Skip to content

P
Postgres FD Implementation
Commit d675226e authored by Tom Lane's avatar Tom Lane
Browse files
Options

Regenerate badly-obsolete README file.

parent 4c80071b
Hide whitespace changes
Inline Side-by-side
Showing with 277 additions and 279 deletions
src/test/regress/README
View file @ d675226e
Regression Tests
The regression tests are a comprehensive set of tests for the SQL Regression Tests
implementation in PostgreSQL. They test standard SQL operations as well as the
extended capabilities of PostgreSQL. From PostgreSQL 6.1 onward, the regression The regression tests are a comprehensive set of tests for the SQL
tests are current for every official release. implementation in PostgreSQL. They test standard SQL operations as
well as the extended capabilities of PostgreSQL.
------------------------------------------------------------------------------- _________________________________________________________________
Running the Tests Running the Tests
The regression test can be run against an already installed and running server, The regression tests can be run against an already installed and
or using a temporary installation within the build tree. Furthermore, there is running server, or using a temporary installation within the build
a "parallel" and a "sequential" mode for running the tests. The sequential tree. Furthermore, there is a "parallel" and a "sequential" mode for
method runs each test script in turn, whereas the parallel method starts up running the tests. The sequential method runs each test script in
multiple server processes to run groups of tests in parallel. Parallel testing turn, whereas the parallel method starts up multiple server processes
gives confidence that interprocess communication and locking are working to run groups of tests in parallel. Parallel testing gives confidence
correctly. For historical reasons, the sequential test is usually run against that interprocess communication and locking are working correctly. For
an existing installation and the parallel method against a temporary historical reasons, the sequential test is usually run against an
installation, but there are no technical reasons for this. existing installation and the parallel method against a temporary
installation, but there are no technical reasons for this.
To run the regression tests after building but before installation, type
To run the regression tests after building but before installation,
gmake check type
gmake check
in the top-level directory. (Or you can change to "src/test/regress" and run
the command there.) This will first build several auxiliary files, such as some in the top-level directory. (Or you can change to "src/test/regress"
sample user-defined trigger functions, and then run the test driver script. At and run the command there.) This will first build several auxiliary
the end you should see something like files, such as some sample user-defined trigger functions, and then
run the test driver script. At the end you should see something like
====================== ======================
All 93 tests passed. All 98 tests passed.
====================== ======================
or otherwise a note about which tests failed. See the Section called Test or otherwise a note about which tests failed. See the section called
Evaluation below for more. Test Evaluation below before assuming that a "failure" represents a
serious problem.
Because this test method runs a temporary server, it will not work when you are
the root user (since the server will not start as root). If you already did the Because this test method runs a temporary server, it will not work
build as root, you do not have to start all over. Instead, make the regression when you are the root user (since the server will not start as root).
test directory writable by some other user, log in as that user, and restart If you already did the build as root, you do not have to start all
the tests. For example over. Instead, make the regression test directory writable by some
other user, log in as that user, and restart the tests. For example
root# chmod -R a+w src/test/regress root# chmod -R a+w src/test/regress
root# chmod -R a+w contrib/spi root# chmod -R a+w contrib/spi
root# su - joeuser root# su - joeuser
joeuser$ cd top-level build directory joeuser$ cd top-level build directory
joeuser$ gmake check joeuser$ gmake check
(The only possible "security risk" here is that other users might be able to (The only possible "security risk" here is that other users might be
alter the regression test results behind your back. Use common sense when able to alter the regression test results behind your back. Use common
managing user permissions.) sense when managing user permissions.)
Alternatively, run the tests after installation. Alternatively, run the tests after installation.
