LinuxChanges

Comprehensible changelog of the linux kernel, inspired by Dragonfly's status. Other places to get news about the linux kernel are Kernel traffic, LWN kernel status, LWN driver porting guide and LWN list of API changes in 2.6 - many articles here link to LWN and much of the 2.6.x changelog was restored from them (you may be be interested in subscribing so Jonathan can keep up the good work ) Changes made to the 2.5.x-2.6.0 development series can be found in the Linux25Changes page. Before adding things here look at the RULES section at the end of the page!!

2.6.14

• Released October 27, 2005 changelog

• Numa-aware slab allocator: It creates slabs on multiple nodes and manages slabs in such a way that locality of allocations is optimized. Each node has its own list of partial, free and full slabs. All object allocations for a node occur from node specific slab lists (commit - benchmarks)

• Lazy page table copies in fork() for VMAs without anonymous pages (the ones with anonymous pages are still copied): Defer copying of ptes until fault time when it is possible to reconstruct the pte from backing store, speeding up fork() greatly specially for processes using lots of shared memory (commit)

• Add /proc/$PID/smaps: This file will shows how much memory is resident in each mapping. Useful for people who want to perform memory consumption analysis (commit)

• Add /proc/$PID/numa_maps: This file will show on which nodes pages reside (commit)

• Lock-free file descriptor look-up (commit) - (commit)

• Four-level page table support for the ppc64 architecture: extends the usable user address range to 44 bits (16T). (commit)

• Support hotplug cpu on 32-bit SMP powermacs: When a cpu is off-lined, it is put into sleep mode with interrupts disabled. It can be on-lined again by asserting its soft-reset pin, which is connected to a GPIO pin (commit)

• Add TASK_NONINTERACTIVE task state bit to the cpu scheduler: It can be used by blocking points to mark the task's wait as "non-interactive". This does not mean the task will be considered a CPU-hog - the wait will simply not have an effect on the waiting task's priority - positive or negative alike (commit)

• PPTP (Point-to-Point Tunneling Protocol) conntrack support and NAT helper added to allow proper management of multiple NAT'd PPTP connections. PPTP, a Microsoft-designed protocol, is used in many places to provide VPN access due to the native support in Microsoft Windows. Wikipedia article (commit)

• DCCP: "Datagram Congestion Control Protocol". Datagram protocol (like UDP), but with a congestion control mechanism. (LWN article) Currently a RFC draft (commit)

• Implement SKB fast cloning: Protocols that make extensive use of SKB cloning, for example TCP, eat at least 2 allocations per packet sent as a result. To cut the kmalloc() count in half, we implement a pre-allocation scheme wherein we allocate 2 sk_buff objects in advance, then use a simple reference count to free up the memory at the correct time (commit)

• Add netlink connector: userspace <-> kernel space easy to use communication module which implements easy to use bidirectional message bus using netlink as its backend (commit), also a "async connector mode" (commit)

• IPV6: Support several new sockopt / ancillary data in Advanced API (RFC3542): IPV6_RECVPKTINFO, IPV6_PKTINFO, IPV6_RECVHOPOPTS, IPV6_HOPOPTS, IPV6_RECVDSTOPTS, IPV6_DSTOPTS, IPV6_RTHDRDSTOPTS, IPV6_RECVRTHDR, IPV6_RTHDR, IPV6_RECVHOPOPTS, IPV6_HOPOPTS, IPV6_{RECV,}TCLASS (commit) - (commit)

• FUSE: Allows to implement a fully functional filesystem in a userspace program (commit)

• 9P support: Linux port of the Plan9's 9P protocol (commit)

• RelayFS: Relayfs is just a bunch of per-cpu kernel buffers that can be efficiently written into from kernel code. These buffers are represented as files which can be mmap'ed and directly read from in user space. The purpose of this setup is to provide the simplest possible mechanism allowing potentially large amounts of data to be logged in the kernel and 'relayed' to user space." (commit)

• securityfs filesystem (LWN article) (commit)

• Suspend support for CIFS filesystem (commit)

• Enable atomic inode security labeling on ext3 and ext2: it enables atomic security labeling of newly created inodes by altering the fs code to invoke a new LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state during the inode creation transaction. This parallels the existing processing for setting ACLs on newly created inodes (commit) - (commit) - (commit)

