- 2014-09-19: kernel 3.14
- 2014-01-08: kernel 3.12
- 2013-10-31: kernel 3.11
- 2013-07-31: kernel 3.10
- 2013-06-12: kernel 3.9
- 2013-02-28: kernel 3.8
- 2012-12-19: kernel 3.7
- 2012-11-05: kernel 3.5
- 2012-06-13: kernel 3.4
- 2012-03-31: kernel 3.3
- 2012-02-29: kernel 3.2
- 2011-11-30: kernel 3.1
- 2011-08-31: kernel 3.0
If you download any software from this web site please be aware of the Warranty Disclaimer and Limitation of Liabilities.
|upstream kernel 3.12||kernel 3.12 - upstream with feature descriptions.|
|linux-3.12-s390-message-catalog.tar.gz / MD5||"Development Stream" kernel 3.12 - message-catalog (optional)|
To download the linux-3.12.tar.gz visit: http://www.kernel.org/pub/linux/kernel/v3.x
The upstream kernel 3.12 contains the following functionality developed by the Linux on System z development team:
- Support of PCI Express
This feature provides support for Peripheral Component Interconnect Express (PCIe) devices and is a prerequisite for RDMA over Converged Ethernet (RoCE), and is available on zEC12 or zBC12 GA2 machines or newer, where 10GbE RoCE Express cards are supported.
See Device Drivers, Features, and Commands, Chapter 'PCI Express Support'.
- Convert interrupt handling to use generic hardirq
With this patch the common code hardirq interface is used to find the interrupt handler for the different interrupt sources.
This simplifies the PCI support for System z. The external, I/O and adapter interrupts are mapped to fixed interrupt numbers, the additional interrupts defined by the PCI/MSIX architecture are allocated dynamically.
- Implement software referenced bits
With this patch the page reference detection is altered from using the hardware referenced bit in the storage key to use a software,
fault based approach. The fault based approach makes page table entries invalid while the page table entry indicates not referenced.
The tradeoff is an increased number of page table invalidations and page faults against versus a reduction in storage key operations.
- Enable DASD ECKD raw_track_access reads without direct I/O
An application that wants to use a DASD in raw_track_access mode will usually use direct I/O to make sure that properly aligned requests are directly submitted to the driver. However, applications that are not aware of this mode, e.g. udev, will encounter I/O errors.
To make the use without direct I/O possible and avoid this kind of alignment errors, we now pad unaligned read requests with a dummy page, so that we can always read full tracks.
Please note that writing is still only possible for full track images that are properly aligned.
- zFCP: enable FCP hardware data router by default
Enabling the data router support by default can increase performance in certain situations.
It is safe to do so and tolerated in LPAR and under z/VM in case there is no data router support in that environment.
You can still disable the feature explicitly by means of the zfcp module parameter zfcp.datarouter=0.
See Device Drivers, Features, and Commands, Chapter 'SCSI-over-Fibre Channel device driver', sections 'zfcp device driver kernel parameters' and 'zfcp module parameters'.
For further information see development stream of kernel 3.2, where the original description can be found.
- The SCSI standalone dump tool and dump no longer require a file system
For more information please see Using the dump tools, Chapter 'Using the SCSI Dump device'.
This patch contains:
- Kernel message catalog.
- Add support for automatic message tags to the printk macro families dev_xyz and pr_xyz. The message tag consists of a component name and a 24 bit hash of the message text. For each message that is documented in the included kernel message catalog a man page can be created with a script (which is included in the patch). The generated man pages contain explanatory text that is intended to help understand the messages.
Note that only s390 specific messages are prepared appropriately and included in the generated message catalog.
This patch is optional as it is very unlikely to be accepted in upstream kernel, but is recommended for all distributions which are built based on the 'Development stream'.