-2007-09-20 Richard Jones <rjones@redhat.com>
+2012-03-06 Richard W.M. Jones <rjones@redhat.com>
- * virt-top/virt_top.ml: qemu:/// URIs (and others) don't support
- virConnectGetHostname. Catch this exception.
+ Version 1.0.7.
- * Makefile.in: Disable virt-df in this release until I have
- time to fix it properly.
- * Makefile.in: Fix 'check-manifest' rule. Added 'make release'.
- * MANIFEST: Updated MANIFEST.
+ Ignore aclocal.m4.
-2007-09-19 Richard Jones <rjones@redhat.com>
+ Document requirements in README file.
- * virt-top/virt_top.ml: Ignore transient error from
- C.num_of_defined_domains when it cannot contact xend; that
- was causing virt-top to exit when xend was restarted.
+2012-03-06 Lai Jiangshan <laijs@cn.fujitsu.com>
-2007-09-18 Richard Jones <rjones@redhat.com>
+ virt-top: correct "virt-top -1" command via cpuacct cgroup
+ Old "virt-top -1" is not correct, its output is generated by guess:
+ use average usage for pinned physical CPUs.
- * virt-df/virt_df.ml: Handle domains with partition-backed
- block devices.
+ example(old "virt-top -1"):
- * virt-top/virt_top.ml: Don't fail on older libvirt which would
- give an error if list_domains or list_defined_domains was called
- with n = 0.
+ PHYCPU %CPU rhel6 Windows
+ 0 0.6 0.1= 0.5=
+ 1 0.6 0.1= 0.5=#
+ 2 0.6 0.1= 0.5=
+ 3 0.6 0.1=# 0.5=
- * ChangeLog: Start tracking changes.
+ The output almost makes no sense(all the value are just average, not real).
+ This is new implement, it use cpuacct cgroup to gain *real* physical usages
+ via cpuacct cgroup by virDomainGetCPUStats() API.
+
+ new result:
+
+ PHYCPU %CPU rhel6 Windows
+ 0 1.3 0.3 1.0
+ 1 2.3 0.3 2.0
+ 2 2.2 0.5 1.7
+ 3 2.5 0.4 2.1
+
+ PHYCPU %CPU rhel6 Windows
+ 0 1.7 0.4 1.3
+ 1 3.6 1.0 2.7
+ 2 1.6 0.4 1.2
+ 3 4.8 3.1 1.7
+
+ Note: average flag(=) is dropped, there is not average value in here.
+ Note: running flag(#) is dropped, because if the value is not empty,
+ it means the guest was once running in the physical CPU in this period
+ between updates.
+
+ Changed from V3:
+ use new virDomainGetCPUStats() libvirt-API.
+ add a new function find_usages_from_stats() to gain cpu usages.
+
+2011-08-15 Richard W.M. Jones <rjones@redhat.com>
+
+ Ensure manpages get built when using native compiler (thanks Laurent Léonard).
+
+2011-08-12 Richard W.M. Jones <rjones@redhat.com>
+
+ Version 1.0.6.
+
+ Update MANIFEST and fix 'make dist'.
+ make dist now copies the generated manual pages into the tarball, so
+ that we don't have to keep these generated files in git.
+
+ Fix ordering of csv_mode and stream_mode in tuple.
+ Note that this bug could have been avoided by using a struct instead
+ of a tuple here.
+
+ This updates commit daf79631f3eff904ba89297148c4125ac7dfee39.
+
+2011-08-11 Richard W.M. Jones <rjones@redhat.com>
+
+ docs: Explain how to debug libvirt initialization problems (RHBZ#680031).
+
+2011-04-02 Richard W.M. Jones <rjones@redhat.com>
+
+ Pull translations from Transifex.
+
+2011-03-08 Richard W.M. Jones <rjones@redhat.com>
+
+ Fix --end-time option with absolute times.
+ Assume the user enters a local time, and convert it to UTC.
+
+ This also adds some clarifying comments to the relative time
+ code (but doesn't change that code).
+
+ Note that the obsolete Calendar1 code is *not* updated by this change.
+
+ docs: Fix documentation for virt-top -c option.
+ Because of changes in libvirt, this no longer connects to Xen by
+ default. Instead the default is to connect to the libvirt default
+ hypervisor (which could be Xen, KVM or others). Update the --help
+ documentation and manual page accordingly.
+
+ Remove generated man pages from git.
+
+2011-03-07 KAMEZAWA Hiroyuki <kamezawa@bluextal.(none)>
+
+ add memory stats to --csv mode.
+ In csv mode, memory usage in kbytes and %mem against total system
+ memory is shown.
+
+2011-03-07 KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
+
+ Record memory statistics information to 'rd' object.
+ This helps to share the calculation (between modes) and
+ will help to show memory statistics in --csv mode
+
+ Signged-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
+
+2011-03-07 Richard W.M. Jones <rjones@redhat.com>
+
+ Import virt-top into transifex.
+ http://www.transifex.net/projects/p/virttop/
+
+2011-02-02 Richard W.M. Jones <rjones@redhat.com>
+
+ Update generated man page files.
+
+2011-02-02 KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
+
+ Add --block-in-bytes option.
+ This option and the 'B' key toggles block I/O stats between
+ requests and bytes.
+
+ Add --stream flag.
+ Output is sent to stdout (similar to the top -b option).
+
+2011-01-24 Richard W.M. Jones <rjones@redhat.com>
+
+ Obey virt-top --end-time down to near millisecond accuracy (RHBZ#637964).
+ Adjust the final delay down so that the --end-time option is obeyed
+ down to near millisecond accuracy.
+
+ The 'usleep' function has been renamed more accurately as 'millisleep'
+ since it sleeps for milliseconds not microseconds.
+
+2011-01-06 Richard W.M. Jones <rjones@redhat.com>
+
+ Version 1.0.5.
+
+ Add aclocal.m4 file to git.
+ This file is sort of generated, but at the moment only has
+ a single line which includes m4/ocaml.m4, that is vital for
+ the build.
+
+ Re-add virt-top/README file.
+ This seems to have been dropped during the hg->git transition.
+
+ Document background noise of RX packets from bridges (RHBZ#647987).
+ For more information see:
+ https://bugzilla.redhat.com/show_bug.cgi?id=647987#c9
+
+ man page: Remove excess blank line.
+
+ main: Record and print full exception stack traces.
+ This turns on stack trace recording unconditionally (the same
+ effect as if OCAMLRUNPARAM=b was always supplied), and also prints
+ stack traces to the log file if there is an exception.
+
+ See also:
+ http://caml.inria.fr/pub/docs/manual-ocaml/libref/Printexc.html
+
+ Note that this requires OCaml >= 3.11.0.
+
+ Fix pad function to work for negative widths (RHBZ#634435).
+ The pad function is supposed to pad a string to a width. Under some
+ circumstances, the width parameter could be negative resulting in a
+ call to String.sub such as:
+
+ String.sub str 0 (-1)
+
+ which caused an exception to be thrown and not caught, causing
+ virt-top to exit. Fix the pad function to return an empty string if
+ width <= 0 instead of throwing an exception.
+
+ See also:
+ https://bugzilla.redhat.com/show_bug.cgi?id=634435
+
+ Update dependencies.
+
+ Update generated man pages.
+
+ Fix virt-top --end-time option when TZ<>UTC (RHBZ#637964).
+ Return the end time as seconds since 1970, and don't adjust it
+ for timezone. In the main module the end time is compared against
+ the result of Unix.gettimeofday, which is also seconds since 1970,
+ so the comparison now works.
+
+ Tested with TZ=UTC-05, TZ=UTC and TZ=UTC+05.
+
+ NB: This has not been tested against ocaml-calendar 1.x, although
+ that version of calendar is obsolete and no one should be using it.
+
+ Update dependencies file.
+
+ man page: memory option shows total guest memory (RHBZ#647991).
+
+ man page: Update copyright date and link to web pages.
+
+ Remove references to xm/xentop manual pages (RHBZ#648186).
+ These Xen tools are obsolete since virsh/virt-top works across
+ all hypervisors.
+
+2010-09-02 Richard Jones <rjones@redhat.com>
+
+ Change order of return values from getyx - fixes display of historical CPU.
+ In this commit:
+ http://cvs.savannah.gnu.org/viewvc/ocaml-tmk/functions.c?root=ocaml-tmk&r1=1.2&r2=1.3
+ Sam changed (fixed) the order that the coordinates are returned
+ from the ncurses 'getyx' call. Since this commit was made 3 years
+ ago it should be safe to assume everyone is using the new ocaml-ncurses
+ by now.
+
+ This meant the historical %CPU was being displayed off the bottom of
+ the screen instead of on the top row at the right. Now it is displayed
+ in the correct place.
+
+2010-06-15 Kiyoto Hashida <khashida@redhat.com>
+
+ Update Japanese translations (RHBZ#493799).
+
+2010-03-05 Yulia <ypoyarko@redhat.com>
+
+ Updated Russian translations (RHBZ#493799).
+
+2010-03-05 Richard Jones <rjones@redhat.com>
+
+ Newer autoconf adds PACKAGE_URL definition.
+
+2010-02-12 Richard Jones <rjones@redhat.com>
+
+ zh_CN: Remove bogus Plural-Forms line.
+
+ Japanese: update bogus 'Plural-Forms: Plural-Forms'.
+
+2010-02-12 Hajime Taira <htaira@redhat.com>
+
+ Update Japanese translations (RHBZ#493799).
+
+2010-01-28 Richard Jones <rjones@redhat.com>
+
+ Rename te_IN.po -> te.po (Piotr Drąg).
+
+ Telugu translation added by Krishna Babu K (RHBZ#493799c32)
+
+ Polish translation updated by Piotr Drag (RHBZ#493799c31)
+
+ Portuguese translation added by Glaucia Cintra (RHBZ#493799c30)
+
+ Chinese translation updated by Wei Liu (RHBZ#493799c29)
+
+ Oriya translation added by Manoj Kumar Giri (RHBZ#493799c28)
+
+ Spanish translation updated by Daniel Cabrera (RHBZ#493799c27)
+
+2009-12-30 Richard Jones <rjones@redhat.com>
+
+ Update to ocaml-autoconf macros 1.1
+
+ Fix detection of ocaml-findlib.
+
+2009-10-06 Richard Jones <rjones@trick.home.annexia.org>
+
+ HACKING file, update (C) messages.
+
+ Update PO files.
+
+ Remove extra spaces after s_ and f_ operators.
+
+ Safer curses functions.
+ Add safer curses functions which don't require use of dangerous
+ ignore() function.
+
+ Replace failwith (sprintf...) with failwithf
+
+ Remove -w s and fix so it compiles without warnings.
+
+ Remove NSIS installer.
+
+ Change over to using ocaml.m4 macros.
+
+2009-10-05 Richard Jones <rjones@trick.home.annexia.org>
+
+ Updated MANIFEST.
+
+ TODO.virt-top renamed as TODO
+
+ Typo fixed.
+
+ Update manpage with --version option.
+
+ Add virt-top --version command line option.
+
+ Add the real virt-top version to help display.
+
+ Version 1.0.4 for release.
+
+ Fix changelog and manifest files.
+
+ gitlog-to-changelog: If the date field is empty, don't barf.
+
+ Added Jim Meyering's gitlog-to-changelog program.
+
+ Add total hardware memory field (RHBZ#521785)
+
+ Don't need to explicitly link against ncurses.
+
+ Update PO files.
+
+ Add Japanese to list of translations.
+
+ Use msgfmt to generate *.mo files correctly.
+
+ Update documentation.
+
+ Update dependencies.
+
+ Rename .hgignore -> .gitignore, update dependencies.