Commit Graph

7158 Commits

Author SHA1 Message Date
Hugo Logmans ed5bf72bcf Fix missing dependency on mruby-print to support 'puts' 2016-02-03 15:05:44 +01:00
Yukihiro "Matz" Matsumoto ce84e42f41 Merge pull request #3093 from retrage/retrage/dev
Add String#rjust to mruby-string-ext
2016-01-31 11:41:01 +09:00
Yukihiro "Matz" Matsumoto ac5649700c Merge pull request #3092 from kenhys/fix-suported
Fix a typo
2016-01-31 11:26:20 +09:00
HAYASHI Kentaro 9442cb9652 Fix a typo
su ported ->
supported
  ^
2016-01-31 00:48:12 +09:00
Akira Moroo cb1a80e741 Add String#rjust test to mruby-string-ext 2016-01-30 14:28:00 +09:00
Akira Moroo e86bc6bb47 Add String#rjust to mruby-string-ext 2016-01-30 14:25:53 +09:00
Yukihiro "Matz" Matsumoto c35a25c51e Merge pull request #3091 from visualsayed/fix_hash_replace_method
protect NoMethodError from calling to_hash in replace
2016-01-28 12:17:51 +09:00
Sayed Abdelhaleem f5f48d9400 protect NoMethodError from calling to_hash in replace 2016-01-27 21:17:33 +02:00
Yukihiro "Matz" Matsumoto 2723b10a01 Merge pull request #3090 from kou/fix-segv-by-stack-extension-in-mrb-get-args
Fix SEGV by stack extension in mrb_get_args()
2016-01-22 00:50:15 +09:00
Kouhei Sutou c77123d20a Fix SEGV by stack extension in mrb_get_args()
mrb_get_args() keeps pointer of the current stack. But address of the
current stack maybe changed by method call.

'i' format character calls #to_i when the argument isn't integer but
has #to_i.

Here is a code that may call #to_i in mrb_get_args():

    case 'i':
      // ...
            default:
              *p = mrb_fixnum(mrb_Integer(mrb, ARGV[arg_i]));
              break;
     // ...

Here is a code #to_i is called:

    class X
      def initialize(i)
        @i = i
      end

      def to_i
        @i
      end
    end

    [][X.new(0), 0] # X#to_i is called

So, mrb_get_args() shouldn't keep pointer and use it. mrb_get_args()
should always refer mrb->ci->stack to use valid address of the current
stack.
2016-01-22 00:20:00 +09:00
Yukihiro "Matz" Matsumoto 8a74e68552 mruby-random: fixed wrong fixnum conversion 2016-01-21 03:31:08 +09:00
Yukihiro "Matz" Matsumoto 8d78c49e9a Merge pull request #3088 from maclover7/patch-1
Small grammar fix
2016-01-20 15:14:46 +09:00
Jon Moss 793a62954e Small grammar fix
[ci skip]
2016-01-19 22:37:44 -05:00
Yukihiro "Matz" Matsumoto 17bd40a566 Merge pull request #3087 from kou/fix-segv-on-rerasing-no-memory-error
Fix SEGV on re-raising NoMemoryError
2016-01-19 17:36:51 +09:00
Kouhei Sutou 1d84b3205a Fix SEGV on re-raising NoMemoryError
Think about the following Ruby script:

segv.rb:

    begin
      lambda do
        lambda do
          "x" * 1000 # NoMemoryError
        end.call
      end.call
    rescue
      raise
    end

If memory can't allocate after `"x" * 1000`, mruby crashes.

Because L_RAISE: block in mrb_vm_exec() calls mrb_env_unshare() via
cipop() and mrb_env_unshare() uses allocated memory without NULL check:

L_RAISE: block:

    L_RAISE:
      // ...
      while (ci[0].ridx == ci[-1].ridx) {
        cipop(mrb);
        // ...
      }

cipop():

    static void
    cipop(mrb_state *mrb)
    {
      struct mrb_context *c = mrb->c;

      if (c->ci->env) {
        mrb_env_unshare(mrb, c->ci->env);
      }

      c->ci--;
    }

mrb_env_unshare():