The parallel regression test starts quite a few processes under your user ID. If you have configured PostgreSQL to install into a location where an
Presently, the maximum concurrency is twenty parallel test scripts, which means older PostgreSQL installation already exists, and you perform gmake
sixty processes: there's a server process, a psql, and usually a shell parent check before installing the new version, you may find that the tests
process for the psql for each test script. So if your system enforces a per- fail because the new programs try to use the already-installed shared
user limit on the number of processes, make sure this limit is at least libraries. (Typical symptoms are complaints about undefined symbols.)
seventy-five or so, else you may get random-seeming failures in the parallel If you wish to run the tests before overwriting the old installation,
test. If you are not in a position to raise the limit, you can cut down the you'll need to build with configure --disable-rpath. It is not
degree of parallelism by setting the MAX_CONNECTIONS parameter. For example, recommended that you use this option for the final installation,
however.
gmake MAX_CONNECTIONS=10 check
The parallel regression test starts quite a few processes under your
runs no more than ten tests concurrently. user ID. Presently, the maximum concurrency is twenty parallel test
scripts, which means sixty processes: there's a server process, a
On some systems, the default Bourne-compatible shell ("/bin/sh") gets confused psql, and usually a shell parent process for the psql for each test
when it has to manage too many child processes in parallel. This may cause the script. So if your system enforces a per-user limit on the number of
parallel test run to lock up or fail. In such cases, specify a different processes, make sure this limit is at least seventy-five or so, else
Bourne-compatible shell on the command line, for example: you may get random-seeming failures in the parallel test. If you are
not in a position to raise the limit, you can cut down the degree of
gmake SHELL=/bin/ksh check parallelism by setting the MAX_CONNECTIONS parameter. For example,
gmake MAX_CONNECTIONS=10 check
If no non-broken shell is available, you may be able to work around the problem
by limiting the number of connections, as shown above. runs no more than ten tests concurrently.
To run the tests after installation, initialize a data area and start the On some systems, the default Bourne-compatible shell ("/bin/sh") gets
server, then type confused when it has to manage too many child processes in parallel.
This may cause the parallel test run to lock up or fail. In such
gmake installcheck cases, specify a different Bourne-compatible shell on the command
line, for example:
The tests will expect to contact the server at the local host and the default gmake SHELL=/bin/ksh check
port number, unless directed otherwise by PGHOST and PGPORT environment
variables. If no non-broken shell is available, you may be able to work around
the problem by limiting the number of connections, as shown above.
-------------------------------------------------------------------------------
To run the tests after installation, initialize a data area and start
Test Evaluation the server, then type
gmake installcheck
Some properly installed and fully functional PostgreSQL installations can
"fail" some of these regression tests due to platform-specific artifacts such or for a parallel test
as varying floating-point representation and time zone support. The tests are gmake installcheck-parallel
currently evaluated using a simple "diff" comparison against the outputs
generated on a reference system, so the results are sensitive to small system The tests will expect to contact the server at the local host and the
differences. When a test is reported as "failed", always examine the default port number, unless directed otherwise by PGHOST and PGPORT
differences between expected and actual results; you may well find that the environment variables.
differences are not significant. Nonetheless, we still strive to maintain _________________________________________________________________
accurate reference files across all supported platforms, so it can be expected
that all tests pass. Test Evaluation
The actual outputs of the regression tests are in files in the "src/test/ Some properly installed and fully functional PostgreSQL installations
regress/results" directory. The test script uses "diff" to compare each output can "fail" some of these regression tests due to platform-specific
file against the reference outputs stored in the "src/test/regress/expected" artifacts such as varying floating-point representation and message
directory. Any differences are saved for your inspection in "src/test/regress/ wording. The tests are currently evaluated using a simple "diff"
regression.diffs". (Or you can run "diff" yourself, if you prefer.) comparison against the outputs generated on a reference system, so the
results are sensitive to small system differences. When a test is
------------------------------------------------------------------------------- reported as "failed", always examine the differences between expected
and actual results; you may well find that the differences are not
Error message differences significant. Nonetheless, we still strive to maintain accurate
reference files across all supported platforms, so it can be expected
Some of the regression tests involve intentional invalid input values. Error that all tests pass.
messages can come from either the PostgreSQL code or from the host platform
system routines. In the latter case, the messages may vary between platforms, The actual outputs of the regression tests are in files in the
but should reflect similar information. These differences in messages will "src/test/regress/results" directory. The test script uses "diff" to
result in a "failed" regression test that can be validated by inspection. compare each output file against the reference outputs stored in the
"src/test/regress/expected" directory. Any differences are saved for
------------------------------------------------------------------------------- your inspection in "src/test/regress/regression.diffs". (Or you can
run "diff" yourself, if you prefer.)
