summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
* Security: do not require executable stack.Kaz Kylheku2016-04-291-0/+12
| | | | | | | | | Ever since jmp.S was added, the txr executable has been building with a bit in the program header indicating that the stack is to be executable. * jmp.S: Define the .note.GNU-stack section to tell the linker we don't need an executable stack.
* Use random padding in PRNG rather than 0xAA.Kaz Kylheku2016-04-283-62/+97
| | | | | | | | | | | | | | | | | | | | | | | | The purpose is to eliminate any biases in the PRNG arising out of the regularity of that pattern, so that the behavior of successive values is good from the beginning. This doesn't solve the problem that a short warm-up period leads to a poor distribution of initial values relative to the seed space. In other words, that similar seeds lead to initially similar sequences. * rand.c (rand_tab): New static array. (make_random_state): Set uninitialized parts of state from the corresponding elements in rand_tab, rather than to the 0xAAAAAAAA values. (rand_compat_fixup): In 139 compatibility mode, clobber rand_tab with 0xAA bytes. * tests/013/maze.expected: Updated. * txr.1: Added some PRNG implementation notes, and also compatibility notes.
* Fix broken bignum to int_ptr_t conversion.Kaz Kylheku2016-04-281-3/+2
| | | | | | | | | | | | | This one affects all platforms. The extra sign check and negation cancels out the one done in mp_get_uintptr, causing a positive value for a negative input value. * mpi/mpi.c (mp_get_intptr): Just coerce the uint_ptr_t result to int_ptr_t. That has the right semantics under that the bits are preserved (under two's complement, in every compiler I've ever used). The unsigned value from mp_get_uintptr already looks like the image of a two's complement value.
* Fix broken bignum to uint_ptr_t conversion.Kaz Kylheku2016-04-281-1/+1
| | | | | | | | | This applies only to some platforms, none of which are current targets on which TXR is tested and released. * mpi/mpi.c (mp_get_uinptr): Do not clobber output accumulator in the loop body: the bits must be OR-ed into it.
* Clean-up rand code; implement configurable warmup.Kaz Kylheku2016-04-284-30/+92
| | | | | | | | | | | | | | | | | | | | | * lib.c (copy): Pass new second argument to make_random_state. * rand.c (random_warmup): New macro. (random_warmup_s): New symbol variable. (make_random_state): New warmup argument. Code restructured to avoid copy flag. Fill code and initialization of r->cur made common. (rand_compat_fixup): Pass second argument to make_random_state. (rand_init): Pass second argument to make_random_state. Initialize random_warmup_s with interned symbol. Register *random-warmup* special variable. Fix registration of make-random-state to two-argument function type. * rand.h (make_random_state): Declaration updated. * txr.1: Document *random-warmup* variable and warup argument of make-random-state.
* Only use getpid for rand seed if HAVE_UNISTD_H.Kaz Kylheku2016-04-281-0/+2
| | | | | * rand.c (make_random_state): Don't use getpid if we aren't on something that looks like Unix.
* Remove bogus code for 128 bit platform.Kaz Kylheku2016-04-281-6/+3
| | | | | | * rand.c (make_random_state): Issue #error if SIZEOF_PTR exceeds 8, rather than including bogus code that neglects to use the remaining upper 32 bits.
* Move PRNG-related initialization to rand.cKaz Kylheku2016-04-282-8/+9
| | | | | | | | | * eval.c (eval_init): Remove registration of make-random-state, random-state-get-vec, random-state-p, random-fixnup, random, and rand. * rand.c (rand_init): Move registrations removed from eval_init here.
* Simple indirection on pattern functions: @(call).Kaz Kylheku2016-04-273-1/+150
| | | | | | | | | | * eval.h (call_s): Declared. * match.c (h_call, v_call): New static function. (dir_tables_init): Register v_call in v_directive_table under call symbol. Likewise h_call in h_directive_table. * txr.1: Documented.
