Code: Select all
xorps xmm5,xmm5
movss [rsi+58],xmm5Any other way of moving 0 to xmm5?
Empyrion XMM Register
Empyrion XMM Register
Ok so this is my code:
I wanted to mov 0 to xmm5, is this the right way of doing it, will it effect other stuff in the game by using the xorps code.
Any other way of moving 0 to xmm5?
Any other way of moving 0 to xmm5?
Re: Empyrion XMM Register
Hm, you've asked variations on this same question more than once now...
Quick answer: unless there's a compare after the instructions you've added or based on the value that you're changing (possibly due to it being copied elsewhere), it's not going to effect other stuff. Side effects for a single instruction tend to be pretty short lived and obviously checked. Also, unless you've written several lines of code to achieve your goal there probably isn't any "better" or "more right" way to do anything, simply other ways that achieve the same effect. So xorps is fine.
If there was another register that always held 0 then you could probably avoid having to allocate memory by changing the bytes of the movss instruction so that it uses that register instead of xmm5 but... I don't know that that's really "better". I've had dozens of scripts that each allocate 1000 memory before and never seen any issues with it, and why should I? After all, games tend to use several gigabytes of memory and are running on processors designed to execute billions of instructions...the relatively infinitesimal modifications scripts make seem very unlikely to cause any performance issues (there are exceptions like in often run loops but in general).
Long Answer:
What have you googled or watched to help figure it out on your own? And yes, I see that you've shown a solution with xorps but...
I found [Link] which provides some good information and explanation but mostly covers floats, while [Link] has a much more comprehensive listing that you could look at and google for more info on each specific instruction like xorps (and you could then google what bitwise operations are and how they work if necessary).
For an instruction that changes the typical flags like add you can usually google something like "assembly flags add" and find a reference page that mentions what flags it changes (and thus could affect other stuff if there's a compare after your injection) like [Link]
I haven't really found that to be the case with SSE instructions however, the compares do tend to be simpler with them in my (somewhat limited) experience though.
Have you found What's a Creel's youtube tutorials yet? [Link]
I know it can be annoying to have people ask "well what have you done yourself already?" but there are 2 main reasons for it:
A) you really _can_ learn more by figuring it out yourself (though not necessarily as quickly). It's more likely to "click" for you if you have to go through the process of encountering the problem, looking up information, testing, looking up more if it didn't work, and finally getting a working solution. And you're more likely to remember the work you did to find the solution the next time you need to do something similar, compared to simply asking and having the answer given to you.
B) By asking you are essentially asking people to spend their finding or writing the answer for you, often with relatively little proof that you've spent much of _your_ time trying to figure it out, which can be annoying
Quick answer: unless there's a compare after the instructions you've added or based on the value that you're changing (possibly due to it being copied elsewhere), it's not going to effect other stuff. Side effects for a single instruction tend to be pretty short lived and obviously checked. Also, unless you've written several lines of code to achieve your goal there probably isn't any "better" or "more right" way to do anything, simply other ways that achieve the same effect. So xorps is fine.
If there was another register that always held 0 then you could probably avoid having to allocate memory by changing the bytes of the movss instruction so that it uses that register instead of xmm5 but... I don't know that that's really "better". I've had dozens of scripts that each allocate 1000 memory before and never seen any issues with it, and why should I? After all, games tend to use several gigabytes of memory and are running on processors designed to execute billions of instructions...the relatively infinitesimal modifications scripts make seem very unlikely to cause any performance issues (there are exceptions like in often run loops but in general).
Long Answer:
What have you googled or watched to help figure it out on your own? And yes, I see that you've shown a solution with xorps but...
I found [Link] which provides some good information and explanation but mostly covers floats, while [Link] has a much more comprehensive listing that you could look at and google for more info on each specific instruction like xorps (and you could then google what bitwise operations are and how they work if necessary).
For an instruction that changes the typical flags like add you can usually google something like "assembly flags add" and find a reference page that mentions what flags it changes (and thus could affect other stuff if there's a compare after your injection) like [Link]
I haven't really found that to be the case with SSE instructions however, the compares do tend to be simpler with them in my (somewhat limited) experience though.
Have you found What's a Creel's youtube tutorials yet? [Link]
I know it can be annoying to have people ask "well what have you done yourself already?" but there are 2 main reasons for it:
A) you really _can_ learn more by figuring it out yourself (though not necessarily as quickly). It's more likely to "click" for you if you have to go through the process of encountering the problem, looking up information, testing, looking up more if it didn't work, and finally getting a working solution. And you're more likely to remember the work you did to find the solution the next time you need to do something similar, compared to simply asking and having the answer given to you.
B) By asking you are essentially asking people to spend their finding or writing the answer for you, often with relatively little proof that you've spent much of _your_ time trying to figure it out, which can be annoying
Re: Empyrion XMM Register
yea I know I asked but i just wanted to know if using this xorps will cause issues in the future that's all, sorry 
Re: Empyrion XMM Register
Hey, I'm trying to cmp byte so i'm doing that:
cmp byte ptr[rsi+21],80
cmp byte ptr[rsi+22],83
cmp byte ptr[rsi+23],107
One of those offset and value are for Player, so I was wondering am I doing it right, the values are in byte form. so am I doign the cmp right?
cmp byte ptr[rsi+21],80
cmp byte ptr[rsi+22],83
cmp byte ptr[rsi+23],107
One of those offset and value are for Player, so I was wondering am I doing it right, the values are in byte form. so am I doign the cmp right?
Re: Empyrion XMM Register
Yeah, you use byte ptr to tell the assembler to check 1 byte rather than the default 4 bytes. 107 however is not a valid 1 byte value, FF is the largest 1 byte value in hexadecimal (each hex digit represents 4 bits, sometimes called a nibble, so 2 represents 1 byte), or 255 in decimal ((int)255 or #255) and remember that the default is hexadecimal...
The values are assuming that it's unsigned, FF/255 is -1 if it's signed, which makes the largest value 7F or 127.
The values are assuming that it's unsigned, FF/255 is -1 if it's signed, which makes the largest value 7F or 127.
Re: Empyrion XMM Register
Thank you
Who is online
Users browsing this forum: No registered users