--- a
+++ b/release-notes/oprofile-0.9.8
@@ -0,0 +1,108 @@
+OProfile is a powerful system-wide profiler for Linux. And now
+with the announcement of the 0.9.8 release, OProfile can also
+be used to profile a specific individual process.
+
+Read more at http://oprofile.sf.net.
+
+OProfile 0.9.8 has been released. OProfile is still in alpha,
+but has been proven stable for many users.
+
+
+Incompatibilities with previous release
+---------------------------------------
+
+- Support for pre-2.6 kernels has been removed
+- Relating to the above, the "--with-kernel-support" configure option is
+  no longer needed nor valid
+- The sample data format has changed (see bug #3309794 below for details)
+- The configure.in file has been renamed to configure.ac (should be a
+  transparent change to users)
+
+
+New features
+------------
+
+- A new 'operf' program is now available that allows non-root users to profile
+  single processes. 'operf' can also be used for system-wide profiling, but
+  root authority is required.  This capability requires a kernel version of
+  2.6.31 or greater.
+
+- New Processors Supports
+    * Tilera tile64
+    * Tilera tilepro
+    * Tilera tile-gx
+    * IBM System z10
+    * IBM System z196
+    * Intel Ivy Bridge
+    * ARMv7 Cortex-A5
+    * ARMv7 Cortex-A15
+    * ARMv7 Cortex-A7
+
+
+Bug fixes
+---------
+
+Filed bug reports:
+-------------------------------------------------------------------------
+|  BUG ID   |  Summary 
+|-----------|------------------------------------------------------------
+|  3483426  |  Fix configure on newer Debian and Ubuntu systems when using
+|           |  --enable-gui=qt4
+|  3448632  |  Make opcontrol --status show session dir
+|  3451782  |  Fix opimport to not try to import .jo ELF files
+|  3400883  |  Fix oparchive so a relative path can be passed in with
+|           |  --session-dir option
+|  3386923  |  Document that 'opannoate -a' needs symbol info
+|  3521382  |  Fix 'make check' warning treated as error when using newer gcc
+|  3448739  |  Change "Nr. event lost ..." field in the oprofiled.log to "Nr.
+|           |  samples lost ..."
+|  3448699  |  Fix how "Nr. non-backtrace samples" from oprofiled.log is
+|           |  calculated
+|  3309794  |  Change type for sample header mtime field to u64
+|  3393079  |  Document non-support for event-based profiling in guest
+|           |  environments
+|  3470511  |  Fix debuginfo processing to handle no symbol info in debuginfo
+|           |  file
+|  2573008  |  Fix debuginfo processing for ppc64 (incorrect source file/line#
+|           |  info)
+|  3405711  |  Fix opreport -X so total samples for the binary does not
+|           |  include any module counts
+-------------------------------------------------------------------------
+
+Bugs without a filed report (e.g., posted to the list):
+---------------
+- OProfile fix to pass string of options from opannotate to objdump
+- Fix build error on Fedora Rawhide (size_t type needed for bfd.h)
+- Fix compile error on Ubuntu 12.10
+- Fix build warning on Fedora 17 with gcc 4.7.0
+- Fix configure to not alter user variables and remove non-working --with-gcc
+  option
+- Fix opjitconv to not end abnormally when no jit dump directory is found
+- Add #include "config.h" before bfd.h (for build error "config.h must be
+  included before this header")
+- Make autogen.sh a bit more discriminating on automake version (fixes problem
+  on Fedora Rawhide)
+- Fix opannotate --assembly to work with prelinked runtime lib and non-prelink
+  debuginfo
+- Discard user context kernel samples for which we have no app_cookie
+- Print correct clock speed when using timer mode
+- Fix oprofile build warnings when using gcc 4.6.1 or newer
+- Fix event name clash in scorpion/scorpionmp
+- Fix debuginfo processing in prelink environments
+
+
+Known problems and limitations
+-------------------------
+- AMD Instruction Based Sampling (IBS) is not currently supported with the new
+  operf program.  Use the "legacy" opcontrol commands for IBS profiling.
+
+- When using operf to profile multiple events, the absolute number of events
+  recorded will usually be substantially fewer than expected. This is due
+  to a bug in the Linux kernel's Performance Events Subsystem that was fixed
+  sometime between Linux kernel version 3.1 and 3.5.
+
+- If NMI watchdog is not disabled (x86), opcontrol may fail to
+  allocate the hardware performance counter(s) it needs (bug #1468916).
+
+- Many Alpha ev67 events do not work (bug #931875).
+