacik <jbacik(a)fb.com> Date: Tue, 22 Aug 2017 16:00:39 -0400 Subject: [PATCH] btrfs: change how we decide to commit transactions during flushing Nikolay reported that generic/273 was failing currently with ENOSPC. Turns out this is because we get to the point where the outstanding reservations are greater than the pinned space on the fs. This is a mistake, previously we used the current reservation amount in may_commit_transaction, not the entire outstanding reservation amount. Fix this to find the minimum byte size needed to make progress in flushing, and pass that into may_commit_transaction. From there we can make a smarter decision on whether to commit the transaction or not. This fixes the failure in generic/273. >From Nikolai, IOW: when we go to the final stage of deciding whether to do trans commit, instead of passing all the reservations from all tickets we just pass the reservation for the current ticket. Otherwise, in case all reservations exceed pinned space, then we don't commit transaction and fail prematurely. Before we passed num_bytes from flush_space, where num_bytes was the sum of all pending reserverations, but now all we do is take the first ticket and commit the trans if we can satisfy that. Fixes: 957780eb2788 ("Btrfs: introduce ticketed enospc infrastructure") Cc: stable(a)vger.kernel.org # 4.8 Reported-by: Nikolay Borisov <nborisov(a)suse.com> Signed-off-by: Josef Bacik <jbacik(a)fb.com> Reviewed-by: Nikolay Borisov <nborisov(a)suse.com> Tested-by: Nikolay Borisov <nborisov(a)suse.com> [ added Nikolai's comment ] Signed-off-by: David Sterba <dsterba(a)suse.com> diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index eccaac17258d..1a6aced00a19 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -4907,6 +4907,13 @@ static void shrink_delalloc(struct btrfs_fs_info *fs_info, u64 to_reclaim, } } +struct reserve_ticket { + u64 bytes; + int error; + struct list_head list; + wait_queue_head_t wait; +}; + /** * maybe_commit_transaction - possibly commit the transaction if its ok to * @root - the root we're allocating for @@ -4918,18 +4925,29 @@ static void shrink_delalloc(struct btrfs_fs_info *fs_info, u64 to_reclaim, * will return -ENOSPC. */ static int may_commit_transaction(struct btrfs_fs_info *fs_info, - struct btrfs_space_info *space_info, - u64 bytes, int force) + struct btrfs_space_info *space_info) { + struct reserve_ticket *ticket = NULL; struct btrfs_block_rsv *delayed_rsv = &fs_info->delayed_block_rsv; struct btrfs_trans_handle *trans; + u64 bytes; trans = (struct btrfs_trans_handle *)current->journal_info; if (trans) return -EAGAIN; - if (force) - goto commit; + spin_lock(&space_info->lock); + if (!list_empty(&space_info->priority_tickets)) + ticket = list_first_entry(&space_info->priority_tickets, + struct reserve_ticket, list); + else if (!list_empty(&space_info->tickets)) + ticket = list_first_entry(&space_info->tickets, + struct reserve_ticket, list); + bytes = (ticket) ? ticket->bytes : 0; + spin_unlock(&space_info->lock); + + if (!bytes) + return 0; /* See if there is enough pinned space to make this reservation */ if (percpu_counter_compare(&space_info->total_bytes_pinned, @@ -4944,8 +4962,12 @@ static int may_commit_transaction(struct btrfs_fs_info *fs_info, return -ENOSPC; spin_lock(&delayed_rsv->lock); + if (delayed_rsv->size > bytes) + bytes = 0; + else + bytes -= delayed_rsv->size; if (percpu_counter_compare(&space_info->total_bytes_pinned, - bytes - delayed_rsv->size) < 0) { + bytes) < 0) { spin_unlock(&delayed_rsv->lock); return -ENOSPC; } @@ -4959,13 +4981,6 @@ static int may_commit_transaction(struct btrfs_fs_info *fs_info, return btrfs_commit_transaction(trans); } -struct reserve_ticket { - u64 bytes; - int error; - struct list_head list; - wait_queue_head_t wait; -}; - /* * Try to flush some data based on policy set by @state. This is only advisory * and may fail for various reasons. The caller is supposed to examine the @@ -5015,8 +5030,7 @@ static void flush_space(struct btrfs_fs_info *fs_info, ret = 0; break; case COMMIT_TRANS: - ret = may_commit_transaction(fs_info, space_info, - num_bytes, 0); + ret = may_commit_transaction(fs_info, space_info); break; default: ret = -ENOSPC;

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] fscrypt: lock mutex before checking for bounce page pool" failed to apply to 4.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From a0b3bc855374c50b5ea85273553485af48caf2f7 Mon Sep 17 00:00:00 2001 From: Eric Biggers <ebiggers(a)google.com> Date: Sun, 29 Oct 2017 06:30:19 -0400 Subject: [PATCH] fscrypt: lock mutex before checking for bounce page pool fscrypt_initialize(), which allocates the global bounce page pool when an encrypted file is first accessed, uses "double-checked locking" to try to avoid locking fscrypt_init_mutex. However, it doesn't use any memory barriers, so it's theoretically possible for a thread to observe a bounce page pool which has not been fully initialized. This is a classic bug with "double-checked locking". While "only a theoretical issue" in the latest kernel, in pre-4.8 kernels the pointer that was checked was not even the last to be initialized, so it was easily possible for a crash (NULL pointer dereference) to happen. This was changed only incidentally by the large refactor to use fs/crypto/. Solve both problems in a trivial way that can easily be backported: just always take the mutex. It's theoretically less efficient, but it shouldn't be noticeable in practice as the mutex is only acquired very briefly once per encrypted file. Later I'd like to make this use a helper macro like DO_ONCE(). However, DO_ONCE() runs in atomic context, so we'd need to add a new macro that allows blocking. Cc: stable(a)vger.kernel.org # v4.1+ Signed-off-by: Eric Biggers <ebiggers(a)google.com> Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> diff --git a/fs/crypto/crypto.c b/fs/crypto/crypto.c index 608f6bbe0f31..472326737717 100644 --- a/fs/crypto/crypto.c +++ b/fs/crypto/crypto.c @@ -410,11 +410,8 @@ int fscrypt_initialize(unsigned int cop_flags) { int i, res = -ENOMEM; - /* - * No need to allocate a bounce page pool if there already is one or - * this FS won't use it. - */ - if (cop_flags & FS_CFLG_OWN_PAGES || fscrypt_bounce_page_pool) + /* No need to allocate a bounce page pool if this FS won't use it. */ + if (cop_flags & FS_CFLG_OWN_PAGES) return 0; mutex_lock(&fscrypt_init_mutex);

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] fscrypt: lock mutex before checking for bounce page pool" failed to apply to 4.1-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.1-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From a0b3bc855374c50b5ea85273553485af48caf2f7 Mon Sep 17 00:00:00 2001 From: Eric Biggers <ebiggers(a)google.com> Date: Sun, 29 Oct 2017 06:30:19 -0400 Subject: [PATCH] fscrypt: lock mutex before checking for bounce page pool fscrypt_initialize(), which allocates the global bounce page pool when an encrypted file is first accessed, uses "double-checked locking" to try to avoid locking fscrypt_init_mutex. However, it doesn't use any memory barriers, so it's theoretically possible for a thread to observe a bounce page pool which has not been fully initialized. This is a classic bug with "double-checked locking". While "only a theoretical issue" in the latest kernel, in pre-4.8 kernels the pointer that was checked was not even the last to be initialized, so it was easily possible for a crash (NULL pointer dereference) to happen. This was changed only incidentally by the large refactor to use fs/crypto/. Solve both problems in a trivial way that can easily be backported: just always take the mutex. It's theoretically less efficient, but it shouldn't be noticeable in practice as the mutex is only acquired very briefly once per encrypted file. Later I'd like to make this use a helper macro like DO_ONCE(). However, DO_ONCE() runs in atomic context, so we'd need to add a new macro that allows blocking. Cc: stable(a)vger.kernel.org # v4.1+ Signed-off-by: Eric Biggers <ebiggers(a)google.com> Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> diff --git a/fs/crypto/crypto.c b/fs/crypto/crypto.c index 608f6bbe0f31..472326737717 100644 --- a/fs/crypto/crypto.c +++ b/fs/crypto/crypto.c @@ -410,11 +410,8 @@ int fscrypt_initialize(unsigned int cop_flags) { int i, res = -ENOMEM; - /* - * No need to allocate a bounce page pool if there already is one or - * this FS won't use it. - */ - if (cop_flags & FS_CFLG_OWN_PAGES || fscrypt_bounce_page_pool) + /* No need to allocate a bounce page pool if this FS won't use it. */ + if (cop_flags & FS_CFLG_OWN_PAGES) return 0; mutex_lock(&fscrypt_init_mutex);

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] fscrypt: lock mutex before checking for bounce page pool" failed to apply to 4.9-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.9-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From a0b3bc855374c50b5ea85273553485af48caf2f7 Mon Sep 17 00:00:00 2001 From: Eric Biggers <ebiggers(a)google.com> Date: Sun, 29 Oct 2017 06:30:19 -0400 Subject: [PATCH] fscrypt: lock mutex before checking for bounce page pool fscrypt_initialize(), which allocates the global bounce page pool when an encrypted file is first accessed, uses "double-checked locking" to try to avoid locking fscrypt_init_mutex. However, it doesn't use any memory barriers, so it's theoretically possible for a thread to observe a bounce page pool which has not been fully initialized. This is a classic bug with "double-checked locking". While "only a theoretical issue" in the latest kernel, in pre-4.8 kernels the pointer that was checked was not even the last to be initialized, so it was easily possible for a crash (NULL pointer dereference) to happen. This was changed only incidentally by the large refactor to use fs/crypto/. Solve both problems in a trivial way that can easily be backported: just always take the mutex. It's theoretically less efficient, but it shouldn't be noticeable in practice as the mutex is only acquired very briefly once per encrypted file. Later I'd like to make this use a helper macro like DO_ONCE(). However, DO_ONCE() runs in atomic context, so we'd need to add a new macro that allows blocking. Cc: stable(a)vger.kernel.org # v4.1+ Signed-off-by: Eric Biggers <ebiggers(a)google.com> Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> diff --git a/fs/crypto/crypto.c b/fs/crypto/crypto.c index 608f6bbe0f31..472326737717 100644 --- a/fs/crypto/crypto.c +++ b/fs/crypto/crypto.c @@ -410,11 +410,8 @@ int fscrypt_initialize(unsigned int cop_flags) { int i, res = -ENOMEM; - /* - * No need to allocate a bounce page pool if there already is one or - * this FS won't use it. - */ - if (cop_flags & FS_CFLG_OWN_PAGES || fscrypt_bounce_page_pool) + /* No need to allocate a bounce page pool if this FS won't use it. */ + if (cop_flags & FS_CFLG_OWN_PAGES) return 0; mutex_lock(&fscrypt_init_mutex);

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] mm/z3fold.c: use kref to prevent page free/compact race" failed to apply to 4.9-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.9-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 5d03a6613957785e94af7a4a6212ad4af66aa5c2 Mon Sep 17 00:00:00 2001 From: Vitaly Wool <vitalywool(a)gmail.com> Date: Fri, 17 Nov 2017 15:26:16 -0800 Subject: [PATCH] mm/z3fold.c: use kref to prevent page free/compact race There is a race in the current z3fold implementation between do_compact() called in a work queue context and the page release procedure when page's kref goes to 0. do_compact() may be waiting for page lock, which is released by release_z3fold_page_locked right before putting the page onto the "stale" list, and then the page may be freed as do_compact() modifies its contents. The mechanism currently implemented to handle that (checking the PAGE_STALE flag) is not reliable enough. Instead, we'll use page's kref counter to guarantee that the page is not released if its compaction is scheduled. It then becomes compaction function's responsibility to decrease the counter and quit immediately if the page was actually freed. Link: http://lkml.kernel.org/r/20171117092032.00ea56f42affbed19f4fcc6c@gmail.com Signed-off-by: Vitaly Wool <vitaly.wool(a)sonymobile.com> Cc: <Oleksiy.Avramchenko(a)sony.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> diff --git a/mm/z3fold.c b/mm/z3fold.c index b2ba2ba585f3..39e19125d6a0 100644 --- a/mm/z3fold.c +++ b/mm/z3fold.c @@ -404,8 +404,7 @@ static void do_compact_page(struct z3fold_header *zhdr, bool locked) WARN_ON(z3fold_page_trylock(zhdr)); else z3fold_page_lock(zhdr); - if (test_bit(PAGE_STALE, &page->private) || - !test_and_clear_bit(NEEDS_COMPACTING, &page->private)) { + if (WARN_ON(!test_and_clear_bit(NEEDS_COMPACTING, &page->private))) { z3fold_page_unlock(zhdr); return; } @@ -413,6 +412,11 @@ static void do_compact_page(struct z3fold_header *zhdr, bool locked) list_del_init(&zhdr->buddy); spin_unlock(&pool->lock); + if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) { + atomic64_dec(&pool->pages_nr); + return; + } + z3fold_compact_page(zhdr); unbuddied = get_cpu_ptr(pool->unbuddied); fchunks = num_free_chunks(zhdr); @@ -753,9 +757,11 @@ static void z3fold_free(struct z3fold_pool *pool, unsigned long handle) list_del_init(&zhdr->buddy); spin_unlock(&pool->lock); zhdr->cpu = -1; + kref_get(&zhdr->refcount); do_compact_page(zhdr, true); return; } + kref_get(&zhdr->refcount); queue_work_on(zhdr->cpu, pool->compact_wq, &zhdr->work); z3fold_page_unlock(zhdr); }

