test-lib: make '--stress' more bisect-friendly
PATCH 1/1 test-lib: make '--stress' more bisect-friendly
Let's suppose that a test somehow becomes flaky between 'master' and 'pu', and tends to fail within the first 50 repetitions when run with '--stress'. In such a case we could use 'git bisect' to find the
culprit: if the test script fails with '--stress', then the commit is definitely bad, but if it survives, say, 300 repetitions, then we could consider it good with reasonable confidence.
Unfortunately, all this could only be done manually, because '--stress' would run the test script repeatedly for all eternity on a good commit, and it would exit with success even when it found a failure on a bad commit.
So let's make '--stress' usable with 'git bisect run':
- Make it exit with failure if a failure is found.
- Add the '--stress-limit=<N>' option to repeat the test script at most N times in each of the parallel jobs, and exit with success when the limit is reached.
And then we could simply run something like:
$ git bisect start origin/pu master
$ git bisect run sh -c 'make && cd t &&
./t1234-foo.sh --stress --stress-limit=300'Sure, as a brand new feature it won't be any useful right now, but in a release or three most cooking topics will already contain this, so we could automatically bisect at least newly introduced flakiness.
---
This is a case when an external stress script works better, as it can easily check commits in the past... if someone has such a script, that is.
Anyway, the approach works:
https://public-inbox.org/git/[email protected]/
https://public-inbox.org/git/[email protected]/
t/README | 5 +++++ t/test-lib.sh | 18 ++++++++++++++++-- 2 files changed, 21 insertions(+), 2 deletions(-)
t/README +5 -0
@@ -202,6 +202,11 @@ appropriately before running "make". '.stress-<nr>' suffix, and the trash directory of the failed test job is renamed to end with a '.stress-failed' suffix. +--stress-limit=<N>:: + When combined with --stress run the test script repeatedly + this many times in each of the parallel jobs or until one of + them fails, whichever comes first. + You can also set the GIT_TEST_INSTALLED environment variable to the bindir of an existing git installation to test that installation. You still need to have built this git sandbox, from which various
t/test-lib.sh +16 -2
@@ -152,6 +152,17 @@ do ;; esac ;; + --stress-limit=*) + stress_limit=${opt#--*=} + case "$stress_limit" in + *[^0-9]*|0*|"") + echo "error: --stress-limit=<N> requires the number of repetitions" >&2 + exit 1 + ;; + *) # Good. + ;; + esac + ;; *) echo "error: unknown test option '$opt'" >&2; exit 1 ;; esac
@@ -237,8 +248,10 @@ then exit 1 ' TERM INT - cnt=0 - while ! test -e "$stressfail" + cnt=1 + while ! test -e "$stressfail" && + { test -z "$stress_limit" || + test $cnt -le $stress_limit ; } do $TEST_SHELL_PATH "$0" "$@" >"$TEST_RESULTS_BASE.stress-$job_nr.out" 2>&1 & test_pid=$!
@@ -261,6 +274,7 @@ then if test -f "$stressfail" then + stress_exit=1 echo "Log(s) of failed test run(s):" for failed_job_nr in $(sort -n "$stressfail") do
Signature · 1 line
2.20.1.940.g8404bb2d1a
Quoted 10 lines · SZEDER Gábor
- Make it exit with failure if a failure is found.
- Add the '--stress-limit=<N>' option to repeat the test script at most N times in each of the parallel jobs, and exit with success when the limit is reached.
[...]This is a case when an external stress script works better, as it can easily check commits in the past... if someone has such a script, that is.
Heh, I literally just implemented this kind of max-count in my own "stress" script[1] to handle this recent t0025 testing. So certainly I think it is a good idea.
Picking an <N> is tough. Too low and you get a false negative, too high and you can wait forever, especially if the script is long. But I don't think there's any real way to auto-scale it, except by seeing a few of the failing cases and watching how long they take.
