Fixed and Known Issues

2 Fixed and Known Issues

This chapter describes the fixed and known issues for the Unbreakable Enterprise Kernel Release 4.

Important:

Run the yum update command regularly to ensure that the latest bug fixes and security errata are installed on your system.

Fixed Issues

The following issues have been fixed in this update.

OFED packages updated to release 1.0.0-8

The packages required for Oracle OFED release 1.0.0-8 have been updated for both Oracle Linux 6 and Oracle Linux 7.
kexec-tools package update to version 2.0.7-38.0.2

The kexec-tools package has been updated to version 2.0.7-38.0.2 to resolve an issue that caused iSCSI boot to fail on some systems.
Intel xHCI controller reset code (xhci) fixed

A workaround fix, provided by Intel, has been applied to the xHCI controller reset code to prevent an intermittent boot hang. (Bug ID 22805913)
Kernel panic when bio_endio() is called for devices with write cache enabled (WCE) fixed

A fix was applied to code in the kernel block IO controller to prevent kernel panic when a write or sync takes place on a device with WCE (Write Cache Enabled). (Bug ID 22820562)
ALSA usb-audio fixed to avoid freeing umidi object twice

A fix was applied to the ALSA usb-audio code which resulted in a double-free flaw in the USB MIDI driver. (Bug ID 22740866)
RDMA operations report success when the bcopy message fails fixed

A fix was applied to delay the signaling of a completion event for the RDMA (Remote Direct Memory Access) operation until the bcopy send event completes. (Bug ID 22847528)
Branch offset adjustment on backjumps after patching ctx expansion fixed

When ctx access is used, the kernel needs to expand instructions so that branch offsets are adjusted for forward and backward jumps. A fix was applied to correctly calculate the delta required for the backward jumps. (Bug ID 22740787)
i40e driver fixed to revert a commit that caused the Ethernet interface to flap when bonded

A fix was applied to the i40e driver to revert an earlier code commit that caused the Ethernet interface to flap when in a bond configuration. (Bug ID 22519254)
XFS data corruption while performing EOF zeroing fixed

A fix was applied to XFS for an issue that resulted in data corruption due to I/O lock management and pending Direct I/O not being drained before EOF zeroing. (Bug ID 22682207)
SCSI/bdi code reverted to fix multiple issues with backing devices

A previous change to the SCSI/bdi code was reverted to resolve multiple issues with backing devices, including a race condition that occurred during bdi reregistration. This fix is in line with changes to the mainline kernel. (Bug ID 22679087)
libfuse I/O error for a chmod on DBFS fixed

A fix was applied to libfuse to resolve an issue that resulted in an I/O error when running a chmod operation on files or directories on DBFS. (Bug ID 22637182)
libfuse fixed for a bug where the numa mount option for DBFS was not shown in /proc/mounts

A fix was applied to libfuse to resolve an issue where the numa mount option was not visible in /proc/mounts if it was used to perform a mount of a DBFS. (Bug ID 22637071)
Block driver I/O errors and performance degradation in some environments fixed

A fix that was applied to resolve a regression in the kernel block driver code that caused performance degradation and I/O errors due to the size of an I/O request at the block driver level. (Bug ID 22611290)
OCFS2 system reboot issue due to a driver error fixed

A fix that was applied to the ocfs2 driver so that when the journal commit receives an error from the device driver, the journal changes to an abort state that stops the commit thread. (Bug ID 22611290)
OCFS2 heartbeat stability and mount failure issue fixed

A fix that was applied to the OCFS2 cluster heartbeat code to increase unsteady iterations and improve heartbeat stability. (Bug ID 22611290)
NFS server crash triggered by 1MB NFS WRITE fixed

A fix has been applied for a bug in NFS server code that caused a general protection fault if an NFS client triggered a 1MB NFS WRITE operation on an NFS/RDMA mount.
Mellanox InfiniBand driver updated to use kvfree() instead of kfree() to free buffers

A fix has been applied to the mlx4 driver to use kvfree() rather than kfree() to free buffers, since kvfree() can handle memory allocated by both kmalloc() and vmalloc().
RDS fencing operation issue fixed for parallel write operations that caused connections to drop

A fix has been applied to the mlx4 driver code that was causing the fencing operation for RDS (Reliable Datagram Sockets) to drop connections as a result of a process that recycled keys too quickly. This problem affected Oracle Database processes using shared memory buffers that had pending RDMA operations.
RDS enhancements for connection tracking and latency tracing

Enhancements to code for RDS connection tracking have been applied to make it easier to debug connection shutdown and reconnect behavior. A tracker variable keeps the latest value of the connection drop source making it easier to identify the reason for a connection drop. Tracking is also applied to determine the length of time that a reconnection attempt takes to fail. An interface has been applied to trace receive message latency.
Block Loop Driver Updated To Log ENOSPC And EDQUOT Errors

