Contents: priorities for 1.2 release Free form list of items to do: Documentation What's left to do (tasks) Stuff that got deferred Full table Data

priorities for 1.2 release [edit section]

Free form list of items to do: [edit section]

• Features:
  • add checkbox and replot to flot plots? (maybe not)
• Bugs:
  • remove or convert all chart_configs to proper files
  • resolve items from release testing:
    • See release-1.2-test-2017-09-27 Notes
  • parse LTP posix test results
• Documentation:
  • document parser:parse_log and parser:common
  • review wiki and fix up all modified functions
• General:
  • review items on CodeFixthisList
  • follow release check list on Release Process
  • make a list of release features

Documentation [edit section]

Here are things needing documentation:

• parser:common
• (done) per_job_build variable in spec files
• (done) run.json - current page is for 1.1 format
• Fuego Object Details
  • update the table
• (done) Fuego charting
  • files:
    • flat_plot_data.txt
    • results.json
    • flot_chart_data.txt
  • chart_config
• (done) dynamic variables - see Test variables
• (done) Dependencies - in fuego_test.sh page?
  • NEED_MEMORY
  • NEED_FREE_STORAGE
  • NEED_KCONFIG
  • NEED_ROOT
• (done) transport connection improvements
  • (done) ov_transport_connect
  • (done) ov_transport_disconnect
  • TARGET_TEARDOWN_LINK - (deprecate)
  • TARGET_SETUP_LINK - (deprecate)
• (done) document new APIS
  • (done) function_is_on_sdk
  • (done) function_processing
  • (done) function_test_pre_check
  • (done) function_test_fetch_results
  • (done) function_cleanup
• (done) document changed APIs
  • (done) function_post_test

What's left to do (tasks) [edit section]

item  ^

priority  ^

status  ^

notes  ^

Edit the data below.

Stuff that got deferred [edit section]

item  ^	priority  ^	status  ^	notes  ^
Benchmark job created without a testplan has no plot link	medium	defer	If you create a Benchmark job (say Benchmark.fuego_check_plots) using ftc add-jobs -b docker -t Benchmark.fuego_check_plots -s default, then build page for this job will have no link to plot.png
support phase reordering	low	defer	May be useful for LAVA integration (don't want to run pre_test before build)
support LAVA transport	high	defer	Required for AGL/JTA integration
support AGL/JTA reports	high	defer	Song Cai is interested in this work
web file transport	low	defer	could be useful for some interesting target cases
support separate build phase	medium	defer	don't want to hold board reserved during build, for lava integration and handling timeouts better
handle set -e differently	medium	defer	handle script errors explicitly, instead of relying on shell
handle timeouts automatically	medium	defer	should be in test and board, not in plan - but depends on spec, transport, build time etc
automatically create a view for each board	medium	defer	use a command line option on fuego add-nodes (--make_view)

Edit the data below.

Full table [edit section]

item  ^	priority  ^	status  ^	notes  ^
Benchmark job created without a testplan has no plot link	medium	defer	If you create a Benchmark job (say Benchmark.fuego_check_plots) using ftc add-jobs -b docker -t Benchmark.fuego_check_plots -s default, then build page for this job will have no link to plot.png
Update wiki documentation for this release	high	closed	In particular, update information about results files, main.sh, spec file format, testplan defaults
support phase reordering	low	defer	May be useful for LAVA integration (don't want to run pre_test before build)
support LAVA transport	high	defer	Required for AGL/JTA integration
support AGL/JTA reports	high	defer	Song Cai is interested in this work
ov_transport_connect/ov_transport_disconnect	high	closed	Sounds like Jan-Simon will do it
get rid of nbench-bytes	medium	closed	This is a dumb duplicate of nbench_bytes (see Fuego naming rules)
modify ftc run-test for base script calling convention	high	closed	call main.sh instead of base script, and set TESTDIR and TESTNAME
web file transport	low	defer	could be useful for some interesting target cases
support unified output format	high	closed	create results.json files for all tests (including functional tests) see Unified_Results_Format_Project
support separate build phase	medium	defer	don't want to hold board reserved during build, for lava integration and handling timeouts better
handle set -e differently	medium	defer	handle script errors explicitly, instead of relying on shell
fix LTP	high	closed	make sure LTP works for docker and bbb
Issue_0048	high	closed	problem with LTP and libaio-dev installation (one toolchain overwrites another toolchains libaio.so symlink (see Issue_0048)
handle timeouts automatically	medium	defer	should be in test and board, not in plan - but depends on spec, transport, build time etc
automatically create a view for each board	medium	defer	use a command line option on fuego add-nodes (--make_view)
fix ftc device dictionary	medium	closed	used for board check and pre-populating options

Data [edit section]

Edit the data here:

item  ^	priority  ^	status  ^	notes  ^
Benchmark job created without a testplan has no plot link	medium	defer	If you create a Benchmark job (say Benchmark.fuego_check_plots) using ftc add-jobs -b docker -t Benchmark.fuego_check_plots -s default, then build page for this job will have no link to plot.png
Update wiki documentation for this release	high	closed	In particular, update information about results files, main.sh, spec file format, testplan defaults
support phase reordering	low	defer	May be useful for LAVA integration (don't want to run pre_test before build)
support LAVA transport	high	defer	Required for AGL/JTA integration
support AGL/JTA reports	high	defer	Song Cai is interested in this work
ov_transport_connect/ov_transport_disconnect	high	closed	Sounds like Jan-Simon will do it
get rid of nbench-bytes	medium	closed	This is a dumb duplicate of nbench_bytes (see Fuego naming rules)
modify ftc run-test for base script calling convention	high	closed	call main.sh instead of base script, and set TESTDIR and TESTNAME
web file transport	low	defer	could be useful for some interesting target cases
support unified output format	high	closed	create results.json files for all tests (including functional tests) see Unified_Results_Format_Project
support separate build phase	medium	defer	don't want to hold board reserved during build, for lava integration and handling timeouts better
handle set -e differently	medium	defer	handle script errors explicitly, instead of relying on shell
fix LTP	high	closed	make sure LTP works for docker and bbb
Issue_0048	high	closed	problem with LTP and libaio-dev installation (one toolchain overwrites another toolchains libaio.so symlink (see Issue_0048)
handle timeouts automatically	medium	defer	should be in test and board, not in plan - but depends on spec, transport, build time etc
automatically create a view for each board	medium	defer	use a command line option on fuego add-nodes (--make_view)
fix ftc device dictionary	medium	closed	used for board check and pre-populating options

Query