* Diagnose empty clauses better in parallel constructs.Kaz Kylheku2016-04-271-86/+119
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In this change we allow the grammar, as such, to express empty clauses in the parallel constructs like @(some), @(cases), @(gather) and so forth. However, we raise an error when these occur. This results in a cleaner diagnostic behavior. In the future, empty clauses may be allowed; the semantics has to be worked out. An empty clause should neither succeed nor fail; the behavior should be as if it is not there. The general strategy in this patch is to eliminate the use of the clauses terminal symbol (completely) and use clauses_opt everywhere, and then diagnose when that produces a nil. Some elems are similarly changed to elems_opt in the horizontal versions of directives. * parser.y (clauses): Nonterminal symbol completely removed. (spec): Empty production rule removed, by using clauses_opt instead of clauses. (all_clause, some_clause, none_clause, maybe_clause, cases_clause, choose_clause, gather_clause): Instead of diagnosing empty clauses via a dedicated grammar rule, diagnose by looking for nil subtree emerging from from a clauses phrase. (gather_parts): Use clauses_opt. Yield nil instead of a cons if this clauses_opt is nil. (additional_gather_parts): Simplify grammar and actions by recursing back to gather_parts. If gather_parts produces nil, diagnose the empty and/or subclause situation. (collect_clause): Use clauses_opt and diagnose empty. Simplify error production, which doesn't have to look at the lookahead token yychar any more to implement the empty diagnostic. (clause_parts): Use clauses_opt instad of clauses. Yield nil instead of a cons if clause_opt yields nil. (additional_parts): Grammar and actions simplified by recursing back to clause_parts. If the recursive clause_parts produces nil, generate empty and/or subclause diagnostic. (elem): Use elems_opt for the various horizontal clauses and implement empty checks for them. Explicit empty productions are eliminated. (clause_parts_h): Changed in way analogous to clause_parts. (additional_parts_h): Changed analogously to additional_parts. (try_clause): Use clauses_opt and catch empties. Eliminate the second error production. (catch_clauses_opt): Diagnose empty catch and finally, unless in compatibility mode with 139 or less. The grammar already uses clauses_opt here; empty catches and finally had been allowed. Fixed messages in error productions for catch and finally to refer to catch and finally rather than try.
* Adding fixnum-min and fixnum-max variables.Kaz Kylheku2016-04-272-0/+17
| | | | | | | * arith.c (arith_init): Registered fixnum-min and fixnum-max variables. * txr.1: Documented.
* Check .metn similarly to .codn and fix.Kaz Kylheku2016-04-262-30/+31
| | | | | | * checkman.txr: Check .metn using logic shared with .codn. * txr.1: Numerous errors fixed.
* Change .meta to single-argument.Kaz Kylheku2016-04-262-10/+10
| | | | | | | | * checkman.txr: Check .meta similarly to .code. * txr.1: Change .meta macro to take only one argument, so that text with spaces must be quoted. Errors found and fixed.
* Check that arg 2 of .codn starts with punctuation.Kaz Kylheku2016-04-262-5/+8
| | | | | | | | * checkman.txr: Constraint on first character of second argument coded as a require constraint, enforced with assert. * txr.1: Newly discovered errors fixed.
* Change .codn and .code to fixed arity.Kaz Kylheku2016-04-262-427/+471
| | | | | | | | | | | | | | | | | | | In the manual, the .code and .codn macros are no longer variadic. They take one an two arguments, respectively. Multiple arguments have to be quoted now. This makes it easier to detect errors in usage, using support in checkman.txr. Numerous errors were found, in fact. * checkman.txr (check-code): New pattern function for doing some checks on .code, codn, cod2 and cod3 macro usage. * txr.1: code and codn macro are no longer variadic. All variadic uses fixes with quotes. Numerous errors discovered and corrected.
* Fix errors found by new checkman.txr.Kaz Kylheku2016-04-261-21/+21
| | | | | | * txr.1: Fix incorrect function and variable description headings with spurious commas and inappropriate plurals. Fix missing .mets in syntax blocks.
