Don't set "defining" until the macro definition is successfully parsed
and we know for sure that we are going to define the macro.
Together with:
a762cd4e54 BR 3392668: preproc: test for macro in TOK_LOCAL_SYMBOL
... this addresses BR 3392668.
Reported-by: <puppet@zju.edu.cn>
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
TOK_LOCAL_SYMBOL is only applicable inside a macro; otherwise error
out just like we do for TOK_MMACRO_PARAM.
This *partially* addresses BR 3392668.
Reported-by: <puppet@zju.edu.cn>
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
Significantly improve the portability/reliability of a general
make command. Build a .pdf.xz version of the documentation,
using an uncompressed pdf as basis.
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
The binary mode has no difference from text mode in
POSIX-compliant operating systems. The two modes are
distinguishable from each other on Windows, and perhaps
on other systems as well.
The binary stream has scalability and other advantages.
Windows treats the standard input stream as text mode by
default. So the code changes it to binary mode.
Also, add a helper function, nasm_set_binary_mode(),
that is OS-agnostic, in the library.
Reported-by: Didier Stevens <didier.stevens@gmail.com>
Suggested-by: Didier Stevens <didier.stevens@gmail.com>
Link: https://bugzilla.nasm.us/show_bug.cgi?id=3392649
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Determining whether we should warn on defining a single-line macro, with a
name and a certain number of parameters, call a helper function,
smacro_defined(). It does not always return the address of the definition
structure.
Fix the code to be cautiously accessing the definition structure.
Fixes: e91f5cc132 ("preproc: fix %undef of macro aliases, and add
%ifdefalias")
Reported-by: Dale Curtis <dalecurtis@chromium.org>
Link: https://bugzilla.nasm.us/show_bug.cgi?id=3392659
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Mistreating the macro-parameter, just equivalent to the given
argument number, leads to casting an unnecessary error. Fix to
assemble the conditional code correctly.
Fixes: de7acc3a46 ("preproc: defer %00, %? and %??
expansion for nested macros, cleanups")
Reported-by: C. Masloch <pushbx@ulukai.org>
Link: https://bugzilla.nasm.us/show_bug.cgi?id=3392660
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
The code looked to be unintentionally always nullifying the
token pointer at first place in handling those macro-parameters.
Remove it to avoid segfault.
Fixes: de7acc3a46 ("preproc: defer %00, %? and %??
expansion for nested macros, cleanups")
Reported-by: C. Masloch <pushbx@ulukai.org>
Link: https://bugzilla.nasm.us/show_bug.cgi?id=3392640
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
The code has been fixed to print the corresponding line
numbers of %rep blocks correctly, but only for the first
iteration. For the subsequent iterations, the current line
number on the expansion needs to be explicitly reset again.
Fixes: ab6f831955 ("listing: when listing lines in macros
and rep blocks, show the actual line")
Reported-by: C. Masloch <pushbx@ulukai.org>
Link: https://bugzilla.nasm.us/show_bug.cgi?id=3392626
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
The binary mode has no difference from text mode in
POSIX-compliant operating systems. The two modes are
distinguishable from each other on Windows, and perhaps
on other systems as well.
The binary stream has scalability and other advantages.
Windows treats the standard input stream as text mode by
default. So the code changes it to binary mode.
Also, add a helper function, nasm_set_binary_mode(),
that is OS-agnostic, in the library.
Reported-by: Didier Stevens <didier.stevens@gmail.com>
Suggested-by: Didier Stevens <didier.stevens@gmail.com>
Link: https://bugzilla.nasm.us/show_bug.cgi?id=3392649
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
The code looked to be unintentionally always nullifying the
token pointer at first place in handling those macro-parameters.
Remove it to avoid segfault.
Fixes: de7acc3a46 ("preproc: defer %00, %? and %??
expansion for nested macros, cleanups")
Reported-by: C. Masloch <pushbx@ulukai.org>
Link: https://bugzilla.nasm.us/show_bug.cgi?id=3392640
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
The code has been fixed to print the corresponding line
numbers of %rep blocks correctly, but only for the first
iteration. For the subsequent iterations, the current line
number on the expansion needs to be explicitly reset again.
Fixes: ab6f831955 ("listing: when listing lines in macros
and rep blocks, show the actual line")
Reported-by: C. Masloch <pushbx@ulukai.org>
Link: https://bugzilla.nasm.us/show_bug.cgi?id=3392626
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Macro aliases can legitimately point to nonexistent
macros. Furthermore, %undef should remove the pointed-at macro, not
the alias. This led to an infinite loop in the existing code; fix
that.
