??? 10/10/08 01:09 Read: times |
#158948 - So smart? Responding to: ???'s previous message |
PS they have actually implemented it rather smartly for the ISRs When you see just ISRs, you may think so. But look at the body code. SiLabs failed on the SFRs assignment for 'F12x, because it was the first attempt. The number of SFRs is no more than 160. These SFRs should fit to just two SFRPAGEs, but SiLabs scattered SFRs over four SFRPAGEs. As the result, we have to switch SFRPAGE everywhere. On the latter MCU family, C8051F36x, SiLabs have improved the assignment. SFRs are grouped into two categories, configuration and operation, and they are assigned to the SFRPAGE of this name, respectively. The SFRs accessed on both phase are assigned to ALL page. While initialization of the code, SFRPAGE stays in the configuration. Before main loop starts, SFRPAGE switches to the operation page. In this way, SFRPAGE switches are reduced significantly. Tsuneo |
Topic | Author | Date |
SFR Paging - advantages ? | 01/01/70 00:00 | |
how else would you do it | 01/01/70 00:00 | |
RE: how else would you do it | 01/01/70 00:00 | |
more options... | 01/01/70 00:00 | |
RE: more options![]() | 01/01/70 00:00 | |
So smart? | 01/01/70 00:00 | |
Not very encouraging....:-( | 01/01/70 00:00 | |
Located this document.. | 01/01/70 00:00 | |
a way to handle it | 01/01/70 00:00 | |
Getting clear on this paging stuff | 01/01/70 00:00 | |
No new concept! | 01/01/70 00:00 | |
maybe, maybe not | 01/01/70 00:00 |