[Fuego] [PATCH] Docs: Convert tbwiki pages into .rst files
Pooja
pooja.sm at pathpartnertech.com
Thu Dec 3 05:18:24 UTC 2020
From: Pooja More <pooja.sm at pathpartnertech.com>
Following are the converted files:
Benchmark.GLMark.rst
Benchmark.bonnie.rst
Benchmark.template.rst
Functional.LTP.rst
Signed-off-by: Pooja More <pooja.sm at pathpartnertech.com>
---
docs/rst_src/Benchmark.GLMark.rst | 69 +++++++++++++
docs/rst_src/Benchmark.bonnie.rst | 89 +++++++++++++++++
docs/rst_src/Benchmark.template.rst | 38 +++++++
docs/rst_src/Functional.LTP.rst | 191 ++++++++++++++++++++++++++++++++++++
4 files changed, 387 insertions(+)
create mode 100644 docs/rst_src/Benchmark.GLMark.rst
create mode 100644 docs/rst_src/Benchmark.bonnie.rst
create mode 100644 docs/rst_src/Benchmark.template.rst
create mode 100644 docs/rst_src/Functional.LTP.rst
diff --git a/docs/rst_src/Benchmark.GLMark.rst b/docs/rst_src/Benchmark.GLMark.rst
new file mode 100644
index 0000000..ac1aee5
--- /dev/null
+++ b/docs/rst_src/Benchmark.GLMark.rst
@@ -0,0 +1,69 @@
+###################
+Benchmark.GLMark
+###################
+
+===============
+Description
+===============
+
+GLMark is a test to determine the performance of OpenGL on your
+system. The program runs a series of tests rendering 2D and 3D
+graphics and animations, and provides a benchmark result indicating
+the number of frames per second for each operation. The it averages
+the different results to generate a score for the GPU.
+
+=============
+Resources
+=============
+
+
+ * `Benchmark graphics card (GPU) performance on Linux with glmark <http://www.binarytides.com/glmark-linux-gpu-performance/>`_,
+ by Silver Moon, Binary Tides, April 2014
+
+ * Newer version: glmark2 - see
+ `<https://github.com/glmark2/glmark2>`_
+ * `Simple DirectMedia Layer wikipedia entry <https://en.wikipedia.org/wiki/Simple_DirectMedia_Layer>`_
+
+===========
+Results
+===========
+
+========
+Tags
+========
+
+ * Graphics
+
+================
+Dependencies
+================
+
+
+ * It requires the following development libraries:
+
+ * sdl: libsdl1.2-dev
+ * gl: libgl1-mesa-dev
+ * glu: libglu1-mesa-dev
+ * glew: libglew1.5-dev
+
+==========
+Status
+==========
+
+
+ * CAN'T BUILD
+
+=========
+Notes
+=========
+
+Compiling for minnowboard inside Docker container, I get lots of
+errors saying that structure members are private. Possibly this is
+due to more stringent error checking by gcc, or some cflag that is
+missing.
+
+All errors are reported from vector.h, included in various cpp files.
+
+On other platforms, the required libraries (SDL) are often not
+available at build or run time.
+
diff --git a/docs/rst_src/Benchmark.bonnie.rst b/docs/rst_src/Benchmark.bonnie.rst
new file mode 100644
index 0000000..2f79a0a
--- /dev/null
+++ b/docs/rst_src/Benchmark.bonnie.rst
@@ -0,0 +1,89 @@
+###################
+Benchmark.bonnie
+###################
+
+===============
+Description
+===============
+
+Bonnie++ is a filesystem benchmark that measures basic speed of
+several operations, including data read and write speed, the number of
+seeks per second, and the number of file metadata operations per
+second.
+
+=============
+Resources
+=============
+
+ * `Using Bonnie++ for filesystem performance benchmarking <https://www.linux.com/news/using-bonnie-filesystem-performance-benchmarking>`_,
+ by Ben Martin, Linux.com, July 2008 (accessed Oct. 2017)
+
+ * `Bonnie++ wikipedia entry <https://en.wikipedia.org/wiki/Bonnie%2B%2B>`_
+ * `Bonnie++ home page <https://www.coker.com.au/bonnie++/>`_
+
+===========
+Results
+===========
+
+=========
+Tags
+=========
+
+ * filesystem
+
+===============
+Dependencies
+===============
+
+Bonnie has no build dependencies.
+
+There is a separate command line option for running it as the 'root'
+user.