Quoted 3 lines
t/README | 5 +++++ t/test-lib.sh | 18 ++++++++++++++++-- 2 files changed, 21 insertions(+), 2 deletions(-)
Patch looks good. A few observations:
t/test-lib.sh then · line 257
Quoted 13 lines
t/test-lib.sh +4
-2@@ -237,8 +248,10 @@ then exit 1 ' TERM INT - cnt=0 - while ! test -e "$stressfail" + cnt=1 + while ! test -e "$stressfail" && + { test -z "$stress_limit" || + test $cnt -le $stress_limit ; } do $TEST_SHELL_PATH "$0" "$@" >"$TEST_RESULTS_BASE.stress-$job_nr.out" 2>&1 & test_pid=$!
You switch to 1-indexing the counts here. I think that makes sense, since otherwise --stress-limit=300 would end at "1.299", etc.
t/test-lib.sh then · line 280
Quoted 8 lines
t/test-lib.sh +1
-0@@ -261,6 +274,7 @@ then if test -f "$stressfail" then + stress_exit=1 echo "Log(s) of failed test run(s):" for failed_job_nr in $(sort -n "$stressfail") do
I think I'd argue that this missing stress_exit is a bug in the original script, and somewhat orthogonal to the limit counter. But I don't think it's worth the trouble to split it out (and certainly the theme of "now you can run this via bisect" unifies the two changes).
-Peff
Quoted 7 lines · Jeff King
Quoted 3 lines
This is a case when an external stress script works better, as it can easily check commits in the past... if someone has such a script, that is.
Heh, I literally just implemented this kind of max-count in my own "stress" script[1] to handle this recent t0025 testing. So certainly I think it is a good idea.
As usual, I forgot my footnote. ;)
It was:
I've actually mostly given up my stress script in favor of --stress. I only used it here because of the bisection issue you mention.
One other thing I've noticed with it: I forget to add my custom --root=/var/ram/git-tests when I invoke it, so my hard disk goes crazy (and the tests often run slower!). I'm not sure if there's a convenient fix.
-Peff
Quoted 4 lines · Jeff King
One other thing I've noticed with it: I forget to add my custom --root=/var/ram/git-tests when I invoke it, so my hard disk goes crazy (and the tests often run slower!). I'm not sure if there's a convenient fix.
OTOH, that could introduce more variance in the timing of the test's commands, thus potentially increasing the chances of a failure. I dunno.
Maybe ./t1234-foo.sh should learn to respect DEFAULT_TEST_OPTS somehow?
Quoted 6 lines · SZEDER Gábor
Quoted 4 lines · Jeff King
One other thing I've noticed with it: I forget to add my custom --root=/var/ram/git-tests when I invoke it, so my hard disk goes crazy (and the tests often run slower!). I'm not sure if there's a convenient fix.
Quoted 6 lines
OTOH, that could introduce more variance in the timing of the test's commands, thus potentially increasing the chances of a failure. I dunno.
Maybe ./t1234-foo.sh should learn to respect
DEFAULT_TEST_OPTSsomehow?
Yeah, that was what I was thinking. On the other hand, I'd actually find that a little bit annoying for the non-stress case. I commonly do "./t1234-foo.sh" in order to dig into a specific breakage, and having the failing trash directory right there is convenient (and I don't care as much about speed, since I'm just running it once).
I may just gut my "stress" script and make it a wrapper for calling the script with "--stress --root=...". :)
-Peff
Quoted 13 lines · Jeff King
Quoted 10 lines · SZEDER Gábor
- Make it exit with failure if a failure is found.
- Add the '--stress-limit=<N>' option to repeat the test script at most N times in each of the parallel jobs, and exit with success when the limit is reached.
[...]This is a case when an external stress script works better, as it can easily check commits in the past... if someone has such a script, that is.
Quoted 8 lines
Heh, I literally just implemented this kind of max-count in my own "stress" script[1] to handle this recent t0025 testing. So certainly I think it is a good idea.
Picking an <N> is tough. Too low and you get a false negative, too high and you can wait forever, especially if the script is long. But I don't think there's any real way to auto-scale it, except by seeing a few of the failing cases and watching how long they take.
So far I've chosen <N> like this: run the test script with --stress 3-5 times to trigger the failure, take the highest repetition count that was necessary for the failure, multiply it by 4-6 to get a round number, and that's a good ballpark for <N>. And once bisect came up with the suspect commit, I double checked it by letting the test script run with --stress on its parent commit for at least 5-10x <N> repetitions.
