Project

General

Profile

Actions

Bug #9157

closed

rb_readlink() calls rb_str_modify_expand() too early

Added by nowacki (Maciek Nowacki) almost 11 years ago. Updated almost 11 years ago.

Status:
Closed
Assignee:
-
Target version:
ruby -v:
ruby 2.0.0p353 (2013-11-22) [x86_64-linux]
[ruby-core:58592]

Description

If rb_readlink() (file.c) is called on a long symlink, Ruby tends to crash. This is present in -trunk. The problem is that the rb_str_modify_expand() statement is incorrectly placed before the statement 'size*=2'. Here is a patch:

--- - 2013-11-25 22:10:59.694183795 -0700
+++ file.c 2013-11-25 22:10:03.076352889 -0700
@@ -2529,8 +2529,8 @@
|| (rv < 0 && errno == ERANGE) /* quirky behavior of GPFS */
#endif
) {

  • rb_str_modify_expand(v, size);
    size *= 2;
  • rb_str_modify_expand(v, size);
    }
    if (rv < 0) {
    rb_str_resize(v, 0);

Files

file.c.patch (310 Bytes) file.c.patch nowacki (Maciek Nowacki), 11/26/2013 03:12 PM

Related issues 1 (0 open1 closed)

Has duplicate Ruby master - Bug #9158: bug in rb_readlink() in file.c (fixed)ClosedActions
Actions #1

Updated by nobu (Nobuyoshi Nakada) almost 11 years ago

  • Status changed from Open to Closed
  • % Done changed from 0 to 100

This issue was solved with changeset r43853.
Maciek, thank you for reporting this issue.
Your contribution to Ruby is greatly appreciated.
May Ruby be with you.


file.c: fix buffer overflow

  • file.c (rb_readlink): fix buffer overflow on a long symlink. since
    rb_str_modify_expand() expands from its length but not its capacity,
    need to set the length properly for each expansion.
    [ruby-core:58592] [Bug #9157]

Updated by nobu (Nobuyoshi Nakada) almost 11 years ago

  • Backport changed from 1.9.3: UNKNOWN, 2.0.0: UNKNOWN to 1.9.3: DONTNEED, 2.0.0: REQUIRED

Updated by nowacki (Maciek Nowacki) almost 11 years ago

nobu (Nobuyoshi Nakada) wrote:

This issue was solved with changeset r43853.

Ah, I didn't realize that rb_str_modify_expand() takes a difference as its argument, not the total buffer length. This works because the function doubles the buffer size, which is the same as adding as much buffer capacity as is already present (size before *=2). My proposed fix simply made the problem less obvious. Interesting.

Updated by vpereira (Victor Pereira) almost 11 years ago

does it deserves a CVE?

Updated by nobu (Nobuyoshi Nakada) almost 11 years ago

No, just a usual bug which aborts by local filesystem access.

Updated by nagachika (Tomoyuki Chikanaga) almost 11 years ago

  • Backport changed from 1.9.3: DONTNEED, 2.0.0: REQUIRED to 1.9.3: DONTNEED, 2.0.0: DONE

r43853 was backported to ruby_2_0_0 branch at r43959.

Actions

Also available in: Atom PDF

Like0
Like0Like0Like0Like0Like0Like0