Simon Butcher | df81924 | 2015-08-19 19:33:27 +0100 | [diff] [blame] | 1 | # mbed TLS Benchmark Example |
Manuel Pégourié-Gonnard | 63e7eba | 2015-07-28 14:17:48 +0200 | [diff] [blame] | 2 | |
| 3 | This application benchmarks the various cryptographic primitives offered by mbed TLS. |
| 4 | |
| 5 | ## Pre-requisites |
| 6 | |
Irit Arkin | cd8c0fb | 2015-08-27 15:44:11 +0100 | [diff] [blame] | 7 | To build and run this example you must have: |
Manuel Pégourié-Gonnard | 63e7eba | 2015-07-28 14:17:48 +0200 | [diff] [blame] | 8 | |
| 9 | * A computer with the following software installed: |
| 10 | * [CMake](http://www.cmake.org/download/). |
| 11 | * [yotta](https://github.com/ARMmbed/yotta). Please note that **yotta has its own set of dependencies**, listed in the [installation instructions](http://armmbed.github.io/yotta/#installing-on-windows). |
| 12 | * [Python](https://www.python.org/downloads/). |
Irit Arkin | cd8c0fb | 2015-08-27 15:44:11 +0100 | [diff] [blame] | 13 | * [The ARM GCC toolchain](https://launchpad.net/gcc-arm-embedded). |
| 14 | * A serial terminal emulator (Like screen, pySerial and cu). |
Simon Butcher | df81924 | 2015-08-19 19:33:27 +0100 | [diff] [blame] | 15 | * An [FRDM-K64F](http://developer.mbed.org/platforms/FRDM-K64F/) development board, or another board supported by mbed OS (in which case you'll have to substitute frdm-k64f-gcc with the appropriate target in the instructions below). |
Manuel Pégourié-Gonnard | 63e7eba | 2015-07-28 14:17:48 +0200 | [diff] [blame] | 16 | * A micro-USB cable. |
| 17 | * If your OS is Windows, please follow the installation instructions [for the serial port driver](https://developer.mbed.org/handbook/Windows-serial-configuration). |
| 18 | |
| 19 | ## Getting started |
| 20 | |
Simon Butcher | df81924 | 2015-08-19 19:33:27 +0100 | [diff] [blame] | 21 | 1. Connect the FRDM-K64F to the computer with the micro-USB cable, being careful to use the "OpenSDA" connector on the target board. |
Manuel Pégourié-Gonnard | 63e7eba | 2015-07-28 14:17:48 +0200 | [diff] [blame] | 22 | |
| 23 | 2. Navigate to the mbedtls directory supplied with your release and open a terminal. |
| 24 | |
| 25 | 3. Set the yotta target: |
| 26 | |
| 27 | ``` |
| 28 | yotta target frdm-k64f-gcc |
| 29 | ``` |
| 30 | |
Simon Butcher | a150050 | 2015-08-19 20:29:31 +0100 | [diff] [blame] | 31 | 4. Build mbedtls and the examples. This may take a long time if this is your first compilation: |
Manuel Pégourié-Gonnard | 63e7eba | 2015-07-28 14:17:48 +0200 | [diff] [blame] | 32 | |
| 33 | ``` |
Simon Butcher | df81924 | 2015-08-19 19:33:27 +0100 | [diff] [blame] | 34 | $ yotta build |
Manuel Pégourié-Gonnard | 63e7eba | 2015-07-28 14:17:48 +0200 | [diff] [blame] | 35 | ``` |
| 36 | |
Brian Daniels | 3c405ae | 2015-08-18 10:52:57 -0700 | [diff] [blame] | 37 | 5. Copy `build/frdm-k64f-gcc/test/mbedtls-test-example-benchmark.bin` to your mbed board and wait until the LED next to the USB port stops blinking. |
Manuel Pégourié-Gonnard | 63e7eba | 2015-07-28 14:17:48 +0200 | [diff] [blame] | 38 | |