Anyway, I doubt that auto-scaling <N> is worth the effort.
Quoted 10 lines
Quoted 3 lines
t/README | 5 +++++ t/test-lib.sh | 18 ++++++++++++++++-- 2 files changed, 21 insertions(+), 2 deletions(-)Patch looks good. A few observations:
Quoted 4 lines
t/test-lib.sh +0
-0@@ -237,8 +248,10 @@ then exit 1 ' TERM INT
Quoted 12 lines
t/test-lib.sh then · line 257
Quoted 9 lines
t/test-lib.sh +4
-2- cnt=0 - while ! test -e "$stressfail" + cnt=1 + while ! test -e "$stressfail" && + { test -z "$stress_limit" || + test $cnt -le $stress_limit ; } do $TEST_SHELL_PATH "$0" "$@" >"$TEST_RESULTS_BASE.stress-$job_nr.out" 2>&1 & test_pid=$!You switch to 1-indexing the counts here. I think that makes sense, since otherwise --stress-limit=300 would end at "1.299", etc.
Yeah, that's exactly why I did it.
Quoted 12 lines
t/test-lib.sh then · line 280
Quoted 9 lines
t/test-lib.sh +1
-0@@ -261,6 +274,7 @@ then if test -f "$stressfail" then + stress_exit=1 echo "Log(s) of failed test run(s):" for failed_job_nr in $(sort -n "$stressfail") doI think I'd argue that this missing
stress_exitis a bug in the original script,
Well, yes, indeed.
Though being able to trigger an elusive test failure is a success in my book ;)
Quoted 5 lines
and somewhat orthogonal to the limit counter. But I don't think it's worth the trouble to split it out (and certainly the theme of "now you can run this via bisect" unifies the two changes).
-Peff
Quoted 5 lines · SZEDER Gábor
Quoted 4 lines
Picking an <N> is tough. Too low and you get a false negative, too high and you can wait forever, especially if the script is long. But I don't think there's any real way to auto-scale it, except by seeing a few of the failing cases and watching how long they take.
Quoted 7 lines
So far I've chosen <N> like this: run the test script with --stress 3-5 times to trigger the failure, take the highest repetition count that was necessary for the failure, multiply it by 4-6 to get a round number, and that's a good ballpark for <N>. And once bisect came up with the suspect commit, I double checked it by letting the test script run with --stress on its parent commit for at least 5-10x <N> repetitions.
Heh. That's exactly my process, too. :)
Quoted 1 line
Anyway, I doubt that auto-scaling <N> is worth the effort.
Yeah, especially because as a concept it exists outside of the script itself (i.e., you have to checkout a failing version and then run the script a bunch of times; that's not something that test-lib.sh should even know about).
So let's go with this for now. It's already a much nicer tool than we had yesterday, so we can take some time to get used to it.
-Peff
PATCH 1/1 test-lib: fix non-portable pattern bracket expressions
In reply to SZEDER GáborUse a '!' character to start a non-matching pattern bracket expression, as specified by POSIX in Shell Command Language section 2.13.1 Patterns Matching a Single Character [1].
I used '^' instead in three places in the previous three commits, to verify that the arguments of the '--stress=' and '--stress-limit=' options and the values of various '*_PORT' environment variables are valid numbers. With certain shells, at least with dash (upstream and in Ubuntu 14.04) and mksh, this led to various undesired behaviors:
# error message in case of a valid number $ ~/src/dash/src/dash ./t3903-stash.sh --stress=8 error: --stress=<N> requires the number of jobs to run
# not the expected error message $ ~/src/dash/src/dash ./t3903-stash.sh --stress=foo ./t3903-stash.sh: 238: test: Illegal number: foo
# no error message at all?! $ mksh ./t3903-stash.sh --stress=foo $ echo $? 0
Some other shells, e.g. Bash (even in posix mode), ksh, dash in Ubuntu 16.04 or later, are apparently happy to accept '^' just as well.