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] MIPS: math-emu: Fix final emulation phase for certain" failed to apply to 4.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 409fcace9963c1e8d2cb0f7ac62e8b34d47ef979 Mon Sep 17 00:00:00 2001 From: Aleksandar Markovic <aleksandar.markovic(a)mips.com> Date: Thu, 2 Nov 2017 12:13:58 +0100 Subject: [PATCH] MIPS: math-emu: Fix final emulation phase for certain instructions Fix final phase of <CLASS|MADDF|MSUBF|MAX|MIN|MAXA|MINA>.<D|S> emulation. Provide proper generation of SIGFPE signal and updating debugfs FP exception stats in cases of any exception flags set in preceding phases of emulation. CLASS.<D|S> instruction may generate "Unimplemented Operation" FP exception. <MADDF|MSUBF>.<D|S> instructions may generate "Inexact", "Unimplemented Operation", "Invalid Operation", "Overflow", and "Underflow" FP exceptions. <MAX|MIN|MAXA|MINA>.<D|S> instructions can generate "Unimplemented Operation" and "Invalid Operation" FP exceptions. The proper final processing of the cases when any FP exception flag is set is achieved by replacing "break" statement with "goto copcsr" statement. With such solution, this patch brings the final phase of emulation of the above instructions consistent with the one corresponding to the previously implemented emulation of other related FPU instructions (ADD, SUB, etc.). Fixes: 38db37ba069f ("MIPS: math-emu: Add support for the MIPS R6 CLASS FPU instruction") Fixes: e24c3bec3e8e ("MIPS: math-emu: Add support for the MIPS R6 MADDF FPU instruction") Fixes: 83d43305a1df ("MIPS: math-emu: Add support for the MIPS R6 MSUBF FPU instruction") Fixes: a79f5f9ba508 ("MIPS: math-emu: Add support for the MIPS R6 MAX{, A} FPU instruction") Fixes: 4e9561b20e2f ("MIPS: math-emu: Add support for the MIPS R6 MIN{, A} FPU instruction") Signed-off-by: Aleksandar Markovic <aleksandar.markovic(a)mips.com> Cc: Ralf Baechle <ralf(a)linux-mips.org> Cc: Douglas Leung <douglas.leung(a)mips.com> Cc: Goran Ferenc <goran.ferenc(a)mips.com> Cc: "Maciej W. Rozycki" <macro(a)imgtec.com> Cc: Miodrag Dinic <miodrag.dinic(a)mips.com> Cc: Paul Burton <paul.burton(a)mips.com> Cc: Petar Jovanovic <petar.jovanovic(a)mips.com> Cc: Raghu Gandham <raghu.gandham(a)mips.com> Cc: linux-mips(a)linux-mips.org Cc: <stable(a)vger.kernel.org> # 4.3+ Patchwork: https://patchwork.linux-mips.org/patch/17581/ Signed-off-by: James Hogan <jhogan(a)kernel.org> diff --git a/arch/mips/math-emu/cp1emu.c b/arch/mips/math-emu/cp1emu.c index 192542dbd972..d2fcb3084279 100644 --- a/arch/mips/math-emu/cp1emu.c +++ b/arch/mips/math-emu/cp1emu.c @@ -1795,7 +1795,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, SPFROMREG(fs, MIPSInst_FS(ir)); SPFROMREG(fd, MIPSInst_FD(ir)); rv.s = ieee754sp_maddf(fd, fs, ft); - break; + goto copcsr; } case fmsubf_op: { @@ -1809,7 +1809,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, SPFROMREG(fs, MIPSInst_FS(ir)); SPFROMREG(fd, MIPSInst_FD(ir)); rv.s = ieee754sp_msubf(fd, fs, ft); - break; + goto copcsr; } case frint_op: { @@ -1834,7 +1834,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, SPFROMREG(fs, MIPSInst_FS(ir)); rv.w = ieee754sp_2008class(fs); rfmt = w_fmt; - break; + goto copcsr; } case fmin_op: { @@ -1847,7 +1847,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, SPFROMREG(ft, MIPSInst_FT(ir)); SPFROMREG(fs, MIPSInst_FS(ir)); rv.s = ieee754sp_fmin(fs, ft); - break; + goto copcsr; } case fmina_op: { @@ -1860,7 +1860,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, SPFROMREG(ft, MIPSInst_FT(ir)); SPFROMREG(fs, MIPSInst_FS(ir)); rv.s = ieee754sp_fmina(fs, ft); - break; + goto copcsr; } case fmax_op: { @@ -1873,7 +1873,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, SPFROMREG(ft, MIPSInst_FT(ir)); SPFROMREG(fs, MIPSInst_FS(ir)); rv.s = ieee754sp_fmax(fs, ft); - break; + goto copcsr; } case fmaxa_op: { @@ -1886,7 +1886,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, SPFROMREG(ft, MIPSInst_FT(ir)); SPFROMREG(fs, MIPSInst_FS(ir)); rv.s = ieee754sp_fmaxa(fs, ft); - break; + goto copcsr; } case fabs_op: @@ -2165,7 +2165,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, DPFROMREG(fs, MIPSInst_FS(ir)); DPFROMREG(fd, MIPSInst_FD(ir)); rv.d = ieee754dp_maddf(fd, fs, ft); - break; + goto copcsr; } case fmsubf_op: { @@ -2179,7 +2179,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, DPFROMREG(fs, MIPSInst_FS(ir)); DPFROMREG(fd, MIPSInst_FD(ir)); rv.d = ieee754dp_msubf(fd, fs, ft); - break; + goto copcsr; } case frint_op: { @@ -2204,7 +2204,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, DPFROMREG(fs, MIPSInst_FS(ir)); rv.l = ieee754dp_2008class(fs); rfmt = l_fmt; - break; + goto copcsr; } case fmin_op: { @@ -2217,7 +2217,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, DPFROMREG(ft, MIPSInst_FT(ir)); DPFROMREG(fs, MIPSInst_FS(ir)); rv.d = ieee754dp_fmin(fs, ft); - break; + goto copcsr; } case fmina_op: { @@ -2230,7 +2230,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, DPFROMREG(ft, MIPSInst_FT(ir)); DPFROMREG(fs, MIPSInst_FS(ir)); rv.d = ieee754dp_fmina(fs, ft); - break; + goto copcsr; } case fmax_op: { @@ -2243,7 +2243,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, DPFROMREG(ft, MIPSInst_FT(ir)); DPFROMREG(fs, MIPSInst_FS(ir)); rv.d = ieee754dp_fmax(fs, ft); - break; + goto copcsr; } case fmaxa_op: { @@ -2256,7 +2256,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, DPFROMREG(ft, MIPSInst_FT(ir)); DPFROMREG(fs, MIPSInst_FS(ir)); rv.d = ieee754dp_fmaxa(fs, ft); - break; + goto copcsr; } case fabs_op:

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] MIPS: Fix odd fp register warnings with MIPS64r2" failed to apply to 4.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From c7fd89a6407ea3a44a2a2fa12d290162c42499c4 Mon Sep 17 00:00:00 2001 From: James Hogan <jhogan(a)kernel.org> Date: Fri, 10 Nov 2017 11:46:54 +0000 Subject: [PATCH] MIPS: Fix odd fp register warnings with MIPS64r2 Building 32-bit MIPS64r2 kernels produces warnings like the following on certain toolchains (such as GNU assembler 2.24.90, but not GNU assembler 2.28.51) since commit 22b8ba765a72 ("MIPS: Fix MIPS64 FP save/restore on 32-bit kernels"), due to the exposure of fpu_save_16odd from fpu_save_double and fpu_restore_16odd from fpu_restore_double: arch/mips/kernel/r4k_fpu.S:47: Warning: float register should be even, was 1 ... arch/mips/kernel/r4k_fpu.S:59: Warning: float register should be even, was 1 ... This appears to be because .set mips64r2 does not change the FPU ABI to 64-bit when -march=mips64r2 (or e.g. -march=xlp) is provided on the command line on that toolchain, from the default FPU ABI of 32-bit due to the -mabi=32. This makes access to the odd FPU registers invalid. Fix by explicitly changing the FPU ABI with .set fp=64 directives in fpu_save_16odd and fpu_restore_16odd, and moving the undefine of fp up in asmmacro.h so fp doesn't turn into $30. Fixes: 22b8ba765a72 ("MIPS: Fix MIPS64 FP save/restore on 32-bit kernels") Signed-off-by: James Hogan <jhogan(a)kernel.org> Cc: Ralf Baechle <ralf(a)linux-mips.org> Cc: Paul Burton <paul.burton(a)imgtec.com> Cc: linux-mips(a)linux-mips.org Cc: <stable(a)vger.kernel.org> # 4.0+: 22b8ba765a72: MIPS: Fix MIPS64 FP save/restore on 32-bit kernels Cc: <stable(a)vger.kernel.org> # 4.0+ Patchwork: https://patchwork.linux-mips.org/patch/17656/ diff --git a/arch/mips/include/asm/asmmacro.h b/arch/mips/include/asm/asmmacro.h index b815d7b3bd27..feb069cbf44e 100644 --- a/arch/mips/include/asm/asmmacro.h +++ b/arch/mips/include/asm/asmmacro.h @@ -19,6 +19,9 @@ #include <asm/asmmacro-64.h> #endif +/* preprocessor replaces the fp in ".set fp=64" with $30 otherwise */ +#undef fp + /* * Helper macros for generating raw instruction encodings. */ @@ -105,6 +108,7 @@ .macro fpu_save_16odd thread .set push .set mips64r2 + .set fp=64 SET_HARDFLOAT sdc1 $f1, THREAD_FPR1(\thread) sdc1 $f3, THREAD_FPR3(\thread) @@ -163,6 +167,7 @@ .macro fpu_restore_16odd thread .set push .set mips64r2 + .set fp=64 SET_HARDFLOAT ldc1 $f1, THREAD_FPR1(\thread) ldc1 $f3, THREAD_FPR3(\thread) @@ -234,9 +239,6 @@ .endm #ifdef TOOLCHAIN_SUPPORTS_MSA -/* preprocessor replaces the fp in ".set fp=64" with $30 otherwise */ -#undef fp - .macro _cfcmsa rd, cs .set push .set mips32r2

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] MIPS: Fix MIPS64 FP save/restore on 32-bit kernels" failed to apply to 4.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 22b8ba765a726d90e9830ff6134c32b04f12c10f Mon Sep 17 00:00:00 2001 From: James Hogan <jhogan(a)kernel.org> Date: Mon, 3 Jul 2017 23:41:47 +0100 Subject: [PATCH] MIPS: Fix MIPS64 FP save/restore on 32-bit kernels 32-bit kernels can be configured to support MIPS64, in which case neither CONFIG_64BIT or CONFIG_CPU_MIPS32_R* will be set. This causes the CP0_Status.FR checks at the point of floating point register save and restore to be compiled out, which results in odd FP registers not being saved or restored to the task or signal context even when CP0_Status.FR is set. Fix the ifdefs to use CONFIG_CPU_MIPSR2 and CONFIG_CPU_MIPSR6, which are enabled for the relevant revisions of either MIPS32 or MIPS64, along with some other CPUs such as Octeon (r2), Loongson1 (r2), XLP (r2), Loongson 3A R2. The suspect code originates from commit 597ce1723e0f ("MIPS: Support for 64-bit FP with O32 binaries") in v3.14, however the code in __enable_fpu() was consistent and refused to set FR=1, falling back to software FPU emulation. This was suboptimal but should be functionally correct. Commit fcc53b5f6c38 ("MIPS: fpu.h: Allow 64-bit FPU on a 64-bit MIPS R6 CPU") in v4.2 (and stable tagged back to 4.0) later introduced the bug by updating __enable_fpu() to set FR=1 but failing to update the other similar ifdefs to enable FR=1 state handling. Fixes: fcc53b5f6c38 ("MIPS: fpu.h: Allow 64-bit FPU on a 64-bit MIPS R6 CPU") Signed-off-by: James Hogan <jhogan(a)kernel.org> Cc: Ralf Baechle <ralf(a)linux-mips.org> Cc: Paul Burton <paul.burton(a)imgtec.com> Cc: linux-mips(a)linux-mips.org Cc: <stable(a)vger.kernel.org> # 4.0+ Patchwork: https://patchwork.linux-mips.org/patch/16739/ diff --git a/arch/mips/include/asm/asmmacro.h b/arch/mips/include/asm/asmmacro.h index 83054f79f72a..b815d7b3bd27 100644 --- a/arch/mips/include/asm/asmmacro.h +++ b/arch/mips/include/asm/asmmacro.h @@ -126,8 +126,8 @@ .endm .macro fpu_save_double thread status tmp -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) sll \tmp, \status, 5 bgez \tmp, 10f fpu_save_16odd \thread @@ -184,8 +184,8 @@ .endm .macro fpu_restore_double thread status tmp -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) sll \tmp, \status, 5 bgez \tmp, 10f # 16 register mode? diff --git a/arch/mips/kernel/r4k_fpu.S b/arch/mips/kernel/r4k_fpu.S index 0a83b1708b3c..8e3a6020c613 100644 --- a/arch/mips/kernel/r4k_fpu.S +++ b/arch/mips/kernel/r4k_fpu.S @@ -40,8 +40,8 @@ */ LEAF(_save_fp) EXPORT_SYMBOL(_save_fp) -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) mfc0 t0, CP0_STATUS #endif fpu_save_double a0 t0 t1 # clobbers t1 @@ -52,8 +52,8 @@ EXPORT_SYMBOL(_save_fp) * Restore a thread's fp context. */ LEAF(_restore_fp) -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) mfc0 t0, CP0_STATUS #endif fpu_restore_double a0 t0 t1 # clobbers t1 @@ -246,11 +246,11 @@ LEAF(_save_fp_context) cfc1 t1, fcr31 .set pop -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) .set push SET_HARDFLOAT -#ifdef CONFIG_CPU_MIPS32_R2 +#ifdef CONFIG_CPU_MIPSR2 .set mips32r2 .set fp=64 mfc0 t0, CP0_STATUS @@ -314,11 +314,11 @@ LEAF(_save_fp_context) LEAF(_restore_fp_context) EX lw t1, 0(a1) -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) .set push SET_HARDFLOAT -#ifdef CONFIG_CPU_MIPS32_R2 +#ifdef CONFIG_CPU_MIPSR2 .set mips32r2 .set fp=64 mfc0 t0, CP0_STATUS

