[llvm-commits] Specification for Run-time Checks

[John Criswell]

On 5/16/12 12:13 PM, Santosh Nagarakatte wrote:
> John,
>
>> I can see how the fastlscheck is useful since it can reduce the
>> instrumentation that SoftBound needs to add.  Additionally, the
>> pool_register optimizations may be indirectly useful.  If an optimization
>> removes a pool_register call for an object, then all of the checks on that
>> object can find the bounds of that object without any further
>> instrumentation.  SoftBound may be able to take advantage of that.
>>
>
> Yes pool_register optimization looks like a good optimization to have.
> Is there any documentation of current pool register optimizations.

There's nothing in the documentation, but you can take a look at 
lib/OptimizeChecks/PoolRegisterElimination.cpp.  It essentially 
determines when all the pointers from a memory object are checked using 
fastlscheck/fastgepcheck instead of lscheck/gepcheck.

>
>
>> I would also argue that lscheck is useful.  The lscheck essentially tells a
>> tool where a load/store check must be performed.  Optimizations can remove
>> them or hoist them out of loops.  Once that is done, SoftBound can add the
>> instrumentation to propagate bounds information and then replace the lscheck
>> with the code that checks to see if the pointer is within bounds.
>>
>> Do you think these ideas will work, or are there problems with them (e.g.,
>> would doing things this way make SoftBound more complicated?)?
>>
> Apart from the checks, SoftBound's overheads are primarily with
> propagation. I don't think lscheck/fastlscheck should necessarily
> complicate SoftBound

Is that the local propagation through SSA virtual register or the 
"through memory" propagation that occurs when a pointer is stored into 
memory?


>
>>> Additionally softboundcets would want to pass extra identifier
>>> metadata with fastlscheck.
>> Is this for the CETS instrumentation?  Can a SoftBound-specific pass easily
>> enhance the fastlscheck with the needed information?
>>
> Yes SoftBoundCETS does lock and key checking, needs identifier
> metadata with each check.

Okay.

-- John T.

>
> Thanks
> Santosh
>