• NTFS: Support journals ($Logfile) which have been modified by chkdsk. This means users can boot into Windows after we marked the volume dirty. The Windows boot will run chkdsk and then reboot. The user can then immediately boot into Linux rather than having to do a full Windows boot first before rebooting into Linux (commit)

• SAS transport class (commit)

• Remove ACPI S4 BIOS support (commit)

• ipw2100 and ipw2200 wireless drivers (intel centrino) (commit) - (commit)

• HostAP: Adds support to work as "Wireless Access Point" (commit)

• Wireless extensions API update to version 19 (mail) (commit)

• ISA DMA suspend for x86/x86-64 (commit) - (commit)

• PCI: Support PCI PM CAP version 3 (commit)

• Keys: Add possessor permissions to keys. It adds extra permission grants to keys for the possessor of a key in addition to the owner, group and other permissions bits. This makes SUID binaries easier to support without going as far as labelling keys and key targets using the LSM facilities (commit)

• Driver for the IBM Hard Drive Active Protection System (HDAPS), an accelerometer found in most modern Thinkpads (LWN article) (commit)

• Add Apple USB touchpad driver for the USB touchpad which can be found on post-February 2005 Apple Powerbooks (commit)

• Spinlock consolidation (commit)

• Largefile support for accounting: The accounting subsystem in the kernel can not correctly handle files larger than 2GB. This fixes it by adding the O_LARGEFILE flag (commit)

• Detect soft lockups, new debug feature (CONFIG_DETECT_SOFTLOCKUP). When enabled then per-CPU watchdog threads are started, which try to run once per second. If they get delayed for more than 10 seconds then a callback from the timer interrupt detects this condition and prints out a warning message and a stack dump (commit)

• Add kzalloc(): new function, it allows to obtain pre-zeroed memory (LWN article) (commit)

• Add schedule_timeout_{,un}interruptible() (LWN article) (commit)

2.6.13

• Released August 29, 2005 changelog

• Execute-in-place support: Traditionally, programs are loaded from disk to memory to be executed. However, the current wave of embedded devices store programs in a ROM/flash chip. XIP allows the kernel executing programs directly from that ROM, without being copied to RAM (saving RAM space), and bypassing the page cache/io scheduler layers (since they're not needed). (LWN article) (commit)

• i386 CPU hotplug support: provides i386 architecture support for safely unregistering and registering processors during runtime (commit)

• Add /proc/sys/fs/suid_dumpable to provide system-wide control over whether or not set-user-ID and set-group-ID processes produce core dumps. The values accepted are: 0 (default) - traditional behaviour, any process which has changed privilege levels or is execute only will not be dumped. 1 - (debug) - all processes dump core when possible. The core dump is owned by the current user and no security is applied. This is intended for system debugging situations only. Ptrace is unchecked. 2 - (suidsafe) - any binary which normally would not be dumped is dumped readable by root only. This allows the end user to remove such a dump but not access it directly. For security reasons core dumps in this mode will not overwrite one another or other files. This mode is appropriate when adminstrators are attempting to debug problems in a normal environment (commit)

• x86 now uses the generic PCI bus setup code for assigning unassigned resources

• Voluntary preemption patches

• build-time configurable clock interrupt frequency: Now HZ defaults to 250 in x86. 1000 is better for "interactivity" (desktops) and 100 is good for performance (servers) and saves some energy in laptops. (LWN article)

• inotify (LWN article) (commit)

• Support for the Xtensa architecture: 32-bit architecture used in embedded devices (commit) - (commit) - (commit) - (commit) - (commit)

• Improved CFQ IO scheduler: With support for I/O priorities (LWN article) (commit)

• kexec and kdump: Kexec allows users to load a new kernel from another running kernel. By preserving the memory contents in a crash scenario, kexec allows to implement kdump. Kdump is able to get a memory dump of the previous kernel, and be used as a debugging tool. (LWN article) (commit)

• Runtime selectable TCP congestion algorithm: Allow using setsockopt to set TCP congestion control to use on a per socket basis. (commit) - (commit) (LWN article)

• Add several TCP congestion modules: H-TCP commit, TCP Hybla (commit), High Speed TCP (HS-TCP) (commit), TCP Westwood (commit), TCP BIC (commit)

• Add Direct Rendering Manager device driver for VIA Unichrome chipsets (commit)

• 32/64-bit DRM ioctl compatibility: 64 bit architectures running 32 bit software have to be compatible with 32-bit ioctl calls (commit)