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] MIPS: Fix MIPS64 FP save/restore on 32-bit kernels" failed to apply to 4.9-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.9-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 22b8ba765a726d90e9830ff6134c32b04f12c10f Mon Sep 17 00:00:00 2001 From: James Hogan <jhogan(a)kernel.org> Date: Mon, 3 Jul 2017 23:41:47 +0100 Subject: [PATCH] MIPS: Fix MIPS64 FP save/restore on 32-bit kernels 32-bit kernels can be configured to support MIPS64, in which case neither CONFIG_64BIT or CONFIG_CPU_MIPS32_R* will be set. This causes the CP0_Status.FR checks at the point of floating point register save and restore to be compiled out, which results in odd FP registers not being saved or restored to the task or signal context even when CP0_Status.FR is set. Fix the ifdefs to use CONFIG_CPU_MIPSR2 and CONFIG_CPU_MIPSR6, which are enabled for the relevant revisions of either MIPS32 or MIPS64, along with some other CPUs such as Octeon (r2), Loongson1 (r2), XLP (r2), Loongson 3A R2. The suspect code originates from commit 597ce1723e0f ("MIPS: Support for 64-bit FP with O32 binaries") in v3.14, however the code in __enable_fpu() was consistent and refused to set FR=1, falling back to software FPU emulation. This was suboptimal but should be functionally correct. Commit fcc53b5f6c38 ("MIPS: fpu.h: Allow 64-bit FPU on a 64-bit MIPS R6 CPU") in v4.2 (and stable tagged back to 4.0) later introduced the bug by updating __enable_fpu() to set FR=1 but failing to update the other similar ifdefs to enable FR=1 state handling. Fixes: fcc53b5f6c38 ("MIPS: fpu.h: Allow 64-bit FPU on a 64-bit MIPS R6 CPU") Signed-off-by: James Hogan <jhogan(a)kernel.org> Cc: Ralf Baechle <ralf(a)linux-mips.org> Cc: Paul Burton <paul.burton(a)imgtec.com> Cc: linux-mips(a)linux-mips.org Cc: <stable(a)vger.kernel.org> # 4.0+ Patchwork: https://patchwork.linux-mips.org/patch/16739/ diff --git a/arch/mips/include/asm/asmmacro.h b/arch/mips/include/asm/asmmacro.h index 83054f79f72a..b815d7b3bd27 100644 --- a/arch/mips/include/asm/asmmacro.h +++ b/arch/mips/include/asm/asmmacro.h @@ -126,8 +126,8 @@ .endm .macro fpu_save_double thread status tmp -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) sll \tmp, \status, 5 bgez \tmp, 10f fpu_save_16odd \thread @@ -184,8 +184,8 @@ .endm .macro fpu_restore_double thread status tmp -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) sll \tmp, \status, 5 bgez \tmp, 10f # 16 register mode? diff --git a/arch/mips/kernel/r4k_fpu.S b/arch/mips/kernel/r4k_fpu.S index 0a83b1708b3c..8e3a6020c613 100644 --- a/arch/mips/kernel/r4k_fpu.S +++ b/arch/mips/kernel/r4k_fpu.S @@ -40,8 +40,8 @@ */ LEAF(_save_fp) EXPORT_SYMBOL(_save_fp) -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) mfc0 t0, CP0_STATUS #endif fpu_save_double a0 t0 t1 # clobbers t1 @@ -52,8 +52,8 @@ EXPORT_SYMBOL(_save_fp) * Restore a thread's fp context. */ LEAF(_restore_fp) -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) mfc0 t0, CP0_STATUS #endif fpu_restore_double a0 t0 t1 # clobbers t1 @@ -246,11 +246,11 @@ LEAF(_save_fp_context) cfc1 t1, fcr31 .set pop -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) .set push SET_HARDFLOAT -#ifdef CONFIG_CPU_MIPS32_R2 +#ifdef CONFIG_CPU_MIPSR2 .set mips32r2 .set fp=64 mfc0 t0, CP0_STATUS @@ -314,11 +314,11 @@ LEAF(_save_fp_context) LEAF(_restore_fp_context) EX lw t1, 0(a1) -#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2) || \ - defined(CONFIG_CPU_MIPS32_R6) +#if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPSR2) || \ + defined(CONFIG_CPU_MIPSR6) .set push SET_HARDFLOAT -#ifdef CONFIG_CPU_MIPS32_R2 +#ifdef CONFIG_CPU_MIPSR2 .set mips32r2 .set fp=64 mfc0 t0, CP0_STATUS

7 years, 7 months

1
0
0 0

Re: [Linux-stable-mirror] WTF: patch "[PATCH] MIPS: pci: Remove KERN_WARN instance inside the mt7620 driver" was seriously submitted to be applied to the 4.9-stable tree?

by Greg KH

On Mon, Nov 27, 2017 at 02:13:00PM +0000, James Hogan wrote: > On Mon, Nov 27, 2017 at 01:56:49PM +0100, Greg KH wrote: > > On Mon, Nov 27, 2017 at 12:40:36PM +0000, James Hogan wrote: > > > Hi Greg, > > > > > > On Mon, Nov 27, 2017 at 01:35:46PM +0100, gregkh(a)linuxfoundation.org wrote: > > > > The patch below was submitted to be applied to the 4.9-stable tree. > > > > > > > > I fail to see how this patch meets the stable kernel rules as found at > > > > Documentation/process/stable-kernel-rules.rst. > > > > > > > > I could be totally wrong, and if so, please respond to > > > > <stable(a)vger.kernel.org> and let me know why this patch should be > > > > applied. Otherwise, it is now dropped from my patch queues, never to be > > > > seen again. > > > > > > I should have adjusted the commit message. KERN_WARN doesn't exist so it > > > actually fixes a build error as well as switching to pr_warn(). > > > > What build error? I have not heard of this breaking the build on 4.9 > > for the past year, is it in some config that no one uses? :) > > The LEDE project has been carrying the patch [1] since February when > they added 4.9 support (their 4.4 support had a slightly earlier version > of the driver added with just a plain printk, no KERN_WARN). > > They have both CONFIG_SOC_MT7620 and CONFIG_PCI=y in their ralink mt7620 > config [2], and they are keeping up to date with stable releases [3], so > I have no doubt they would appreciate having the patch applied to > upstream stable to reduce their delta. > > The only defconfigs in mainline which enable this platform > (CONFIG_SOC_MT7620) are omega2p_defconfig and vocore2_defconfig, which > were added in August by Harvey to help widen our internal continuous > build & boot test coverage. Neither defconfig enables CONFIG_PCI yet > which is required to see the build failure below, but regardless it is a > valid configuration which LEDE is actively using. > > arch/mips/pci/pci-mt7620.c: In function ‘wait_pciephy_busy’: > arch/mips/pci/pci-mt7620.c:123:11: error: ‘KERN_WARN’ undeclared (first use in this function) > printk(KERN_WARN "PCIE-PHY retry failed.\n"); > ^~~~~~~~~ > > John: I'm not familiar with the hardware, but would it be appropriate to > add CONFIG_PCI=y to either of those 2 defconfigs (omega2p_defconfig and > vocore2_defconfig) so this driver gets some upstream build[/boot] > testing? > > Anyway, hopefully that helps allay stable backport concerns. Yes, thanks, that explains it a lot better, now queued up. greg k-h

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] [-stable PATCH] sched/rt: Simplify the IPI based RT balancing logic

