closes#8274
The parent commit tries a different approach, but that fails on Apple
Clang version:
Apple LLVM version 10.0.0 (clang-1000.11.45.5)
Target: x86_64-apple-darwin17.7.0
which somehow compiles the check_c_source_compiles() check, but then
complains during later compilation that __fpclassify is not defined
(regardless of "#include <math.h>").
Note about shada.c:
- shada_read_next_item_start was intentionally shadowing `unpacked` and
`i` because many of the macros (e.g. ADDITIONAL_KEY) implicitly
depended on those variable names.
- Macros were changed to parameterize `unpacked` (but not `i`). Macros
like CLEAR_GA_AND_ERROR_OUT do control-flow (goto), so any other
approach is messy.
In file included from ../src/nvim/eval/encode.c:974:
../src/nvim/eval/typval_encode.c.h:390:40: error: equality comparison with extraneous parentheses [-Werror,-Wparentheses-equality]
do { if ((tv->vval.v_special == kSpecialVarTrue)) { msgpack_pack_true(packer); } else { msgpack_pack_false(packer); } } while (0);
~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~
Problem: Cannot make Vim fail on an internal error.
Solution: Add IEMSG() and IEMSG2(). (Domenique Pelle) Avoid reporting an
internal error without mentioning where.
95f096030e
Signed-off-by: Michael Schupikov <michael@schupikov.de>
This makes gdb backtraces much more meaningful: specifically I now know at which
line it crashes in place of seeing that it crashes at
TYPVAL_ENCODE_DEFINE_CONV_FUNCTIONS macros invocation.
In order to not conflict with new error codes that Vim adds, all Neovim
error codes should be above 5000. The three existing sub-5000 error
codes (E926, E951, and E952) are now E50003, E5004, and E5005
respectively.
E953 was removed in 6167ce6df2, so just
remove it from the help.
This ought to prevent stack overflow, but I do not see this actually working:
*lua* code crashes with stack overflow when trying to deserialize msgpack from
Neovim, Neovim is fine even if nesting level is increased 100x (though test
becomes very slow); not sure how recursive function may survive this. So it
looks like there are currently only two positive effects:
1. NULL lists are returned as empty (#4596).
2. Functional tests are slightly more fast. Very slightly. Checked for Release
build for test/functional/eval tests because benchmarking of debug mode is
not very useful.
Adds two undocumented v: variables: _null_list and _null_dict because I do not
know a reproducible way to get such lists (though I think I heard about this)
and dictionaries (do not remember hearing about them). NULL strings are obtained
using $XXX_UNEXISTENT_VAR_XXX.
Fixes crash in json_encode($XXX_UNEXISTENT_VAR_XXX). Other added tests worked
fine before this commit.
