New memeasm commands by babyygemperor · Pull Request #83 · kammt/MemeAssembly

babyygemperor · 2023-06-11T09:07:59Z

Pull Request Changelog

Alright, let's take a quick squiz at what I've done in this pull request.

"knock knock, who's there? {p}" - This function is asking the universe for a value at a certain memory location. It moves the value at the address pointed to by {p} into rax.
"why don't you come on in, {p}" - This function moves the value from rax into the memory location pointed to by {p}.
"big brain time {p} {p}" - This moves the value of the second parameter {1} into the memory location pointed to by the first parameter {0}.
"execute order 66 {p}" - This function moves the number 66 into rax, and then moves rax into the memory location pointed to by {p}.
"I don't feel so good" - This function sets all registers to zero.
"just a little switcheroo {p} {p}" - This one swaps the contents of the two registers {0} and {1} using only logical operators.
"deja vu" - This function jumps back to the start.
"why are we still here, just to suffer" - This function generates a divide-by-zero error and crash the program.
"you're in the wrong neighbourhood {p}" - Generates a random number and jumps to that address.
"stop, you violated the law" - This function stops the program execution.
"into the multiverse {p}" - This function takes an integer as an input and then performs a series of operations that essentially multiply the input by 2, subtract 2, and then check if the result is 0. If it's not zero, it goes back to the start and repeats the process. This is like entering multiple universes where the same actions keep happening.
"you're invited to suffer {p}" - This function takes an integer as an input and compares it with the number 666 in a loop. If it's not 666, it increments the input and repeats the process. This creates a cycle of suffering until the input becomes 666.
"hollup, let him cook {p}" - It initialises eax, ebx, and ecx to 0 and enters a "cook" loop where it increments eax, ebx, and ecx by one. It then moves the value of eax into edx and compares edx to {p}, if they are not equal it jumps back to "cook".
"I'm feeling lucky {p}" - It triggers the interrupt associated with {p}.

Enjoy lol

Bumps [actions/upload-artifact](https://github.com/actions/upload-artifact) from 2 to 3. - [Release notes](https://github.com/actions/upload-artifact/releases) - [Commits](actions/upload-artifact@v2...v3) --- updated-dependencies: - dependency-name: actions/upload-artifact dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com>

Bumps [actions/checkout](https://github.com/actions/checkout) from 2 to 3. - [Release notes](https://github.com/actions/checkout/releases) - [Changelog](https://github.com/actions/checkout/blob/main/CHANGELOG.md) - [Commits](actions/checkout@v2...v3) --- updated-dependencies: - dependency-name: actions/checkout dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com>

…ons/checkout-3

…ons/upload-artifact-3

aengelke

Did you, like, you know, test these commands? 😅

aengelke · 2023-06-11T09:38:04Z

compiler/compiler.c

+            .usedParameters = 1,
+            .allowedParamTypes = {PARAM_REG | PARAM_DECIMAL},
+            .analysisFunction = NULL,
+            .translationPattern = "mov rax, [rip + {0}]"


Not going to work with a register (also on other occasions).

aengelke · 2023-06-11T09:39:00Z

compiler/compiler.c

+            .usedParameters = 2,
+            .allowedParamTypes = {PARAM_REG, PARAM_REG | PARAM_DECIMAL},
+            .analysisFunction = NULL,
+            .translationPattern = "mov [rip + {0}], {1}"


Same here. You might want to make .text rwx, too..

aengelke · 2023-06-11T09:42:10Z

compiler/compiler.c

+            .usedParameters = 1,
+            .allowedParamTypes = {PARAM_REG64 | PARAM_REG32 | PARAM_REG16},
+            .analysisFunction = NULL,
+            .translationPattern = "rdrand {0}\n\tjmp {0}"


jmp r/m only supports 16/64 bit opsize, and 16-bit opsize is useless as Linux by default refuses to map to the lowest 64 kiB.

aengelke · 2023-06-11T09:44:08Z

compiler/compiler.c

+            .usedParameters = 1,
+            .allowedParamTypes = {PARAM_REG64 | PARAM_REG32 | PARAM_REG16},
+            .analysisFunction = NULL,
+            .translationPattern = "mov ecx, {0}\nmultiverse:\nadd ecx, ecx\nsub ecx, 2\ncmp ecx, 0\njnz multiverse\ninc ecx\ndec ecx\nmov {0}, ecx\npushad\npopad\nmov eax, 2\ndiv eax"


pushad/popad don't exist in 64-bit mode.

