Related Weaknesses
| CWE-ID |
Weakness Name |
Source |
| CWE-367 |
Time-of-check Time-of-use (TOCTOU) Race Condition
The product checks the state of a resource before using that resource, but the resource's state can change between the check and the use in a way that invalidates the results of the check. This can cause the product to perform invalid actions when the resource is in an unexpected state. |
|
Metrics
| Metrics |
Score |
Severity |
CVSS Vector |
Source |
| V3.0 |
7.8 |
HIGH |
CVSS:3.0/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
Base: Exploitabilty MetricsThe Exploitability metrics reflect the characteristics of the thing that is vulnerable, which we refer to formally as the vulnerable component. Attack Vector This metric reflects the context by which vulnerability exploitation is possible. A vulnerability exploitable with Local access means that the vulnerable component is not bound to the network stack, and the attacker's path is via read/write/execute capabilities. In some cases, the attacker may be logged in locally in order to exploit the vulnerability, otherwise, she may rely on User Interaction to execute a malicious file. Attack Complexity This metric describes the conditions beyond the attacker's control that must exist in order to exploit the vulnerability. Specialized access conditions or extenuating circumstances do not exist. An attacker can expect repeatable success against the vulnerable component. Privileges Required This metric describes the level of privileges an attacker must possess before successfully exploiting the vulnerability. The attacker is unauthorized prior to attack, and therefore does not require any access to settings or files to carry out an attack. User Interaction This metric captures the requirement for a user, other than the attacker, to participate in the successful compromise of the vulnerable component. Successful exploitation of this vulnerability requires a user to take some action before the vulnerability can be exploited. For example, a successful exploit may only be possible during the installation of an application by a system administrator. Base: Scope MetricsAn important property captured by CVSS v3.0 is the ability for a vulnerability in one software component to impact resources beyond its means, or privileges. Scope Formally, Scope refers to the collection of privileges defined by a computing authority (e.g. an application, an operating system, or a sandbox environment) when granting access to computing resources (e.g. files, CPU, memory, etc). These privileges are assigned based on some method of identification and authorization. In some cases, the authorization may be simple or loosely controlled based upon predefined rules or standards. For example, in the case of Ethernet traffic sent to a network switch, the switch accepts traffic that arrives on its ports and is an authority that controls the traffic flow to other switch ports. An exploited vulnerability can only affect resources managed by the same authority. In this case the vulnerable component and the impacted component are the same. Base: Impact MetricsThe Impact metrics refer to the properties of the impacted component. Confidentiality Impact This metric measures the impact to the confidentiality of the information resources managed by a software component due to a successfully exploited vulnerability. There is total loss of confidentiality, resulting in all resources within the impacted component being divulged to the attacker. Alternatively, access to only some restricted information is obtained, but the disclosed information presents a direct, serious impact. For example, an attacker steals the administrator's password, or private encryption keys of a web server. Integrity Impact This metric measures the impact to integrity of a successfully exploited vulnerability. Integrity refers to the trustworthiness and veracity of information. There is a total loss of integrity, or a complete loss of protection. For example, the attacker is able to modify any/all files protected by the impacted component. Alternatively, only some files can be modified, but malicious modification would present a direct, serious consequence to the impacted component. Availability Impact This metric measures the impact to the availability of the impacted component resulting from a successfully exploited vulnerability. There is total loss of availability, resulting in the attacker being able to fully deny access to resources in the impacted component; this loss is either sustained (while the attacker continues to deliver the attack) or persistent (the condition persists even after the attack has completed). Alternatively, the attacker has the ability to deny some availability, but the loss of availability presents a direct, serious consequence to the impacted component (e.g., the attacker cannot disrupt existing connections, but can prevent new connections; the attacker can repeatedly exploit a vulnerability that, in each instance of a successful attack, leaks a only small amount of memory, but after repeated exploitation causes a service to become completely unavailable). Temporal MetricsThe Temporal metrics measure the current state of exploit techniques or code availability, the existence of any patches or workarounds, or the confidence that one has in the description of a vulnerability. Environmental Metrics
|
[email protected] |
| V2 |
9.3 |
|
AV:N/AC:M/Au:N/C:C/I:C/A:C |
[email protected] |
EPSS
EPSS is a scoring model that predicts the likelihood of a vulnerability being exploited.
