xref: /linux/tools/memory-model/scripts/README (revision 200323768787a0ee02e01c35c1aff13dc9d77dde)
1			============
2			LKMM SCRIPTS
3			============
4
5
6These scripts are run from the tools/memory-model directory.
7
8checkalllitmus.sh
9
10	Run all litmus tests in the litmus-tests directory, checking
11	the results against the expected results recorded in the
12	"Result:" comment lines.
13
14checkghlitmus.sh
15
16	Run all litmus tests in the https://github.com/paulmckrcu/litmus
17	archive that are C-language and that have "Result:" comment lines
18	documenting expected results, comparing the actual results to
19	those expected.
20
21checklitmushist.sh
22
23	Run all litmus tests having .litmus.out files from previous
24	initlitmushist.sh or newlitmushist.sh runs, comparing the
25	herd7 output to that of the original runs.
26
27checklitmus.sh
28
29	Check a single litmus test against its "Result:" expected result.
30	Not intended to for manual use.
31
32checktheselitmus.sh
33
34	Check the specified list of litmus tests against their "Result:"
35	expected results.  This takes optional parseargs.sh arguments,
36	followed by "--" followed by pathnames starting from the current
37	directory.
38
39cmplitmushist.sh
40
41	Compare output from two different runs of the same litmus tests,
42	with the absolute pathnames of the tests to run provided one
43	name per line on standard input.  Not normally run manually,
44	provided instead for use by other scripts.
45
46initlitmushist.sh
47
48	Run all litmus tests having no more than the specified number
49	of processes given a specified timeout, recording the results
50	in .litmus.out files.
51
52judgelitmus.sh
53
54	Given a .litmus file and its herd7 output, check the output file
55	against the .litmus file's "Result:" comment to judge whether
56	the test ran correctly.  Not normally run manually, provided
57	instead for use by other scripts.
58
59newlitmushist.sh
60
61	For all new or updated litmus tests having no more than the
62	specified number of processes given a specified timeout, run
63	and record the results in .litmus.out files.
64
65parseargs.sh
66
67	Parse command-line arguments.  Not normally run manually,
68	provided instead for use by other scripts.
69
70runlitmushist.sh
71
72	Run the litmus tests whose absolute pathnames are provided one
73	name per line on standard input.  Not normally run manually,
74	provided instead for use by other scripts.
75
76README
77
78	This file
79
80Testing a change to LKMM might go as follows:
81
82	# Populate expected results without that change, and
83	# runs for about an hour on an 8-CPU x86 system:
84	scripts/initlitmushist.sh --timeout 10m --procs 10
85	# Incorporate the change:
86	git am -s -3 /path/to/patch # Or whatever it takes.
87
88	# Test the new version of LKMM as follows...
89
90	# Runs in seconds, good smoke test:
91	scripts/checkalllitmus.sh
92
93	# Compares results to those produced by initlitmushist.sh,
94	# and runs for about an hour on an 8-CPU x86 system:
95	scripts/checklitmushist.sh --timeout 10m --procs 10
96
97	# Checks results against Result tags, runs in minutes:
98	scripts/checkghlitmus.sh --timeout 10m --procs 10
99
100The checkghlitmus.sh should not report errors in cases where the
101checklitmushist.sh script did not also report a change.  However,
102this check is nevertheless valuable because it can find errors in the
103original version of LKMM.  Note however, that given the above procedure,
104an error in the original LKMM version that is fixed by the patch will
105be reported both as a mismatch by checklitmushist.sh and as an error
106by checkghlitmus.sh.  One exception to this rule of thumb is when the
107test fails completely on the original version of LKMM and passes on the
108new version.  In this case, checklitmushist.sh will report a mismatch
109and checkghlitmus.sh will report success.  This happens when the change
110to LKMM introduces a new primitive for which litmus tests already existed.
111