aengelke · 2023-06-11T09:44:29Z

compiler/compiler.c

+            .usedParameters = 1,
+            .allowedParamTypes = {PARAM_REG64 | PARAM_REG32 | PARAM_REG16},
+            .analysisFunction = NULL,
+            .translationPattern = "mov eax, {0}\nsuffer:\ncmp eax, 666\nje end_suffer\nadd eax, 1\njmp suffer\nend_suffer:\npush eax\npop eax\nxor eax, eax\ninc eax\nadd eax, 2\ndec eax\ndec eax\nmov {0}, eax"


push r32 doesn't exist in 64-bit mode...

aengelke · 2023-06-11T09:45:04Z

compiler/compiler.c

+            .usedParameters = 1,
+            .allowedParamTypes = {PARAM_REG64 | PARAM_REG32 | PARAM_REG16},
+            .analysisFunction = NULL,
+            .translationPattern = "mov eax, 0\nmov ebx, 0\nmov ecx, 0\ncook:\ninc eax\ninc ebx\ninc ecx\nmov edx, eax\ncmp edx, {0}\njne cook\n"


Only works with REG32.

aengelke · 2023-06-11T09:45:20Z

compiler/compiler.c

+            .usedParameters = 1,
+            .allowedParamTypes = {PARAM_REG | PARAM_DECIMAL},
+            .analysisFunction = NULL,
+            .translationPattern = "int {0}"


Only works with integer between 0 and 255.

kammt · 2023-06-11T13:50:53Z

compiler/compiler.c

+                .pattern = "deja vu",
+                .usedParameters = 0,
+                .analysisFunction = NULL,
+                .translationPattern = "jmp main"


FYI: MemeASM only requires a main-function when you create an executable. When you create an assembly file (-S) or an object file (-O), using this command would result in the generated code having a potentially undefined reference (Since you could link it with something that defines start, but not main)

kammt · 2023-06-11T14:35:24Z

First of all, thank you for your contribution!
Could you change the PR title to something more fitting? "New features" is quite meaningless.

Regarding the proposed commands, here are my initial thoughts:

So this is just an alias for rax is brilliant, but I like {p} to you know de wey? I don't really see why you'd need a special command for that 😅. Also, this is not a Meme.
Same here
Same here. Also, what does the meme have to do with the command's behavior?
I like the command idea (=the meme), but not what it does. Maybe I can come up with a better behavior, I'll have a think about that one (Maybe it could call a function called order66 that the programmer can define?)
I'd change it to Mr. Stark, I don't feel so good, but I like the idea
This is also not a Meme
See my review comment. I also wouldn't call that a Meme. ah shit, here we go again would be better imo (That command was in v0.1 with exactly that behavior). It would probably be a better implementation if the command jumped back to the start of the current function (That would require an update to the translation patterns in translator.c to support that though)
Sure, why not
I like to idea, but why would you require a register parameter for that? push rax, generate a random number, move it to .Ltmp64, pop rax, jump
Well, I wouldn't call it stopping, it essentially generates a GP-Fault. You could also describe command number 8 as "stops program execution". Not sure about that one, since we already have you shall not pass! and there are already two commands that crash the program.
That's not a Meme? Also, I don't really see the point behind that one 😅
Hm, I don't know about that one. Also, writing it out as a sentence kind of destroys the reference to the meme (something like you're invited to: suffer would be more fitting imo). A bit like number 4: I like the idea (= the meme that was used), but not what the command actually does
How did you come up with that behavior? This command could be used as an implementation for wait(2) when I come to implementing a fork-command (yes, that is planned). Your suggested behavior seems very random
I don't think that the meme is fitting here. But it's true that access to int could be a good feature.

TL;DR: This is not something I'd merge right away. I'll leave it open and have a think about what makes sense and what doesn't. Right now, I would adapt (i.e. merge with a few modifications) 5, 8, 9 and maybe 10. @xarantolus what are your thoughts on this?