by Ingo Molnar

hi Greg, Please include the commit below in -stable, once it gets upstream. Maybe also add: Fixes: b6366f048e0c: ("sched/rt: Use IPI to trigger RT task push migration instead of pulling") because it fixes a pretty old bug. Thanks, Ingo ----- Forwarded message from "tip-bot for Steven Rostedt (Red Hat)" <tipbot(a)zytor.com> ----- Date: Tue, 10 Oct 2017 04:02:17 -0700 From: "tip-bot for Steven Rostedt (Red Hat)" <tipbot(a)zytor.com> To: linux-tip-commits(a)vger.kernel.org Cc: rostedt(a)goodmis.org, williams(a)redhat.com, mingo(a)kernel.org, tglx(a)linutronix.de, peterz(a)infradead.org, linux-kernel(a)vger.kernel.org, jkacur(a)redhat.com, bristot(a)redhat.com, torvalds(a)linux-foundation.org, efault(a)gmx.de, hpa(a)zytor.com, swood(a)redhat.com Subject: [tip:sched/core] sched/rt: Simplify the IPI based RT balancing logic Commit-ID: 4bdced5c9a2922521e325896a7bbbf0132c94e56 Gitweb: https://git.kernel.org/tip/4bdced5c9a2922521e325896a7bbbf0132c94e56 Author: Steven Rostedt (Red Hat) <rostedt(a)goodmis.org> AuthorDate: Fri, 6 Oct 2017 14:05:04 -0400 Committer: Ingo Molnar <mingo(a)kernel.org> CommitDate: Tue, 10 Oct 2017 11:45:40 +0200 sched/rt: Simplify the IPI based RT balancing logic When a CPU lowers its priority (schedules out a high priority task for a lower priority one), a check is made to see if any other CPU has overloaded RT tasks (more than one). It checks the rto_mask to determine this and if so it will request to pull one of those tasks to itself if the non running RT task is of higher priority than the new priority of the next task to run on the current CPU. When we deal with large number of CPUs, the original pull logic suffered from large lock contention on a single CPU run queue, which caused a huge latency across all CPUs. This was caused by only having one CPU having overloaded RT tasks and a bunch of other CPUs lowering their priority. To solve this issue, commit: b6366f048e0c ("sched/rt: Use IPI to trigger RT task push migration instead of pulling") changed the way to request a pull. Instead of grabbing the lock of the overloaded CPU's runqueue, it simply sent an IPI to that CPU to do the work. Although the IPI logic worked very well in removing the large latency build up, it still could suffer from a large number of IPIs being sent to a single CPU. On a 80 CPU box, I measured over 200us of processing IPIs. Worse yet, when I tested this on a 120 CPU box, with a stress test that had lots of RT tasks scheduling on all CPUs, it actually triggered the hard lockup detector! One CPU had so many IPIs sent to it, and due to the restart mechanism that is triggered when the source run queue has a priority status change, the CPU spent minutes! processing the IPIs. Thinking about this further, I realized there's no reason for each run queue to send its own IPI. As all CPUs with overloaded tasks must be scanned regardless if there's one or many CPUs lowering their priority, because there's no current way to find the CPU with the highest priority task that can schedule to one of these CPUs, there really only needs to be one IPI being sent around at a time. This greatly simplifies the code! The new approach is to have each root domain have its own irq work, as the rto_mask is per root domain. The root domain has the following fields attached to it: rto_push_work - the irq work to process each CPU set in rto_mask rto_lock - the lock to protect some of the other rto fields rto_loop_start - an atomic that keeps contention down on rto_lock the first CPU scheduling in a lower priority task is the one to kick off the process. rto_loop_next - an atomic that gets incremented for each CPU that schedules in a lower priority task. rto_loop - a variable protected by rto_lock that is used to compare against rto_loop_next rto_cpu - The cpu to send the next IPI to, also protected by the rto_lock. When a CPU schedules in a lower priority task and wants to make sure overloaded CPUs know about it. It increments the rto_loop_next. Then it atomically sets rto_loop_start with a cmpxchg. If the old value is not "0", then it is done, as another CPU is kicking off the IPI loop. If the old value is "0", then it will take the rto_lock to synchronize with a possible IPI being sent around to the overloaded CPUs. If rto_cpu is greater than or equal to nr_cpu_ids, then there's either no IPI being sent around, or one is about to finish. Then rto_cpu is set to the first CPU in rto_mask and an IPI is sent to that CPU. If there's no CPUs set in rto_mask, then there's nothing to be done. When the CPU receives the IPI, it will first try to push any RT tasks that is queued on the CPU but can't run because a higher priority RT task is currently running on that CPU. Then it takes the rto_lock and looks for the next CPU in the rto_mask. If it finds one, it simply sends an IPI to that CPU and the process continues. If there's no more CPUs in the rto_mask, then rto_loop is compared with rto_loop_next. If they match, everything is done and the process is over. If they do not match, then a CPU scheduled in a lower priority task as the IPI was being passed around, and the process needs to start again. The first CPU in rto_mask is sent the IPI. This change removes this duplication of work in the IPI logic, and greatly lowers the latency caused by the IPIs. This removed the lockup happening on the 120 CPU machine. It also simplifies the code tremendously. What else could anyone ask for? Thanks to Peter Zijlstra for simplifying the rto_loop_start atomic logic and supplying me with the rto_start_trylock() and rto_start_unlock() helper functions. Signed-off-by: Steven Rostedt (VMware) <rostedt(a)goodmis.org> Signed-off-by: Peter Zijlstra (Intel) <peterz(a)infradead.org> Cc: Clark Williams <williams(a)redhat.com> Cc: Daniel Bristot de Oliveira <bristot(a)redhat.com> Cc: John Kacur <jkacur(a)redhat.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mike Galbraith <efault(a)gmx.de> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Scott Wood <swood(a)redhat.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Link: http://lkml.kernel.org/r/20170424114732.1aac6dc4@gandalf.local.home Signed-off-by: Ingo Molnar <mingo(a)kernel.org> --- kernel/sched/rt.c | 316 ++++++++++++++++++------------------------------ kernel/sched/sched.h | 24 ++-- kernel/sched/topology.c | 6 + 3 files changed, 138 insertions(+), 208 deletions(-) diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index 0af5ca9..fda2799 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -73,10 +73,6 @@ static void start_rt_bandwidth(struct rt_bandwidth *rt_b) raw_spin_unlock(&rt_b->rt_runtime_lock); } -#if defined(CONFIG_SMP) && defined(HAVE_RT_PUSH_IPI) -static void push_irq_work_func(struct irq_work *work); -#endif - void init_rt_rq(struct rt_rq *rt_rq) { struct rt_prio_array *array; @@ -96,13 +92,6 @@ void init_rt_rq(struct rt_rq *rt_rq) rt_rq->rt_nr_migratory = 0; rt_rq->overloaded = 0; plist_head_init(&rt_rq->pushable_tasks); - -#ifdef HAVE_RT_PUSH_IPI - rt_rq->push_flags = 0; - rt_rq->push_cpu = nr_cpu_ids; - raw_spin_lock_init(&rt_rq->push_lock); - init_irq_work(&rt_rq->push_work, push_irq_work_func); -#endif #endif /* CONFIG_SMP */ /* We start is dequeued state, because no RT tasks are queued */ rt_rq->rt_queued = 0; @@ -1875,241 +1864,166 @@ static void push_rt_tasks(struct rq *rq) } #ifdef HAVE_RT_PUSH_IPI + /* - * The search for the next cpu always starts at rq->cpu and ends - * when we reach rq->cpu again. It will never return rq->cpu. - * This returns the next cpu to check, or nr_cpu_ids if the loop - * is complete. + * When a high priority task schedules out from a CPU and a lower priority + * task is scheduled in, a check is made to see if there's any RT tasks + * on other CPUs that are waiting to run because a higher priority RT task + * is currently running on its CPU. In this case, the CPU with multiple RT + * tasks queued on it (overloaded) needs to be notified that a CPU has opened + * up that may be able to run one of its non-running queued RT tasks. + * + * All CPUs with overloaded RT tasks need to be notified as there is currently + * no way to know which of these CPUs have the highest priority task waiting + * to run. Instead of trying to take a spinlock on each of these CPUs, + * which has shown to cause large latency when done on machines with many + * CPUs, sending an IPI to the CPUs to have them push off the overloaded + * RT tasks waiting to run. + * + * Just sending an IPI to each of the CPUs is also an issue, as on large + * count CPU machines, this can cause an IPI storm on a CPU, especially + * if its the only CPU with multiple RT tasks queued, and a large number + * of CPUs scheduling a lower priority task at the same time. + * + * Each root domain has its own irq work function that can iterate over + * all CPUs with RT overloaded tasks. Since all CPUs with overloaded RT + * tassk must be checked if there's one or many CPUs that are lowering + * their priority, there's a single irq work iterator that will try to + * push off RT tasks that are waiting to run. + * + * When a CPU schedules a lower priority task, it will kick off the + * irq work iterator that will jump to each CPU with overloaded RT tasks. + * As it only takes the first CPU that schedules a lower priority task + * to start the process, the rto_start variable is incremented and if + * the atomic result is one, then that CPU will try to take the rto_lock. + * This prevents high contention on the lock as the process handles all + * CPUs scheduling lower priority tasks. + * + * All CPUs that are scheduling a lower priority task will increment the + * rt_loop_next variable. This will make sure that the irq work iterator + * checks all RT overloaded CPUs whenever a CPU schedules a new lower + * priority task, even if the iterator is in the middle of a scan. Incrementing + * the rt_loop_next will cause the iterator to perform another scan. * - * rq->rt.push_cpu holds the last cpu returned by this function, - * or if this is the first instance, it must hold rq->cpu. */ static int rto_next_cpu(struct rq *rq) { - int prev_cpu = rq->rt.push_cpu; + struct root_domain *rd = rq->rd; + int next; int cpu; - cpu = cpumask_next(prev_cpu, rq->rd->rto_mask); - /* - * If the previous cpu is less than the rq's CPU, then it already - * passed the end of the mask, and has started from the beginning. - * We end if the next CPU is greater or equal to rq's CPU. + * When starting the IPI RT pushing, the rto_cpu is set to -1, + * rt_next_cpu() will simply return the first CPU found in + * the rto_mask. + * + * If rto_next_cpu() is called with rto_cpu is a valid cpu, it + * will return the next CPU found in the rto_mask. + * + * If there are no more CPUs left in the rto_mask, then a check is made + * against rto_loop and rto_loop_next. rto_loop is only updated with + * the rto_lock held, but any CPU may increment the rto_loop_next + * without any locking. */ - if (prev_cpu < rq->cpu) { - if (cpu >= rq->cpu) - return nr_cpu_ids; + for (;;) { - } else if (cpu >= nr_cpu_ids) { - /* - * We passed the end of the mask, start at the beginning. - * If the result is greater or equal to the rq's CPU, then - * the loop is finished. - */ - cpu = cpumask_first(rq->rd->rto_mask); - if (cpu >= rq->cpu) - return nr_cpu_ids; - } - rq->rt.push_cpu = cpu; + /* When rto_cpu is -1 this acts like cpumask_first() */ + cpu = cpumask_next(rd->rto_cpu, rd->rto_mask); - /* Return cpu to let the caller know if the loop is finished or not */ - return cpu; -} + rd->rto_cpu = cpu; -static int find_next_push_cpu(struct rq *rq) -{ - struct rq *next_rq; - int cpu; + if (cpu < nr_cpu_ids) + return cpu; - while (1) { - cpu = rto_next_cpu(rq); - if (cpu >= nr_cpu_ids) - break; - next_rq = cpu_rq(cpu); + rd->rto_cpu = -1; + + /* + * ACQUIRE ensures we see the @rto_mask changes + * made prior to the @next value observed. + * + * Matches WMB in rt_set_overload(). + */ + next = atomic_read_acquire(&rd->rto_loop_next); - /* Make sure the next rq can push to this rq */ - if (next_rq->rt.highest_prio.next < rq->rt.highest_prio.curr) + if (rd->rto_loop == next) break; + + rd->rto_loop = next; } - return cpu; + return -1; } -#define RT_PUSH_IPI_EXECUTING 1 -#define RT_PUSH_IPI_RESTART 2 +static inline bool rto_start_trylock(atomic_t *v) +{ + return !atomic_cmpxchg_acquire(v, 0, 1); +} -/* - * When a high priority task schedules out from a CPU and a lower priority - * task is scheduled in, a check is made to see if there's any RT tasks - * on other CPUs that are waiting to run because a higher priority RT task - * is currently running on its CPU. In this case, the CPU with multiple RT - * tasks queued on it (overloaded) needs to be notified that a CPU has opened - * up that may be able to run one of its non-running queued RT tasks. - * - * On large CPU boxes, there's the case that several CPUs could schedule - * a lower priority task at the same time, in which case it will look for - * any overloaded CPUs that it could pull a task from. To do this, the runqueue - * lock must be taken from that overloaded CPU. Having 10s of CPUs all fighting - * for a single overloaded CPU's runqueue lock can produce a large latency. - * (This has actually been observed on large boxes running cyclictest). - * Instead of taking the runqueue lock of the overloaded CPU, each of the - * CPUs that scheduled a lower priority task simply sends an IPI to the - * overloaded CPU. An IPI is much cheaper than taking an runqueue lock with - * lots of contention. The overloaded CPU will look to push its non-running - * RT task off, and if it does, it can then ignore the other IPIs coming - * in, and just pass those IPIs off to any other overloaded CPU. - * - * When a CPU schedules a lower priority task, it only sends an IPI to - * the "next" CPU that has overloaded RT tasks. This prevents IPI storms, - * as having 10 CPUs scheduling lower priority tasks and 10 CPUs with - * RT overloaded tasks, would cause 100 IPIs to go out at once. - * - * The overloaded RT CPU, when receiving an IPI, will try to push off its - * overloaded RT tasks and then send an IPI to the next CPU that has - * overloaded RT tasks. This stops when all CPUs with overloaded RT tasks - * have completed. Just because a CPU may have pushed off its own overloaded - * RT task does not mean it should stop sending the IPI around to other - * overloaded CPUs. There may be another RT task waiting to run on one of - * those CPUs that are of higher priority than the one that was just - * pushed. - * - * An optimization that could possibly be made is to make a CPU array similar - * to the cpupri array mask of all running RT tasks, but for the overloaded - * case, then the IPI could be sent to only the CPU with the highest priority - * RT task waiting, and that CPU could send off further IPIs to the CPU with - * the next highest waiting task. Since the overloaded case is much less likely - * to happen, the complexity of this implementation may not be worth it. - * Instead, just send an IPI around to all overloaded CPUs. - * - * The rq->rt.push_flags holds the status of the IPI that is going around. - * A run queue can only send out a single IPI at a time. The possible flags - * for rq->rt.push_flags are: - * - * (None or zero): No IPI is going around for the current rq - * RT_PUSH_IPI_EXECUTING: An IPI for the rq is being passed around - * RT_PUSH_IPI_RESTART: The priority of the running task for the rq - * has changed, and the IPI should restart - * circulating the overloaded CPUs again. - * - * rq->rt.push_cpu contains the CPU that is being sent the IPI. It is updated - * before sending to the next CPU. - * - * Instead of having all CPUs that schedule a lower priority task send - * an IPI to the same "first" CPU in the RT overload mask, they send it - * to the next overloaded CPU after their own CPU. This helps distribute - * the work when there's more than one overloaded CPU and multiple CPUs - * scheduling in lower priority tasks. - * - * When a rq schedules a lower priority task than what was currently - * running, the next CPU with overloaded RT tasks is examined first. - * That is, if CPU 1 and 5 are overloaded, and CPU 3 schedules a lower - * priority task, it will send an IPI first to CPU 5, then CPU 5 will - * send to CPU 1 if it is still overloaded. CPU 1 will clear the - * rq->rt.push_flags if RT_PUSH_IPI_RESTART is not set. - * - * The first CPU to notice IPI_RESTART is set, will clear that flag and then - * send an IPI to the next overloaded CPU after the rq->cpu and not the next - * CPU after push_cpu. That is, if CPU 1, 4 and 5 are overloaded when CPU 3 - * schedules a lower priority task, and the IPI_RESTART gets set while the - * handling is being done on CPU 5, it will clear the flag and send it back to - * CPU 4 instead of CPU 1. - * - * Note, the above logic can be disabled by turning off the sched_feature - * RT_PUSH_IPI. Then the rq lock of the overloaded CPU will simply be - * taken by the CPU requesting a pull and the waiting RT task will be pulled - * by that CPU. This may be fine for machines with few CPUs. - */ -static void tell_cpu_to_push(struct rq *rq) +static inline void rto_start_unlock(atomic_t *v) { - int cpu; + atomic_set_release(v, 0); +} - if (rq->rt.push_flags & RT_PUSH_IPI_EXECUTING) { - raw_spin_lock(&rq->rt.push_lock); - /* Make sure it's still executing */ - if (rq->rt.push_flags & RT_PUSH_IPI_EXECUTING) { - /* - * Tell the IPI to restart the loop as things have - * changed since it started. - */ - rq->rt.push_flags |= RT_PUSH_IPI_RESTART; - raw_spin_unlock(&rq->rt.push_lock); - return; - } - raw_spin_unlock(&rq->rt.push_lock); - } +static void tell_cpu_to_push(struct rq *rq) +{ + int cpu = -1; - /* When here, there's no IPI going around */ + /* Keep the loop going if the IPI is currently active */ + atomic_inc(&rq->rd->rto_loop_next); - rq->rt.push_cpu = rq->cpu; - cpu = find_next_push_cpu(rq); - if (cpu >= nr_cpu_ids) + /* Only one CPU can initiate a loop at a time */ + if (!rto_start_trylock(&rq->rd->rto_loop_start)) return; - rq->rt.push_flags = RT_PUSH_IPI_EXECUTING; + raw_spin_lock(&rq->rd->rto_lock); + + /* + * The rto_cpu is updated under the lock, if it has a valid cpu + * then the IPI is still running and will continue due to the + * update to loop_next, and nothing needs to be done here. + * Otherwise it is finishing up and an ipi needs to be sent. + */ + if (rq->rd->rto_cpu < 0) + cpu = rto_next_cpu(rq); - irq_work_queue_on(&rq->rt.push_work, cpu); + raw_spin_unlock(&rq->rd->rto_lock); + + rto_start_unlock(&rq->rd->rto_loop_start); + + if (cpu >= 0) + irq_work_queue_on(&rq->rd->rto_push_work, cpu); } /* Called from hardirq context */ -static void try_to_push_tasks(void *arg) +void rto_push_irq_work_func(struct irq_work *work) { - struct rt_rq *rt_rq = arg; - struct rq *rq, *src_rq; - int this_cpu; + struct rq *rq; int cpu; - this_cpu = rt_rq->push_cpu; + rq = this_rq(); - /* Paranoid check */ - BUG_ON(this_cpu != smp_processor_id()); - - rq = cpu_rq(this_cpu); - src_rq = rq_of_rt_rq(rt_rq); - -again: + /* + * We do not need to grab the lock to check for has_pushable_tasks. + * When it gets updated, a check is made if a push is possible. + */ if (has_pushable_tasks(rq)) { raw_spin_lock(&rq->lock); - push_rt_task(rq); + push_rt_tasks(rq); raw_spin_unlock(&rq->lock); } - /* Pass the IPI to the next rt overloaded queue */ - raw_spin_lock(&rt_rq->push_lock); - /* - * If the source queue changed since the IPI went out, - * we need to restart the search from that CPU again. - */ - if (rt_rq->push_flags & RT_PUSH_IPI_RESTART) { - rt_rq->push_flags &= ~RT_PUSH_IPI_RESTART; - rt_rq->push_cpu = src_rq->cpu; - } + raw_spin_lock(&rq->rd->rto_lock); - cpu = find_next_push_cpu(src_rq); + /* Pass the IPI to the next rt overloaded queue */ + cpu = rto_next_cpu(rq); - if (cpu >= nr_cpu_ids) - rt_rq->push_flags &= ~RT_PUSH_IPI_EXECUTING; - raw_spin_unlock(&rt_rq->push_lock); + raw_spin_unlock(&rq->rd->rto_lock); - if (cpu >= nr_cpu_ids) + if (cpu < 0) return; - /* - * It is possible that a restart caused this CPU to be - * chosen again. Don't bother with an IPI, just see if we - * have more to push. - */ - if (unlikely(cpu == rq->cpu)) - goto again; - /* Try the next RT overloaded CPU */ - irq_work_queue_on(&rt_rq->push_work, cpu); -} - -static void push_irq_work_func(struct irq_work *work) -{ - struct rt_rq *rt_rq = container_of(work, struct rt_rq, push_work); - - try_to_push_tasks(rt_rq); + irq_work_queue_on(&rq->rd->rto_push_work, cpu); } #endif /* HAVE_RT_PUSH_IPI */ diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index a81c978..8aa24b4 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -505,7 +505,7 @@ static inline int rt_bandwidth_enabled(void) } /* RT IPI pull logic requires IRQ_WORK */ -#ifdef CONFIG_IRQ_WORK +#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_SMP) # define HAVE_RT_PUSH_IPI #endif @@ -527,12 +527,6 @@ struct rt_rq { unsigned long rt_nr_total; int overloaded; struct plist_head pushable_tasks; -#ifdef HAVE_RT_PUSH_IPI - int push_flags; - int push_cpu; - struct irq_work push_work; - raw_spinlock_t push_lock; -#endif #endif /* CONFIG_SMP */ int rt_queued; @@ -641,6 +635,19 @@ struct root_domain { struct dl_bw dl_bw; struct cpudl cpudl; +#ifdef HAVE_RT_PUSH_IPI + /* + * For IPI pull requests, loop across the rto_mask. + */ + struct irq_work rto_push_work; + raw_spinlock_t rto_lock; + /* These are only updated and read within rto_lock */ + int rto_loop; + int rto_cpu; + /* These atomics are updated outside of a lock */ + atomic_t rto_loop_next; + atomic_t rto_loop_start; +#endif /* * The "RT overload" flag: it gets set if a CPU has more than * one runnable RT task. @@ -658,6 +665,9 @@ extern void init_defrootdomain(void); extern int sched_init_domains(const struct cpumask *cpu_map); extern void rq_attach_root(struct rq *rq, struct root_domain *rd); +#ifdef HAVE_RT_PUSH_IPI +extern void rto_push_irq_work_func(struct irq_work *work); +#endif #endif /* CONFIG_SMP */ /* diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c index f51d123..e50450c 100644 --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -268,6 +268,12 @@ static int init_rootdomain(struct root_domain *rd) if (!zalloc_cpumask_var(&rd->rto_mask, GFP_KERNEL)) goto free_dlo_mask; +#ifdef HAVE_RT_PUSH_IPI + rd->rto_cpu = -1; + raw_spin_lock_init(&rd->rto_lock); + init_irq_work(&rd->rto_push_work, rto_push_irq_work_func); +#endif + init_dl_bw(&rd->dl_bw); if (cpudl_init(&rd->cpudl) != 0) goto free_rto_mask; ----- End forwarded message -----