Locale differences
If for some reason a particular platform generates a "failure" for a
If you run the tests against an already-installed server that was initialized given test, but inspection of the output convinces you that the result
with a collation-order locale other than C, then there may be differences due is valid, you can add a new comparison file to silence the failure
to sort order and follow-up failures. The regression test suite is set up to report in future test runs. See the section called Variant Comparison
handle this problem by providing alternative result files that together are Files for details.
known to handle a large number of locales. For example, for the char test, the _________________________________________________________________
expected file "char.out" handles the C and POSIX locales, and the file
"char_1.out" handles many other locales. The regression test driver will Error message differences
automatically pick the best file to match against when checking for success and
for computing failure differences. (This means that the regression tests cannot Some of the regression tests involve intentional invalid input values.
detect whether the results are appropriate for the configured locale. The tests Error messages can come from either the PostgreSQL code or from the
will simply pick the one result file that works best.) host platform system routines. In the latter case, the messages may
vary between platforms, but should reflect similar information. These
If for some reason the existing expected files do not cover some locale, you differences in messages will result in a "failed" regression test that
can add a new file. The naming scheme is testname_digit.out. The actual digit can be validated by inspection.
is not significant. Remember that the regression test driver will consider all _________________________________________________________________
such files to be equally valid test results. If the test results are platform-
specific, the technique described in the Section called Platform-specific Locale differences
comparison files should be used instead.
If you run the tests against an already-installed server that was
------------------------------------------------------------------------------- initialized with a collation-order locale other than C, then there may
be differences due to sort order and follow-up failures. The
Date and time differences regression test suite is set up to handle this problem by providing
alternative result files that together are known to handle a large
A few of the queries in the "horology" test will fail if you run the test on number of locales.
the day of a daylight-saving time changeover, or the day after one. These _________________________________________________________________
queries expect that the intervals between midnight yesterday, midnight today
and midnight tomorrow are exactly twenty-four hours --- which is wrong if Date and time differences
daylight-saving time went into or out of effect meanwhile.
Most of the date and time results are dependent on the time zone
Note: Because USA daylight-saving time rules are used, this problem environment. The reference files are generated for time zone PST8PDT
always occurs on the first Sunday of April, the last Sunday of (Berkeley, California), and there will be apparent failures if the
October, and their following Mondays, regardless of when daylight- tests are not run with that time zone setting. The regression test
saving time is in effect where you live. Also note that the problem driver sets environment variable PGTZ to PST8PDT, which normally
appears or disappears at midnight Pacific time (UTC-7 or UTC-8), not ensures proper results.
midnight your local time. Thus the failure may appear late on _________________________________________________________________
Saturday or persist through much of Tuesday, depending on where you
live. Floating-point differences
Most of the date and time results are dependent on the time zone environment. Some of the tests involve computing 64-bit floating-point numbers
The reference files are generated for time zone PST8PDT (Berkeley, California), (double precision) from table columns. Differences in results
and there will be apparent failures if the tests are not run with that time involving mathematical functions of double precision columns have been
zone setting. The regression test driver sets environment variable PGTZ to observed. The float8 and geometry tests are particularly prone to
PST8PDT, which normally ensures proper results. However, your operating system small differences across platforms, or even with different compiler
must provide support for the PST8PDT time zone, or the time zone-dependent optimization options. Human eyeball comparison is needed to determine
tests will fail. To verify that your machine does have this support, type the the real significance of these differences which are usually 10 places
following: to the right of the decimal point.
env TZ=PST8PDT date Some systems display minus zero as -0, while others just show 0.