• Add ACL support for NFSv3 (commit) - (commit) and NFSv4 (commit) - (commit)

• New driver for the "trusted computing" (TPM) crap^Wchip (commit)

• Removal of the devfs configuration option

2.6.12

• Released June 17, 2005 changelog

• Pageout throttling: With silly pageout testcases it is possible to place huge amounts of memory under I/O. With a large request queue (CFQ uses 8192 requests) it is possible to place _all_ memory under I/O at the same time. This means that all memory is pinned and unreclaimable and the VM gets upset and goes oom. This patch limits the amount of memory which is under pageout writeout to be a little more than the amount of memory at which balance_dirty_pages() callers will synchronously throttle. This means that heavy pageout activity can starve heavy writeback activity completely, but heavy writeback activity will not cause starvation of pageout (commit)

• Address space randomization: With these patches applied, each process's stack will begin at a random location, and the beginning of the memory area used for mmap() (which is where shared libraries go, among other things) will be randomized as well (LWN article) (commit) - (commit) - (commit) - (commit) - (commit) - (commit) - (commit) - (commit)

• Cpusets (LWN article) (commit)

• "resource limits": RLIMIT_NICE and RLIMIT_RTPRIO added to allow non-root tasks to raise nice and rt priorities. Defaults to traditional behavior (LWN article) (commit)

• Multilevel security implementation for SELinux

• Support for Cyrix MediaGX (aka GEODE) CPUs. Linux and GCC treat this chip as a 586TSC with some extended instructions and alignment reqirements (commit)

• I/O barrier support for serial ATA drives

• Block I/O barrier rewrite (enables full barrier support on serial ATA drives)

• Annotate /proc/$PID/maps with [heap]/[stack]/[vdso] markers (commit)

• SuperHyway bus support

• Device mapper multipath support (LWN article)

• Hot-pluggable parallel ports (commit)

• Updated FAT attributes: This updates the FAT attributes as well as corrects the handling of VFAT ctime (commit)

• Handle MS_SYNCHRONOUS flag in FAT: FAT filesystem has been ignoring the "sync" mount option for ages. This patches fixes this, but (obviously) degrades performance unless you mount your FAT filesystem as asynchronous ("async mount option) (commit)

• Add timing information to printk messages (commit)

• Allow admin to enable only some of the Magic-Sysrq functions (commit)

• Loglevel boot option (commit)

• Remove IPV6 "experimental" status

2.6.11

• Released March 2, 2005 changelog

• Conversion to 4-level page tables (LWN article) (commit)

• New Pipe implementation (LWN article)

• "Big Kernel Semaphore": Turns the Big Kernel Lock into a semaphore (LWN article) (commit)

• Introduces the CAP_AUDIT_WRITE and CAP_AUDIT_CONTROL capabilities (commit)

• Simplify readahead code (commit) - (commit)

• Support for Extended Attributes in the body of large inode in ext3: saves space and improves performance in some cases (commit)

• Add /proc/sys/kernel/bootloader_type (commit)

• Remove bitmaps from the buddy allocator (commit)

• Enhanced I/O and memory accounting (commit) - (commit)

• AMD Dual-core support (commit)

• Add Fujitsu FR-V CPU architecture (commit)

• TCP port randomization (commit)

• SATA support for Intel ICH7

• DebugFS (LWN article)

• Infiniband support (LWN article)

• Remove UMSDOS (commit)

• Major problems with TCP/IP BIC (the default congestion control) are resolved

2.6.10

• Released December 24, 2004 changelog

• Accounting: report single record for multithreaded process. In kernels before 2.6.10, a separate process accounting record was written for each thread created using NTPL. Since 2.6.10, a single accounting record is written for the entire process (commit)

• x86-64 clustered APIC support (commit)

• Make rlimit settings per-process instead of per-thread for POSIX compliance (commit)

• POSIX compliant CPU clocks (commit)

• Show aggregate per-process counters in /proc/$PID/stat: Add up resource usage counters for live and dead threads to show aggregate per-process usage. This mirrors the new getrusage() semantics. /proc/$PID/task/$TID/stat still has the per-thread usage (commit)

• Report the pages used for pagetables in /proc/$PID/status (commit)

• Display committed memory limit (per the current overcommit ratio) and the amount of memory remaining under this limit in meminfo (commit)

