Timeline for Make your language unusable
Current License: CC BY-SA 3.0
15 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 5, 2015 at 16:40 | comment | added | Joshua | I've been pondering, what would happen if the following code scanned memory to find the routines the IVT is supposed to point to. It can take as much time as it needs (we note that an IVT value of zero causes the CPU to discard the interrupt). | |
| Oct 26, 2015 at 21:57 | comment | added | Blacklight Shining |
@NateEldredge My understanding is that that requirement is meant to prevent things like exit(0); /* insert code here */. Not all of the inserted code has to execute, and an error can occur as soon as execution starts. The system just has to try to start executing it.
|
|
| Oct 24, 2015 at 14:15 | comment | added | Ruslan |
I'd rather go with something like this, not rely on the first instruction executed to not be cli. This of course does have workarounds too (stack-based), but not as straightforward, and can also be circumvented by supervisor.
|
|
| Oct 22, 2015 at 15:00 | comment | added | Matteo Italia | @kirbyfan64sos: yes; this should make a mess on all the calls to DOS and BIOS routines, plus whenever an interrupt (keyboard, timer, ...) is invoked. | |
| Oct 22, 2015 at 14:09 | comment | added | kirbyfan64sos | So this clears the interrupt table? (Not a DOS programmer!) | |
| Oct 22, 2015 at 5:54 | comment | added | Matteo Italia |
@NateEldredge: the rules aren't really clear on this, and if you look around most answers actually consist in modifications to the environment that generate a runtime error of some type on trivial instructions (the JS clear(this);, the memory limit in PHP, the recursion limit in Python, the sandboxed environment in Python and many others), I wouldn't see this as a problem.
|
|
| Oct 22, 2015 at 5:50 | comment | added | Matteo Italia | @JerryCoffin: of course, but any interrupt (I'm thinking keyboard and timer) should quickly break havoc. | |
| Oct 21, 2015 at 20:28 | comment | added | Jerry Coffin | At least as it stands, this won't stop a program from writing directly to the screen buffer (which was pretty common under DOS). | |
| Oct 21, 2015 at 19:23 | comment | added | Nate Eldredge | I actually thought about something a little similar. But I wondered about that as soon as your code attempts to make its first privileged instruction, system call, etc, you'll get a trap, and then the rest of the code will not in fact execute, which the rules seem to require. Unless you have supervisor code to manually restart at the next instruction - but then I think you have to be able to parse all instructions to determine their length, which is a pain. | |
| Oct 21, 2015 at 19:19 | comment | added | Matteo Italia |
@NateEldredge: the next step is to jail the rest of the code to ring 3 with no trampoline back to ring 0; I'll see if I manage to put together a working example (another possibility would be to scan the whole address space and NOP-out all the cli (and inp and outp just for good measure), but I don't know if that would be allowed by the rules.
|
|
| Oct 21, 2015 at 16:22 | comment | added | Nate Eldredge |
So what if the first instruction of my inserted code is cli, and then I fix the interrupt table, and go on to compute some primes, etc?
|
|
| Oct 21, 2015 at 14:52 | history | edited | Matteo Italia | CC BY-SA 3.0 |
Forgot xor di,di
|
| Oct 20, 2015 at 12:57 | comment | added | Luminous | Sir, if I may interrupt you for a few cycles so I may fini.... | |
| Oct 20, 2015 at 12:26 | history | edited | Matteo Italia | CC BY-SA 3.0 |
added 79 characters in body
|
| Oct 20, 2015 at 12:20 | history | answered | Matteo Italia | CC BY-SA 3.0 |