Fix warnings
objnam.c: In function ‘wizterrainwish’:
objnam.c:3536:54: warning: variable ‘didblock’ set but not used [-Wunused-but-set-variable]
3536 | boolean madeterrain = FALSE, badterrain = FALSE, didblock, is_dbridge;
| ^~~~~~~~
We've had reports of a couple of issues building against musl libc.
Issues reported:
- build procedures utilize col for Guidebook-creation, and col
is deprecated in distros that use musl libc
- some of the CRASHREPORT code is using library functions that
are not available in the musl libc environment. The reported
functions were backtrace() and backtrace_symbols(), which use
header file /usr/include/execinfo.h.
So we'll try to accommodate this. Since we don't have a means of
autodetecting the musl libc situation during the build (as of yet), the
builder will have to specify 'make musl=1' on the make command line.
Specifying 'musl=1' on the make command line will:
1. ensure that NOCRASHREPORT gets defined in the C preprocessor.
2. set COLCMD to be '../util/stripbs' instead of 'col -bx'.
Related to GitHub #1393
We've had reports of a couple of issues building against musl libc.
Issues reported:
- build procedures utilize cat for Guidebook-creation, and cat
is deprecated in distros that use musl libc.
- some of the CRASHREPORT code is using library functions that
are not available in the musl libc environment. The reported
functions were backtrace() and backtrace_symbols(), which use
header file /usr/include/execinfo.h.
So we'll try to accommodate this. Since we don't have a means of
autodetecting the musl libc situation during the build (as of yet), the
builder will have to specify 'make musl=1' on the make command line.
Specifying 'musl=1' on the make command line will:
1. ensure that NOCRASHREPORT gets defined in the C preprocessor.
2. set COLCMD to be '../util/stripbs' instead of 'col -bx'.
Closes#1393
Bug description of #1386 by @copperwater on GitHub:
"When generating a random monster from a class using des.monster(),
the G_NOGEN in their statblock is suppressed, but because every monster
of this class has frequency 0, none of them are actually eligible to get
picked. mkclass ends up returning a null pointer and create_monster has
to pick a random monster instead.
This affects the following levels (all the ones that use random sea monsters):
Healer quest start
Healer quest locate
Plane of Water (difficult to notice, since it has lots of specific sea monsters and only 5 random ones)
This can be pretty easily viewed by going to the Healer quest start and
detecting monsters: there is a shark and a giant eel, which are
specifically defined, but the remaining random sea monster that should
be there is absent."
Add a tracking array mclass_maxf[MAXMCLASSES] (about 61 entries, the
first not being used), and fill it one time in init_mongen_order() with
the maximum frequency value seen of any monster in that class.
Any mclass_maxf[] entry of zero represents that entire class of monsters
having no positive frequency value.
Detect that in mkclass_aligned(), and use it to work around the situation
to produce the monster being sought by the Lua level description file.
The array was ending up ordered the same on different qsort
implementations.
This incorporates the mlet value into the sort value for comparison.
That guarantees that everything stays ordered by mlet, followed by
difficulty (and would even if something ever got misplaced in
monsters.h). It means the "they are equal" zero return for differing
mlet values is not required, and has been removed.
This removes the "they are equal" zero returns for G_NOGEN | G_UNIQ
monsters, so they will still get sorted rather than left at the
whatever array element they happened to be at (which I don't think
should be an issue?).
Noticed while testing an earlier fix - after your pet consumes a large mimic
corpse don't have it remain leashed once it begins to mimic something
that isn't leashable
Some variants were already using a similar approach
using a struct called 'ebones', so adopt the same naming
so NetHack-3.7, hardfought, and some variants are using
the same name.
As before there are fields in the struct that are not
currently used by NetHack-3.7, but the intent is that
hardfought save and bones files can be loaded by
NetHack-3.7 without code modification, for debugging
bug reports.
This invalidates existing save and bones files.
There are two hardfought code additions that render save and bones files incompatible
with the upstream NetHack-3.7, and that makes testing with hardfought
save and bones files more challenging than it needs to be, when
investigating and troubleshooting bug reports.
Add some unused fields to advance towards achieving save file parity with
hardfought, which is a significant source of play-testing for NetHack-3.7.
1) the elbereth field addition to u_conduct
This adds an unused placeholder field named 'hf_reserved1', at the appropriate
place in u_conduct to achieve struct field parity with the one in use on
hardfought.
2) hardfought adds a field to struct monst:
char former_rank[25]; /* for bones' ghost rank in their former life */
Instead of adding that to every monst, this adds a new mextra struct
named 'former', which currently contains the equivalent 25-character
field called 'rank' which can hold the content that was in the
former_rank[25] field. That way, the field will only be added when it
is needed.
A pull request https://github.com/k21971/NetHack37/pull/2 has been
done on hardfought to do it the same way (untested there as of yet).
Even though NetHack-3.7 does not utilize that information presently,
this will be a further step toward allowing hardfought-generated save
and bones files to be used for troubleshooting, without modification,
on a similar architecture running stock NetHack-3.7 code.
That savefile parity won't be achieved until the after the
hardfought pull-request mentioned above (or equivalent) is merged.
As this change will not be compatible with existing save and bones
files, it will be accompanied with an EDITLEVEL increment.
Commit 1acc2727 helped ensure that the which_armor(mtmp, W_SADDLE)
test at the top of put_saddle_on_mon() wouldn't lead to an obj
leak.
This commit covers off the adjacent can_saddle() test in
put_saddle_on_mon(), because if that failed, it could also lead
to a memory leak of the saddle obj passed by the caller.
- have put_saddle_on_mon() create and use its own saddle obj
if a NULL saddle obj is passed, instead of having to do that
in the caller.
- where an existing saddle obj needs to be passed from the caller,
ensure that the caller has done its own can_saddle(mon) check prior
to calling put_saddle_on_mon(), so that the can_saddle() test
in put_saddle_on_mon() won't fail.
- lastly, add an impossible() to put_saddle_on_mon() to catch
a failure when a saddle obj is passed from the caller and either
test has failed, just in case. That should not happen with any of
the existing cases now, but it will provide some bullet-proofing
for new code, new callers.
In file included from ../include/config.h:723:0,
from ../include/hack.h:10,
from files.c:8:
../include/global.h:519:24: error: expected ')' before '<=' token
#define unctrl(c) ((c) <= C('z') ? (0x60 | (c)) : (c))
^
../lib/pdcursesmod/curses.h:1686:16: note: in expansion of macro 'unctrl'
PDCEX char *unctrl(chtype);
^~~~~~
Define a macro NH_C to provide a shorter & simpler way to test for
which C standard the build is being carried out under (c99 or c23).
NH_C > 202300L Being compiled under C23 or greater
NH_C > 199900L Being compiled under C99 or greater
NH_C > 198900L Being compiled under C89 or greater,
or C std could not be determined.
While NetHack only requires c99, we've been taking advantage
of some c23 features (attributes), if they are available,
to allow the use of ATTRNORETURN/NORETURN and FALLTHROUGH on
compilers other than gcc.
Also add some comment documentation to tradstdc.h about NetHack's
use of c99.
The sys/unix/Makefile.top change overcomes a warning in the
Makefile-generated nhlua.h. That warning arises under some compilers
that rely on attribute [[noreturn]] ahead of a declaration
(NetHack macro ATTRNORETURN), rather than the trailing gcc
__attribute((noreturn)) (NetHack macro NORETURN). The sed command
is modified to include ATTRNORETURN at the start of the declaration
in addition to the NORETURN at the end of the declaration, in the
generated file. That's the same combination that's used for the
declaration of other functions that don't return.
nhmall