• Thrashing control tuning (commit)

• Big Kernel Lock preemption

• IRQ subsystem code rework (commit)

• Removal of the "BIO walking" helper functions

• Generic circular buffer type (LWN article) (commit)

• Ext3 block reservation (LWN article) (commit)

• Ext3 online resizing patches (LWN article) (commit)

• Add I/O error handling to journal operations in reiserfs (commit)

• sysfs backing store (LWN article)

• I/O space write barriers (LWN article) (commit)

• Modular, on-the-fly switchable I/O schedulers (LWN article) (commit)

• CFQ v2 update (commit)

• BSD secure levels module (commit)

• In-kernel cryptographic key management (commit)

• DVD+RW support (commit)

• CDRW packet writing support (commit)

• kernel events notification mechanism

• Lock initializater unification (commit)

• Panick blinking: Makes the keyboard blink when a oops happens - useful to know if the kernel has really oopsed or not (ej: X.org bugs) (commit)

• Reworking of spin lock initialization (LWN article)

• Configurable /proc/kcore (commit)

2.6.9

• Released October 19, 2004 changelog

• Token-based thrashing control (LWN article) (commit)

• Concurrent O_SYNC write support: In databases it is common to have multiple threads or processes performing O_SYNC writes against different parts of the same file. (commit)

• Reiserfs v3 barrier support: Add reiserfs support for flush barriers, mount with -o barrier=flush to enable them. Barriers are triggered on fsync and for log commits (commit)

• ext3 barrier support: Mount with "mount -o barrier=1" to enable barriers (commit)

• Support of disk barriers: (commit)

• m32r architecture support

• Scheduler statistics: adds lots of CPU scheduler stats in /proc/$PID/stat (commit)

• RLIMIT_MEMLOCK semantics change: In Linux kernels before 2.6.9, only privileged processes (CAP_IPC_LOCK) could lock memory, and the RLIMIT_MEMLOCK soft resource limit placed an upper limit on the number of bytes that a privileged process can lock. Since kernel 2.6.9: no limits are placed on the amount of memory that a privileged process can lock; and an unprivileged process is now able to lock memory up to the soft limit defined by RLIMIT_MEMLOC (commit)

• Tunable "max sectors" limit for block I/O requests (can help latency reduction)

• New prctl() option allowing programs to change their name (PR_GET_NAME)

• AMD dual-core support

• Out-of-line spinlocks (LWN article) (commit)

• Support for POSIX's waitid() Also, now linux now supports the POSIX specification that SIGCHLD is sent to the parent when one of its children resumes as a consequence of receiving a SIGCONT signal. Also, WCONTINUED flag is added to waitpid() and waitid() (commit).

• Shared memory scalability improvements

• "flex mmap" user-space memory layout (LWN article) (commit)

• x86 PAE swapspace expansion. PAE is artificially limited in terms of swapspace to the same bitsplit as ordinary i386, a 5/24 split (32 swapfiles, 64GB max swapfile size), when a 5/27 split (32 swapfiles, 512GB max swapfile size) is feasible. This patch transparently removes that limitation by using more of the space available in PAE's wider ptes for swap ptes (commit)

• Show Active/Inactive on per-node meminfo: The patch below enable to display the size of Active/Inactive pages on per-node meminfo (/sys/devices/system/node/node%d/meminfo) like /proc/meminfo (commit)

• Change in TCP ICMP source quench behavior

• USB "on the go" support

• New USB storage driver

• Support for more than eight partitions on BSD-labeled disks

• Ethtool support in the loopback driver

• NETIF_F_LLTX interface (LWN article)

• DSCP decapsulation for IPsec (commit)

• Removal of the ancient "busmouse" driver

• Infrastructure for cluster-wide file locking

• DRM subsystem cleanups

• "fake NUMA" mode for x86-64 testing

• Small-footprint tmpfs implementation

• Support for scheduler profiling (seeing where context switches come from)

• Automatic TCP window scaling calculation

• Some VFS interface improvements

• Executable support in hugetlb mappings

• The Whirlpool digest algorithm

• Removal of the very last suser() call

2.6.8

• Released August 14, 2004 changelog

• Allow x86 to reenable interrupts on lock contention (commit)