The block loop driver has been modified to write masking ENOSPC and EDQUOT errors before they get masked to EIO, to assist with debugging on loop devices.
btrfs quota enable command allows correct file space usage calculation

A bug in the btrfs quota enable command which caused existing files not to count toward space usage on a non-empty file system has been fixed. (Bug ID 16569350)
The btrfs quota rescan command has been implemented and is working

The btrfs quota rescan command has been implemented in this release. The command correctly performs a rescan and returns a message to indicate whether the scan was successful. (Bug ID 16569350)

Known Issues

This section describes the known issues in this update.

btrfs

The btrfs filesystem balance command does not warn that the RAID level can be changed under certain circumstances, and does not provide the choice of cancelling the operation. (Bug ID 16472824)
The copy-on-write nature of btrfs means that every operation on the file system initially requires disk space. It is possible that you cannot execute any operation on a disk that has no space left; even removing a file might not be possible. The workaround is to run sync before retrying the operation. If this does not help, remount the file system with the -o nodatacow option and delete some files to free up space. See https://btrfs.wiki.kernel.org/index.php/ENOSPC.
When you overwrite data in a file, starting somewhere in the middle of the file, the overwritten space is counted twice in the space usage numbers that btrfs qgroup show displays. Using the btrfs quota rescan does not help fix this issue either. (Bug ID 16609467)
If you use the -s option to specify a sector size to mkfs.btrfs that is different from the page size, the created file system cannot be mounted. By default, the sector size is set to be the same as the page size. (Bug ID 17087232)
The btrfs-progs and btrfs-progs-devel packages for use with UEK R4 are made available in the ol6_x86_64_UEKR4 and ol7_x86_64_UEKR4 ULN channels and the ol6_UEKR4 and ol7_UEKR4 Oracle Linux yum server repositories. In UEK R3, these packages were made available in the ol6_x86_64_latest and ol7_x86_64_latest ULN channels and the ol6_latest and ol7_latest Oracle Linux yum server repositories.

DIF/DIX is not supported for ext file systems

The Data Integrity Field (DIF) and Data Integrity Extension (DIX) features that have been added to the SCSI standard are dependent on a file system that is capable of correctly handling attempts by the memory management system to change data in the buffer while it is queued for a write.

The ext2, ext3 and ext4 file system drivers do not prevent pages from being modified during I/O which can cause checksum failures and a "Logical block guard check failed" error. Other file systems such as XFS are supported. (Bug ID 24361968)

Docker

Running yum install within a container on an overlayfs file system can fail with the following error:
```
Rpmdb checksum is invalid: dCDPT(pkg checksums): package_name
```
This error breaks Dockerfile builds and is a known issue upstream (see https://github.com/docker/docker/issues/10180.)

The workaround is to run touch /var/lib/rpm/* before installing the package. (Bug ID 21804564)

DTrace

Argument declarations for probe definitions cannot be declared with derived types such as enum, struct, or union.
The following compiler warning can be ignored for probe definition arguments of type string (which is a D type but not a C type):
```
provider_def.h:line#: warning: parameter names (without types) in function declaration
```
Multi-threaded processes under ustack(), usym(), uaddr() and umod() which perform dlopen() in threads other than the first thread may not have accurate symbol resolution for symbols introduced by the dlopen(). (Bug ID 20045149)

LXC

The lxc-net service does not always start immediately after installation on Oracle Linux 6

The lxc-net service does not always start immediately after installation on Oracle Linux 6, even though this action is specified as part of the RPM post-installation script. This can prevent the lxcbr0 interface from coming up. If this interface is not up after installation, you can manually start it by running service lxc-net start. (Bug ID 23177405)
LXC read-only ip_local_port_range parameter

With lxc-1.1 or later and UEK R4, ip_local_port_range is a read-writable parameter under /proc/sys/net/ipv4 in an Oracle Linux container rather than being read-only. (Bug ID 21880467)

Console Appears to Hang when Booting

When booting Oracle Linux 6 on hardware with an ASPEED graphics controller, the console may appear to hang during the boot process after starting udev. However, the system does boot properly and is accessible. The workaround is to add nomodeset as a kernel boot parameter in /etc/grub.conf. (Bug ID 22389972)

OFED iSER feature causes Out-Of-Memory error on Oracle Linux 6

An Oracle Linux 6 system with the oracle-ofed-release packages installed and an iSER (iSCSI Extensions for RDMA) target configured, hangs and generates an Out-Of-Memory error when the initiator logs into the iSER target disk. The login fails and the target system hangs. The following error is displayed on the remote console:

Out of memory: Kill process ...

The cause is under investigation and there is no known workaround at this time. (Bug ID 22230403)

XFS quotas are disabled after a read-only remount on Oracle Linux 6

Quotas are disabled on XFS if the file system is remounted with read-only permissions on Oracle Linux 6. (Bug ID 22908906)

Open File Description (OFD) locks are not supported on NFSv4 mounts

NFS is not designed to handle OFD locking. (Bug ID 22948696).