EPSS Score
The EPSS model produces a probability score between 0 and 1 (0 and 100%). The higher the score, the greater the probability that a vulnerability will be exploited.
EPSS Percentile
The percentile is used to rank CVE according to their EPSS score. For example, a CVE in the 95th percentile according to its EPSS score is more likely to be exploited than 95% of other CVE. Thus, the percentile is used to compare the EPSS score of a CVE with that of other CVE.
Exploit information
Exploit Database EDB-ID : 43464
Publication date : 2018-01-07 23h00 +00:00
Author : Google Security Research
EDB Verified : Yes
The MemoryIntArray class allows processes to share an in-memory array of integers backed by an "ashmem" file descriptor. As the class implements the Parcelable interface, it can be inserted into a Parcel, and optionally placed in a Bundle and transferred via binder to remote processes.
Instead of directly tracking the size of the shared memory region, the MemoryIntArray class calls the ASHMEM_GET_SIZE ioctl on the ashmem descriptor to retrieve it on-demand. Previously, the code made a single call to ASHMEM_GET_SIZE in order to retrieve the region's size, both before mapping it, and before unmapping it. Since the region's size could be set via ASHMEM_SET_SIZE until the region has been mapped, this opened the possibility for race conditions where an attacker alters the size in-between the first size retrieval and the mapping operation.
This issue has since been addressed (CVE-2017-0412), using the following pattern:
(see http://androidxref.com/8.0.0_r4/xref/frameworks/base/core/jni/android_util_MemoryIntArray.cpp#69)
1. int ashmemSize = ashmem_get_size_region(fd);
2. if (ashmemSize <= 0) {
3. jniThrowException(env, "java/io/IOException", "bad ashmem size");
4. return -1;
5. }
6.
7. // IMPORTANT: Ashmem allows the caller to change its size until
8. // it is memory mapped for the first time which lazily creates
9. // the underlying VFS file. So the size we get above may not
10. // reflect the size of the underlying shared memory region. Therefore,
11. // we first memory map to set the size in stone an verify if
12. // the underlying ashmem region has the same size as the one we
13. // memory mapped. This is critical as we use the underlying
14. // ashmem size for boundary checks and memory unmapping.
15. int protMode = owner ? (PROT_READ | PROT_WRITE) : PROT_READ;
16. void* ashmemAddr = mmap(NULL, ashmemSize, protMode, MAP_SHARED, fd, 0);
17. if (ashmemAddr == MAP_FAILED) {
18. jniThrowException(env, "java/io/IOException", "cannot mmap ashmem");
19. return -1;
20. }
21.
22. // Check if the mapped size is the same as the ashmem region.
23. int mmapedSize = ashmem_get_size_region(fd);
24. if (mmapedSize != ashmemSize) {
25. munmap(reinterpret_cast<void *>(ashmemAddr), ashmemSize);
26. jniThrowException(env, "java/io/IOException", "bad file descriptor");
27. return -1;
28. }
As we can see above, the code verifies that the size retrieved prior to mapping and after performing the mapping operation are equal, thus attempting to eliminate the race condition. However, looking at the ashmem driver, the following code is used to implement the ASHMEM_SET_SIZE ioctl:
(see http://androidxref.com/kernel_3.18/xref/drivers/staging/android/ashmem.c#753)
a. case ASHMEM_SET_SIZE:
b. ret = -EINVAL;
c. if (!asma->file) {
d. ret = 0;
e. asma->size = (size_t) arg;
f. }
g. break;
The ioctl does not acquire the "ashmem_mutex" to perform the ioctl itself. Therefore, an "mmap" operation could be in-flight, while the ASHMEM_SET_SIZE ioctl is being processed. This opens up the possibility to the following schedule, triggering a race condition:
[Process A]:
1. Attacker sends a MemoryIntArray with a crafted ashmem file descriptor in a Bundle, and with a small size
[System Server]:
2. Target process (e.g., system_server) unparcels the bundle with the MemoryIntArray, instantiating it
3. This triggers the code path above, executing lines 1-16
[Process A]:
4. Attacker calls ASHMEM_SET_SIZE, either during or before the mmap call
4.1. Lines a-c are executed, asma->file is still NULL
[System Server]:
5. Target process continues executing lines 16-24
5.1. Target process sees the old size, as the ASHMEM_SET_SIZE operation didn't complete yet
5.2. Therefore, the condition at line 24 is not satisfied
[Process A]:
6. Lines d-f are executed, setting the size to a new value
[System Server]:
7. Some time later, target process runs the finalizer, which retrieves the new size, and uses it to munmap the descriptor
7.1. This causes an inter-process munmap with an attacker-controller size
This issue can be exploited similarly to the previous ashmem bugs -- once a larger "munmap" is performed in the target process, it can be used to "free" a data structure such as a thread's stack, allowing the attacker to replace it with their own controlled contents.
