Chakracore: Heap overflow in CreateTTDDirectoryAsNeeded

[rajatd]

Repro:
build chakracore on linux:
./build.sh --cc=/usr/bin/clang --cxx=/usr/bin/clang++ --arch=amd64 --debug --static -j 8 --sanitize=address,undefined,signed-integer-overflow

run ./ch - from /out/Debug/bin/ch

output:

==67392==ERROR: AddressSanitizer: heap-buffer-overflow on address 0x60200000eec4 at pc 0x7f4ec99064bf bp 0x7ffcc6c0ab30 sp 0x7ffcc6c0ab28
READ of size 2 at 0x60200000eec4 thread T0
#0 0x7f4ec99064be in PAL_wcslen /home/chakraut/ChakraCore/pal/src/cruntime/wchar.cpp:1212:12
#1 0x7f4ec9845b78 in Helpers::CreateTTDDirectoryAsNeeded(unsigned long, char, char const, char16_t const) /home/chakraut/ChakraCore/bin/ch/Helpers.cpp:440:41
#2 0x7f4ec98462a0 in Helpers::GetTTDDirectory(char16_t const, unsigned long, char*, unsigned long) /home/chakraut/ChakraCore/bin/ch/Helpers.cpp:499:5
#3 0x7f4ec9834a79 in main /home/chakraut/ChakraCore/bin/ch/ch.cpp:986:13
#4 0x7f4ec776c2b0 in __libc_start_main (/lib/x86_64-linux-gnu/libc.so.6+0x202b0)
#5 0x7f4ec975cf79 in _start (/home/chakraut/ChakraCore/out/Debug/bin/ch/ch+0xb60f79)

0x60200000eec4 is located 12 bytes to the left of 10-byte region [0x60200000eed0,0x60200000eeda)
allocated by thread T0 here:
#0 0x7f4ec97fb578 in __interceptor_malloc (/home/chakraut/ChakraCore/out/Debug/bin/ch/ch+0xbff578)
#1 0x7f4ec98cc577 in CorUnix::InternalMalloc(unsigned long) /home/chakraut/ChakraCore/pal/src/cruntime/malloc.cpp:114:20
#2 0x7f4ec98cc544 in PAL_malloc /home/chakraut/ChakraCore/pal/src/cruntime/malloc.cpp:98:12
#3 0x7f4ec9839204 in utf8::malloc_allocator::allocate(unsigned long) /home/chakraut/ChakraCore/lib/Common/Codex/Utf8Helper.h:128:53
#4 0x7f4ec9838cc6 in int utf8::NarrowStringToWide)(unsigned long), char const, unsigned long, char16_t, unsigned long, unsigned long) /home/chakraut/ChakraCore/lib/Common/Codex/Utf8Helper.h:72:37
#5 0x7f4ec9838c4e in int utf8::NarrowStringToWide(char const, unsigned long, char16_t, unsigned long, unsigned long) /home/chakraut/ChakraCore/lib/Common/Codex/Utf8Helper.h:122:16
#6 0x7f4ec98373a1 in utf8::NarrowStringToWideDynamic(char const, char16_t**) /home/chakraut/ChakraCore/lib/Common/Codex/Utf8Helper.h:142:16
#7 0x7f4ec9834422 in main /home/chakraut/ChakraCore/bin/ch/ch.cpp:900:9
#8 0x7f4ec776c2b0 in __libc_start_main (/lib/x86_64-linux-gnu/libc.so.6+0x202b0)

SUMMARY: AddressSanitizer: heap-buffer-overflow /home/chakraut/ChakraCore/pal/src/cruntime/wchar.cpp:1212:12 in PAL_wcslen
Shadow bytes around the buggy address:
0x0c047fff9d80: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0c047fff9d90: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0c047fff9da0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0c047fff9db0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0c047fff9dc0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
=>0x0c047fff9dd0: fa fa fa fa fa fa 04 fa[fa]fa 00 02 fa fa 00 00
0x0c047fff9de0: fa fa 00 00 fa fa 06 fa fa fa 02 fa fa fa 02 fa
0x0c047fff9df0: fa fa fd fd fa fa 00 fa fa fa 04 fa fa fa 00 fa
0x0c047fff9e00: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0c047fff9e10: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0c047fff9e20: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
Shadow byte legend (one shadow byte represents 8 application bytes):
Addressable: 00
Partially addressable: 01 02 03 04 05 06 07
Heap left redzone: fa
Heap right redzone: fb
Freed heap region: fd
Stack left redzone: f1
Stack mid redzone: f2
Stack right redzone: f3
Stack partial redzone: f4
Stack after return: f5
Stack use after scope: f8
Global redzone: f9
Global init order: f6
Poisoned by user: f7
Container overflow: fc
Array cookie: ac
Intra object redzone: bb
ASan internal: fe
Left alloca redzone: ca
Right alloca redzone: cb
==67392==ABORTING

[mrkmarron]

It looks like this is a bug in the PAL implementation of wcsstr or something in the string conversion.

Getting to a call to GetTTDDirectory requires a check of the form:

else if(wcsstr(argv[i], _u("-TTRecord=")) == argv[i])

The documentation for wcsstr says it returns "A pointer to the first occurrence in wcs1 of the entire sequence of characters specified in wcs2, or a null pointer if the sequence is not present in wcs1." However, if the input is ./ch - then this test should definitely fail.

[kfarnung]

This looks similar to a bug I was investigating, let me take a look.

[kfarnung]

This is because the PAL implementation of wcsstr doesn't handle the case where the string reaches a null character before strCharSet has fully matched.

https://github.com/Microsoft/ChakraCore/blob/master/pal/src/cruntime/wchar.cpp#L1496

This means that

wchar_t* foo = "-";
wcsstr(foo, "-TTDRecord=") == foo // this should be false, but it's true

[kfarnung]

Looks like the same bug exists in coreclr as well: https://github.com/dotnet/coreclr/blob/master/src/pal/src/cruntime/wchar.cpp

[mike-kaufman]

Fixed with https://github.com/Microsoft/ChakraCore/pull/3549

[mike-kaufman]

Looks like the same bug exists in coreclr as well: https://github.com/dotnet/coreclr/blob/master/src/pal/src/cruntime/wchar.cpp

Fixed in coreclr with PR https://github.com/dotnet/coreclr/pull/13504.