| Russell Bull 08/04/07 21:05 Read: 714 times Melbourne Australia |
#142776 - Similar findings Responding to: Michael Karas's previous message |
Since I use Modbus a lot (CRC-16), my original code used the bit serial method initially. Then I did some benchmarks when I designed a product using the aduc834 which is only a 12clocker cpu. The table method was substantially faster and since I had no shortage of code space, the table method won hands down. I'd guess at around 5 times faster and the 'c' table method was faster than bit serial in assembler. My library code has an ifdef so you can choose at compile time what method I want to use. Michael - nice beavers in Beaverton? |
| Topic | Author | Date |
| How to create a crc table? | John Myers | 08/04/07 02:33 |
| simple | Jan Waclawek | 08/04/07 08:10 |
| RE: simple | John Myers | 08/04/07 15:34 |
| I said it\'s simple | Jan Waclawek | 08/04/07 17:33 |
| Thank you | John Myers | 08/04/07 18:27 |
| Protocols | Andy Neil | 08/06/07 07:48 |
| examples | Jan Waclawek | 08/06/07 08:05 |
| Runtime Computed Table?? | Michael Karas | 08/04/07 09:14 |
| RE: Runtime...? | John Myers | 08/04/07 16:02 |
| table is not necessary | Jan Waclawek | 08/04/07 17:41 |
| Benchmark | John Myers | 08/04/07 18:16 |
| I did benchmarks as such on AVR | Michael Karas | 08/04/07 19:38 |
| Similar findings | Russell Bull | 08/04/07 21:05 |
| no beavers that I've tended to!! | Michael Karas | 08/04/07 22:50 |
| not bitwise.... | Jan Waclawek | 08/05/07 02:45 |
| Better is a strong word | Neil Kurzman | 08/05/07 21:19 |
| Also comment about SMBus PEC Code... | Michael Karas | 08/05/07 21:30 |
| The Table | Neil Kurzman | 08/06/07 11:19 |
| The Nibble Table for Poly 107 | Neil Kurzman | 08/06/07 11:20 |
| The formula | Neil Kurzman | 08/06/07 11:22 |
| Have you seen this? | Richard Erlacher | 08/06/07 18:42 |
| Now I have | Neil Kurzman | 08/06/07 21:26 |
Yes, that is a Key Document! | Andy Neil | 08/07/07 02:45 |