• VFS shrinkage tuning: This adds /proc/sys/vm/vfs_cache_pressure, which tunes the vfs cache versus pagecache scanning pressure. At vfs_cache_pressure=0 we don't shrink dcache and icache at all, at vfs_cache_pressure=100 there is no change in behaviour, at vfs_cache_pressure > 100 we reclaim dentries and inodes harder (commit)

• Conversion to the new symbolic link resolution code (which will eventually allow an increase in the maximum link depth) (commit)

• Add O_NOATIME open flag support (GNU extension): If this bit is set, read will not update the access time of the file. It is useful if you want to do something with the file atime (for instance, moving files that have not been accessed in a while to somewhere else, or something like Debian's popularity-contest) but you also want to read all files periodically (for instance, tripwire or debsums) (commit)

• MNT_EXPIRE for umount(): Intrinsic automount and mountpoint degradation support. This adds support for a filesystem (such as kAFS) to perform automounting intrinsically without the need for a userspace daemon. It also adds support for such mountpoints to be degraded at the filesystem's behest until they've been untouched long enough that they'll be removed (commit)

• Process Accounting Version 3 format added (commit)

• RLIMIT_SIGPENDING added, introducing per user rlimits for both queued signals and POSIX message queues (commit), and remove unused queued_signals global accounting - rtsig-max and rtsig-nr /proc interfaces went away (commit)

• support for 64-bit Super-H hardware

• Removal of the PC9800 subarchitecture (lack of mainteinance) (commit)

• Preemptible kernel support for the PPC64 architecture (commit)

• Oprofile support for ppc32 (commit)

• Support for new Apple Powerbooks (commit)

• x86 no-execute support (commit)

• Asynchronous I/O support for USB gadgets

• HPET (High Precision Event Timer) support (commit)

• Reworked symbolic link lookup implementation (LWN article)

• Lockless loopback (commit)

• New "CPU mask" implementation

• RCU performance improvements (commit) - (commit)

• New wait_event_interruptible_exclusive() macro

• Sysfs knobs for tuning the CFQ I/O scheduler (commit)

• Mirroring, snapshot and dm-zero targets for the device mapper (commit) - (commit) - (commit)

• Reiserfs data=journal support (commit)

• Added permissions checking on raw SCSI commands from user space (LWN article) (commit)

• Removal of the fcntl() file operations method

• New internal infrastructure for handling file positioning and seekability (commit)

• Removal of the (non-functional) "fastroute" networking option

• TEA, XTEA, Khazad crypto algorithms (commit) - (commit)

• Add deb-pkg build option (commit)

• TCP/IP congestion control changes from Reno to BIC

2.6.7

• Released June 16, 2004 changelog:

• Scheduling Domains (LWN article) (commit)

• Full object-based reverse-mapping scheme and removal of the per-page PTE chains (LWN article)

• Filtered Wakeups (LWN article)

• Ability to re-enable interrupts while waiting in spin_lock_irqsave() (for all architectures now)

• msleep() function for millisecond-scale waits

• del_singleshot_timer(): Deleting timers quickly (LWN article)

• shrink "dentry" structure

• Speedup readahead for seeky workloads (commit)

• Support for quotas, extended attributes, ACLs, SELinux in reiserfs (commit) - (commit) - (commit) - (commit)

• Removal of the Intermezzo filesystem (lack of mainteinance) (commit)

• Remove IDE PATA TCQ support: It's been disabled some time ago, PATA TCQ has so many technical short comings, that it was never really interesting (commit)

• Dynamic addition of virtual disks on PPC64 iSeries (commit)

• Implement separate per-cpu stacks for processing interrupts and softirqs, along the lines of the CONFIG_4KSTACKS stuff on x86 (commit)

• PowerPC 750GX support (commit)

• new API for NUMA systems

• Debugging option to put data symbols in kallsyms (commit)

2.6.6

• Released May 10, 2004 changelog

• Removal of (struct page)->list and page dirty/clean/free lists, replaced by radix-tree tagging (LWN article)

• Network packet timestamping optimization

• Binary Increase Control (BIC) TCP developed by NCSU. It is yet another TCP congestion control algorithm for handling big fat pipes. For normal size congestion windows it behaves the same as existing TCP Reno, but when window is large it uses additive increase to ensure fairness and when window is small it uses binary search increase (commit)

• POSIX message queues (commit)

• fsync() and fdatasync() speed improvements to ext2/3 (commit)

• Addition of the fcntl() method to the file_operations structure (LWN article)

