I was hoping to find on-line manuals that I could print and read on the john. So far no luck. But what I really need is how to declare variables in C.
For example, I have a sample code that has the statement:
unsigned char
and
unsigned long
What do these mean and do they limit the declaration to integers. What do i declare if I want a floating point?
Also, someone told me I cant do floating point with a Keil Compiler because of licensing issues (I got my copy from Silicon Labs and have just registered it). I will eventually need to do trig functions. What do i have to get (buy) to complete my project?
You don't need an on-line manual, you need actual literature on the basics of C programming. The C51 compiler manual is not a substitute for a book about C, nor is it intended to be one.
K&R (i.e. "The C programming language" by B. W. Kernighan and D. M. Ritchie) has already been suggested, and I strongly second that recommendation. After all, it's _the_ book about C written by the guys who invented C in the first place.
I will eventually need to do trig functions.
Using trigonometric functions does not require use of floating point arithmetics. In fact, part of the skillset of an embedded developer is coming up with solutions that fit the capabilities of the target hardware - and the '51 is very much unsuited for floating point math (even though the C51 compiler will not try to keep the programmer from shooting himself in the foot there).
There are other solutions, like using lookup tables and interpolation.
The embedded developer should also have the skill to assess whether the use of a professionally written, optimised and debugged floating point math library such as that supplied with the C51 compiler would either positively or negatively impact his project.
Reinventing an already nicely developed wheel if not absolutely necessary would, frankly, be plain stupid.
In my experience (only 25 years, I'm afraid), it is standard practice that the tools supplied with "dev kits" are just "evaluation" versions.
And free "evaluation" versions are always "crippled" in some way or another.
(in fact, I think the SiLabs kit gives you a slightly less limited version than the "standard" free evaluation?)
The exception is the single-source proprietary architectures like PIC and AVR - where the chip maker knows that the tools are useless with any chips other than their own. Therefore the tools can simply be considered as part of the marketing budget!
And free "evaluation" versions are always "crippled" in some way or another.<p>
"Ain't no such thing as a free lunch."
;)
And so we live in two different worlds. I have learned much. Thanks.
Good point. I said nothing about free. It cost me 79 buck-a-roos. ;)
and my comments about "waste" and "inefficiencies" are my opinion only and based on my beliefs. This was going to be my first C based project.
I think everybody moving from asm to C had such beliefs initially. After a while you learn that C is the way to go where nominal "waste and inefficiencies" do not matter and the percentage where (you think) it matters (and you write an assembly module) will be constantly shrinking, but - in my opinion, never go to zero.
Erik
"I think everybody moving from asm to C had such beliefs initially"
and it is interesting to see the same beliefs emerging in people moving from C to C++...
Lou; It still applies, "read the small print". The Keil tools are fully functional but offset the 4K and restricted to 4K of linked code. There are free 8051 tools for 'C' and 'Pascal' for the "standard" 8051 but I'm sure that you chose the Silabs device for some non-standard features. Without questioning your programming abilities and your 8051 experience, I challange you to write some simple programs for the Silabs device using some of the non-standard featues. Write in both 'C' and assembly. Use some of the examples such as 'Blinky', etc. I'll bet the stripped Keil tools will make better ASSEMBLY code for these devices than you can write until you get up the learning curve on the SiLabs "deviates" as Erik calls them. As you have probably discovered by now, the Silabs IDE and tools have a fully functional, no restrictions assembler for their devices. One other thing, I recommend that you download the eval versions from this website. The tools that come on the eval CD are several versions old. The eval versions will have the same 4K restrictions. No,I do not work for Keil. Al Bradfrod Bradford
I can't even spell my own name but I have an excuse. I'm typing with a broken hand. Bradford
Fully functional? Part of my "what crap" expression was also because I read somewhere in the last two days that Keil was limited to 2K/4K and the floating point, including trig functions, dont work.
So actually, it appeared that I had "crap". Bunch of misinformation or misunderstanding, I guess.
So far my little C program has consumed 3K of the 4K limitation so I am VERY dissappointed with the efficiencies I see so far.
With regards to Blinky, come on, it's just initializing a port, setting a loop for a timer, and toggling a bit. But then again I dont know what you mean by "stripped Keil tools" but I cant imagine they can do better. Yeh, I know, there are probably some registers that need to be set (ie clk registers) but that cant be much. Yea, and you're right, there's a learning curve with assembly and, for me, C.
BTW, I chose SiLab because it was cheap enough, had 12 bit A/D (2 channel), a UART, some I/O, and the plus was -40 to +125 deg operation, and small footprint. Built-in clk, too. Dont know if all that is standard or special but if someone can recommend something else I'll listen.
Thanks for pointing out the unlimited assembler. No, I hadn't discovered it but since I am sure I will exceed the 4K limit by tomorrow, I'll need it. Not even doing floating point or trig yet.
Have you modified the compilers optimization settings yet? As default, the compiler produces completely un-optimized (i.e. large and slow) code that's easy to debug.
The Keil evaluation limitations are here: http://www.keil.com/demo/limits.asp
Both the 2K limit and the lack of floating-point are clearly stated.
Your "little" C program doesn't use printf, does it?
printf is a very powerful and, therefore, very large function - it can easily eat up most of the demo code limit!
If it weren't for the fact that I suspect you have a rather rigid idea of what 'nominal' means I'd have congratulated you for taking a balanced view...
Yes. The conversation usually goes like this:
Have you used C++ on the 8051?
No! It's for PCs. Too slow. Too much overhead.
Have you actually tried it?
[silence]
You may well find that a small program that occupies 3k written in 'C' would occupy 300 bytes an assembly. That doesn't mean that a 600 byte assembly program would occupy 6k in 'C'. If you don't understand why, ask and I'll explain.
Unless you are pushing your 8051 derivative to the limit in terms of code size or speed than 'C' is almost certainly a good choice. I would suggest you persist with it but you will have to accept that you will have to buy the real version of the toolchain.
The code size limit you are approaching is designed, for obvious reasons, to prevent you from doing any serious work with it.
View all questions in Keil forum