* Adding manual page checker.Kaz Kylheku2016-04-262-1/+80
| | | | | | | * Makefile (txr-manpage.pdf): Invoke checkman.txr on the man page source. * checkman.txr: New file.
* Check man page for errors when generating PDF.Kaz Kylheku2016-04-261-1/+1
| | | | | | * Makefile (txr-manpage.pdf): Add -ww option to pdfroff to report all errors, such as uses of undefined macros.
* Pretty print filename in assertion.Kaz Kylheku2016-04-251-2/+2
| | | | | * match.c (v_assert, h_assert): Use ~a rather than ~s for c->curfile in assert message.
* Version 139.txr-139Kaz Kylheku2016-04-226-428/+471
| | | | | | | | | | * RELNOTES: Updated. * configure, txr.1: Bumped version and date. * share/txr/stdlib/ver.tl: Likewise. * txr.vim, tl.vim: Regenerated.
* Spring cleaning: find wrong wrong troff macros.Kaz Kylheku2016-04-221-60/+61
| | | | | | * txr.1: Hunted down and fixed incorrect .code, .codn, .meta and .metn invocations. Added missing .desc macros after .coNP Variables ...
* Recycle temporary list-of-lists in mapping functions.Kaz Kylheku2016-04-221-3/+9
| | | | | | | | * eval.c (mapcarv, mappendv, mapdov): When done, we can recycle the conses used for the temporary copy of the list-of-lists, whose car-s are used for iterating over the lists in paralle. This is safe because the temporary list's conses aren't shared with any other function.
* Support list of paths in ftw.Kaz Kylheku2016-04-222-0/+59
| | | | | | | * ftw.c (ftw_wrap): Handle case when dirpath is a list, by recursion. * txr.1: Documented.
* Optimization in path-private-to-me-p.Kaz Kylheku2016-04-221-8/+8
| | | | | | | | | * share/txr/stdlib/path-test.tl (path-private-to-me-p): Use mlet to initialize the g variable lazily, so that getgrgid is only called when g is used. getgrid is a fairly heavy-weight function, and it's only used to resolve the case when a file is writable to the group owner (is the user the only member of that group?)
* Don't quote path in load and @(load) diagnostics.Kaz Kylheku2016-04-222-6/+6
| | | | | | | * eval.c (sys_load): Use ~a format specifier rather than ~s for the load path (if already known to be a string). * match.c (v_load): Likewise.
* Strengthen against resource leaks upon exceptions.Kaz Kylheku2016-04-216-23/+44
| | | | | | | | | | | | | | | | | | | | | | | * glob.c (glob_wrap): Perform argument conversions that might throw before allocating UTF-8 string. * parser.y (text): In the action for SPACE, the lexeme is not needed so free($1) right away. If regex_compile were to throw an exception, that lexeme will leak. * socket.c (getaddrinfo_wrap): Harden against leakage of node_u8 and service_u8 strings with an unwind block. For instance, the hints structure could contain bad values which cause addrinfo_in to throw. * stream.c (make_string_byte_input_stream): Perform possibly throwing argument conversions before allocating resources. * sysif.c (mkdir_wrap, mknod_wrap, chmod_wrap, symlink_wrap, link_wrap, setenv_wrap, crypt_wrap): Likewise. * syslog.c (openlog_wrap, syslog_wrapv): Likewise.
* Rename misleadingly named out_str_pretty function.Kaz Kylheku2016-04-211-5/+5
| | | | | | | | * lib.c (out_str_pretty): Rename to out_str_readable. This function is called from obj_print_impl precisely when the pretty flag is false: pretty means output textual things as the text they denote, rather than machine-readable Lisp. (out_lazy_str, obj_print_impl): Calls updated.