While the exploitable condition is present in MemoryIntArray, I believe a fix should also be applied to the kernel to prevent such conditions from occurring in other contexts. Namely, the ashmem driver should acquire the "ashmem_mutex" during the ASHMEM_SET_SIZE operation, in order to guarantee that no races with ongoing "mmap" operations are possible. In addition, MemoryIntArray should not rely on multiple calls to ASHMEM_GET_SIZE, but should rather perform a single ASHMEM_GET_SIZE operation and store the returned size for both the "mmap" and "munmap" operations.
To demonstrate the race condition, I've added a busy loop to the ashmem driver between lines c. and d., increasing the race window to allow for easier demonstration of the schedule above.
I've attached a PoC which triggers this race condition and causes system_server to call munmap on a large memory region. To reproduce the issue, apply the diff in "ashmem_delay.diff" to the ashmem driver, then run the attached program. Doing so should result in a large "munmap" operation in system_server, causing it to crash.
The issue can also be exploited from the "isolated_app" SELinux context (and perhaps from the Chrome sandbox?), as all that's required to leverage the attack is the ability to issue ashmem syscalls, and to interact with the ActivityManager service (which is exposed to "isolated_app").
Proof of Concept:
https://gitlab.com/exploit-database/exploitdb-bin-sploits/-/raw/main/bin-sploits/43464.zip
Exploit Database EDB-ID : 41355
Publication date : 2017-02-13 23h00 +00:00
Author : Google Security Research
EDB Verified : Yes
Source: https://bugs.chromium.org/p/project-zero/issues/detail?id=1002
The MemoryIntArray class allows processes to share an in-memory array of integers by transferring an ashmem file descriptor. As the class implements the Parcelable interface, it can be passed within a Parcel or a Bundle and transferred via binder to remote processes.
Instead of directly tracking the size of the shared memory region, the MemoryIntArray class calls the ASHMEM_GET_SIZE ioctl on the ashmem descriptor to retrieve it on demand. This opens up a variety of race conditions when using MemoryIntArray, as the size of the ashmem descriptor can be modified (via ASHMEM_SET_SIZE) so long as the descriptor itself has not yet been mapped.
To illustrate this, here is a snippet from the native function called when a MemoryIntArray is first mapped in:
1. static jlong android_util_MemoryIntArray_open(JNIEnv* env, jobject clazz, jint fd,
2. jboolean owner, jboolean writable)
3. {
4. if (fd < 0) {
5. jniThrowException(env, "java/io/IOException", "bad file descriptor");
6. return -1;
7. }
8.
9. int ashmemSize = ashmem_get_size_region(fd);
10. if (ashmemSize <= 0) {
11. jniThrowException(env, "java/io/IOException", "bad ashmem size");
12. return -1;
13. }
14.
15. int protMode = (owner || writable) ? (PROT_READ | PROT_WRITE) : PROT_READ;
16. void* ashmemAddr = mmap(NULL, ashmemSize, protMode, MAP_SHARED, fd, 0);
17. ...
18.}
If an attacker can call ASHMEM_SET_SIZE on the shared ashmem descriptor during the execution of lines 10-15, he may modify the internal size of the descriptor, causing a mismatch between the mapped-in size and the underlying size of the descriptor.
As the MemoryIntArray class uses the size reported by the ashmem descriptor to perform all bounds checks (see http://androidxref.com/7.0.0_r1/xref/frameworks/base/core/java/android/util/MemoryIntArray.java#217), this allows an attacker to cause out-of-bounds accesses to the mapped in buffer via subsequent calls to the "get" and "set" methods.