Add an %ifdefalias directive to test for the existence of an alias.
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
Add the -Lw option to flush the list file after every line
output. This is handy for debugging if nasm hangs.
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
1. %xdefine was broken because the code used i as a loop, which is
a standard use for the name i. To avoid that confusion in the
future, use "op" rather than "i" to hold the directive constant.
2. Once (1) was fixed, the smacro expansion code would fail because of
parameter token numbers being indistinguishable between the ongoing
expansion and the %xdefine parameters. In a first pass, mark the
parameters with a new TOK_XDEF_PARAM token number, and change them
to proper parameter token numbers in a second pass, which is now
moved into define_smacro() which is where it arguably belongs.
3. Add a few tests for token pasting and xdefine with and without
parameters.
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
Technically, this is not necessary, because make_tok_qstr_len()
doesn't rely on NUL termination, and in fact it *can't*, since the
string might contain embedded NULs, but tacking on a NUL is good for
debugging if nothing else. That means reserving space for it!
Reported-by: C. Masloch <pushbx@ulukai.org>
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
The only way in which
%xdefine(xxx) yyyy zzzz
differs from
%define(xxx) yyyy %[zzzz]
is that in the former case macro arguments get preserved, even if
they are macros defined elsewhere. Revert to that behavior.
Reported-by: C. Masloch <pushbx@ulukai.org>
Link: https://bugzilla.nasm.us/show_bug.cgi?id=3392623
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
Add support for complex data (Dx) statement expressions involving both
initialized and uninitialized data. In addition, we have support for
overriding the size of each element on an individual item and/or list
basis.
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
%exitrep should should stop emitting code immediately, not just
terminate the loop when we hit %endrep. There is a bunch of hacky code
that special-cases that using istk->in_progress == 0.
The handling of the tail of %exitrep, %include and non-emitting
conditionals using entirely different mechanisms is just dumb. They
need to be unified.
Link: https://bugzilla.nasm.us/show_bug.cgi?id=3392612
Reported-by: Jason Hood <jadoxa@yahoo.com.au>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
It turns out that we need a Fontmap file after all, *and* -I. to make
gs find it. Inconsistent results came from stray Fontmap files from
previous debug attempts.
Now generate both fontpath and Fontmap, and hopefully at least one of
them should work. We might, in fact, need both, one for gs to know
where the files are and one for gs to know it is allowed to read them.
The core problem seems to be that gs will find OTF fonts by its normal
discovery mechanisms, but for some reason don't seem to use them
unless it can find them in a Fontmap, Font directory, of CIDFont
directory.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
We need to create a separate paragraph if the help text had used \c
anyway. Putting the enabled/disabled separately for all entries makes
it read a lot cleaner anyway.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
rdsrc.pl requires blank lines around \c paragraph, but warnings.pl
would strip them. Create a *!- prefix to force a blank line.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
config/config.h as generated by autoconf has #undef directives
commented out, but the autoheader-generated template contains them,
and config/unconfig.h should not contain them. Re-introduce
config/config.h.in, and postprocess it to generate config/unconfig.h
by commenting out all the #undef directives, just as configure does.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
<config/unconfig.h> is generated by autoconf (autoheader), but just
like the autoconf/helpers directory contains files for people who
don't have the proper tools, check this file in.
It should even be less annoying change-wise than the helpers...
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
\\?\ is supposed to override the hard-coded path limit, but it has
other effects as well, such as not working with relative paths and
paths containing /. On Windows 10 it is possible to set a registry key
to override this option anyway.
Maybe one day we can just use fopen() like on normal systems, even...
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Link: https://bugzilla.nasm.us/show_bug.cgi?id=3392614
Reported-by: Iouri Kharon <bc-info@styx.cabel.net>
The previous code would fail to process any directive if the directive
list was NULL. However, we also need to process the default name
passed to search_pragma_list() (e.g. "elf32"), as well as the global
name (e.g. "output") and call the default handler in that case.
In the process, improve the handling such that if one handler returns
DIRR_UNKNOWN, try calling subsequent handlers in the list.
Finally, factor out as much as possible to generic handlers.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
The inlining of preprocessor tokens has benchmarked about 15% faster in any
realistic scenario than the old code. For an artificial case where
over 50% of symbols are longer than 47 characters there was a slowdown
of about 15%, but that represents an incredibly unlikely scenario in
practice: even if labels are that long (e.g. C++ name mangling) the
majority of tokens should be instructions and registers, which are
uniformly short.
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