7 years, 7 months

3
8
0 0

[Linux-stable-mirror] Patch "sched/rt: Simplify the IPI based RT balancing logic" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sched/rt: Simplify the IPI based RT balancing logic to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sched-rt-simplify-the-ipi-based-rt-balancing-logic.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 4bdced5c9a2922521e325896a7bbbf0132c94e56 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (Red Hat)" <rostedt(a)goodmis.org> Date: Fri, 6 Oct 2017 14:05:04 -0400 Subject: sched/rt: Simplify the IPI based RT balancing logic From: Steven Rostedt (Red Hat) <rostedt(a)goodmis.org> commit 4bdced5c9a2922521e325896a7bbbf0132c94e56 upstream. When a CPU lowers its priority (schedules out a high priority task for a lower priority one), a check is made to see if any other CPU has overloaded RT tasks (more than one). It checks the rto_mask to determine this and if so it will request to pull one of those tasks to itself if the non running RT task is of higher priority than the new priority of the next task to run on the current CPU. When we deal with large number of CPUs, the original pull logic suffered from large lock contention on a single CPU run queue, which caused a huge latency across all CPUs. This was caused by only having one CPU having overloaded RT tasks and a bunch of other CPUs lowering their priority. To solve this issue, commit: b6366f048e0c ("sched/rt: Use IPI to trigger RT task push migration instead of pulling") changed the way to request a pull. Instead of grabbing the lock of the overloaded CPU's runqueue, it simply sent an IPI to that CPU to do the work. Although the IPI logic worked very well in removing the large latency build up, it still could suffer from a large number of IPIs being sent to a single CPU. On a 80 CPU box, I measured over 200us of processing IPIs. Worse yet, when I tested this on a 120 CPU box, with a stress test that had lots of RT tasks scheduling on all CPUs, it actually triggered the hard lockup detector! One CPU had so many IPIs sent to it, and due to the restart mechanism that is triggered when the source run queue has a priority status change, the CPU spent minutes! processing the IPIs. Thinking about this further, I realized there's no reason for each run queue to send its own IPI. As all CPUs with overloaded tasks must be scanned regardless if there's one or many CPUs lowering their priority, because there's no current way to find the CPU with the highest priority task that can schedule to one of these CPUs, there really only needs to be one IPI being sent around at a time. This greatly simplifies the code! The new approach is to have each root domain have its own irq work, as the rto_mask is per root domain. The root domain has the following fields attached to it: rto_push_work - the irq work to process each CPU set in rto_mask rto_lock - the lock to protect some of the other rto fields rto_loop_start - an atomic that keeps contention down on rto_lock the first CPU scheduling in a lower priority task is the one to kick off the process. rto_loop_next - an atomic that gets incremented for each CPU that schedules in a lower priority task. rto_loop - a variable protected by rto_lock that is used to compare against rto_loop_next rto_cpu - The cpu to send the next IPI to, also protected by the rto_lock. When a CPU schedules in a lower priority task and wants to make sure overloaded CPUs know about it. It increments the rto_loop_next. Then it atomically sets rto_loop_start with a cmpxchg. If the old value is not "0", then it is done, as another CPU is kicking off the IPI loop. If the old value is "0", then it will take the rto_lock to synchronize with a possible IPI being sent around to the overloaded CPUs. If rto_cpu is greater than or equal to nr_cpu_ids, then there's either no IPI being sent around, or one is about to finish. Then rto_cpu is set to the first CPU in rto_mask and an IPI is sent to that CPU. If there's no CPUs set in rto_mask, then there's nothing to be done. When the CPU receives the IPI, it will first try to push any RT tasks that is queued on the CPU but can't run because a higher priority RT task is currently running on that CPU. Then it takes the rto_lock and looks for the next CPU in the rto_mask. If it finds one, it simply sends an IPI to that CPU and the process continues. If there's no more CPUs in the rto_mask, then rto_loop is compared with rto_loop_next. If they match, everything is done and the process is over. If they do not match, then a CPU scheduled in a lower priority task as the IPI was being passed around, and the process needs to start again. The first CPU in rto_mask is sent the IPI. This change removes this duplication of work in the IPI logic, and greatly lowers the latency caused by the IPIs. This removed the lockup happening on the 120 CPU machine. It also simplifies the code tremendously. What else could anyone ask for? Thanks to Peter Zijlstra for simplifying the rto_loop_start atomic logic and supplying me with the rto_start_trylock() and rto_start_unlock() helper functions. Signed-off-by: Steven Rostedt (VMware) <rostedt(a)goodmis.org> Signed-off-by: Peter Zijlstra (Intel) <peterz(a)infradead.org> Cc: Clark Williams <williams(a)redhat.com> Cc: Daniel Bristot de Oliveira <bristot(a)redhat.com> Cc: John Kacur <jkacur(a)redhat.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mike Galbraith <efault(a)gmx.de> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Scott Wood <swood(a)redhat.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Link: http://lkml.kernel.org/r/20170424114732.1aac6dc4@gandalf.local.home Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- kernel/sched/rt.c | 316 +++++++++++++++++------------------------------- kernel/sched/sched.h | 24 ++- kernel/sched/topology.c | 6 3 files changed, 138 insertions(+), 208 deletions(-) --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -74,10 +74,6 @@ static void start_rt_bandwidth(struct rt raw_spin_unlock(&rt_b->rt_runtime_lock); } -#if defined(CONFIG_SMP) && defined(HAVE_RT_PUSH_IPI) -static void push_irq_work_func(struct irq_work *work); -#endif - void init_rt_rq(struct rt_rq *rt_rq) { struct rt_prio_array *array; @@ -97,13 +93,6 @@ void init_rt_rq(struct rt_rq *rt_rq) rt_rq->rt_nr_migratory = 0; rt_rq->overloaded = 0; plist_head_init(&rt_rq->pushable_tasks); - -#ifdef HAVE_RT_PUSH_IPI - rt_rq->push_flags = 0; - rt_rq->push_cpu = nr_cpu_ids; - raw_spin_lock_init(&rt_rq->push_lock); - init_irq_work(&rt_rq->push_work, push_irq_work_func); -#endif #endif /* CONFIG_SMP */ /* We start is dequeued state, because no RT tasks are queued */ rt_rq->rt_queued = 0; @@ -1876,241 +1865,166 @@ static void push_rt_tasks(struct rq *rq) } #ifdef HAVE_RT_PUSH_IPI + /* - * The search for the next cpu always starts at rq->cpu and ends - * when we reach rq->cpu again. It will never return rq->cpu. - * This returns the next cpu to check, or nr_cpu_ids if the loop - * is complete. + * When a high priority task schedules out from a CPU and a lower priority + * task is scheduled in, a check is made to see if there's any RT tasks + * on other CPUs that are waiting to run because a higher priority RT task + * is currently running on its CPU. In this case, the CPU with multiple RT + * tasks queued on it (overloaded) needs to be notified that a CPU has opened + * up that may be able to run one of its non-running queued RT tasks. + * + * All CPUs with overloaded RT tasks need to be notified as there is currently + * no way to know which of these CPUs have the highest priority task waiting + * to run. Instead of trying to take a spinlock on each of these CPUs, + * which has shown to cause large latency when done on machines with many + * CPUs, sending an IPI to the CPUs to have them push off the overloaded + * RT tasks waiting to run. + * + * Just sending an IPI to each of the CPUs is also an issue, as on large + * count CPU machines, this can cause an IPI storm on a CPU, especially + * if its the only CPU with multiple RT tasks queued, and a large number + * of CPUs scheduling a lower priority task at the same time. + * + * Each root domain has its own irq work function that can iterate over + * all CPUs with RT overloaded tasks. Since all CPUs with overloaded RT + * tassk must be checked if there's one or many CPUs that are lowering + * their priority, there's a single irq work iterator that will try to + * push off RT tasks that are waiting to run. + * + * When a CPU schedules a lower priority task, it will kick off the + * irq work iterator that will jump to each CPU with overloaded RT tasks. + * As it only takes the first CPU that schedules a lower priority task + * to start the process, the rto_start variable is incremented and if + * the atomic result is one, then that CPU will try to take the rto_lock. + * This prevents high contention on the lock as the process handles all + * CPUs scheduling lower priority tasks. + * + * All CPUs that are scheduling a lower priority task will increment the + * rt_loop_next variable. This will make sure that the irq work iterator + * checks all RT overloaded CPUs whenever a CPU schedules a new lower + * priority task, even if the iterator is in the middle of a scan. Incrementing + * the rt_loop_next will cause the iterator to perform another scan. * - * rq->rt.push_cpu holds the last cpu returned by this function, - * or if this is the first instance, it must hold rq->cpu. */ static int rto_next_cpu(struct rq *rq) { - int prev_cpu = rq->rt.push_cpu; + struct root_domain *rd = rq->rd; + int next; int cpu; - cpu = cpumask_next(prev_cpu, rq->rd->rto_mask); - /* - * If the previous cpu is less than the rq's CPU, then it already - * passed the end of the mask, and has started from the beginning. - * We end if the next CPU is greater or equal to rq's CPU. + * When starting the IPI RT pushing, the rto_cpu is set to -1, + * rt_next_cpu() will simply return the first CPU found in + * the rto_mask. + * + * If rto_next_cpu() is called with rto_cpu is a valid cpu, it + * will return the next CPU found in the rto_mask. + * + * If there are no more CPUs left in the rto_mask, then a check is made + * against rto_loop and rto_loop_next. rto_loop is only updated with + * the rto_lock held, but any CPU may increment the rto_loop_next + * without any locking. */ - if (prev_cpu < rq->cpu) { - if (cpu >= rq->cpu) - return nr_cpu_ids; + for (;;) { - } else if (cpu >= nr_cpu_ids) { - /* - * We passed the end of the mask, start at the beginning. - * If the result is greater or equal to the rq's CPU, then - * the loop is finished. - */ - cpu = cpumask_first(rq->rd->rto_mask); - if (cpu >= rq->cpu) - return nr_cpu_ids; - } - rq->rt.push_cpu = cpu; + /* When rto_cpu is -1 this acts like cpumask_first() */ + cpu = cpumask_next(rd->rto_cpu, rd->rto_mask); - /* Return cpu to let the caller know if the loop is finished or not */ - return cpu; -} + rd->rto_cpu = cpu; -static int find_next_push_cpu(struct rq *rq) -{ - struct rq *next_rq; - int cpu; + if (cpu < nr_cpu_ids) + return cpu; - while (1) { - cpu = rto_next_cpu(rq); - if (cpu >= nr_cpu_ids) - break; - next_rq = cpu_rq(cpu); + rd->rto_cpu = -1; + + /* + * ACQUIRE ensures we see the @rto_mask changes + * made prior to the @next value observed. + * + * Matches WMB in rt_set_overload(). + */ + next = atomic_read_acquire(&rd->rto_loop_next); - /* Make sure the next rq can push to this rq */ - if (next_rq->rt.highest_prio.next < rq->rt.highest_prio.curr) + if (rd->rto_loop == next) break; + + rd->rto_loop = next; } - return cpu; + return -1; } -#define RT_PUSH_IPI_EXECUTING 1 -#define RT_PUSH_IPI_RESTART 2 +static inline bool rto_start_trylock(atomic_t *v) +{ + return !atomic_cmpxchg_acquire(v, 0, 1); +} -/* - * When a high priority task schedules out from a CPU and a lower priority - * task is scheduled in, a check is made to see if there's any RT tasks - * on other CPUs that are waiting to run because a higher priority RT task - * is currently running on its CPU. In this case, the CPU with multiple RT - * tasks queued on it (overloaded) needs to be notified that a CPU has opened - * up that may be able to run one of its non-running queued RT tasks. - * - * On large CPU boxes, there's the case that several CPUs could schedule - * a lower priority task at the same time, in which case it will look for - * any overloaded CPUs that it could pull a task from. To do this, the runqueue - * lock must be taken from that overloaded CPU. Having 10s of CPUs all fighting - * for a single overloaded CPU's runqueue lock can produce a large latency. - * (This has actually been observed on large boxes running cyclictest). - * Instead of taking the runqueue lock of the overloaded CPU, each of the - * CPUs that scheduled a lower priority task simply sends an IPI to the - * overloaded CPU. An IPI is much cheaper than taking an runqueue lock with - * lots of contention. The overloaded CPU will look to push its non-running - * RT task off, and if it does, it can then ignore the other IPIs coming - * in, and just pass those IPIs off to any other overloaded CPU. - * - * When a CPU schedules a lower priority task, it only sends an IPI to - * the "next" CPU that has overloaded RT tasks. This prevents IPI storms, - * as having 10 CPUs scheduling lower priority tasks and 10 CPUs with - * RT overloaded tasks, would cause 100 IPIs to go out at once. - * - * The overloaded RT CPU, when receiving an IPI, will try to push off its - * overloaded RT tasks and then send an IPI to the next CPU that has - * overloaded RT tasks. This stops when all CPUs with overloaded RT tasks - * have completed. Just because a CPU may have pushed off its own overloaded - * RT task does not mean it should stop sending the IPI around to other - * overloaded CPUs. There may be another RT task waiting to run on one of - * those CPUs that are of higher priority than the one that was just - * pushed. - * - * An optimization that could possibly be made is to make a CPU array similar - * to the cpupri array mask of all running RT tasks, but for the overloaded - * case, then the IPI could be sent to only the CPU with the highest priority - * RT task waiting, and that CPU could send off further IPIs to the CPU with - * the next highest waiting task. Since the overloaded case is much less likely - * to happen, the complexity of this implementation may not be worth it. - * Instead, just send an IPI around to all overloaded CPUs. - * - * The rq->rt.push_flags holds the status of the IPI that is going around. - * A run queue can only send out a single IPI at a time. The possible flags - * for rq->rt.push_flags are: - * - * (None or zero): No IPI is going around for the current rq - * RT_PUSH_IPI_EXECUTING: An IPI for the rq is being passed around - * RT_PUSH_IPI_RESTART: The priority of the running task for the rq - * has changed, and the IPI should restart - * circulating the overloaded CPUs again. - * - * rq->rt.push_cpu contains the CPU that is being sent the IPI. It is updated - * before sending to the next CPU. - * - * Instead of having all CPUs that schedule a lower priority task send - * an IPI to the same "first" CPU in the RT overload mask, they send it - * to the next overloaded CPU after their own CPU. This helps distribute - * the work when there's more than one overloaded CPU and multiple CPUs - * scheduling in lower priority tasks. - * - * When a rq schedules a lower priority task than what was currently - * running, the next CPU with overloaded RT tasks is examined first. - * That is, if CPU 1 and 5 are overloaded, and CPU 3 schedules a lower - * priority task, it will send an IPI first to CPU 5, then CPU 5 will - * send to CPU 1 if it is still overloaded. CPU 1 will clear the - * rq->rt.push_flags if RT_PUSH_IPI_RESTART is not set. - * - * The first CPU to notice IPI_RESTART is set, will clear that flag and then - * send an IPI to the next overloaded CPU after the rq->cpu and not the next - * CPU after push_cpu. That is, if CPU 1, 4 and 5 are overloaded when CPU 3 - * schedules a lower priority task, and the IPI_RESTART gets set while the - * handling is being done on CPU 5, it will clear the flag and send it back to - * CPU 4 instead of CPU 1. - * - * Note, the above logic can be disabled by turning off the sched_feature - * RT_PUSH_IPI. Then the rq lock of the overloaded CPU will simply be - * taken by the CPU requesting a pull and the waiting RT task will be pulled - * by that CPU. This may be fine for machines with few CPUs. - */ -static void tell_cpu_to_push(struct rq *rq) +static inline void rto_start_unlock(atomic_t *v) { - int cpu; + atomic_set_release(v, 0); +} - if (rq->rt.push_flags & RT_PUSH_IPI_EXECUTING) { - raw_spin_lock(&rq->rt.push_lock); - /* Make sure it's still executing */ - if (rq->rt.push_flags & RT_PUSH_IPI_EXECUTING) { - /* - * Tell the IPI to restart the loop as things have - * changed since it started. - */ - rq->rt.push_flags |= RT_PUSH_IPI_RESTART; - raw_spin_unlock(&rq->rt.push_lock); - return; - } - raw_spin_unlock(&rq->rt.push_lock); - } +static void tell_cpu_to_push(struct rq *rq) +{ + int cpu = -1; - /* When here, there's no IPI going around */ + /* Keep the loop going if the IPI is currently active */ + atomic_inc(&rq->rd->rto_loop_next); - rq->rt.push_cpu = rq->cpu; - cpu = find_next_push_cpu(rq); - if (cpu >= nr_cpu_ids) + /* Only one CPU can initiate a loop at a time */ + if (!rto_start_trylock(&rq->rd->rto_loop_start)) return; - rq->rt.push_flags = RT_PUSH_IPI_EXECUTING; + raw_spin_lock(&rq->rd->rto_lock); - irq_work_queue_on(&rq->rt.push_work, cpu); + /* + * The rto_cpu is updated under the lock, if it has a valid cpu + * then the IPI is still running and will continue due to the + * update to loop_next, and nothing needs to be done here. + * Otherwise it is finishing up and an ipi needs to be sent. + */ + if (rq->rd->rto_cpu < 0) + cpu = rto_next_cpu(rq); + + raw_spin_unlock(&rq->rd->rto_lock); + + rto_start_unlock(&rq->rd->rto_loop_start); + + if (cpu >= 0) + irq_work_queue_on(&rq->rd->rto_push_work, cpu); } /* Called from hardirq context */ -static void try_to_push_tasks(void *arg) +void rto_push_irq_work_func(struct irq_work *work) { - struct rt_rq *rt_rq = arg; - struct rq *rq, *src_rq; - int this_cpu; + struct rq *rq; int cpu; - this_cpu = rt_rq->push_cpu; + rq = this_rq(); - /* Paranoid check */ - BUG_ON(this_cpu != smp_processor_id()); - - rq = cpu_rq(this_cpu); - src_rq = rq_of_rt_rq(rt_rq); - -again: + /* + * We do not need to grab the lock to check for has_pushable_tasks. + * When it gets updated, a check is made if a push is possible. + */ if (has_pushable_tasks(rq)) { raw_spin_lock(&rq->lock); - push_rt_task(rq); + push_rt_tasks(rq); raw_spin_unlock(&rq->lock); } - /* Pass the IPI to the next rt overloaded queue */ - raw_spin_lock(&rt_rq->push_lock); - /* - * If the source queue changed since the IPI went out, - * we need to restart the search from that CPU again. - */ - if (rt_rq->push_flags & RT_PUSH_IPI_RESTART) { - rt_rq->push_flags &= ~RT_PUSH_IPI_RESTART; - rt_rq->push_cpu = src_rq->cpu; - } + raw_spin_lock(&rq->rd->rto_lock); - cpu = find_next_push_cpu(src_rq); + /* Pass the IPI to the next rt overloaded queue */ + cpu = rto_next_cpu(rq); - if (cpu >= nr_cpu_ids) - rt_rq->push_flags &= ~RT_PUSH_IPI_EXECUTING; - raw_spin_unlock(&rt_rq->push_lock); + raw_spin_unlock(&rq->rd->rto_lock); - if (cpu >= nr_cpu_ids) + if (cpu < 0) return; - /* - * It is possible that a restart caused this CPU to be - * chosen again. Don't bother with an IPI, just see if we - * have more to push. - */ - if (unlikely(cpu == rq->cpu)) - goto again; - /* Try the next RT overloaded CPU */ - irq_work_queue_on(&rt_rq->push_work, cpu); -} - -static void push_irq_work_func(struct irq_work *work) -{ - struct rt_rq *rt_rq = container_of(work, struct rt_rq, push_work); - - try_to_push_tasks(rt_rq); + irq_work_queue_on(&rq->rd->rto_push_work, cpu); } #endif /* HAVE_RT_PUSH_IPI */ --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -502,7 +502,7 @@ static inline int rt_bandwidth_enabled(v } /* RT IPI pull logic requires IRQ_WORK */ -#ifdef CONFIG_IRQ_WORK +#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_SMP) # define HAVE_RT_PUSH_IPI #endif @@ -524,12 +524,6 @@ struct rt_rq { unsigned long rt_nr_total; int overloaded; struct plist_head pushable_tasks; -#ifdef HAVE_RT_PUSH_IPI - int push_flags; - int push_cpu; - struct irq_work push_work; - raw_spinlock_t push_lock; -#endif #endif /* CONFIG_SMP */ int rt_queued; @@ -638,6 +632,19 @@ struct root_domain { struct dl_bw dl_bw; struct cpudl cpudl; +#ifdef HAVE_RT_PUSH_IPI + /* + * For IPI pull requests, loop across the rto_mask. + */ + struct irq_work rto_push_work; + raw_spinlock_t rto_lock; + /* These are only updated and read within rto_lock */ + int rto_loop; + int rto_cpu; + /* These atomics are updated outside of a lock */ + atomic_t rto_loop_next; + atomic_t rto_loop_start; +#endif /* * The "RT overload" flag: it gets set if a CPU has more than * one runnable RT task. @@ -655,6 +662,9 @@ extern void init_defrootdomain(void); extern int sched_init_domains(const struct cpumask *cpu_map); extern void rq_attach_root(struct rq *rq, struct root_domain *rd); +#ifdef HAVE_RT_PUSH_IPI +extern void rto_push_irq_work_func(struct irq_work *work); +#endif #endif /* CONFIG_SMP */ /* --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -269,6 +269,12 @@ static int init_rootdomain(struct root_d if (!zalloc_cpumask_var(&rd->rto_mask, GFP_KERNEL)) goto free_dlo_mask; +#ifdef HAVE_RT_PUSH_IPI + rd->rto_cpu = -1; + raw_spin_lock_init(&rd->rto_lock); + init_irq_work(&rd->rto_push_work, rto_push_irq_work_func); +#endif + init_dl_bw(&rd->dl_bw); if (cpudl_init(&rd->cpudl) != 0) goto free_rto_mask; Patches currently in stable-queue which might be from rostedt(a)goodmis.org are queue-4.14/sched-make-resched_cpu-unconditional.patch queue-4.14/sched-rt-simplify-the-ipi-based-rt-balancing-logic.patch