* Harmonize rules for string and character printing.Kaz Kylheku2016-04-211-9/+24
| | | | | | | | | | | | | | | | | In this patch we change which characters objects are printed using hex escapes, and which characters are printed as hex when printing string literals. * lib.c (obj_print_impl): Add DEL (U+7F) to the list of character objects which are printed as hex. In a string literal, it's already printed as \x7F. Use upper case hex rather than lower case. (out_str_char): Copy the rules used by obj_print_impl for deciding what string constituents to print as hex and how to print it. So for instance, U+80 to U+A0 will now print in hex as well as the U+D800 to U+DFFF range, rather than just U+DC00 to U+DCFF, and the BOM code U+FFFE, U+FFFF and anything higher.
* Handle non-UTF-8 byte in regex scanned from string.Kaz Kylheku2016-04-211-0/+6
| | | | | | | | The current behavior is that there is no lex rule for this, so such a byte gets echoed. parser.l (grammar): Add fallback rule to match one byte in SREGEX state and turn it into 0xDCxx character.
* Regex syntax errors trigger exception.Kaz Kylheku2016-04-211-1/+5
| | | | | * parser.c (regex_parse): Don't return nil if there were errors; throw a syntax-error exception.
* Diagnose out-of-range char to UTF-8 conversion.Kaz Kylheku2016-04-211-1/+3
| | | | | * utf8.c (utf8_encode): Don't return zero in the out-of-range case, but throw.
* Better job of diagnosing out-of-range char escapes.Kaz Kylheku2016-04-211-2/+9
| | | | | | * parser.l (num_esc): Check for converted value being out of the range of wchar_t or beyond 0x10FFFF, whichever is less.
* Integrating fnmatch.Kaz Kylheku2016-04-203-0/+127
| | | | | | | | | | | | * configure: new test for fnmatch, introducing HAVE_FNMATCH configure variable into config/config.h. * sysif.c (fnmatch_wrap): New function. (sysif_init): Register intrinsic variables fnm-pathname, fnm-noescape, fnm-period, fnm-leading-dir, fnm-casefold and fnm-extmatch. Register intrinsic function fnmatch. * txr.1: Documented.
* linenoise: g++ signed/unsigned warnings.Kaz Kylheku2016-04-201-4/+5
| | | | | | | | * linenoise/linenoise.c (free_completions): Use size_t for iterating over completions vector. (complete_line): Likewise. (edit_insert, edit_insert_str): Cast size_t value to int before comparing with l->dlen.
* ftw.c affected by g++/glibc issue __USE_GNU is defined.Kaz Kylheku2016-04-201-0/+2
| | | | | | | | | When compiling as C++, __USE_GNU is defined even though we use -ansi and don't specify -D_GNU_SOURCE anywhere. Glibc headers don't seem to respond to the C++ ANSI options. ftw.c: Workaround for compiler warning: only define __USE_GNU if it is not defined already.
* Recycle conses in unget-char and read-until-match.Kaz Kylheku2016-04-203-5/+9
| | | | | | | | | | | | * regex.c (ead_until_match): Use rcyc_pop instead of pop to move the conses to the recycle list. We know these are not shared with anything. Adding additional logic to completely recycle the stack. * socket.c (dgram_get_char): Use rcyc_pop to get the character from the push-back list. * stream.c (stdio_get_char): Likewise.
* Mechanism for recycling conses outside of GC.Kaz Kylheku2016-04-203-2/+52
| | | | | | | | | | | | * lib.c (recycled_conses): New static variable. (rcyc_pop, rcyc_cons, rcyc_list, rcyc_empty): New functions. (cons): Take a recycled cons, if available. * lib.h (rcyc_pop, rcyc_cons, rcyc_list, rcyc_empty): Declared. * gc.c (gc): Call rcyc_empty to make recycle_list unreachable.
* Bugfix: optional arg defaulting in get-string.Kaz Kylheku2016-04-201-1/+2
| | | | | * stream.c (get_string): The optional stream argument must be treated accordingly.
