Android’s defense-in-depth technique applies not solely to the Android OS working on the Software Processor (AP) but in addition the firmware that runs on gadgets. We significantly prioritize hardening the mobile baseband given its distinctive mixture of working in an elevated privilege and parsing untrusted inputs which are remotely delivered into the gadget.
This submit covers the right way to use two high-value sanitizers which might forestall particular courses of vulnerabilities discovered inside the baseband. They’re structure agnostic, appropriate for bare-metal deployment, and needs to be enabled in current C/C++ code bases to mitigate unknown vulnerabilities. Past safety, addressing the problems uncovered by these sanitizers improves code well being and general stability, lowering assets spent addressing bugs sooner or later.
As we outlined beforehand, safety analysis centered on the baseband has highlighted a constant lack of exploit mitigations in firmware. Baseband Distant Code Execution (RCE) exploits have their very own categorization in well-known third-party marketplaces with a comparatively low payout. This means baseband bugs might probably be plentiful and/or not too complicated to search out and exploit, and their outstanding inclusion within the market demonstrates that they’re helpful.
Baseband safety and exploitation has been a recurring theme in safety conferences for the final decade. Researchers have additionally made a dent on this space in well-known exploitation contests. Most not too long ago, this space has turn into outstanding sufficient that it’s widespread to search out sensible baseband exploitation trainings in high safety conferences.
Acknowledging this pattern, mixed with the severity and obvious abundance of those vulnerabilities, final 12 months we launched updates to the severity tips of Android’s Vulnerability Rewards Program (VRP). For instance, we contemplate vulnerabilities permitting Distant Code Execution (RCE) within the mobile baseband to be of CRITICAL severity.
Frequent courses of vulnerabilities may be mitigated by way of the usage of sanitizers offered by Clang-based toolchains. These sanitizers insert runtime checks towards widespread courses of vulnerabilities. GCC-based toolchains may present some stage of assist for these flags as properly, however won’t be thought of additional on this submit. We encourage you to verify your toolchain’s documentation.
Two sanitizers included in Undefined Habits Sanitizer (UBSan) can be our focus – Integer Overflow Sanitizer (IntSan) and BoundsSanitizer (BoundSan). These have been extensively deployed in Android userspace for years following a data-driven method. These two are properly suited to bare-metal environments such because the baseband since they don’t require assist from the OS or particular structure options, and so are typically supported for all Clang targets.
Integer Overflow Sanitizer (IntSan)
IntSan causes signed and unsigned integer overflows to abort execution except the overflow is made express. Whereas unsigned integer overflows are technically outlined habits, it may possibly usually result in unintentional habits and vulnerabilities – particularly after they’re used to index into arrays.
As each intentional and unintentional overflows are possible current in most code bases, IntSan might require refactoring and annotating the code base to forestall intentional or benign overflows from trapping (which we contemplate a false optimistic for our functions). Overflows which must be addressed may be uncovered through testing (see the Deploying Sanitizers part)
BoundsSanitizer (BoundSan)
BoundSan inserts instrumentation to carry out bounds checks round some array accesses. These checks are solely added if the compiler can not show at compile time that the entry can be secure and if the dimensions of the array can be identified at runtime, in order that it may be checked towards. Word that this won’t cowl all array accesses as the dimensions of the array is probably not identified at runtime, resembling operate arguments that are arrays.
So long as the code is accurately written C/C++, BoundSan ought to produce no false positives. Any violations found when first enabling BoundSan is not less than a bug, if not a vulnerability. Resolving even these which aren’t exploitable can drastically enhance stability and code high quality.
Modernize your toolchains
Adopting fashionable mitigations additionally means adopting (and sustaining) fashionable toolchains. The advantages of this transcend using sanitizers nevertheless. Sustaining an outdated toolchain just isn’t free and entails hidden alternative prices. Toolchains comprise bugs that are addressed in subsequent releases. Newer toolchains deliver new efficiency optimizations, priceless within the extremely constrained bare-metal atmosphere that basebands function in. Safety points may even exist within the generated code of out-of-date compilers.
Sustaining a contemporary up-to-date toolchain for the baseband entails some prices when it comes to upkeep, particularly at first if the toolchain is especially outdated, however over time the advantages, as outlined above, outweigh the prices.
Each BoundSan and IntSan have a measurable efficiency overhead. Though we had been in a position to considerably scale back this overhead previously (for instance to lower than 1% in media codecs), even very small will increase in CPU load can have a considerable impression in some environments.
Enabling sanitizers over all the codebase offers essentially the most profit, however enabling them in security-critical assault surfaces can function a primary step in an incremental deployment. For instance:
- Features parsing messages delivered over the air in 2G, 3G, 4G, and 5G (particularly features dealing with pre-authentication messages that may be injected with a false/malicious base station)
- Libraries encoding/decoding complicated codecs (e.g. ASN.1, XML, DNS, and so forth…)
- IMS, TCP and IP stacks
- Messaging features (SMS, MMS)
Within the specific case of 2G, the perfect technique is to disable the stack altogether by supporting Android’s “2G toggle”. Nevertheless, 2G continues to be a obligatory cell entry expertise in sure elements of the world and a few customers may must have this legacy protocol enabled.
Having a transparent plan for deployment of sanitizers saves plenty of effort and time. We consider the deployment course of as having three phases:
- Detecting (and fixing) violations
- Measuring and lowering overhead
- Soaking in pre-production
We additionally introduce two modes during which sanitizers needs to be run: diagnostics mode and trapping mode. These can be mentioned within the following sections, however briefly: diagnostics mode recovers from violations and offers priceless debug data, whereas trapping mode actively mitigates vulnerabilities by trapping execution on violations.
Detecting (and Fixing) Violations
To efficiently ship these sanitizers, any benign integer overflows have to be made express and unintended out-of-bounds accesses have to be addressed. These should be uncovered by way of testing. The upper the code protection your assessments present, the extra points you’ll be able to uncover at this stage and the better deployment can be in a while.
To diagnose violations uncovered in testing, sanitizers can emit calls to runtime handlers with debug data such because the file, line quantity, and values resulting in the violation. Sanitizers can optionally proceed execution after a violation has occurred, permitting a number of violations to be found in a single check run. We seek advice from utilizing the sanitizers on this method as working them in “diagnostics mode”. Diagnostics mode just isn’t meant for manufacturing because it offers no safety advantages and provides excessive overhead.
Diagnostics mode for the sanitizers may be set utilizing the next flags:
-fsanitize=signed-integer-overflow,unsigned-integer-overflow,bounds -fsanitize-recover=all
Since Clang doesn’t present a UBSan runtime for bare-metal targets, a runtime will must be outlined and offered at hyperlink time:
// integer overflow handlers __ubsan_handle_add_overflow(OverflowData *knowledge, ValueHandle lhs, ValueHandle rhs) __ubsan_handle_sub_overflow(OverflowData *knowledge, ValueHandle lhs, ValueHandle rhs) __ubsan_handle_mul_overflow(OverflowData *knowledge, ValueHandle lhs, ValueHandle rhs) __ubsan_handle_divrem_overflow(OverflowData *knowledge, ValueHandle lhs, ValueHandle rhs) __ubsan_handle_negate_overflow(OverflowData *knowledge, ValueHandle old_val) // boundsan handler __ubsan_handle_out_of_bounds_overflow(OverflowData *knowledge, ValueHandle old_val)
For example, see the default Clang implementation; the Linux Kernels implementation may be illustrative.
With the runtime outlined, allow the sanitizer over all the baseband codebase and run all obtainable assessments to uncover and deal with any violations. Vulnerabilities needs to be patched, and overflows ought to both be refactored or made express by way of the usage of checked arithmetic builtins which don’t set off sanitizer violations. Sure features that are anticipated to have intentional overflows (resembling cryptographic features) may be preemptively excluded from sanitization (see subsequent part).
Other than uncovering safety vulnerabilities, this stage is extremely efficient at uncovering code high quality and stability bugs that would end in instability on person gadgets.
As soon as violations have been addressed and assessments are now not uncovering new violations, the following stage can start.
Measuring and Decreasing Overhead
As soon as shallow violations have been addressed, benchmarks may be run and the overhead from the sanitizers (efficiency, code measurement, reminiscence footprint) may be measured.
Measuring overhead have to be performed utilizing manufacturing flags – particularly “trapping mode”, the place violations trigger execution to abort. The diagnostics runtime used within the first stage carries vital overhead and isn’t indicative of the particular efficiency sanitizer overhead.
Trapping mode may be enabled utilizing the next flags:
-fsanitize=signed-integer-overflow,unsigned-integer-overflow,bounds -fsanitize-trap=all
Many of the overhead is probably going attributable to a small handful of “sizzling features”, for instance these with tight long-running loops. High-quality-grained per-function efficiency metrics (just like what Simpleperf offers for Android), permits evaluating metrics earlier than and after sanitizers and offers the simplest means to determine sizzling features. These features can both be refactored or, after guide inspection to confirm that they’re secure, have sanitization disabled.
Sanitizers may be disabled both inline within the supply or by way of the usage of ignorelists and the -fsanitize-ignorelist flag.
The bodily layer code, with its extraordinarily tight efficiency margins and decrease likelihood of exploitable vulnerabilities, could also be a very good candidate to disable sanitization wholesale if preliminary efficiency appears prohibitive.
Soaking in Pre-production
With overhead minimized and shallow bugs resolved, the ultimate stage is enabling the sanitizers in trapping mode to mitigate vulnerabilities.
We strongly suggest a protracted interval of inner soak in pre-production with check populations to uncover any remaining violations not found in testing. The extra thorough the check protection and size of the soak interval, the much less danger there can be from undiscovered violations.
As above, the configuration for trapping mode is as follows:
-fsanitize=signed-integer-overflow,unsigned-integer-overflow,bounds -fsanitize-trap=all
Having infrastructure in place to gather bug studies which consequence from any undiscovered violations may also help reduce the danger they current.
The advantages from deploying sanitizers in your current code base are tangible, nevertheless in the end they deal with solely the bottom hanging fruit and won’t end in a code base freed from vulnerabilities. Different courses of reminiscence security vulnerabilities stay unaddressed by these sanitizers. A long run answer is to start transitioning right this moment to memory-safe languages resembling Rust.
Rust is prepared for bare-metal environments such because the baseband, and we’re already utilizing it in different bare-metal parts in Android. There isn’t any must rewrite all the things in Rust, as Rust offers a robust C FFI assist and simply interfaces with current C codebases. Simply writing new code in Rust can quickly scale back the variety of reminiscence security vulnerabilities. Rewrites needs to be restricted/prioritized just for essentially the most important parts, resembling complicated parsers dealing with untrusted knowledge.
The Android staff has developed a Rust coaching meant to assist skilled builders rapidly ramp up Rust fundamentals. A complete day for bare-metal Rust is included, and the course has been translated to quite a few totally different languages.
Whereas the Rust compiler might not explicitly assist your bare-metal goal, as a result of it’s a front-end for LLVM, any goal supported by LLVM may be supported in Rust by way of customized goal definitions.
Because the high-level working system turns into a harder goal for attackers to efficiently exploit, we count on that decrease stage parts such because the baseband will entice extra consideration. Through the use of fashionable toolchains and deploying exploit mitigation applied sciences, the bar for attacking the baseband may be raised as properly. If in case you have any questions, tell us – we’re right here to assist!