970285 - Crash [@ __memcpy_ssse3_rep] through [@ js::SizedTypeRepresentation::initInstance]

Status: VERIFIED FIXED

(Core :: JavaScript Engine, defect)

Description

[Christian Holler (:decoder)]

The following testcase crashes on mozilla-central revision ecf20a2484b6 (run with --fuzzing-safe):


var AA = TypedObject.uint8.array(2147483647).array();
var aa = new AA(-1);

Comment 1

[Christian Holler (:decoder)]

Attachment: [crash-signature] Machine-readable crash signature

Comment 2

[Christian Holler (:decoder)]

This looks like a bad write to me:

Program received signal SIGSEGV, Segmentation fault.
__memcpy_ssse3_rep () at ../sysdeps/i386/i686/multiarch/memcpy-ssse3-rep.S:1310
#0  __memcpy_ssse3_rep () at ../sysdeps/i386/i686/multiarch/memcpy-ssse3-rep.S:1310
#1  0x080cd376 in js::SizedTypeRepresentation::initInstance (this=0x9500a38, rt=0x9471ff8, mem=0x776ff008 "", length=4294967295) at /usr/include/bits/string3.h:52
#2  0x080db58a in js::TypedObject::createZeroed (cx=0x9488e98, descr=..., length=-1) at js/src/builtin/TypedObject.cpp:2223
#3  0x08103b85 in js::TypedObject::construct (cx=0x9488e98, argc=1, vp=0x9482e10) at js/src/builtin/TypedObject.cpp:2269
#4  0x0858dbd6 in js::CallJSNative (cx=0x9488e98, native=0x8103ac0 <js::TypedObject::construct(JSContext*, unsigned int, JS::Value*)>, args=...) at js/src/jscntxtinlines.h:220
#5  0x0859042b in js::CallJSNativeConstructor (cx=0x9488e98, native=0x8103ac0 <js::TypedObject::construct(JSContext*, unsigned int, JS::Value*)>, args=...) at js/src/jscntxtinlines.h:253
#6  0x0857e213 in js::InvokeConstructor (cx=0x9488e98, args=...) at js/src/vm/Interpreter.cpp:573
#7  0x08575e47 in Interpret (cx=0x9488e98, state=...) at js/src/vm/Interpreter.cpp:2605
eax     0x77800021      2004877345
ebx     0xf7e5cff4      -135933964
ecx     0x7fefef66      2146430822
edx     0xf77fffa0      -142606432
esi     0xf76ff007      -143659001
edi     0x2     2
ebp     0xffffbfc8      4294950856
esp     0xffffbf78      4294950776
eip     0xf7df1186 <__memcpy_ssse3_rep+3510>
=> 0xf7df1186 <__memcpy_ssse3_rep+3510>:        movntdq %xmm6,0x60(%edx)


Marking s-s and sec-critical for now, based on the trace.

Comment 3

[Niko Matsakis [:nmatsakis]]

Goofy. Looks like it'll be an easy fix.

Comment 4

[Niko Matsakis [:nmatsakis]]

Attachment: Bug970285.diff

Comment 5

[Shu-yu Guo [:shu]]

Comment on attachment 8375176 [details] [diff] [review]
Bug970285.diff

Review of attachment 8375176 [details] [diff] [review]:
-----------------------------------------------------------------

derp

::: js/src/builtin/TypedObject.cpp
@@ +2257,5 @@
>        case TypeRepresentation::UnsizedArray:
>          // First argument is a length.
> +        if (nextArg >= argc
> +            || !args[nextArg].isInt32()
> +            || args[nextArg].toInt32() < 0)

Style nit: || at end of each line instead of beginning

Comment 6

[Niko Matsakis [:nmatsakis]]

(In reply to Shu-yu Guo [:shu] from comment #5)
> Style nit: || at end of each line instead of beginning

That's how I prefer it anyway... thought I saw some other code doing it at the beginning.

Comment 7

[Al Billings [:abillings - ex-MoCo]]

This doesn't need sec-approval+ to go in since it only affects trunk. Just get it in.

Comment 8

[Christian Holler (:decoder)]

JSBugMon: Bisection requested, result:
=== Tinderbox Build Bisection Results by autoBisect ===

The "good" changeset has the timestamp "20140220102129" and the hash "7a5cbe4dadf8".
The "bad" changeset has the timestamp "20140220102430" and the hash "cc73b1f7a47d".

Likely regression window: https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=7a5cbe4dadf8&tochange=cc73b1f7a47d

Comment 9

[Niko Matsakis [:nmatsakis]]

Argh, forgot I didn't land this. Do so right now.

Comment 10

[Niko Matsakis [:nmatsakis]]

Actually I think I fixed this as part of bug 898356, but I'd still like to land the test.

Comment 11

[Niko Matsakis [:nmatsakis]]

My mistake, not fixed. Fixing now.

Comment 12

[Niko Matsakis [:nmatsakis]]

Attachment: Bug970285.diff

Rebased. Carrying over r+.

Comment 13

[Niko Matsakis [:nmatsakis]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/afbf1c788ada

Comment 14

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/afbf1c788ada

Comment 15

[Christian Holler (:decoder)]

JSBugMon: This bug has been automatically verified fixed.