SFR的地址定义在哪里？

HI，连同XC32编译器的安装，有几个汇编程序代码的例子：程序文件（x86）Microchip xC32 v1.43 示例CalulySO示例……包括一个程序，该程序执行一些端口控制操作：还需要几个汇编指令。迈西尔



以下为原文

Hi,
Together with XC32 Compiler installation, there are a few examples of Assembler code program:
Program Files (x86) Microchip xc32 v1.43 examples assembly_examples ...

Including a program that do some Port control operations:
#include
#define IOPORT_BIT_7 (1 << 7)
...
ori a0, zero, IOPORT_BIT_7
la s0, LATACLR
sw a0, 0(s0) /* clear specified bits */
There are several assembler directives also needed.

Mysil

嗨，谢谢你的帮助。我看到了这些例子，它们是有意义的，但是他们使用宏指令（LA，里县等），这些文件在任何地方都没有记录（甚至在IMGTEC网站上）。更基本的是，我想看看我是否能手动编码所有的东西，我不知道如何访问刚才的地址。C定义了上面列出的，如果我使用C地址运算符和PoBB，C中给出定义的地址，也会得到错误，并且该行没有被正确地处理。例如，如果我使用PoTB的地址：Lui V0，0xBF86110＆Gt；G16.16ORI V0，0xBF86110＆0xFFFFI将喜欢T。这是一个学术性的练习，不会为实际的实际发展而这样做。



以下为原文

Hi, thank you for the help.

I saw those examples and they make sense but they use the macro instructions(LA, LI, etc) which aren't documented anywhere(even on imgtec website).

More fundamentally though I wanted to see if I could code everything manually, I am unsure how to access just the address of the C defines listed above, if I used the C address operator &PORTB which in C would give the address of the definition I also get an error and the line isn't processed correctly.

As an example if I use the address of PORTB:

LUI v0, 0xBF886110>>16
ORI v0, 0xBF886110 & 0xFFFF

I would like to be able to do that with a define.

This is an academic exercise, wouldn't be doing this for actual real development.

好的，我发现SFR定义了，它们在PIC32 LIBS/PROC//LT；Primor文件夹和gt；//lt；处理器s＆gt；汇编文件。GeeZe，两个星期后，我现在准备编写一个纯ASM程序。Microchip似乎对这些步骤不太感兴趣。



以下为原文

Ok, I found the SFR defines, they are in the pic32-libs/proc/ / assembly file.

Geeze, after two weeks I am now setup to write a pure ASM program. Microchip doesn't seem too interested in detailing these steps.

我一直在使用GNU工具链，只是使用数据表中的信息来维护我自己的定义。当然，也有一些像Pinguino这样可以提供可用文件的东西，虽然我不知道它们是否是Microchip提供的其他文件或其他东西。至于不同的指令，我似乎记得LA显然不是真正的指令。Falm是一个伪指令，它产生Li，但它的加载地址源于一个标签，它是由汇编程序查找的，而不是一个作为操作数立即可用的常数。但是，像GNU这样的东西，不管怎么说，都是可以互换的。我倾向于使用常量偏移量来传递任何给定的基寄存器的SET、CLR和IV变量。所以，像这样…显然，我必须定义CLR是适当的偏移量。



以下为原文

I've been doing this using the GNU toolchain and just maintaining my own #defines using the information from the datasheets. Of course, there are also things like Pinguino that provide usable files for this purpose, although I don't know if they are the same files that Microchip provides or something else.

As for the different instructions, I seem to recall that lais apparently not a genuine instruction as such: it is effectively a pseudo-instruction producing libut with the loaded address originating from a label looked up by the assembler, as opposed to a constant immediately usable as an operand. But things like GNU asseem to treat them interchangeably anyway.

I tend to use the constant offset to communicate the SET, CLR and INV variants of any given base register. So, something like this...

la $s0, LATA
sw $a0, CLR($s0)

Obviously, I have to #define CLR to be the appropriate offset.