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] sched/rt: Simplify the IPI based RT balancing logic" failed to apply to 4.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 4bdced5c9a2922521e325896a7bbbf0132c94e56 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (Red Hat)" <rostedt(a)goodmis.org> Date: Fri, 6 Oct 2017 14:05:04 -0400 Subject: [PATCH] sched/rt: Simplify the IPI based RT balancing logic When a CPU lowers its priority (schedules out a high priority task for a lower priority one), a check is made to see if any other CPU has overloaded RT tasks (more than one). It checks the rto_mask to determine this and if so it will request to pull one of those tasks to itself if the non running RT task is of higher priority than the new priority of the next task to run on the current CPU. When we deal with large number of CPUs, the original pull logic suffered from large lock contention on a single CPU run queue, which caused a huge latency across all CPUs. This was caused by only having one CPU having overloaded RT tasks and a bunch of other CPUs lowering their priority. To solve this issue, commit: b6366f048e0c ("sched/rt: Use IPI to trigger RT task push migration instead of pulling") changed the way to request a pull. Instead of grabbing the lock of the overloaded CPU's runqueue, it simply sent an IPI to that CPU to do the work. Although the IPI logic worked very well in removing the large latency build up, it still could suffer from a large number of IPIs being sent to a single CPU. On a 80 CPU box, I measured over 200us of processing IPIs. Worse yet, when I tested this on a 120 CPU box, with a stress test that had lots of RT tasks scheduling on all CPUs, it actually triggered the hard lockup detector! One CPU had so many IPIs sent to it, and due to the restart mechanism that is triggered when the source run queue has a priority status change, the CPU spent minutes! processing the IPIs. Thinking about this further, I realized there's no reason for each run queue to send its own IPI. As all CPUs with overloaded tasks must be scanned regardless if there's one or many CPUs lowering their priority, because there's no current way to find the CPU with the highest priority task that can schedule to one of these CPUs, there really only needs to be one IPI being sent around at a time. This greatly simplifies the code! The new approach is to have each root domain have its own irq work, as the rto_mask is per root domain. The root domain has the following fields attached to it: rto_push_work - the irq work to process each CPU set in rto_mask rto_lock - the lock to protect some of the other rto fields rto_loop_start - an atomic that keeps contention down on rto_lock the first CPU scheduling in a lower priority task is the one to kick off the process. rto_loop_next - an atomic that gets incremented for each CPU that schedules in a lower priority task. rto_loop - a variable protected by rto_lock that is used to compare against rto_loop_next rto_cpu - The cpu to send the next IPI to, also protected by the rto_lock. When a CPU schedules in a lower priority task and wants to make sure overloaded CPUs know about it. It increments the rto_loop_next. Then it atomically sets rto_loop_start with a cmpxchg. If the old value is not "0", then it is done, as another CPU is kicking off the IPI loop. If the old value is "0", then it will take the rto_lock to synchronize with a possible IPI being sent around to the overloaded CPUs. If rto_cpu is greater than or equal to nr_cpu_ids, then there's either no IPI being sent around, or one is about to finish. Then rto_cpu is set to the first CPU in rto_mask and an IPI is sent to that CPU. If there's no CPUs set in rto_mask, then there's nothing to be done. When the CPU receives the IPI, it will first try to push any RT tasks that is queued on the CPU but can't run because a higher priority RT task is currently running on that CPU. Then it takes the rto_lock and looks for the next CPU in the rto_mask. If it finds one, it simply sends an IPI to that CPU and the process continues. If there's no more CPUs in the rto_mask, then rto_loop is compared with rto_loop_next. If they match, everything is done and the process is over. If they do not match, then a CPU scheduled in a lower priority task as the IPI was being passed around, and the process needs to start again. The first CPU in rto_mask is sent the IPI. This change removes this duplication of work in the IPI logic, and greatly lowers the latency caused by the IPIs. This removed the lockup happening on the 120 CPU machine. It also simplifies the code tremendously. What else could anyone ask for? Thanks to Peter Zijlstra for simplifying the rto_loop_start atomic logic and supplying me with the rto_start_trylock() and rto_start_unlock() helper functions. Signed-off-by: Steven Rostedt (VMware) <rostedt(a)goodmis.org> Signed-off-by: Peter Zijlstra (Intel) <peterz(a)infradead.org> Cc: Clark Williams <williams(a)redhat.com> Cc: Daniel Bristot de Oliveira <bristot(a)redhat.com> Cc: John Kacur <jkacur(a)redhat.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mike Galbraith <efault(a)gmx.de> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Scott Wood <swood(a)redhat.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Link: http://lkml.kernel.org/r/20170424114732.1aac6dc4@gandalf.local.home Signed-off-by: Ingo Molnar <mingo(a)kernel.org> diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index 0af5ca9e3e3f..fda27991699a 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -73,10 +73,6 @@ static void start_rt_bandwidth(struct rt_bandwidth *rt_b) raw_spin_unlock(&rt_b->rt_runtime_lock); } -#if defined(CONFIG_SMP) && defined(HAVE_RT_PUSH_IPI) -static void push_irq_work_func(struct irq_work *work); -#endif - void init_rt_rq(struct rt_rq *rt_rq) { struct rt_prio_array *array; @@ -96,13 +92,6 @@ void init_rt_rq(struct rt_rq *rt_rq) rt_rq->rt_nr_migratory = 0; rt_rq->overloaded = 0; plist_head_init(&rt_rq->pushable_tasks); - -#ifdef HAVE_RT_PUSH_IPI - rt_rq->push_flags = 0; - rt_rq->push_cpu = nr_cpu_ids; - raw_spin_lock_init(&rt_rq->push_lock); - init_irq_work(&rt_rq->push_work, push_irq_work_func); -#endif #endif /* CONFIG_SMP */ /* We start is dequeued state, because no RT tasks are queued */ rt_rq->rt_queued = 0; @@ -1875,241 +1864,166 @@ static void push_rt_tasks(struct rq *rq) } #ifdef HAVE_RT_PUSH_IPI + /* - * The search for the next cpu always starts at rq->cpu and ends - * when we reach rq->cpu again. It will never return rq->cpu. - * This returns the next cpu to check, or nr_cpu_ids if the loop - * is complete. + * When a high priority task schedules out from a CPU and a lower priority + * task is scheduled in, a check is made to see if there's any RT tasks + * on other CPUs that are waiting to run because a higher priority RT task + * is currently running on its CPU. In this case, the CPU with multiple RT + * tasks queued on it (overloaded) needs to be notified that a CPU has opened + * up that may be able to run one of its non-running queued RT tasks. + * + * All CPUs with overloaded RT tasks need to be notified as there is currently + * no way to know which of these CPUs have the highest priority task waiting + * to run. Instead of trying to take a spinlock on each of these CPUs, + * which has shown to cause large latency when done on machines with many + * CPUs, sending an IPI to the CPUs to have them push off the overloaded + * RT tasks waiting to run. + * + * Just sending an IPI to each of the CPUs is also an issue, as on large + * count CPU machines, this can cause an IPI storm on a CPU, especially + * if its the only CPU with multiple RT tasks queued, and a large number + * of CPUs scheduling a lower priority task at the same time. + * + * Each root domain has its own irq work function that can iterate over + * all CPUs with RT overloaded tasks. Since all CPUs with overloaded RT + * tassk must be checked if there's one or many CPUs that are lowering + * their priority, there's a single irq work iterator that will try to + * push off RT tasks that are waiting to run. + * + * When a CPU schedules a lower priority task, it will kick off the + * irq work iterator that will jump to each CPU with overloaded RT tasks. + * As it only takes the first CPU that schedules a lower priority task + * to start the process, the rto_start variable is incremented and if + * the atomic result is one, then that CPU will try to take the rto_lock. + * This prevents high contention on the lock as the process handles all + * CPUs scheduling lower priority tasks. + * + * All CPUs that are scheduling a lower priority task will increment the + * rt_loop_next variable. This will make sure that the irq work iterator + * checks all RT overloaded CPUs whenever a CPU schedules a new lower + * priority task, even if the iterator is in the middle of a scan. Incrementing + * the rt_loop_next will cause the iterator to perform another scan. * - * rq->rt.push_cpu holds the last cpu returned by this function, - * or if this is the first instance, it must hold rq->cpu. */ static int rto_next_cpu(struct rq *rq) { - int prev_cpu = rq->rt.push_cpu; + struct root_domain *rd = rq->rd; + int next; int cpu; - cpu = cpumask_next(prev_cpu, rq->rd->rto_mask); - /* - * If the previous cpu is less than the rq's CPU, then it already - * passed the end of the mask, and has started from the beginning. - * We end if the next CPU is greater or equal to rq's CPU. + * When starting the IPI RT pushing, the rto_cpu is set to -1, + * rt_next_cpu() will simply return the first CPU found in + * the rto_mask. + * + * If rto_next_cpu() is called with rto_cpu is a valid cpu, it + * will return the next CPU found in the rto_mask. + * + * If there are no more CPUs left in the rto_mask, then a check is made + * against rto_loop and rto_loop_next. rto_loop is only updated with + * the rto_lock held, but any CPU may increment the rto_loop_next + * without any locking. */ - if (prev_cpu < rq->cpu) { - if (cpu >= rq->cpu) - return nr_cpu_ids; + for (;;) { - } else if (cpu >= nr_cpu_ids) { - /* - * We passed the end of the mask, start at the beginning. - * If the result is greater or equal to the rq's CPU, then - * the loop is finished. - */ - cpu = cpumask_first(rq->rd->rto_mask); - if (cpu >= rq->cpu) - return nr_cpu_ids; - } - rq->rt.push_cpu = cpu; + /* When rto_cpu is -1 this acts like cpumask_first() */ + cpu = cpumask_next(rd->rto_cpu, rd->rto_mask); - /* Return cpu to let the caller know if the loop is finished or not */ - return cpu; -} + rd->rto_cpu = cpu; -static int find_next_push_cpu(struct rq *rq) -{ - struct rq *next_rq; - int cpu; + if (cpu < nr_cpu_ids) + return cpu; - while (1) { - cpu = rto_next_cpu(rq); - if (cpu >= nr_cpu_ids) - break; - next_rq = cpu_rq(cpu); + rd->rto_cpu = -1; + + /* + * ACQUIRE ensures we see the @rto_mask changes + * made prior to the @next value observed. + * + * Matches WMB in rt_set_overload(). + */ + next = atomic_read_acquire(&rd->rto_loop_next); - /* Make sure the next rq can push to this rq */ - if (next_rq->rt.highest_prio.next < rq->rt.highest_prio.curr) + if (rd->rto_loop == next) break; + + rd->rto_loop = next; } - return cpu; + return -1; } -#define RT_PUSH_IPI_EXECUTING 1 -#define RT_PUSH_IPI_RESTART 2 +static inline bool rto_start_trylock(atomic_t *v) +{ + return !atomic_cmpxchg_acquire(v, 0, 1); +} -/* - * When a high priority task schedules out from a CPU and a lower priority - * task is scheduled in, a check is made to see if there's any RT tasks - * on other CPUs that are waiting to run because a higher priority RT task - * is currently running on its CPU. In this case, the CPU with multiple RT - * tasks queued on it (overloaded) needs to be notified that a CPU has opened - * up that may be able to run one of its non-running queued RT tasks. - * - * On large CPU boxes, there's the case that several CPUs could schedule - * a lower priority task at the same time, in which case it will look for - * any overloaded CPUs that it could pull a task from. To do this, the runqueue - * lock must be taken from that overloaded CPU. Having 10s of CPUs all fighting - * for a single overloaded CPU's runqueue lock can produce a large latency. - * (This has actually been observed on large boxes running cyclictest). - * Instead of taking the runqueue lock of the overloaded CPU, each of the - * CPUs that scheduled a lower priority task simply sends an IPI to the - * overloaded CPU. An IPI is much cheaper than taking an runqueue lock with - * lots of contention. The overloaded CPU will look to push its non-running - * RT task off, and if it does, it can then ignore the other IPIs coming - * in, and just pass those IPIs off to any other overloaded CPU. - * - * When a CPU schedules a lower priority task, it only sends an IPI to - * the "next" CPU that has overloaded RT tasks. This prevents IPI storms, - * as having 10 CPUs scheduling lower priority tasks and 10 CPUs with - * RT overloaded tasks, would cause 100 IPIs to go out at once. - * - * The overloaded RT CPU, when receiving an IPI, will try to push off its - * overloaded RT tasks and then send an IPI to the next CPU that has - * overloaded RT tasks. This stops when all CPUs with overloaded RT tasks - * have completed. Just because a CPU may have pushed off its own overloaded - * RT task does not mean it should stop sending the IPI around to other - * overloaded CPUs. There may be another RT task waiting to run on one of - * those CPUs that are of higher priority than the one that was just - * pushed. - * - * An optimization that could possibly be made is to make a CPU array similar - * to the cpupri array mask of all running RT tasks, but for the overloaded - * case, then the IPI could be sent to only the CPU with the highest priority - * RT task waiting, and that CPU could send off further IPIs to the CPU with - * the next highest waiting task. Since the overloaded case is much less likely - * to happen, the complexity of this implementation may not be worth it. - * Instead, just send an IPI around to all overloaded CPUs. - * - * The rq->rt.push_flags holds the status of the IPI that is going around. - * A run queue can only send out a single IPI at a time. The possible flags - * for rq->rt.push_flags are: - * - * (None or zero): No IPI is going around for the current rq - * RT_PUSH_IPI_EXECUTING: An IPI for the rq is being passed around - * RT_PUSH_IPI_RESTART: The priority of the running task for the rq - * has changed, and the IPI should restart - * circulating the overloaded CPUs again. - * - * rq->rt.push_cpu contains the CPU that is being sent the IPI. It is updated - * before sending to the next CPU. - * - * Instead of having all CPUs that schedule a lower priority task send - * an IPI to the same "first" CPU in the RT overload mask, they send it - * to the next overloaded CPU after their own CPU. This helps distribute - * the work when there's more than one overloaded CPU and multiple CPUs - * scheduling in lower priority tasks. - * - * When a rq schedules a lower priority task than what was currently - * running, the next CPU with overloaded RT tasks is examined first. - * That is, if CPU 1 and 5 are overloaded, and CPU 3 schedules a lower - * priority task, it will send an IPI first to CPU 5, then CPU 5 will - * send to CPU 1 if it is still overloaded. CPU 1 will clear the - * rq->rt.push_flags if RT_PUSH_IPI_RESTART is not set. - * - * The first CPU to notice IPI_RESTART is set, will clear that flag and then - * send an IPI to the next overloaded CPU after the rq->cpu and not the next - * CPU after push_cpu. That is, if CPU 1, 4 and 5 are overloaded when CPU 3 - * schedules a lower priority task, and the IPI_RESTART gets set while the - * handling is being done on CPU 5, it will clear the flag and send it back to - * CPU 4 instead of CPU 1. - * - * Note, the above logic can be disabled by turning off the sched_feature - * RT_PUSH_IPI. Then the rq lock of the overloaded CPU will simply be - * taken by the CPU requesting a pull and the waiting RT task will be pulled - * by that CPU. This may be fine for machines with few CPUs. - */ -static void tell_cpu_to_push(struct rq *rq) +static inline void rto_start_unlock(atomic_t *v) { - int cpu; + atomic_set_release(v, 0); +} - if (rq->rt.push_flags & RT_PUSH_IPI_EXECUTING) { - raw_spin_lock(&rq->rt.push_lock); - /* Make sure it's still executing */ - if (rq->rt.push_flags & RT_PUSH_IPI_EXECUTING) { - /* - * Tell the IPI to restart the loop as things have - * changed since it started. - */ - rq->rt.push_flags |= RT_PUSH_IPI_RESTART; - raw_spin_unlock(&rq->rt.push_lock); - return; - } - raw_spin_unlock(&rq->rt.push_lock); - } +static void tell_cpu_to_push(struct rq *rq) +{ + int cpu = -1; - /* When here, there's no IPI going around */ + /* Keep the loop going if the IPI is currently active */ + atomic_inc(&rq->rd->rto_loop_next); - rq->rt.push_cpu = rq->cpu; - cpu = find_next_push_cpu(rq); - if (cpu >= nr_cpu_ids) + /* Only one CPU can initiate a loop at a time */ + if (!rto_start_trylock(&rq->rd->rto_loop_start)) return; - rq->rt.push_flags = RT_PUSH_IPI_EXECUTING; + raw_spin_lock(&rq->rd->rto_lock); + + /* + * The rto_cpu is updated under the lock, if it has a valid cpu + * then the IPI is still running and will continue due to the + * update to loop_next, and nothing needs to be done here. + * Otherwise it is finishing up and an ipi needs to be sent. + */ + if (rq->rd->rto_cpu < 0) + cpu = rto_next_cpu(rq); - irq_work_queue_on(&rq->rt.push_work, cpu); + raw_spin_unlock(&rq->rd->rto_lock); + + rto_start_unlock(&rq->rd->rto_loop_start); + + if (cpu >= 0) + irq_work_queue_on(&rq->rd->rto_push_work, cpu); } /* Called from hardirq context */ -static void try_to_push_tasks(void *arg) +void rto_push_irq_work_func(struct irq_work *work) { - struct rt_rq *rt_rq = arg; - struct rq *rq, *src_rq; - int this_cpu; + struct rq *rq; int cpu; - this_cpu = rt_rq->push_cpu; + rq = this_rq(); - /* Paranoid check */ - BUG_ON(this_cpu != smp_processor_id()); - - rq = cpu_rq(this_cpu); - src_rq = rq_of_rt_rq(rt_rq); - -again: + /* + * We do not need to grab the lock to check for has_pushable_tasks. + * When it gets updated, a check is made if a push is possible. + */ if (has_pushable_tasks(rq)) { raw_spin_lock(&rq->lock); - push_rt_task(rq); + push_rt_tasks(rq); raw_spin_unlock(&rq->lock); } - /* Pass the IPI to the next rt overloaded queue */ - raw_spin_lock(&rt_rq->push_lock); - /* - * If the source queue changed since the IPI went out, - * we need to restart the search from that CPU again. - */ - if (rt_rq->push_flags & RT_PUSH_IPI_RESTART) { - rt_rq->push_flags &= ~RT_PUSH_IPI_RESTART; - rt_rq->push_cpu = src_rq->cpu; - } + raw_spin_lock(&rq->rd->rto_lock); - cpu = find_next_push_cpu(src_rq); + /* Pass the IPI to the next rt overloaded queue */ + cpu = rto_next_cpu(rq); - if (cpu >= nr_cpu_ids) - rt_rq->push_flags &= ~RT_PUSH_IPI_EXECUTING; - raw_spin_unlock(&rt_rq->push_lock); + raw_spin_unlock(&rq->rd->rto_lock); - if (cpu >= nr_cpu_ids) + if (cpu < 0) return; - /* - * It is possible that a restart caused this CPU to be - * chosen again. Don't bother with an IPI, just see if we - * have more to push. - */ - if (unlikely(cpu == rq->cpu)) - goto again; - /* Try the next RT overloaded CPU */ - irq_work_queue_on(&rt_rq->push_work, cpu); -} - -static void push_irq_work_func(struct irq_work *work) -{ - struct rt_rq *rt_rq = container_of(work, struct rt_rq, push_work); - - try_to_push_tasks(rt_rq); + irq_work_queue_on(&rq->rd->rto_push_work, cpu); } #endif /* HAVE_RT_PUSH_IPI */ diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index a81c9782e98c..8aa24b41f652 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -505,7 +505,7 @@ static inline int rt_bandwidth_enabled(void) } /* RT IPI pull logic requires IRQ_WORK */ -#ifdef CONFIG_IRQ_WORK +#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_SMP) # define HAVE_RT_PUSH_IPI #endif @@ -527,12 +527,6 @@ struct rt_rq { unsigned long rt_nr_total; int overloaded; struct plist_head pushable_tasks; -#ifdef HAVE_RT_PUSH_IPI - int push_flags; - int push_cpu; - struct irq_work push_work; - raw_spinlock_t push_lock; -#endif #endif /* CONFIG_SMP */ int rt_queued; @@ -641,6 +635,19 @@ struct root_domain { struct dl_bw dl_bw; struct cpudl cpudl; +#ifdef HAVE_RT_PUSH_IPI + /* + * For IPI pull requests, loop across the rto_mask. + */ + struct irq_work rto_push_work; + raw_spinlock_t rto_lock; + /* These are only updated and read within rto_lock */ + int rto_loop; + int rto_cpu; + /* These atomics are updated outside of a lock */ + atomic_t rto_loop_next; + atomic_t rto_loop_start; +#endif /* * The "RT overload" flag: it gets set if a CPU has more than * one runnable RT task. @@ -658,6 +665,9 @@ extern void init_defrootdomain(void); extern int sched_init_domains(const struct cpumask *cpu_map); extern void rq_attach_root(struct rq *rq, struct root_domain *rd); +#ifdef HAVE_RT_PUSH_IPI +extern void rto_push_irq_work_func(struct irq_work *work); +#endif #endif /* CONFIG_SMP */ /* diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c index f51d123f9fe1..e50450c2fed8 100644 --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -268,6 +268,12 @@ static int init_rootdomain(struct root_domain *rd) if (!zalloc_cpumask_var(&rd->rto_mask, GFP_KERNEL)) goto free_dlo_mask; +#ifdef HAVE_RT_PUSH_IPI + rd->rto_cpu = -1; + raw_spin_lock_init(&rd->rto_lock); + init_irq_work(&rd->rto_push_work, rto_push_irq_work_func); +#endif + init_dl_bw(&rd->dl_bw); if (cpudl_init(&rd->cpudl) != 0) goto free_rto_mask;