• Laptop Mode (LWN article) (commit)

• Oprofile for s390 (commit)

• 4KB kernel stacks option for the i386 architecture (commit)

• Non-executable stack support for several architectures (commit)

• Make sysfs configurable (commit)

• Reiserfs updates: data=ordered support, space preallocation, laptop mode support, logging rework, support for nested transactions (commit) - (commit) - (commit) - (commit) - (commit)

• Reiserfs and ext3 "commit=0 support": Restores commit interval to the default value (commit) - (commit)

• Ext3 journalled quotas (commit)

• NFS v4: Implement server-side reboot recovery (mostly) (commit)

• IPv6 support in SELinux (commit)

• The lightweight auditing framework (LWN article) (commit)

• A mechanism which allows block drivers to respond to queries about the congestion state of their queues

• "per-device unplugging patch" (LWN article)

• CFQ scheduler (commmit)

• External module support (commit)

• Generic snapshot support code for filesystems (taken from XFS) (commit)

2.6.5

• Released April 4, 2004 changelog

• Adaptative lazy readahead (commit)

• CDROMREADAUDIO dma support: support DMA for extrating audio (commit)

• Netpoll infrastructure (LWN article)

• New "kref" reference counting mechanism (LWN article)

• AIO support for reiserfs (commit)

• Read-only support for UFS2 (commit)

• Display number of slab, mapped and pagetable pages in the sysrq-M output (commit)

2.6.4

• Released March 11, 2004 changelog

• NGROUPS_MAX, which sets the maximum number of groups a user can belong to has increased the value to 65536 (was previously 32). This limit is also visible via the read-only /proc/sys/kernel/ngroups_max file. (commit)

• HFS rewrite and HFS+ support (commit) - (commit)

• Add SOCK_SEQPACKET for Unix domain sockets (commit)

• Support for the Intel "ia32e" arch

• PPC64 iommu and TBL flush rework (commit) - (commit)

• UTF-8 tty mode (commit)

• Dynamic PTY allocation (up to a million PTY devices)

• Sysfs support for SCSI tapes and bluetooth devices

• ARC4 crypto module Support for large numbers of groups

• Generic kernel thread infrastructure (LWN article)

• Groundwork for the hotplug CPU code

• ARC4 crypto module (commit)

• Add dm-crypto crypto module (commit)

• Enable coredumps > 2 GB (commit)

• Add -mregparm=3 config option (commit)

2.6.3

• Released February 18, 2004 changelog

• G5 support (commit)

• Support up to 255 char columns in virtual terminals (commit)

• Set HZ to 1000 in PPC32 (commit)

• Removal of the USB scanner code: moved support to userspace (libusb)

• New DMA pool abstraction (LWN article)

• "context mount" support for SELinux

2.6.2

• Released February 4, 2004 changelog

• RAID 6 implementation (commit)

• One-shot support for epoll (commit)

• Add support for m68knommu (commit)

• Schedulers interactivity improvements (commit)

• New Qlogic SCSI driver

2.6.1

• Released January 9, 2004 changelog

• Message Signaled Interrupt support - MSI (LWN article) (commit)

• Add 32bit a.out support for x86-64 (commit)

• Add `gcc -Os' config option (commit)

• Extensible Firmware Interface (EFI) support. (commit)

2.6.0

• Released December 18, 2003 changelog

Rules:

• Name of the new feature/description of the change, including a small explanation if possible

• If its a feature and the feature has some web page add a link to it

• If there's a paper (OLS!) or an article add a link to it. Or write one

• If there's a interesting mail (benchmarks, announcements if there's no web page, etc) about it in the mailing list, get a link in one of the list's archives (at http://marc.theaimsgroup.com/?l=linux-kernel please) and add the link here

• Add commit's link if possible. For commits add a "(commit)" word and add the link there. Don't add the commit link to another word (please).

• Searching commits for a given feature is easy. Just look for it in the shortlog, or try to find a file related with that change, then search it in the GIT tree representation and click the "history" link to see all the changes made which affects that file. The release dates will help you too.

• There're two main GIT trees where you can search for commits: the one which has all the stuff commited after 2.6.12-rc2 and the one which has all the stuff BEFORE 2.6.12-rc2

TODO:

• Import relevant data from davej's post halloween document or remaining stuff from Kernelnewbies status

• Find links to corresponding changesets (!!)

• Keep track of what gets changed