[1] http://pubs.opengroup.org/onlinepubs/009695399/utilities/xcu_chap02.html#tag_02_13
---
Should go on top of 'sg/stress-test'.
t/test-lib-functions.sh | 2 +- t/test-lib.sh | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-)
t/test-lib-functions.sh +1 -1
@@ -1289,7 +1289,7 @@ test_set_port () { port=$(($port + 10000)) fi ;; - *[^0-9]*|0*) + *[!0-9]*|0*) error >&7 "invalid port number: $port" ;; *)
t/test-lib.sh +2 -2
@@ -144,7 +144,7 @@ do --stress=*) stress=${opt#--*=} case "$stress" in - *[^0-9]*|0*|"") + *[!0-9]*|0*|"") echo "error: --stress=<N> requires the number of jobs to run" >&2 exit 1 ;;
@@ -155,7 +155,7 @@ do --stress-limit=*) stress_limit=${opt#--*=} case "$stress_limit" in - *[^0-9]*|0*|"") + *[!0-9]*|0*|"") echo "error: --stress-limit=<N> requires the number of repetitions" >&2 exit 1 ;;
Signature · 1 line
2.21.0.rc0.405.g6226d63c89
Re: [PATCH] test-lib: fix non-portable pattern bracket expressions
Quoted 5 lines · SZEDER Gábor
Use a '!' character to start a non-matching pattern bracket expression, as specified by POSIX in Shell Command Language section 2.13.1 Patterns Matching a Single Character [1].
I used '^' instead in three places in the previous three commits, to
Ah, thanks for catching them. We should have caught these during review.
Quoted 1 line
Should go on top of 'sg/stress-test'.
Thanks.
Re: [PATCH] test-lib: fix non-portable pattern bracket expressions
In reply to SZEDER GáborQuoted 3 lines · SZEDER Gábor
Use a '!' character to start a non-matching pattern bracket expression, as specified by POSIX in Shell Command Language section 2.13.1 Patterns Matching a Single Character [1].
Just when I think I know every little gotcha in the shell, I learn another one. :) Thanks for fixing this, and for digging up the POSIX reference.
-Peff
Re: [PATCH] test-lib: fix non-portable pattern bracket expressions
Quoted 7 lines · Jeff King
Quoted 3 lines · SZEDER Gábor
Use a '!' character to start a non-matching pattern bracket expression, as specified by POSIX in Shell Command Language section 2.13.1 Patterns Matching a Single Character [1].
Just when I think I know every little gotcha in the shell,
I've already gave up on that :)
Quoted 3 lines
I learn
another one. :) Thanks for fixing this, and for digging up the POSIX reference.
I had to, that 16.04's dash worked, but neither dash in older LTS nor newer upstream version did was particularly puzzling. Turns out that in 0.5.8-1 Ubuntu (Debian? dunno.) started to configure dash with '--enable-fnmatch', which makes it understand '^' as well.
Re: [PATCH] test-lib: fix non-portable pattern bracket expressions
Quoted 8 lines · SZEDER Gábor
Quoted 3 lines
I learn
another one. :) Thanks for fixing this, and for digging up the POSIX reference.I had to, that 16.04's dash worked, but neither dash in older LTS nor newer upstream version did was particularly puzzling. Turns out that in 0.5.8-1 Ubuntu (Debian? dunno.) started to configure dash with '--enable-fnmatch', which makes it understand '^' as well.
Ah, doubly puzzling. It works fine in my Debian dash (but fnmatch was enabled in 0.5.8-1 there, too).
-Peff
Re: [PATCH] test-lib: fix non-portable pattern bracket expressions
Quoted 3 lines · Jeff King
Ah, doubly puzzling. It works fine in my Debian dash (but fnmatch was enabled in 0.5.8-1 there, too).
FWIW, it is reported by both checkbashisms and shellcheck if "linting" could be considered of help to track these kind of issues.
other "portability" issues seem safe, mainly because they are either silly (like the use of local) or meant to be used only with bash (as on the mingw specific functions or the bash version check that was broken recently for NetBSD) but the functions run_with_limited_{cmdline,stack} that use ulimit -s might be worth looking into
Carlo