Additionally, MemoryIntArray uses the ashmem-reported size when unmapping the shared memory buffer, like so:
1. static void android_util_MemoryIntArray_close(JNIEnv* env, jobject clazz, jint fd,
2. jlong ashmemAddr, jboolean owner)
3. {
4. ...
5. int ashmemSize = ashmem_get_size_region(fd);
6. if (ashmemSize <= 0) {
7. jniThrowException(env, "java/io/IOException", "bad ashmem size");
8. return;
9. }
10. int unmapResult = munmap(reinterpret_cast<void *>(ashmemAddr), ashmemSize);
11. ...
12.}
This allows an attacker to trigger an inter-process munmap with a controlled size by modifying the underlying ashmem size to a size larger than the mapped in buffer's size. Doing so will cause the finalizer to call munmap with the new size, thus forcibly freeing memory directly after the buffer. After the memory is freed, the attacker can attempt to re-capture it using controlled data.
I've attached a PoC which triggers this race condition and causes system_server to call munmap on a large memory region. Running it should cause system_server to crash.
Note that simply modifying the size of the ashmem file descriptor is insufficient. This is due to the fact that Parcel objects keep track of the size of the ashmem descriptors passed through them using an unsigned variable (http://androidxref.com/7.0.0_r1/xref/frameworks/native/libs/binder/Parcel.cpp#216). When a descriptor object is released, the size variable is decremented according to the reported size of the descriptor. Although this variable is not used in any meaningful way, increasing the size of the ashmem descriptor between the creation and destruction of a Parcel would cause the size variable to underflow. As system_server is compiled with UBSAN, this triggers an abort (thus preventing us from using the exploit). To get around this, I've added an additional descriptor to the Parcel, whose size is appropriately reduced before increasing the size of the MemoryIntArray's descriptor (thus keeping the size variable from underflowing).
################################################################################
Attaching another version of the PoC, adjusted for Android 7.1 (since MemoryIntArray's fields have slightly changed). This version also contains a few additional tricks to make it slightly more reliable:
-Waits for the ASHMEM_SET_SIZE ioctl to fail to more closely control the race by waiting for MemoryIntArray's constructor (allowing for a smaller chance of hitting the UBSAN check before the additional descriptor's size is decreased)
-Uses an ugly hack to avoid constructing MemoryIntArray instances in the attacking process (modifies the marshalled class in the Parcel manually). This prevents the PoC process from crashing due to the same UBSAN checks.
Note that if the race is lost (that is - if Parcel.recycle() is called before we manage to call ASHMEM_SET_SIZE on the additional descriptor), the UBSAN abort will be triggered, resulting in a SIGABRT. If that happens, just try and run the PoC again.
Here is a sample crash from a successful execution of the PoC:
11-22 14:38:43.137 9328 10749 F libc : Fatal signal 11 (SIGSEGV), code 1, fault addr 0x7ffedcbfa690 in tid 10749 (RenderThread)
11-22 14:38:43.137 1183 1183 W : debuggerd: handling request: pid=9328 uid=1000 gid=1000 tid=10749
11-22 14:38:43.137 9328 9336 F libc : Fatal signal 11 (SIGSEGV), code 1, fault addr 0x7ffedca00000 in tid 9336 (FinalizerDaemon)
11-22 14:38:43.137 9328 9336 I libc : Another thread contacted debuggerd first; not contacting debuggerd.
11-22 14:38:43.151 10816 10816 F DEBUG : *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** ***
11-22 14:38:43.152 10816 10816 F DEBUG : Build fingerprint: 'Android/sdk_google_phone_x86_64/generic_x86_64:7.1.1/NPF10D/3354678:userdebug/test-keys'
11-22 14:38:43.152 10816 10816 F DEBUG : Revision: '0'
11-22 14:38:43.152 10816 10816 F DEBUG : ABI: 'x86_64'
11-22 14:38:43.152 10816 10816 F DEBUG : pid: 9328, tid: 10749, name: RenderThread >>> system_server <<<
11-22 14:38:43.152 10816 10816 F DEBUG : signal 11 (SIGSEGV), code 1 (SEGV_MAPERR), fault addr 0x7ffedcbfa690
11-22 14:38:43.152 10816 10816 F DEBUG : rax 00007ffed6f4fce0 rbx 00007ffedcbfa680 rcx 00007ffef5955547 rdx 0000000000000004
11-22 14:38:43.152 10816 10816 F DEBUG : rsi 00007ffed7ad14c4 rdi 00007ffedcbfa680
11-22 14:38:43.152 10816 10816 F DEBUG : r8 00007ffee8801ca0 r9 0000000000000000 r10 00007ffef58eed20 r11 0000000000000206
11-22 14:38:43.152 10816 10816 F DEBUG : r12 00007ffeea5b8598 r13 00007ffedbb30a18 r14 00007ffef66fe610 r15 00007ffedef974a0
11-22 14:38:43.152 10816 10816 F DEBUG : cs 0000000000000033 ss 000000000000002b
11-22 14:38:43.152 10816 10816 F DEBUG : rip 00007ffee88efe87 rbp 00007ffed7ad1760 rsp 00007ffed7ad16d0 eflags 0000000000000202
11-22 14:38:43.156 10816 10816 F DEBUG :
11-22 14:38:43.156 10816 10816 F DEBUG : backtrace:
11-22 14:38:43.157 10816 10816 F DEBUG : #00 pc 0000000000001e87 /system/lib64/libOpenglSystemCommon.so (_ZN14HostConnection10gl2EncoderEv+7)
11-22 14:38:43.157 10816 10816 F DEBUG : #01 pc 0000000000008434 /system/lib64/egl/libGLESv2_emulation.so (glCreateProgram+36)
11-22 14:38:43.157 10816 10816 F DEBUG : #02 pc 000000000007c9ec /system/lib64/libhwui.so
11-22 14:38:43.157 10816 10816 F DEBUG : #03 pc 000000000007d58f /system/lib64/libhwui.so
11-22 14:38:43.157 10816 10816 F DEBUG : #04 pc 000000000005e36e /system/lib64/libhwui.so
11-22 14:38:43.157 10816 10816 F DEBUG : #05 pc 0000000000099ddd /system/lib64/libhwui.so
11-22 14:38:43.157 10816 10816 F DEBUG : #06 pc 00000000000a4674 /system/lib64/libhwui.so
11-22 14:38:43.157 10816 10816 F DEBUG : #07 pc 0000000000037373 /system/lib64/libhwui.so
11-22 14:38:43.157 10816 10816 F DEBUG : #08 pc 0000000000036b5d /system/lib64/libhwui.so
11-22 14:38:43.157 10816 10816 F DEBUG : #09 pc 0000000000039745 /system/lib64/libhwui.so
11-22 14:38:43.157 10816 10816 F DEBUG : #10 pc 000000000003f128 /system/lib64/libhwui.so (_ZN7android10uirenderer12renderthread12RenderThread10threadLoopEv+136)
11-22 14:38:43.157 10816 10816 F DEBUG : #11 pc 0000000000012c39 /system/lib64/libutils.so (_ZN7android6Thread11_threadLoopEPv+313)
11-22 14:38:43.157 10816 10816 F DEBUG : #12 pc 00000000000aa613 /system/lib64/libandroid_runtime.so (_ZN7android14AndroidRuntime15javaThreadShellEPv+99)
11-22 14:38:43.157 10816 10816 F DEBUG : #13 pc 00000000000897b1 /system/lib64/libc.so (_ZL15__pthread_startPv+177)
11-22 14:38:43.157 10816 10816 F DEBUG : #14 pc 0000000000029a6b /system/lib64/libc.so (__start_thread+11)
11-22 14:38:43.157 10816 10816 F DEBUG : #15 pc 000000000001cae5 /system/lib64/libc.so (__bionic_clone+53)
Proofs of Concept:
https://gitlab.com/exploit-database/exploitdb-bin-sploits/-/raw/main/bin-sploits/41355.zip
Products Mentioned
Configuraton 0
Google>>Android >> Version 7.0
Google>>Android >> Version 7.1.0
Google>>Android >> Version 7.1.1
References