J-MR-T · 2023-06-11T16:39:50Z

I'd change it to Mr. Stark, I don't feel so good, but I like the idea

I also like that mnemonic, but I think setting all registers to 0 is almost a bit too useful for this project. What about sending a random signal to the parent process or maybe the process itself (that would of course only work on POSIX adjacent OSes, I'm not sure if execution on windows is supported?)?
Only if there isn't already a similar command ofc, I have a vague recollection of you telling me about something along those lines @kammt?

10/14.

I actually think that access to int via Stop, you vialoated the law! {p} would be quite fitting, as that comes as close to a meme interrupting something as I can think of.

kammt · 2023-06-12T10:32:36Z

I also like that mnemonic, but I think setting all registers to 0 is almost a bit too useful for this. What about sending a random signal to the parent process or maybe the process itself

Well, it should revolve about the program becoming unstable. This would work with zeroing all registers, since it also zeroes the Stack Pointer (however, this is also achieved with Houston, we have a problem, so that's a little boring).
The problem I see with sending a random signal is that most of the time, this would just result in the program getting killed by the signal, with no "slow deterioration" involved. What could be done is that after that command has been read, the compiler will add instructions which alter the program's state (Stack Pointer, GP-Registers, or sending a signal). That would ofc be more work on the development side than just sending a random signal.

I'm not sure if execution on windows is supported?

Well, you can theoretically run it on Windows, but everything involving Signals until now (i.e. Martyrdom) is not supported

Only if there isn't already a similar command ofc, I have a vague recollection of you telling me about something along those lines @kammt?

I think you're talking about Martyrdom, where the MemeASM program sends SIGKILL to the parent process when it receives SIGINT

I actually think that access to int via Stop, you vialoated the law! {p} would be quite fitting, as that comes as close to a meme interrupting something as I can think of.

That's a good idea!

J-MR-T · 2023-06-13T05:22:48Z

Well, it should revolve about the program becoming unstable.

I see, that makes sense. Maybe I'm a bit obsessed with signals atm, because all my ideas seem to be based on them, but here I go anyways ^^: What about the system as a whole becoming unstable with the use of that command? Sending a random signal to a random (user) process might be an option. Could kill the x server, file manager deamon, or similar, or do almost nothing ^^.

But on second thought, that is a bit detached from the meme itself.

kammt and others added 17 commits May 7, 2023 14:30

Merge pull request kammt#80 from kammt/develop

873036a

publish-aur.yml: Repo name as variable

65120aa

publish-aur.yml: Allow manual dispatch

5530f8d

Merge pull request kammt#81 from kammt/dependabot/github_actions/acti…

dc75a29

…ons/checkout-3

Merge pull request kammt#82 from kammt/dependabot/github_actions/acti…

d7208b5

…ons/upload-artifact-3

publish-aur.yml: Hopefully fixed automatic and manual run

21ede48

publish-aur.yml: pkgrel as variable, change ' to " in sed

13619a7

publish-aur.yml: Typo

7fa4c05

PKGBUILD: Correct usage of DESTDIR

a41c928

nfpm: Simplified uninstallation script

d01d340

nfpm.yaml: Update package name

b7dfaed

create_installers.yml: Update package name, removed Windows compilation

4b94378

Added instructions on contribution

6962376

Added a fuck tonne of new commands

4864001

Added guidelines to PR Branch

58a8bd7

aengelke suggested changes Jun 11, 2023

View reviewed changes

kammt changed the base branch from main to develop June 11, 2023 13:23

kammt reviewed Jun 11, 2023

View reviewed changes

kammt added the enhancement New feature or request label Jun 11, 2023

babyygemperor changed the title ~~New features~~ New memeasm commands Jun 12, 2023

Conversation

babyygemperor commented Jun 11, 2023

Pull Request Changelog

Uh oh!

aengelke left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kammt commented Jun 11, 2023

Uh oh!

J-MR-T commented Jun 11, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

kammt commented Jun 12, 2023

Uh oh!

J-MR-T commented Jun 13, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

J-MR-T commented Jun 11, 2023 •

edited

Loading

J-MR-T commented Jun 13, 2023 •

edited

Loading