xref: /aosp_15_r20/external/ltp/doc/old/Maintainer-Patch-Review-Checklist.asciidoc (revision 49cdfc7efb34551c7342be41a7384b9c40d7cab7)
1# Patch Review
2
3Anyone can and should review patches. It's the only way to get good at
4patch review and for the project to scale.
5
6## Goals of patch review
7
81. Prevent false positive test results
92. Prevent false negative test results
103. Keep the code as simple as possible, but no simpler
11
12## How to find clear errors
13
14A clear error is one where there is unlikely to be any argument if you
15provide evidence of it. Evidence being an error trace or logical proof
16that an error will occur in a common situation.
17
18The following are examples and may not be appropriate for all tests.
19
20* Merge the patch locally. It should apply cleanly to master.
21* Compile the patch with default and non-default configurations.
22  - Use sanitizers e.g. undefined behaviour, address.
23  - Compile on non-x86
24  - Compile on x86 with -m32
25  - Compile testing patches with GitHub Actions in LTP repo fork can cover
26    various distros/architectures
27* Use `make check`
28* Run effected tests in a VM
29  - Use single vCPU
30  - Use many vCPUs and enable NUMA
31  - Restrict RAM to < 1GB.
32* Run effected tests on an embedded device
33* Run effected tests on non-x86 machine in general
34* Run reproducers on a kernel where the bug is present
35* Run tests with "-i0" and "-i2"
36* Compare usage of system calls with man page descriptions
37* Compare usage of system calls with kernel code
38* Double check commit message
39* Search the LTP library for existing helper functions
40* Check doc formatting, i.e. make doc && chromium docparse/metadata.html
41
42## How to find subtle errors
43
44A subtle error is one where you can expect some argument because you
45do not have clear evidence of an error. It is best to state these as
46questions and not make assertions if possible.
47
48Although if it is a matter of style or "taste" then senior maintainers
49can assert what is correct to avoid bike shedding.
50
51* Ask what happens if there is an error, could it be debugged just
52  with the test output?
53* Are we testing undefined behavior?
54  - Could future kernel behaviour change without "breaking userland"?
55  - Does the kernel behave differently depending on hardware?
56  - Does it behave differently depending on kernel configuration?
57  - Does it behave differently depending on the compiler?
58  - Would it behave differently if the order of checks on syscall parameters
59    changed in the kernel?
60* Will it scale to tiny and huge systems?
61  - What happens if there are 100+ CPUs?
62  - What happens if each CPU core is very slow?
63  - What happens if there are 2TB of RAM?
64* Are we repeating a pattern that can be turned into a library function?
65* Is a single test trying to do too much?
66* Could multiple similar tests be merged?
67* Race conditions
68  - What happens if a process gets preempted?
69  - Could checkpoints or fuzzsync by used instead?
70  - Note, usually you can insert a sleep to prove a race condition
71    exists however finding them is hard
72* Is there a simpler way to achieve the same kernel coverage?
73
74## How to get patches merged
75
76Once you think a patch is good enough you should add your Reviewed-by
77and/or Tested-by tags. This means you will get some credit for getting
78the patch merged. Also some blame if there are problems.
79
80If you ran the test you can add the Tested-by tag. If you read the
81code or used static analysis tools on it, you can add the Reviewed-by
82tag.
83
84In addition you can expect others to review your patches and add their
85tags. This will speed up the process of getting your patches merged.
86
87## Maintainers Checklist
88
89Patchset should be tested locally and ideally also in maintainer's fork in
90GitHub Actions on GitHub.
91
92NOTE: GitHub Actions do only build testing, passing the CI means only that
93      the test compiles fine on variety of different distributions and releases.
94
95The test should be executed at least once locally and should PASS as well.
96
97Commit messages should have
98
99* Author's `Signed-off-by` tag
100* Committer's `Reviewed-by` or `Signed-off-by` tag
101* Check also mailing lists for other reviewers / testers tags, notes and failure reports
102* `Fixes: hash` if it fixes particular LTP commit
103* `Fixes: #N` if it fixes github issue number N, so it's automatically closed
104* LTP WIKI (git: https://github.com/linux-test-project/ltp.wiki.git) should be kept up to date.
105
106After patch is accepted or rejected, set correct state and archive in
107https://patchwork.ozlabs.org/project/ltp/list/[LTP patchwork instance].
108
109## New tests
110New test should
111
112* Have a record in runtest file
113* Test should work fine with more than one iteration
114  (e.g. run with `-i 100`)
115* Run with `-i 0` to check that setup and cleanup are coded properly (no test is being run)
116* Have a brief description
117* License: the default license for new tests is GPL v2 or later, use
118  GPL-2.0-or-later; the licence for test (e.g. GPL-2.0) should not change
119  unless test is completely rewritten
120* Old copyrights should be kept unless test is completely rewritten
121
122### C tests
123* Use new https://github.com/linux-test-project/ltp/wiki/Test-Writing-Guidelines#22-writing-a-test-in-c[C API]
124* Test binaries are added into corresponding `.gitignore` files
125* Check coding style with `make check`
126  (more in https://github.com/linux-test-project/ltp/wiki/Test-Writing-Guidelines#21-c-coding-style[C coding style])
127* Docparse documentation
128* If a test is a regression test it should include tags
129  (more in https://github.com/linux-test-project/ltp/wiki/Test-Writing-Guidelines#2238-test-tags[Test tags])
130
131### Shell tests
132* Use new https://github.com/linux-test-project/ltp/wiki/Test-Writing-Guidelines#23-writing-a-testcase-in-shell[shell API]
133* Check coding style with `make check`
134  (more in https://github.com/linux-test-project/ltp/wiki/Test-Writing-Guidelines#132-shell-coding-style[Shell coding style])
135* If a test is a regression test it should include related kernel or glibc commits as a comment
136
137## LTP library
138For patchset touching library please check also
139https://github.com/linux-test-project/ltp/wiki/LTP-Library-API-Writing-Guidelines[LTP Library API Writing Guidelines].
140