Irit Arkin | cd8c0fb | 2015-08-27 15:44:11 +0100 | [diff] [blame] | 39 | 6. Start the serial terminal emulator and connect to the virtual serial port presented by FRDM-K64F. |
Manuel Pégourié-Gonnard | 63e7eba | 2015-07-28 14:17:48 +0200 | [diff] [blame] | 40 | |
Irit Arkin | cd8c0fb | 2015-08-27 15:44:11 +0100 | [diff] [blame] | 41 | Use the following settings: |
| 42 | |
| 43 | * 115200 baud (not 9600). |
| 44 | * 8N1. |
| 45 | * No flow control. |
| 46 | |
| 47 | 7. Press the Reset button on the board. |
Manuel Pégourié-Gonnard | 63e7eba | 2015-07-28 14:17:48 +0200 | [diff] [blame] | 48 | |
Brian Daniels | 3c405ae | 2015-08-18 10:52:57 -0700 | [diff] [blame] | 49 | 8. The output in the terminal window should look like: |
Manuel Pégourié-Gonnard | 63e7eba | 2015-07-28 14:17:48 +0200 | [diff] [blame] | 50 | |
| 51 | ``` |
| 52 | {{timeout;150}} |
| 53 | {{host_test_name;default}} |
| 54 | {{description;mbed TLS benchmark program}} |
| 55 | {{test_id;MBEDTLS_BENCHMARK}} |
| 56 | {{start}} |
| 57 | |
| 58 | |
| 59 | SHA-1 : 3644 Kb/s, 32 cycles/byte |
| 60 | SHA-256 : 1957 Kb/s, 59 cycles/byte |
| 61 | SHA-512 : 587 Kb/s, 200 cycles/byte |
| 62 | AES-CBC-128 : 1359 Kb/s, 86 cycles/byte |
| 63 | AES-CBC-192 : 1183 Kb/s, 99 cycles/byte |
| 64 | AES-CBC-256 : 1048 Kb/s, 111 cycles/byte |
| 65 | AES-GCM-128 : 421 Kb/s, 279 cycles/byte |
| 66 | AES-GCM-192 : 403 Kb/s, 292 cycles/byte |
| 67 | AES-GCM-256 : 385 Kb/s, 305 cycles/byte |
| 68 | AES-CCM-128 : 542 Kb/s, 216 cycles/byte |
| 69 | AES-CCM-192 : 484 Kb/s, 242 cycles/byte |
| 70 | AES-CCM-256 : 437 Kb/s, 268 cycles/byte |
| 71 | CTR_DRBG (NOPR) : 1002 Kb/s, 117 cycles/byte |
| 72 | CTR_DRBG (PR) : 705 Kb/s, 166 cycles/byte |
| 73 | HMAC_DRBG SHA-1 (NOPR) : 228 Kb/s, 517 cycles/byte |
| 74 | HMAC_DRBG SHA-1 (PR) : 210 Kb/s, 561 cycles/byte |
| 75 | HMAC_DRBG SHA-256 (NOPR) : 212 Kb/s, 557 cycles/byte |
| 76 | HMAC_DRBG SHA-256 (PR) : 185 Kb/s, 637 cycles/byte |
| 77 | RSA-2048 : 41 ms/ public |
| 78 | RSA-2048 : 1349 ms/private |
| 79 | RSA-4096 : 134 ms/ public |
| 80 | RSA-4096 : 7149 ms/private |
| 81 | ECDSA-secp384r1 : 640 ms/sign |
| 82 | ECDSA-secp256r1 : 387 ms/sign |
| 83 | ECDSA-secp384r1 : 1233 ms/verify |
| 84 | ECDSA-secp256r1 : 751 ms/verify |
| 85 | ECDHE-secp384r1 : 1191 ms/handshake |
| 86 | ECDHE-secp256r1 : 730 ms/handshake |
| 87 | ECDHE-Curve25519 : 611 ms/handshake |
| 88 | ECDH-secp384r1 : 584 ms/handshake |
| 89 | ECDH-secp256r1 : 365 ms/handshake |
| 90 | ECDH-Curve25519 : 303 ms/handshake |
| 91 | |
| 92 | {{success}} |
| 93 | {{end}} |
| 94 | ``` |
Simon Butcher | 1404ddb | 2016-03-17 16:00:13 +0000 | [diff] [blame^] | 95 | |
| 96 | Any performance data generated by this example application are indicative only of the performance of the mbed TLS module on the platform it's executed on. |
| 97 | |
| 98 | Differences in the integration of mbed TLS into the platform, such as whether all available hardware accelerators have been used or not, can lead to significant differences in performance, and so results from the program are not intended to be used to meaningfully compare platforms. |
| 99 | |
| 100 | The figures may also slightly change from execution to execution due to variations in the timing functions. |