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] WTF: patch "[PATCH] timekeeping: Eliminate the stale declaration of" was seriously submitted to be applied to the 4.14-stable tree?

by gregkh＠linuxfoundation.org

The patch below was submitted to be applied to the 4.14-stable tree. I fail to see how this patch meets the stable kernel rules as found at Documentation/process/stable-kernel-rules.rst. I could be totally wrong, and if so, please respond to <stable(a)vger.kernel.org> and let me know why this patch should be applied. Otherwise, it is now dropped from my patch queues, never to be seen again. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 8a7a8e1eab929eb3a5b735a788a23b9731139046 Mon Sep 17 00:00:00 2001 From: Dou Liyang <douly.fnst(a)cn.fujitsu.com> Date: Mon, 13 Nov 2017 13:49:04 +0800 Subject: [PATCH] timekeeping: Eliminate the stale declaration of ktime_get_raw_and_real_ts64() Commit ba26621e63ce got rid of ktime_get_raw_and_real_ts64(), but left its declaration behind. Remove it. Fixes: ba26621e63ce ("time: Remove duplicated code in ktime_get_raw_and_real()") Signed-off-by: Dou Liyang <douly.fnst(a)cn.fujitsu.com> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: Christopher S. Hall <christopher.s.hall(a)intel.com> Cc: joelaf(a)google.com Cc: arnd(a)arndb.de Cc: gregkh(a)linuxfoundation.org Cc: john.stultz(a)linaro.org Cc: deepa.kernel(a)gmail.com Cc: stable(a)vger.kernel.org Link: https://lkml.kernel.org/r/1510552144-20831-1-git-send-email-douly.fnst@cn.f… diff --git a/include/linux/timekeeping.h b/include/linux/timekeeping.h index 0021575fe871..51293e1aa4da 100644 --- a/include/linux/timekeeping.h +++ b/include/linux/timekeeping.h @@ -272,12 +272,6 @@ extern bool timekeeping_rtc_skipresume(void); extern void timekeeping_inject_sleeptime64(struct timespec64 *delta); -/* - * PPS accessor - */ -extern void ktime_get_raw_and_real_ts64(struct timespec64 *ts_raw, - struct timespec64 *ts_real); - /* * struct system_time_snapshot - simultaneous raw/real time capture with * counter value

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] Patch "dm bufio: fix integer overflow when limiting maximum cache size" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled dm bufio: fix integer overflow when limiting maximum cache size to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: dm-bufio-fix-integer-overflow-when-limiting-maximum-cache-size.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 74d4108d9e681dbbe4a2940ed8fdff1f6868184c Mon Sep 17 00:00:00 2001 From: Eric Biggers <ebiggers(a)google.com> Date: Wed, 15 Nov 2017 16:38:09 -0800 Subject: dm bufio: fix integer overflow when limiting maximum cache size From: Eric Biggers <ebiggers(a)google.com> commit 74d4108d9e681dbbe4a2940ed8fdff1f6868184c upstream. The default max_cache_size_bytes for dm-bufio is meant to be the lesser of 25% of the size of the vmalloc area and 2% of the size of lowmem. However, on 32-bit systems the intermediate result in the expression (VMALLOC_END - VMALLOC_START) * DM_BUFIO_VMALLOC_PERCENT / 100 overflows, causing the wrong result to be computed. For example, on a 32-bit system where the vmalloc area is 520093696 bytes, the result is 1174405 rather than the expected 130023424, which makes the maximum cache size much too small (far less than 2% of lowmem). This causes severe performance problems for dm-verity users on affected systems. Fix this by using mult_frac() to correctly multiply by a percentage. Do this for all places in dm-bufio that multiply by a percentage. Also replace (VMALLOC_END - VMALLOC_START) with VMALLOC_TOTAL, which contrary to the comment is now defined in include/linux/vmalloc.h. Depends-on: 9993bc635 ("sched/x86: Fix overflow in cyc2ns_offset") Fixes: 95d402f057f2 ("dm: add bufio") Signed-off-by: Eric Biggers <ebiggers(a)google.com> Signed-off-by: Mike Snitzer <snitzer(a)redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/md/dm-bufio.c | 15 ++++++--------- 1 file changed, 6 insertions(+), 9 deletions(-) --- a/drivers/md/dm-bufio.c +++ b/drivers/md/dm-bufio.c @@ -937,7 +937,8 @@ static void __get_memory_limit(struct dm buffers = c->minimum_buffers; *limit_buffers = buffers; - *threshold_buffers = buffers * DM_BUFIO_WRITEBACK_PERCENT / 100; + *threshold_buffers = mult_frac(buffers, + DM_BUFIO_WRITEBACK_PERCENT, 100); } /* @@ -1856,19 +1857,15 @@ static int __init dm_bufio_init(void) memset(&dm_bufio_caches, 0, sizeof dm_bufio_caches); memset(&dm_bufio_cache_names, 0, sizeof dm_bufio_cache_names); - mem = (__u64)((totalram_pages - totalhigh_pages) * - DM_BUFIO_MEMORY_PERCENT / 100) << PAGE_SHIFT; + mem = (__u64)mult_frac(totalram_pages - totalhigh_pages, + DM_BUFIO_MEMORY_PERCENT, 100) << PAGE_SHIFT; if (mem > ULONG_MAX) mem = ULONG_MAX; #ifdef CONFIG_MMU - /* - * Get the size of vmalloc space the same way as VMALLOC_TOTAL - * in fs/proc/internal.h - */ - if (mem > (VMALLOC_END - VMALLOC_START) * DM_BUFIO_VMALLOC_PERCENT / 100) - mem = (VMALLOC_END - VMALLOC_START) * DM_BUFIO_VMALLOC_PERCENT / 100; + if (mem > mult_frac(VMALLOC_TOTAL, DM_BUFIO_VMALLOC_PERCENT, 100)) + mem = mult_frac(VMALLOC_TOTAL, DM_BUFIO_VMALLOC_PERCENT, 100); #endif dm_bufio_default_cache_size = mem; Patches currently in stable-queue which might be from ebiggers(a)google.com are queue-4.9/lib-mpi-call-cond_resched-from-mpi_powm-loop.patch queue-4.9/dm-bufio-fix-integer-overflow-when-limiting-maximum-cache-size.patch

7 years, 7 months

1
0
0 0

[Linux-stable-mirror] Patch "dm: allocate struct mapped_device with kvzalloc" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled dm: allocate struct mapped_device with kvzalloc to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: dm-allocate-struct-mapped_device-with-kvzalloc.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 856eb0916d181da6d043cc33e03f54d5c5bbe54a Mon Sep 17 00:00:00 2001 From: Mikulas Patocka <mpatocka(a)redhat.com> Date: Tue, 31 Oct 2017 19:33:02 -0400 Subject: dm: allocate struct mapped_device with kvzalloc From: Mikulas Patocka <mpatocka(a)redhat.com> commit 856eb0916d181da6d043cc33e03f54d5c5bbe54a upstream. The structure srcu_struct can be very big, its size is proportional to the value CONFIG_NR_CPUS. The Fedora kernel has CONFIG_NR_CPUS 8192, the field io_barrier in the struct mapped_device has 84kB in the debugging kernel and 50kB in the non-debugging kernel. The large size may result in failure of the function kzalloc_node. In order to avoid the allocation failure, we use the function kvzalloc_node, this function falls back to vmalloc if a large contiguous chunk of memory is not available. This patch also moves the field io_barrier to the last position of struct mapped_device - the reason is that on many processor architectures, short memory offsets result in smaller code than long memory offsets - on x86-64 it reduces code size by 320 bytes. Note to stable kernel maintainers - the kernels 4.11 and older don't have the function kvzalloc_node, you can use the function vzalloc_node instead. Signed-off-by: Mikulas Patocka <mpatocka(a)redhat.com> Signed-off-by: Mike Snitzer <snitzer(a)redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/md/dm-core.h | 3 ++- drivers/md/dm.c | 7 ++++--- 2 files changed, 6 insertions(+), 4 deletions(-) --- a/drivers/md/dm-core.h +++ b/drivers/md/dm-core.h @@ -29,7 +29,6 @@ struct dm_kobject_holder { * DM targets must _not_ deference a mapped_device to directly access its members! */ struct mapped_device { - struct srcu_struct io_barrier; struct mutex suspend_lock; /* @@ -127,6 +126,8 @@ struct mapped_device { struct blk_mq_tag_set *tag_set; bool use_blk_mq:1; bool init_tio_pdu:1; + + struct srcu_struct io_barrier; }; void dm_init_md_queue(struct mapped_device *md); --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -21,6 +21,7 @@ #include <linux/delay.h> #include <linux/wait.h> #include <linux/pr.h> +#include <linux/vmalloc.h> #define DM_MSG_PREFIX "core" @@ -1511,7 +1512,7 @@ static struct mapped_device *alloc_dev(i struct mapped_device *md; void *old_md; - md = kzalloc_node(sizeof(*md), GFP_KERNEL, numa_node_id); + md = vzalloc_node(sizeof(*md), numa_node_id); if (!md) { DMWARN("unable to allocate device, out of memory."); return NULL; @@ -1605,7 +1606,7 @@ bad_io_barrier: bad_minor: module_put(THIS_MODULE); bad_module_get: - kfree(md); + kvfree(md); return NULL; } @@ -1624,7 +1625,7 @@ static void free_dev(struct mapped_devic free_minor(minor); module_put(THIS_MODULE); - kfree(md); + kvfree(md); } static void __bind_mempools(struct mapped_device *md, struct dm_table *t) Patches currently in stable-queue which might be from mpatocka(a)redhat.com are queue-4.9/dm-allocate-struct-mapped_device-with-kvzalloc.patch

7 years, 7 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror