ref #3215
If a function (such as mrb_read_irep_file()) is declared without MRB_API
in header file (such as include/mruby/dump.h), implementation of the
function in source file (such as src/load.c) should also defined without
MRB_API.
If MRB_API is mismatch, Visual C++ reports link error with C2375 error
code: https://msdn.microsoft.com/en-us/library/5k6kw95a.aspx
mrb_vm_define_class() may realloc() mrb->c->stack because it calls
mrb_funcall() for inherited hook. If mrb->c->stack is realloc()-ed, regs
refers orphan address.
GitHub: fix#3122
It reverts #3126. #3126 fixes the segmentation fault but generates
broken backtrace.
This change fixes the segmentation fault and generates correct
backtrace. The strategy of this change is "generating backtrace while
irep is alive".
/tmp/test.rb:
def gen
e0 = nil
begin
1.times {
raise 'foobar'
}
rescue => e
e0 = e
end
e0
end
e = gen
GC.start
gen
GC.start
puts e.backtrace.join("\n")
Run:
% bin/mruby /tmp/test.rb
/tmp/test.rb:5:in Object.gen
/home/kou/work/ruby/mruby.kou/mrblib/numeric.rb:77:in Integral#times
/tmp/test.rb:4:in Object.gen
/tmp/test.rb:13
FYI:
% ruby -v /tmp/test.rb
ruby 2.3.0p0 (2015-12-25) [x86_64-linux-gnu]
/tmp/test.rb:5:in `block in gen'
/tmp/test.rb:4:in `times'
/tmp/test.rb:4:in `gen'
/tmp/test.rb:13:in `<main>'
It looks like the logic to reallocate the backtrace was flawed,
based on the wrong variable (loc_raw->i, which, as I have verified,
decreases from 16 to 0 instead of increasing)
I am not sure if this is the correct fix
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.
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
The code to iterate over backtrace locations was changed in #3065, but
unfortunately output_backtrace was not correctly updated to forward the
callback.
Yukihiro "Matz" Matsumoto