The command above should have returned the current system time in the PST8PDT Some systems signal errors from pow() and exp() differently from the
time zone. If the PST8PDT time zone is not available, then your system may have mechanism expected by the current PostgreSQL code.
returned the time in UTC. If the PST8PDT time zone is missing, you can set the _________________________________________________________________
time zone rules explicitly:
Row ordering differences
PGTZ='PST8PDT7,M04.01.0,M10.05.03'; export PGTZ
You might see differences in which the same rows are output in a
There appear to be some systems that do not accept the recommended syntax for different order than what appears in the expected file. In most cases
explicitly setting the local time zone rules; you may need to use a different this is not, strictly speaking, a bug. Most of the regression test
PGTZ setting on such machines. scripts are not so pedantic as to use an ORDER BY for every single
SELECT, and so their result row orderings are not well-defined
Some systems using older time-zone libraries fail to apply daylight-saving according to the letter of the SQL specification. In practice, since
corrections to dates before 1970, causing pre-1970 PDT times to be displayed in we are looking at the same queries being executed on the same data by
PST instead. This will result in localized differences in the test results. the same software, we usually get the same result ordering on all
platforms, and so the lack of ORDER BY isn't a problem. Some queries
------------------------------------------------------------------------------- do exhibit cross-platform ordering differences, however. When testing
against an already-installed server, ordering differences can also be
Floating-point differences caused by non-C locale settings or non-default parameter settings,
such as custom values of work_mem or the planner cost parameters.
Some of the tests involve computing 64-bit floating-point numbers (double
precision) from table columns. Differences in results involving mathematical Therefore, if you see an ordering difference, it's not something to
functions of double precision columns have been observed. The float8 and worry about, unless the query does have an ORDER BY that your result
geometry tests are particularly prone to small differences across platforms, or is violating. But please report it anyway, so that we can add an ORDER
even with different compiler optimization options. Human eyeball comparison is BY to that particular query and thereby eliminate the bogus "failure"
needed to determine the real significance of these differences which are in future releases.
usually 10 places to the right of the decimal point.
You might wonder why we don't order all the regression test queries
Some systems display minus zero as -0, while others just show 0. explicitly to get rid of this issue once and for all. The reason is
that that would make the regression tests less useful, not more, since
Some systems signal errors from pow() and exp() differently from the mechanism they'd tend to exercise query plan types that produce ordered results
expected by the current PostgreSQL code. to the exclusion of those that don't.
_________________________________________________________________
-------------------------------------------------------------------------------
The "random" test
Row ordering differences
The random test script is intended to produce random results. In rare
You might see differences in which the same rows are output in a different cases, this causes the random regression test to fail. Typing
order than what appears in the expected file. In most cases this is not, diff results/random.out expected/random.out
strictly speaking, a bug. Most of the regression test scripts are not so
pedantic as to use an ORDER BY for every single SELECT, and so their result row should produce only one or a few lines of differences. You need not
orderings are not well-defined according to the letter of the SQL worry unless the random test fails repeatedly.
specification. In practice, since we are looking at the same queries being _________________________________________________________________
executed on the same data by the same software, we usually get the same result
ordering on all platforms, and so the lack of ORDER BY isn't a problem. Some Variant Comparison Files
queries do exhibit cross-platform ordering differences, however. (Ordering
differences can also be triggered by non-C locale settings.) Since some of the tests inherently produce environment-dependent
results, we have provided ways to specify alternative "expected"
Therefore, if you see an ordering difference, it's not something to worry result files. Each regression test can have several comparison files
about, unless the query does have an ORDER BY that your result is violating. showing possible results on different platforms. There are two
But please report it anyway, so that we can add an ORDER BY to that particular independent mechanisms for determining which comparison file is used
query and thereby eliminate the bogus "failure" in future releases. for each test.
You might wonder why we don't order all the regression test queries explicitly The first mechanism allows comparison files to be selected for
to get rid of this issue once and for all. The reason is that that would make specific platforms. There is a mapping file,
the regression tests less useful, not more, since they'd tend to exercise query "src/test/regress/resultmap", that defines which comparison file to
plan types that produce ordered results to the exclusion of those that don't. use for each platform. To eliminate bogus test "failures" for a
particular platform, you first choose or make a variant result file,
------------------------------------------------------------------------------- and then add a line to the "resultmap" file.
