Skip to content

Latest commit

 

History

History
146 lines (112 loc) · 6.23 KB

README.md

File metadata and controls

146 lines (112 loc) · 6.23 KB

HPCTool

Features:

  • Monitors hardware performance counters
  • Features per-process filtering
  • Features two modes:
    1. POLLING mode reads performance counter values at a desired state. We instrument source code using software trap (int 2e) to indicate the desired state.
    2. SAMPLING mode reads performance counter values at a performance monitoring interrupt (PMI), which is triggered by setting a threshold value on an event. In our code, we set a threshold value on instruction retired event.

Requirements:

  1. The HPCTool has been tested on 32 bit, Windows 7, both bare-metal and virtualized.

    • To use the HPCTool in a virtualized environment, verify that the virtual performance counters are enabled. For example, in VMWare Fusion one needs to modify following settings:
      • Enable virtual CPU Performance counters (e.g., on VMware -> settings: VMware->Hardware)
      • In .vmx file add below lines (on VMware):
        • vpmc.enable = "TRUE"
        • vpmc.freezeMode="guest"
  2. To build the kernel driver, install Windows Driver Kit version 7 available at Microsoft. You will need at least the "Full Development Environment" edition.

Building the HPCtool:

  1. Open drv/HPCTestDrv.c and make following changes

    a. Set the mode of operation to "POLLING_MODE" OR "SAMPLING_MODE".

    	#define SAMPLING_MODE

    b. Set threshold values for generating PMI.

    • In SAMPLING_MODE, set pmiThreshold as desired.
    • In POLLING_MODE, set pmiThreshold = 0.
    	#ifdef SAMPLING_MODE
    		INT32 pmiThreshold = -50000;
    	#else	//polling mode
    		INT32 pmiThreshold = 0;
    	#endif

    c. Change TEST_APP to a process/application that has to be monitored.

    	#define TEST_APP "test.exe"

    d. Modify LOG_FILE, to reflect your environment e.g. "C:\Users\Sanjeev\Desktop\hpcoutput.csv"

    	#define LOG_FILE L"\\DosDevices\\C:\\Users\\Sanjeev\\Desktop\\hpcoutput.csv"

    e. Set the event type of the performance counters EVENT0, EVENT1, EVENT2, EVENT3 to measure the events of interests.

    • By default only user space events are monitored for the following events:
      • EVENT0: Number of branches retired,
      • EVENT1: Number of mis-predicted branches retired,
      • EVENT2: Number of last level cache references,
      • EVENT3: Number of last level cache misses.
    	#define EVENT0	0x004100C4		//Branch instruction retired
    	#define EVENT1	0x004100C5		//Mispredicted branch instructions
    	#define EVENT2	0x00414F2E		//LLC cache reference
    	#define EVENT3	0x0041412E		//LLC misses

    E.g., to monitor Branch Instruction retired (Event Num. = C4H, Umask Value = 00H), in user space: we set EVENT0 = 0x4100C4 in kernel space: we set EVENT0 = 0x4200C4

    Refer to our tutorial for further details on configuring each event.

    Refer to Intel Manual, Chapter 19 for more information event num and umask value for performance counter events.

  2. Open x86 Checked Build Environment command prompt with Administrator privilege (right click -> Run as Administrator)

  3. Change the current directory to the path containing the kernel driver source:

    	cd PATH-TO-SRC-DIR 
    	builddrv.bat

    If the build is successful, HPCTestDrv.sys will be copied to Windows driver directory (e.g., C:\Windows\System32\drivers)

Preparing the instrumented binary:

  1. The binary that will be measured needs to be instrumented in the source code as shown in the test example in testcode\test.c For C/C++ code, insert the following instrumentation trigger to generate the software trap before and after the section of code to be profiled:

    	__asm __volatile{ 
    		mov eax, 19h 
    		int 0x2e 
    	}
  2. Compile the source code using C compiler and run HPC tool on the compiled binary.

Running the HPCTool:

  1. Disable the driver signing In the privileged command prompt opened in step 2) type:

    	bcdedit.exe -set loadoptions DDISABLE_INTEGRITY_CHECKS
    	bcdedit.exe -set TESTSIGNING ON
  2. Install the kernel driver - you need to do this just once

    Double click InstallTestDrv.reg and click to accept the registry of the kernel driver. Restart the system.

  3. Create an empty file in the location specified in LOG_FILE

  4. Open a command prompt with Administrator privileges and run the test program using the sample runtest.bat script. It starts the HPC driver, executes the test program and finally stops the driver.

    	runtest.bat

Output: HPC output is logged in the file specified as LOG_FILE at compile time.

Output:

The output comprises of collection of samples.

  • Each sample contains the measurement of 7 events -- 3 fixed and 4 programmable/configurable events.

  • 3 Fixed events: No. of instructions retired, logical cycles, reference cycles

  • 4 programmable events: No. of branches retired, mis-predicted branches retired, LLC cache references, LLC misses.

  • The four programmable events can be changed to address various profiling goals. However, changing the events measured requires re-compiling the kernel driver.

  • The data collected using the performance counters is written to a file in a comma separated value (CSV) format. The order of the fields is as follows:

     	#instructions retired, #logical-cycles, #reference-cycles, #event0, #event1, #event2, #event3
  • In the sampling mode, a data point is generated every pmiThreshold instructions retired.

  • In the polling mode there is only one data point collected after the second instrumentation trigger is invoked.

Cite as:

If you use this tool, please cite as:

Das, S., Werner, J., Antonakakis, M., Polychronakis, M. and Monrose, F., 2019, May. SoK: The Challenges, Pitfalls, and Perils of Using Hardware Performance Counters for Security. To appear in Proceedings of the 40th IEEE Symposium on Security and Privacy (S&P).