    MRB_API void
    mrb_env_unshare(mrb_state *mrb, struct REnv *e)
    {
      size_t len = (size_t)MRB_ENV_STACK_LEN(e);
      // p is NULL in this case
      mrb_value *p = (mrb_value *)mrb_malloc(mrb, sizeof(mrb_value)*len);

      MRB_ENV_UNSHARE_STACK(e);
      if (len > 0) {
        stack_copy(p, e->stack, len); // p is NULL but used. It causes SEGV.
      }
      e->stack = p;
      mrb_write_barrier(mrb, (struct RBasic *)e);
    }

To solve the SEGV, this change always raises NoMemoryError even when
realloc() is failed after the first NoMemoryError in
mrb_realloc(). mrb_unv_unshare() doesn't need to check NULL with this
change.

But it causes infinite loop in the following while:

    L_RAISE:
      // ...
      while (ci[0].ridx == ci[-1].ridx) {
        cipop(mrb);
        // ...
      }

Because cipop() never pops ci.

This change includes cipop() change. The change pops ci even when
mrb_unv_unshare() is failed by NoMemoryError.

This case can be reproduced by the following program:

    #include <stdlib.h>
    #include <mruby.h>
    #include <mruby/compile.h>

    static void *
    allocf(mrb_state *mrb, void *ptr, size_t size, void *ud)
    {
      static mrb_bool always_fail = FALSE;

      if (size == 1001) {
        always_fail = TRUE;
      }
      if (always_fail) {
        return NULL;
      }

      if (size == 0) {
        free(ptr);
        return NULL;
      } else {
        return realloc(ptr, size);
      }
    }

    int
    main(int argc, char **argv)
    {
      mrb_state *mrb;
      mrbc_context *c;
      FILE *file;

      mrb = mrb_open_allocf(allocf, NULL);
      c = mrbc_context_new(mrb);
      file = fopen(argv[1], "r");
      mrb_load_file_cxt(mrb, file, c);
      fclose(file);
      mrbc_context_free(mrb, c);
      mrb_close(mrb);

      return EXIT_SUCCESS;
    }

Try the following command lines:

    % cc -I include -L build/host/lib -O0 -g3 -o no-memory no-memory.c -lmruby -lm
    % ./no-memory segv.rb
2016-01-19 16:37:42 +09:00
Yukihiro "Matz" Matsumoto 70d24a9e5f Merge pull request #3086 from pra85/2016
Update license year range to 2016
2016-01-18 13:08:17 +09:00
Prayag Verma 70283a94a5 Update license year range to 2016 2016-01-18 08:55:12 +05:30
Yukihiro "Matz" Matsumoto 2e49a13163 Merge pull request #3084 from syohex/all-zeros
Fix passing all zero string to Kernel#Integer
2016-01-14 16:38:58 +09:00
Syohei YOSHIDA 787685a0b9 Fix all zero string case 2016-01-14 11:07:07 +09:00
Syohei YOSHIDA 2a661ac612 Add tests which passes all zero string to Kernel#Integer 2016-01-14 11:05:03 +09:00
Yukihiro "Matz" Matsumoto 4924856949 Merge pull request #3082 from jefffederman/feature/install-mruby-with-ruby-install
Update README to mention ruby version management tools
2016-01-14 00:49:25 +09:00
Jeff Federman d2961203b5 Update README to mention ruby version management tools 2016-01-13 10:30:15 -05:00
Yukihiro "Matz" Matsumoto 2f7efda346 Merge pull request #3081 from val00274/fix_toolchains_params
Fix build error in ruby 1.8.
2016-01-13 17:02:38 +09:00
asatou d0668cf9c9 Fix build error in ruby 1.8.
Apply change of #2978 to :clang, :gcc, :visualcpp toolchains
2016-01-13 16:02:58 +09:00
Yukihiro "Matz" Matsumoto 8143285956 Merge pull request #3080 from kou/fix-class-variable-in-module
Fix class variable reference in module
2016-01-11 10:00:34 +09:00
Kouhei Sutou ad492eb9f3 Fix class variable reference in module
Fix #3079
2016-01-11 00:34:15 +09:00
Yukihiro "Matz" Matsumoto fb2e0b721f fixed a problem with upvar access from instance_eval; ref #3072 2016-01-08 09:07:58 +09:00
Yukihiro "Matz" Matsumoto 0000192710 Merge pull request #3077 from sgnr/output-backtrace-segfault
Fix segfault on mrb_exc_backtrace.
2016-01-08 08:40:30 +09:00
Simon Génier 146a9eeb38 Fix segfault on mrb_exc_backtrace.
The code to iterate over backtrace locations was changed in #3065, but
unfortunately output_backtrace was not correctly updated to forward the
callback.
2016-01-07 11:24:16 -05:00
Yukihiro "Matz" Matsumoto f3cce24035 replace mrb_toplevel_run() by mrb_top_run() 2016-01-07 22:48:22 +09:00
Yukihiro "Matz" Matsumoto 258cb2093c mruby-eval: fixed receiver value in eval context; close #3072 2016-01-07 22:37:58 +09:00
Yukihiro "Matz" Matsumoto f7afe1d82a change mrb_run related API names; compatibility macros provided 2016-01-07 22:36:29 +09:00
Yukihiro "Matz" Matsumoto 725b3ca5a7 move KHASH_DECLARE(ht..) to mruby/hash.h; close #3073 2016-01-07 12:46:38 +09:00
Yukihiro "Matz" Matsumoto 7c82bfa7fe mruby-sprintf to use mrb_int formatting macros; ref #3076 2016-01-07 12:23:17 +09:00
Yukihiro "Matz" Matsumoto 3cf5af681a Merge pull request #3074 from syohex/precision-parameter-type
printf precision parameter must be 'int' type
2016-01-07 11:19:43 +09:00
Syohei YOSHIDA 931cc0359a printf precision parameter must be 'int' type
There is a problem when MRB_INT64 is enabled.
2016-01-07 11:06:20 +09:00
Yukihiro "Matz" Matsumoto ec711914a4 Merge pull request #3071 from ksss/symbol
symname_p support `!~`
2016-01-06 13:45:26 +09:00
Yukihiro "Matz" Matsumoto 20ebe9c5eb Merge pull request #3070 from kazuho/kazuho/printf-specifiers
provide macro to ease using printf mrb_int
2016-01-06 12:17:05 +09:00
Kazuho Oku cd131d18a7 define MRB_PRI? that map to PRI?NN depending on the size of mrb_int 2016-01-06 12:00:56 +09:00
ksss af21b365d2 symname_p support !~ 2016-01-06 11:59:09 +09:00
Yukihiro "Matz" Matsumoto 9145aed85e bytes2chars() conversion to fail if target byte offset is not on the character boundary; ref #3067
that means String#index matches first byte of a multi-byte character. this behavior is different
from CRuby, but a compromise for mruby which does not have encoding stuffs.
2016-01-05 18:05:25 +09:00
Yukihiro "Matz" Matsumoto aec825a64c stack_extend before eval_under() 2016-01-04 10:22:38 +09:00
Yukihiro "Matz" Matsumoto aa1f668b80 instance_eval should pass the receiver as a block parameter; close #3029 2016-01-02 22:01:00 +09:00
Yukihiro "Matz" Matsumoto 879e27d0f1 mirb: enlarge code buffer size to 4KB 2016-01-02 16:58:42 +09:00
Yukihiro "Matz" Matsumoto bd462c5edc mruby-fiber: fiber_switch() to use nesting VM when it's called from C API or mrb_funcall(); close #3056 2016-01-02 13:48:45 +09:00
Yukihiro "Matz" Matsumoto 4683b89b84 Merge pull request #3068 from kazuho/kazuho/osx-_setjmp
use _setjmp/_longjmp on OS X
2016-01-01 10:15:55 +09:00
Kazuho Oku 121c6e30aa use _setjmp/_longjmp on other BSD flavors 2016-01-01 06:48:53 +09:00
Kazuho Oku f4bb329c74 use _setjmp/_longjmp on OS X 2016-01-01 06:24:49 +09:00
Yukihiro "Matz" Matsumoto 38b9c2d308 mruby-fiber: add Fiber.yield description; close #3066
Fiber.yield cannot be called from #initialize which is called by
mrb_funcall().  It is mruby limitation.
2015-12-31 21:25:53 +09:00
Yukihiro "Matz" Matsumoto 31b8469bff Merge pull request #3067 from ksss/use-memchr
Use memchr for performance
2015-12-31 21:20:51 +09:00