The "random" test Each line in the mapping file is of the form
testname/platformpattern=comparisonfilename
There is at least one case in the random test script that is intended to
produce random results. This causes random to fail the regression test once in The test name is just the name of the particular regression test
a while (perhaps once in every five to ten trials). Typing module. The platform pattern is a pattern in the style of the Unix
tool "expr" (that is, a regular expression with an implicit ^ anchor
diff results/random.out expected/random.out at the start). It is matched against the platform name as printed by
"config.guess" followed by :gcc or :cc, depending on whether you use
should produce only one or a few lines of differences. You need not worry the GNU compiler or the system's native compiler (on systems where
unless the random test always fails in repeated attempts. (On the other hand, there is a difference). The comparison file name is the base name of
if the random test is *never* reported to fail even in many trials of the the substitute result comparison file.
regression tests, you probably *should* worry.)
For example: some systems interpret very small floating-point values
------------------------------------------------------------------------------- as zero, rather than reporting an underflow error. This causes a few
differences in the "float8" regression test. Therefore, we provide a
Platform-specific comparison files variant comparison file, "float8-small-is-zero.out", which includes
the results to be expected on these systems. To silence the bogus
Since some of the tests inherently produce platform-specific results, we have "failure" message on OpenBSD platforms, "resultmap" includes
provided a way to supply platform-specific result comparison files. Frequently, float8/i.86-.*-openbsd=float8-small-is-zero
the same variation applies to multiple platforms; rather than supplying a
separate comparison file for every platform, there is a mapping file that which will trigger on any machine for which the output of
defines which comparison file to use. So, to eliminate bogus test "failures" "config.guess" matches i.86-.*-openbsd. Other lines in "resultmap"
for a particular platform, you must choose or make a variant result file, and select the variant comparison file for other platforms where it's
then add a line to the mapping file, which is "src/test/regress/resultmap". appropriate.
Each line in the mapping file is of the form The second selection mechanism for variant comparison files is much
more automatic: it simply uses the "best match" among several supplied
testname/platformpattern=comparisonfilename comparison files. The regression test driver script considers both the
standard comparison file for a test, testname.out, and variant files
The test name is just the name of the particular regression test module. The named testname_digit.out (where the "digit" is any single digit 0-9).
platform pattern is a pattern in the style of the Unix tool "expr" (that is, a If any such file is an exact match, the test is considered to pass;
regular expression with an implicit ^ anchor at the start). It is matched otherwise, the one that generates the shortest diff is used to create
against the platform name as printed by "config.guess" followed by :gcc or :cc, the failure report. (If "resultmap" includes an entry for the
depending on whether you use the GNU compiler or the system's native compiler particular test, then the base "testname" is the substitute name given
(on systems where there is a difference). The comparison file name is the name in "resultmap".)
of the substitute result comparison file.
For example, for the char test, the comparison file "char.out"
For example: some systems using older time zone libraries fail to apply contains results that are expected in the C and POSIX locales, while
daylight-saving corrections to dates before 1970, causing pre-1970 PDT times to the file "char_1.out" contains results sorted as they appear in many
be displayed in PST instead. This causes a few differences in the "horology" other locales.
regression test. Therefore, we provide a variant comparison file, "horology-no-
DST-before-1970.out", which includes the results to be expected on these The best-match mechanism was devised to cope with locale-dependent
systems. To silence the bogus "failure" message on HPUX platforms, "resultmap" results, but it can be used in any situation where the test results
includes cannot be predicted easily from the platform name alone. A limitation
of this mechanism is that the test driver cannot tell which variant is
horology/.*-hpux=horology-no-DST-before-1970 actually "correct" for the current environment; it will just pick the
variant that seems to work best. Therefore it is safest to use this
which will trigger on any machine for which the output of "config.guess" mechanism only for variant results that you are willing to consider
includes -hpux. Other lines in "resultmap" select the variant comparison file equally valid in all contexts.
for other platforms where it's appropriate.
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment