Two bugs. First, multifire() checks the condition only for surface
ships, not for submarines. Second, multifire() neglects to write back
the ship after retreating it. The player is told the ship retreats,
but it actually stays where it is.
Broken since retreat was added in Chainsaw. Previous fixes (commit
8065fe8, v4.3.1 and commit de2651e, v4.3.19) "fixed" only the
bulletin.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
It's not firing, yet.
While there, trim an unwanted blank line before reporting the first
sector ready.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
Can't actually happen with the current damage formulas. If it could,
then the special treatment would be inconsistent with sectors and land
units.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
We check all necessary conditions for being able to fire before
prompting for a target. Except for land unit guns. Clean that up.
fort_fire(), shp_fire() or lnd_fire() fail only when the fort, ship or
land unit can't fire. If that happens, our checking is incomplete.
Oops then.
We recheck some of the necessary conditions after getting the target.
However, because the command fails when the firing sector, ship or
land unit has changed since the first check, these rechecks can't
fail. Drop them.
Note that the rechecks were just as useless before commit 66165f3
fixed fire to fail on change, because they rechecked the unchanged
cached copy instead of the possibly changed original.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
This was a lame attempt at limiting the ripple effect of adding /
removing commands. Better means are now available, so drop it.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
Changed xdump output is too painful to review. final.xdump should
still catch changes that are visible only in xdump.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
Adding or removing a command to/from a test has unfortunate effects:
* Before the previous commit: if the command consumes pseudorandom
numbers, all subsequent users of pseudorandom numbers get different
ones. This has always been a major headache.
* Since the previous commit: all subsequent users of pseudorandom
numbers get different ones whether the command consumes any or not.
That's even worse.
* If the command uses BTUs, subsequent prompts are changed. Not
nearly as bad as the above, but still annoying.
Create a new command __cmd to allow compensating for adding/removing
commands for tests. Throw in the ability to compensate treasury
changes for good measure. Three arguments: command count, BTU use,
money use.
Usage example: say you add a convert command to a test, and it uses 3
BTUs and $15. Then you compensate by adding "__cmd added 1 3 15"
right after it.
The command must not be available unless running_test_suite is on, of
course. Make it require the new player command capability TESTING,
and give that to all players when running_test_suite is on.
The command is intentionally not documented in info. Switch
running_test_suite off for info-test, to hide it (and any future
TESTING commands) from info-test.
Suppress the command counter increment for TESTING commands, so they
can be used without upsetting pseudorandom numbers
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
Tests need repeatable pseudorandom numbers to yield repeatable
results. We seed the pseudorandom number generator with a fixed value
(emp_server -R) to make it produce the same sequence of numbers every
time. But whenever code exercised by a test is changed to consume
fewer or more of them, all subsequent users get different numbers
regardless. The ensuing test result changes are extremely tedious to
review.
To address this problem, reseed the PRNG with the count of commands
right before executing a command when running_test_suite is on. This
way, the effect of perturbing the PRN sequence lasts only until the
next command.
Note that the next command could be another player's. Doesn't matter.
Adding or removing commands now upsets the PRN sequence even for
commands that don't consume PRNs. The next commit will take care of
that.
Perturbs test results across the board. Hopefully, that'll happen
much less frequently now.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
For now, it just logs "Configured for testing" on startup, and prints
a scary warning on player login.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
Command prompts show nat_timeused rounded down to minutes. They need
to be normalized, or else tests can fail when they take too long, or
cross midnight. Formatted prompts are normalized correctly (not
actually used since commit 9ca3fa9), but the journal contains raw
prompts. Normalize them, too.
Smoke test's journal.log is affected, because the server charges at
least 15s per login, which adds up into minutes in this test.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
To make setup scripting more similar to test scripting. Also permits
use of countries other than POGO there, but that isn't necessary right
now.
Setup scripts renamed from init_script to setup-POGO.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
Instead of relative to build directory. Shorter, and independent of
the path from build to source directory.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
The other test output files have fixed names, and having just one with
a name that varies with the test name complicates things with no gain.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
Only actofgod-test wants it. Enable it there. The others either
don't want it (fire-test), or don't care.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
Warn the first time a server is started. Incorrect for info-test with
POSIX threads, so suppress the warning there.
Improve the warning message a bit while we're at it.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
Commit 37ff377 stripped trailing white space in test output, except
for journal.log, where it stripped it only from player output. This
misses the space preceeding player empty output, and doesn't cover
equally annoying trailing white space elsewhere, such as the space
preceeding empty input and trailing white space in prompts. Testing
the latter could be marginally useful, but let's keep things simple,
and strip all trailing white space for now.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
Empire 2 settled on this formula for the stack size:
stacksize = 100000 +
/* finish_sects */ WORLD_X * WORLD_Y * (2 * sizeof(double) +
sizeof(char *));
Obviously attempts to provide space for a known configuration-
dependent stack hog. The hog went away when finish_sects()'s arrays
became dynamically allocated in 4.2.0.
Adjusting for that by dropping the extra term might well do (I observe
only a few KiB of stack used on my system). But let's set it to 512
KiB instead to be on the safe side.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
I observed a stack overflow in news command on my x86_64 system
running Fedora 18.
Empire 2 settled on this formula for the stack size:
stacksize = 100000
/* budget */ + MAX(WORLD_SZ() * sizeof(int) * 7,
/* power */ MAXNOC * sizeof(struct powstr));
Obviously attempts to provide space for known configuration-dependent
stack hogs.
The first hog is allegedly budget. Bogus since day one: its large
arrays were static in Empire 2, and became dynamically allocated in
Empire 3.
The second one makes some sense: powe() has a struct powstr[MAXNOC].
It also has an int[MAXNOC], which isn't accounted for.
Except for ridiculously small worlds, the second term is smaller, and
only the (bogus) first term matters.
Two hogs are missing: head() has a struct histstr[MAXNOC][MAXNOC], and
news() has a short[MAXNOC][MAXNOC]. It also calls head().
I looked for more hogs with "gcc -fstack-usage", and found none.
On my x86_64 system, a news command needs almost 107KiB of stack.
Only slightly less when compiled for 32 bit. Stack overrun for worlds
with fewer than some 320 sectors, thus unlikely to bite in real games.
Increase player stack size to 1 MiB. Using MAXNOC to size the stack
isn't worth the trouble.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
Normally, git's pre-commit hook protects us from them. However, when
expected test output contains trailing white space, we have to bypass
commit hooks. Unwanted space can then slip in if you don't pay
attention. I obviously didn't; clean up.
The previous commit should reduce the need for such hook suppression.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
feed_input drops lines starting with a bar character '|', so they can
serve as comments. Syntax chosen because such lines shouldn't be
needed in tests. In particular, the server already ignores such lines
when it reads commands, because they get parsed as empty command with
a pipeline, and empty commands get ignored, regardless of
redirections.
Signed-off-by: Markus Armbruster <armbru@pond.sub.org>