+
+Bonnie uses the following test variables at runtime:
+
+ * ``BENCHMARK_BONNIE_MOUNT_BLOCKDEV`` - name of block device where
+ filesystem to be tested in located, or "ROOT"
+
+ * ``BENCHMARK_BONNIE_MOUNT_POINT`` - directory name where the filesystem
+ should be mounted (if needed), and the tests run.
+
+ * ``BENCHMARK_BONNIE_SIZE`` - specifies the size, in megabytes, of the
+ files used for IO performance measurements
+
+ * This is the parameter to the -s command line option
+
+ * ``BENCHMARK_BONNIE_RAM`` - specifies the size of the board's RAM in
+ megabytes, or "0" if ram size sanity checks should be disabled
+
+ * This is the parameter to the -r command line option
+
+ * ``BENCHMARK_BONNIE_NUM_FILES`` - is a colon-separate 4-tuple indicating
+ the number of files, the file max size, the file min size, and the
+ number of directories to spread the files into, for metadata tests
+
+ * The default value, if not specified is: "16:0:0:1". This results in 16K
+ files with maximum and minimum size 0, in 1 directory.
+
+ * ``BENCHMARK_BONNIE_ROOT`` - should be set to "true" if the bonnie should
+ try to execute as the root user on the board.
+
+==========
+Status
+==========
+
+ * OK
+
+=========
+Notes
+=========
+
+.. note::
+
+ If a test executes too quickly, bonnie does not report the
+ result, and instead produces '+++++'s in the entries for those tests.
+ Specifically, bonnie will emit this if a test result was less than .5
+ (MinTime in the source code). If this happens for you, consider using
+ or writing a spec that increases the size of the files, or the number
+ of files used for tests.
+
diff --git a/docs/rst_src/Benchmark.template.rst b/docs/rst_src/Benchmark.template.rst
new file mode 100644
index 0000000..9ccb0b7
--- /dev/null
+++ b/docs/rst_src/Benchmark.template.rst
@@ -0,0 +1,38 @@
+#####################
+Benchmark.template
+#####################
+
+===============
+Description
+===============
+
+==============
+Resources
+==============
+
+ * First external resource (web site, tutorial, article, etc)
+
+===========
+Results
+===========
+
+=========
+Tags
+=========
+
+ * First tag
+
+================
+Dependencies
+================
+
+==========
+Status
+==========
+
+ * OK
+
+=========
+Notes
+=========
+
diff --git a/docs/rst_src/Functional.LTP.rst b/docs/rst_src/Functional.LTP.rst
new file mode 100644
index 0000000..95c86fc
--- /dev/null
+++ b/docs/rst_src/Functional.LTP.rst
@@ -0,0 +1,191 @@
+###################
+Functional.LTP
+###################
+
+===============
+Description
+===============
+
+LTP is the Linux Test Project test, which is actually a "super-test"
+which contains literally thousands of sub-tests in all kinds of
+categories.
+
+=============
+Resources
+=============
+
+ * `Project source on github <https://github.com/linux-test-project/>`_
+ * `LTP page on Wikipedia <https://en.wikipedia.org/wiki/Linux_Test_Project>`_
+ * `Kernel test automation with LTP <https://lwn.net/Articles/625969/>`_ -
+ by Cyril Hrubis LWN.net, December 2014
+
+
+============
+Results
+============
+
+LTP has thousands of sub-tests, which themselves can be grouped into
+collections.
+
+Tests results are organized into testsets, that correspond to these
+collections. A Fuego spec can be used to select what collections to
+execute, and also what sub-test cases to skip.
+
+========
+Tags
+========
+
+ * kernel, posix, realtime
+
+================
+Dependencies
+================
+
+This test has no dependencies for either the build, or the running
+system, in the current configuration used by Fuego.
+
+===========
+Status
+===========
+
+ * Can be run, but has lot of bugs
+ * see :ref:`Functional.LTP_issues`
+
+=========
+Notes
+=========
+
+For notes about LTP internals, see :ref:`LTP_Notes`
+
+A number of fuego-specific features are available for dealing with
+LTP. Since it is quite large, the build and deployment of LTP can
+take a substantial amount of time. Also, LTP is often already
+pre-installed on the target board (along with the distribution), so
+building LTP is not required.
+
+Special Fuego variables and specs are available to deal with different
+LTP installation and usage scenarios.
+
+Variables
+============
+
+ * ``FUNCTIONAL_LTP_TESTS`` (spec variable 'tests') - list the LTP test
+ collections to execute
+
+ * Usually specified in the spec
+
+ * ``FUNCTIONAL_LTP_SKIPLIST`` (spec variable 'skiplist') - list the
+ individual LTP tests to not execute
+
+ * May be specified in a spec or in the board file
+
+ * ``FUNCTIONAL_LTP_BOARD_SKIPLIST`` - list individual LTP tests to
+ not execute
+
+ * Should only be specified in the board file
+
+ * Is used to avoid overwriting the spec 'skiplist'
+
+ * ``FUNCTIONAL_LTP_HOMEDIR`` - specifies a non-default, persistent, location
+ where LTP resides on the board
+
+ * This is placed in the board file for a board that has LTP pre-installed
+
+
+Specs
+============
+
+ * Special Fuego LTP specs:
+
+ * **install**
+
+ * Use this to deploy LTP, but not run any tests
+
+ * **make_pkg**
+
+ * Use this to create a tarfile which can be manually installed
+ on the target
+
+Special build and deployment scenarios
+=============================================
+
+The ``Functional.LTP`` test in Fuego supports 2 major different usage
+scenarios, with some helper specs to accomplish setting up for them.
+
+Normally, Fuego installs all test materials needed for a test,
+executes the test, then removes all the test materials. However, in
+the case of LTP, the materials are so big that putting them on the
+target for every test may be too much overhead.
+
+Also, sometime LTP is already installed on a board, and there's no
+need for Fuego to build and deploy LTP to the board in the first
+place.
+
+Here are the 2 main scenarios:
+
+ 1 - Build LTP, deploy it, and run the requested tests
+
+ * This is the default Fuego scenario, and if no other variables
+ are used, this is what will occur
+
+ 2 - Check for LTP already on target, and run the requested tests
+
+ * If the variable FUNCTIONAL_LTP_HOMEDIR is set in the board file, then
+ ``Functional.LTP`` will skip deploying LTP to the target
+ * Fuego will still install a few custom scripts and files for each test
+ * LTP tests will run relatively quickly (because build and deploy are omitted)
+
+ * However, LTP test data will take up a lot of space on the target
+ (on my system, about 500M)
+
+
+There are special test specs for setting up LTP persistently on the target:
+
+ * spec: **install** - builds and deploys LTP to the test directory
+
+ * You should set ``FUNCTIONAL_LTP_HOMEDIR`` in the board file. This is the
+ persistent directory where LTP will be installed
+
+ * spec: **make_pkg** - builds LTP, and then creates a tar file that can
+ be manually installed on the target
+
+ * The tar file is called ltp.tar.gz, and is left in the log directory
+ for the test
+ * A link should be provided (called 'tar.gz') to download this tar file
+ from the Jenkins server, on the job page for this.
+
+Steps
+--------------
+
+Here are exact steps to follow, to build Fuego's LTP, and install it on target
+
+ * decide where to install LTP (recommendation is /opt/ltp)
+
+ * set the ``FUNCTIONAL_LTP_HOMEDIR`` variable in your board file
+
+ * It should have quotes (like other board variables) and look something like this:
+
+::
+
+ FUNCTIONAL_LTP_HOMEDIR="/opt/ltp"
+
+
+ * Make directory ``/opt/ltp`` on the target board
+ * Create a job for the board, with spec "install"
+
+ * 'ftc add-job -b myboard -t Functional.LTP -s install'
+
+ * Execute the job in Jenkins or from ftc
+
+ * 'ftc build-job myboard.install.Functional.LTP'
+
+ * Verify on the target that the materials are present
+
+ * Login, and check that ``/opt/ltp`` is populated
+
+ * Create and execute other LTP jobs (with different specs)
+
+ * e.g. 'ftc add-job -b myboard -t Functional.LTP -s quickhit'
+ * e.g. 'ftc add-job -b myboard -t Functional.LTP -s selectionwithrt'
+ * Build them using Jenkins or ftc
+
--
2.7.4
--
This
message contains confidential information and is intended only
for the
individual(s) named. If you are not the intended
recipient, you are
notified that disclosing, copying, distributing or taking any
action in
reliance on the contents of this mail and attached file/s is strictly
prohibited. Please notify the
sender immediately and delete this e-mail
from your system. E-mail transmission
cannot be guaranteed to be secured or
error-free as information could be
intercepted, corrupted, lost, destroyed,
arrive late or incomplete, or contain
viruses. The sender therefore does
not accept liability for any errors or
omissions in the contents of this
message, which arise as a result of e-mail
transmission.
More information about the Fuego
mailing list