OSDev.org

Hi I am trying to convert decimal into character I have noticed a pattern in the binary

0x01 = 00000001b
1 = 00110001b

My question is how can I change the last four bits of 0x01 into 0011 so it can be seen as a character?

If you could also give some examples that would be great thanks. I have looked into this already and can't find an answer

Many thanks

You convert an ascii encoded digit into its corresponding integer by the simple means of subtracting 48 from it (and vice versa).

Ok thanks, Do you know how I would change individual bits just for future reference?

Do this has anything to do with OS development?

Of course it does read the title. My question was about changing bits individually or merging registers which is a processor question and it is important for OS development

computertrick wrote:Of course it does read the title. My question was about changing bits individually or merging registers which is a processor question and it is important for OS development

That is the sort of thing any assembly language or C programmer should know, long before he/she even thinks of embarking upon OS development.

Look up the assembly language OR, AND, XOR and NOT instructions, or the C ~, |, &, ^ operators.

I am aware of all of those instructions my question was is their an instruction in 8086 to change a single bit that's what I was trying to get across

computertrick wrote:I am aware of all of those instructions my question was is their an instruction in 8086 to change a single bit that's what I was trying to get across

No, that is absolutely not possible.

computertrick wrote:I am aware of all of those instructions my question was is their an instruction in 8086 to change a single bit that's what I was trying to get across

Those instructions do precisely that, and more besides. For example, xor ax, 4 would flip bit 2 of the ax register.

For that matter, you can do it even without those instructions. add ax, 0x80 would set bit 7, and sub ax, 0x80 would clear it. That is elementary, and, if you haven't got that far, it will be a long time before you are ready for osdev.

computertrick wrote:Of course it does read the title. My question was about changing bits individually or merging registers which is a processor question and it is important for OS development

Summon the triple /faceplam...

While OS development require excellent skills of all aspect, including flipping a bit, I suggest you read the Required_Knowledge.

BTS, BTR (80386 and later).

For that matter, you can do it even without those instructions. add ax, 0x80 would set bit 7, and sub ax, 0x80 would clear it. That is elementary, and, if you haven't got that far, it will be a long time before you are ready for osdev.

That is incorrect. What if the bit is already set (or clear)? Adding (or subtracting) would not have the desired result.

iansjack wrote:

That is incorrect. What if the bit is already set (or clear)? Adding (or subtracting) would not have the desired result.

What has that got to do with anything? or ax, 0x80 wouldn't change a bit which was already set, and "and ax, 0xff7f" wouldn't clear a bit which was already clear. That is in the nature of bitwise operations.

Casm wrote:

iansjack wrote:

That is incorrect. What if the bit is already set (or clear)? Adding (or subtracting) would not have the desired result.

What has that got to do with anything? or ax, 0x80 wouldn't change a bit which was already set, and "and ax, 0xff7f" wouldn't clear a bit which was already clear. That is in the nature of bitwise operations.

Read the bit of your post that I quoted again. add is not and, sub is not or. add would clear a set bit (and affect other bits), sub would set a clear bit (and affect other bits). That is the nature of arithmetic operations.

iansjack wrote: Read the bit of your post that I quoted again. add is not and, sub is not or. add would clear a set bit (and affect other bits), sub would set a clear bit (and affect other bits). That is the nature of arithmetic operations.

If I wanted to convert a digit into its corresponding integer value, I could either do and ax, 0xcf or sub ax, 48, and which I did would depend upon the mood I was in at the time. Similarly, and ax, 0xdf and sub ax, 32 will both convert a lower case letter into an upper case letter (or clear the archive bit in an MS-DOS fille attribute byte).

I don't remember saying that sub was the same as or (or and).