* read-until-match can optionally keep matched text.Kaz Kylheku2016-04-205-38/+50
| | | | | | | | | | | | | | | | | | | | * regex.c (read_until_match): New argument, include_match. Three times repeated termination code refactored into block reached by forward goto. (regex_init): Registration of read-until-match updated. * regex.h (read_until_match): Declaration updated. * stream.c (struct record_adapter_base): New member, include_match. (record_adapter_get_line): Pass match to read_until_match as new argument. (record_adapater): New argument, include_match. (stream_init): Update registration of record-adapter. * stream.h (record_adapter): Declaration updated. * txr.1: Updated.
* Fix broken read_until_match.Kaz Kylheku2016-04-191-17/+51
| | | | | * regex.c (read_until_match): Completely rewrite broken, unsalvageable, garbage logic.
* Allow unlimited character pushback in unget-char.Kaz Kylheku2016-04-193-22/+25
| | | | | | | | | | | | | | | | Fixing read_until_match will require this feature. * socket.c (dgram_get_char): Treat unget_c as a cons-based stack; pop a character from it if available. (dgram_unget_char): Push the character onto unget_c rather than storing the characer into unget_c. * stream.c (stdio_get_char, stdio_unget_char): Closely analogous changes to the ones in dgram_get_char and dgram_unget_char, respectively. * txr.1: Documentation improved and updated.
* Incorrect format args in string stream code.Kaz Kylheku2016-04-191-1/+1
| | | | | * stream.c (string_in_unget_char): Missing argument in uw_throwf call.
* Fix broken unget_char over string input streams.Kaz Kylheku2016-04-191-1/+1
| | | | | | * stream.c (string_in_unget_char): Store the updated position into the stream, rather than the original value, which does nothing.
* Bugfix: allow newline in regex parsing from string.Kaz Kylheku2016-04-181-1/+7
| | | | | | | * parser.l (grammar): The newline character is incorrectly handled by the same rule under the SREGEX and REGEX states. In the SREGEX state, just return it as a REGCHAR, not forgetting to increment the line number.
* Trailing whitespace.Kaz Kylheku2016-04-181-1/+1
| | | | * parser.l: Remove trailing whitespace.
* Bugfix: @(if expr) not macro-expanding expr.Kaz Kylheku2016-04-181-2/+2
| | | | | * parser.y (if_clause, elif_clauses_opt): Add missing expand calls.
* Fix inconsistency of Lisp var visibility in TXR.Kaz Kylheku2016-04-172-41/+69
| | | | | | | | | | | | | | | | | | | Numerous places in match.c are using assoc(sym, bindings) logic to access a variable, which doesn't see the Lisp globals, as we would like. For example, if foo is defined using (defvar foo), @(set foo "A") doesn't work. This is subject to the compatibility option. * match.c (tx_lookup_var): New static function. (dest_set, h_var, h_coll, h_parallel, h_fun, v_next, v_parallel, v_gather, v_collect, v_flatten, v_cat, v_output, v_filter, v_fun, match_filter): Use tx_lookup_var instead of assoc for all lookups that see the full variable scope. Only variable lists known to be locally consed up are scanned with just assoc. * txr.1: Documented new rules and added compatibility notes.
* Bugfix: @(output) not expanding some Lisp exprs.Kaz Kylheku2016-04-171-1/+1
| | | | | | | * parser.y (make_expr): This function is part of a hack for converting some hard-coded syntax like @(if ...) or @(and ...) in an @(output) block to a to a Lisp expression. Alas, the crucial step of expanding the Lisp form was neglected.
* Valgrind support in continuations.Kaz Kylheku2016-04-161-1/+18
| | | | | | | * unwind.c (revive_cont): When scanning the stack of the revived continuation to fix up pointers, we can save the Valgrind validity bits of each word, mark it defined, and then restore the validity bits.
* Put notes into cadr sources that they are generated.Kaz Kylheku2016-04-164-0/+12
| | | | | | * gencadr.txr: Insert notice about generation. * cadr.c, cadr.h, share/txr/stdlib/cadr.tl: Regenerated.
generated by cgit v1.2.3 (git 2.25.1) at 2025-04-08 15:03:37 +0000